Abstract
Over the past decade many new and important discoveries have been made at numerous Early Bronze Age I (EB I) sites in central but especially southern Israel which have direct bear- ing on the earliest Egyptian-Canaanite interconnections attested thus far (cf., e.g., Brandl 1992; Gophna 1992; Kempinski 1992; Kempinski and Gilead 1991; Oren 1989; Oren and Yekutieli

References



Anatolia Kültepe 2 — Özgüç 1986: pl. 6:70

Acemhöyük 1 — Erkanal 1977: 22, pl. 6:70

Syria Ebla 1 Tomb Q.78.B1 Matthiae 1980: fig. 11

Ugarit 20 Tombs Shaeffer 1949: figs. 18, 25; 1962:
fig. 26, 1978: fig. 9:1

Tell Sukas 2 No context Buhl 1983: pl. XXII:367–68

Baghouz 11 Tombs Du Mesnil du Buisson 1948: pls. XLV–LVII

Mari 1 — Parrot 1959: 85, pl. 33:999

Hama 2 Tomb G.I Fugmann 1958: pl. X:5A812–813

Kedesh 3 No context Petrie 1917: pl. VI:169–71

Tell el-Tin 1 Tomb Oren 1971: 122; Gautier 1895

Yabrud 4 Tomb 4 Assaf 1967

Lebanon Byblos 2 Tomb III Montet 1928: 247–48, pl. CXLIX:940–41

Byblos 4 Hoards Dunand 1950: 389:10645–6;
Seeden 1980: pl. 126:4–7

Lebea 1 Tomb 1 Guigues 1937: 39

Kfar Jarrah 1 Tomb 57 Guigues 1938: 30–34

Sin el-Fil 2 Tomb Chehab 1939

Palestine ºEin Saªed 1 No context Epstein and Gutman 1972: 290

Shamir 1 Surface find Miron 1992: 54

Tel Dan 1 Glacis Biran 1994: 63–65; Ilan 1992

Safed 2 — Miron 1992: 54, Bahat excavations

Safed 2 Tomb Damati and Stepanski 1987–88

Meron 1 Tomb 1 Tpilinsky 1962: 25; Miron 1992: 53

ash-
Shajeriyeh

1 Chance find Maxwell-Hyslop 1949: 121

Kabri 1 Tomb 990 Gershuny 1989

Nahariya 1 Chance find Miron 1992: 53

Tell Kurdaneh 1 Tomb Maisler 1939: 154; Miron 1992: 54

Gesher 3 Three tombs Garfinkel and Bonfil 1990

Beit Shean 1 Tomb 92 Oren 1971; 1973: 61–67

Tel Rehov 1 Tomb 2 Yogev 1985: 104–105

ºEin es-
Samiyeh

2 Cemetery Dever 1975: 30; Miron 1992: 54, from the
market

Aphek 2 Two tombs Miron 1992: 54

Ashkelon 1 No context Maxwell-Hyslop 1949: 121

Egypt Tell el-Dabªa 1 Built tomb Bietak 1991: 56–58

Cyprus ? Few No context Maxwell-Hyslop 1949: 121



Total 31 sites 81

oi.uchicago.edu





145



WARRIOR BURIAL CUSTOMS IN THE LEVANT



In the cemetery fourteen graves were unearthed (fig. 8.1)



1



. They had been cut into the soft
sediment of the Jordan Valley. This sediment had caved in over time, so it was impossible to
discern the layout of the graves. However, the heaps of stones which were found near the skel-
etons, together with the grave goods, suggest that the cemetery at Gesher consists of shaft
tombs, and that the stones had originally sealed the burial chambers. Similar Middle Bronze
II tombs have been reported from other sites in the southern Levant, such as Jericho (Kenyon
1960: figs. 139, 159; 1965: figs. 132, 138, 211) and Tel Aviv (Kaplan 1955: pl. I:1).

The most common burial pattern at Gesher is of one body in each tomb. The burial is pri-
mary, and the corpse was placed with its head facing east and its legs extending to the west.
The grave goods included a number of clay vessels and occasionally bronze items. Weapons
were recovered among the other grave goods in four tombs, which were thus identified as
warrior tombs:

Grave 2 (fig. 8.2). This is an individual primary burial. The corpse was laid in a semi-
flexed position on an east-west axis, its head facing east and its legs west. The knees were bent
northward. The head was placed on two flat basalt stones. At the feet of the deceased a flat



1. The “grave” labeled number 6 on figure 8.1 turned out to be only a group of stones.

FIGURE 8.1. The Middle Bronze IIA cemetery at Gesher.

oi.uchicago.edu

[image: image43.tif]



146



YOSEF GARFINKEL



stone was placed, sealing the corpse on the west. A bronze duck-bill ax (pl. 8.1a) and a bowl
were found next to the head. Next to the bowl, animal bones, which had served as food offer-
ings, were discovered. A large jar and a bronze spearhead were placed by the corpse’s legs. To
the north of the skeleton a rectangular stone heap was discovered, sealing the entrance to the
burial chamber.

Grave 12 (fig. 8.3). This is an individual primary burial. The corpse was laid in a semi-
flexed position on an east-west axis, its head facing east and its legs west. The skeleton was
found close to the modern ground surface, and little of it was preserved. Next to the head a
jar was found with a bronze duck-bill ax (pl. 8.1b) close by. Ten centimeters from the ax was
a small bronze nail. North of the corpse a massive row of stones, five to six courses high, was
unearthed.



FIGURE 8.2. Grave 2 at Gesher.

spread one pica long

oi.uchicago.edu

[image: image44.tif]



147



WARRIOR BURIAL CUSTOMS IN THE LEVANT



FIGURE 8.3. Grave 12 at Gesher.

FIGURE 8.4. Grave 13 at Gesher.

oi.uchicago.edu

[image: image45.tif][image: image46.tif]



148



YOSEF GARFINKEL



Grave 13 (fig. 8.4). This is an individual primary burial. The corpse was laid in a semi-
flexed position on an east-west axis, its head facing east and its legs west. The face and knees
are turned to the north. The skull is laid on a flat basalt stone. Just to the east of the head a
bronze duck-bill ax was found (pl. 8.2a). A bronze spearhead was recovered beneath the elbows.
North of the corpse’s breast were a jar and a bowl. Animal bones, which had served as food
offerings, were found by the bowl.

Grave 14 (fig. 8.5). This is an individual primary burial. The corpse was laid in a semi-
flexed position on a north-south axis, its head to the south and its legs to the north. This
was the only burial in which the deceased was not laid on an east-west axis. The legs are
flexed and the knees face east. A jug and a bowl decorated with four knobs were found
close to the head. Inside the bowl lay a narrow, elongated socket ax (pl. 8.2b, c). A stone
heap, composed of two rows of stones, was found to the east of the corpse. East of the
stone heap, about 35 cm higher than the level of the skeleton, three more clay vessels were
discovered: a bowl turned upside down and two juglets. It seems that these items were



FIGURE 8.5. Grave 14 at Gesher.

oi.uchicago.edu

[image: image47.tif]



149



WARRIOR BURIAL CUSTOMS IN THE LEVANT



placed in the tomb shaft and not in the chamber. A similar case has been reported from
Jericho (Kenyon 1965: figs. 132, 139). It appears that these items had been added later and
were not part of the original burial equipment. These three items are thus not included in
table 8.3 and in fig. 8.11.



Warrior Burials at Baghouz



The site of Baghouz lies 10 km south of Mari, near the Syrian-Iraqi border. A total of 320
tombs were excavated (Du Mesnil du Buisson 1948; Hrouda 1990) in an area of about 1 sq km
(fig. 8.6). In Baghouz, as shown in table 8.2, eleven warrior tombs were reported, and in nine
cases the exact location of the tomb on the site can be traced (the report fails to record the loca-
tion of Tombs 141 and 305). The nine tombs whose location is clear are grouped together in a



FIGURE 8.6. General plan of the Baghouz cemetery (after Du Mesnil du Buisson 1948: pl. VI).

oi.uchicago.edu

[image: image48.tif]



150



YOSEF GARFINKEL



limited area on “Mamelon II” (fig. 8.7). It should be stated that Tomb 309 is not marked on the
original map of “Mamelon II” (Du Mesnil du Buisson 1948: pl. XXXIX), and its position is
based on information gathered from the general map (pl. VI, Mamelon II bis).

A careful study of the map of “Mamelon II” shows that forty-eight tombs were found in
this area, of which forty are numbered and another eight are not. Such a discrepancy is not sur-
prising for an excavation that took place in the 1930s. It is quite likely that Tombs 141 and 305
are also located in this area, since tombs with consecutive numbers such as 309 or 143 are
noted on “Mamelon II.” Be that as it may, if we only take into account the forty tombs marked
on the map, the nine warrior burials constitute 22.5% of the graves. If eleven warrior tombs
are calculated as a percentage of forty-eight tombs, they constitute 23%.



FIGURE 8.7. Detailed plan of the tombs at “Mamelon II” (after Du Mesnil du Buisson 1948: pls. VI, XXXIX).

oi.uchicago.edu

[image: image49.tif]



151



WARRIOR BURIAL CUSTOMS IN THE LEVANT



The warrior burials were found inside stone-built tumuli, as were other tombs at the site.
Inside the tombs individual burials were found. The corpses were laid out in a flexed position,
on wooden biers that were preserved because of the dry desert climate. Various offerings were
found next to the deceased, including duck-bill axes, spearheads, daggers, clay vessels, wood
and bone tools, and animal bones. The plans of these warrior burials, as published in the exca-
vation report, are presented together, at the same scale in figure 8.8.



FIGURE 8.8. Warrior tombs at Baghouz (after Du Mesnil du Buisson 1948: pls. XLV, XLVIII, LI, LII, LIV,
LVII).

oi.uchicago.edu

[image: image50.tif]



152



YOSEF GARFINKEL



Table 8.2.



Data on Warrior Burials from the
Baghouz Excavation Report



No. Tomb no. Plan Location



1. 67 XLII Mamelon II

2. 95 XLV Mamelon II

3. 102 — Mamelon II

4. 103b — Mamelon II

5. 121 XLVIII Mamelon II

6. 122 LI Mamelon II

7. 123 LII Mamelon II

8. 141 LVII —

9. 143 LIV Mamelon II

10. 305 — —

11. 309 — Mamelon II



FIGURE 8.9. Tomb 2 at Rehov (after Yogev
1985: 93).

FIGURE 8.10. Tomb 990 at Kabri (after Gershuny 1989).

oi.uchicago.edu

[image: image51.tif][image: image52.tif]



153



WARRIOR BURIAL CUSTOMS IN THE LEVANT



Warrior Burial from Rehov



Tel Rehov lies in the central Jordan Valley, near Beit Shean. Nine tombs dated to the fol-
lowing periods were excavated near the tell: Middle Bronze I, Middle Bronze IIA, and Middle
Bronze IIB (Yogev 1985). In Tomb 2, dated to the Middle Bronze IIA, a single skeleton in a
flexed position was unearthed (fig. 8.9). The excavators described it as follows:



In the burial chamber, about half a meter from the corridor, an intact jar was uncovered
standing on the floor. Beyond were the remains of a skeleton, laid from S to N in a flexed
position, facing E. To the right of the pelvic region (the bones of which were not
preserved) was a dagger and an alabaster crescentic pommel. A few centimeters to the
right of the skull was a duck-bill ax-head, with its cutting-edge facing the skull. Two
spear-heads were uncovered in the northern part of the chamber beyond the limbs.
(Yogev 1985: 93)



Warrior Burial from Kabri



One warrior burial (Tomb 990) was discovered at the site of Kabri in western Galilee
(fig. 8.10). The skeleton was found in a flexed position together with a duck-bill ax and clay
vessels (Gershuny 1989). Only an artist’s reconstruction of the burial was published in a pre-
liminary report, without a conventional tomb plan (Gershuny 1989: fig. 3). The ceramic assem-
blage includes a bowl decorated with knobs and a juglet near the head, a juglet near the pelvis,
and two closed vessels (a jug and a small jar with no handles) near the legs (Gershuny 1989:
14, fig. 14). The juglet near the head does not appear in the drawing of the tomb (Gershuny
1989: fig. 3) nor in the drawing of the pottery (Gershuny 1989: fig. 14), and is only mentioned
in the text (Gershuny 1989: 14). A second excavation season revealed that another person had
been interred together with the first corpse (Gershuny, pers. comm.).

DISCUSSION



Treatment of the Corpse



Despite the differences in shape of the tombs in the various sites—tumuli at Baghouz, a
rock-hewn cave at Rehov, and shaft tombs at Gesher—similar patterns can be observed in the
treatment of the bodies and the offerings at these sites:

1. Primary burial. The bones are found in anatomical order, suggesting no special
retreatment of the corpse.

2. Flexed burial. The dead were placed with their legs folded.
3. Individual burial. Only in Kabri Tomb 990 is another skeleton associated with the

warrior.



Composition of the Offerings



Details of the weapons and clay vessels (mainly jars, jugs, and bowls) placed as offerings
in the warrior tombs of Baghouz, Gesher, Rehov, and Kabri are presented in table 8.3. These
do not include the wooden tables and beds found at Baghouz since such items could not have
been preserved in the other sites. One should not rule out the possibility that wooden items
were also deposited in the other three sites since wooden items were preserved in Middle
Bronze tombs at Jericho, thanks to the dry climate (Kenyon 1960: 327, 340, 382–90, 462–66).

oi.uchicago.edu





154



YOSEF GARFINKEL



It is evident that the offerings in warrior tombs consisted of the following items:



Closed vessels:



A total of sixteen closed vessels (jars, jugs, and juglets) were discovered
in the thirteen tombs. Each of the burials was provided with one jar or one jug, probably con-
taining beer (Maeir and Garfinkel 1992; Gates 1988: 69–73). Two small juglets were reported
in only one case, at Kabri, in addition to a jar and a jug. Perhaps they could be interpreted as
offerings for the second corpse buried there.



Axes:



Thirteen axes were reported from thirteen tombs. Thus every warrior was buried
with one ax. Most axes are of the duck-bill type and only one (in Gesher Grave 14) was a nar-
row, elongated socket ax.



Bowls:



Thirteen bowls were reported from ten tombs. At Baghouz some of the bowls were
made of wood. Seven tombs contained one bowl for each warrior, and three tombs contained
two bowls for each person. It should be noted that in Grave 14 at Gesher and in Tomb 990 at
Kabri the bowls were decorated with knobs. This kind of bowl has often been reported from
tombs where weapons were found, such as Tomb 1 at Ginnosar (Epstein 1974: fig. 7:3, 15)
and Level 3 of Tomb IV at Tell Sukas (Thrane 1978: 25–26, figs. 32–33, 78, 86).



Spearheads:



Thirteen spearheads were reported from ten tombs. Eight of the burials had
one spearhead each, while the warrior buried in Grave 2 from Rehov had two, and the warrior
interred in Tomb Z-95 at Baghouz had three.



Daggers:



Two daggers were reported, each from a different tomb.



Table 8.3.



Weapons and Vessels Placed as Offerings in Warrior Burials



Tomb Jar/Jug Ax Bowl Spearhead Dagger Juglet



Baghouz Z-67 1 1 1 1 — —

Baghouz Z-95 1 1 — 3 — —

Baghouz Z-121 1 1 1 1 — —

Baghouz Z-122 1 1 2 1 1 —

Baghouz Z-123 1 1 2 1 — —

Baghouz Z-141 1 1 2 1 — —

Baghouz Z-143 l l l l — —

Gesher 2 1 1 1 1 — —

Gesher 12 1 1 — — — —

Gesher 13 1 1 1 1 — —

Gesher 14 1 1 1 — — —

Rehov 2 1 1 — 2 1 —

Kabri 990 2 1 1 — — 2

Number of items 14 13 13 13 2 2

Number of tombs 13 13 10 10 2 1

oi.uchicago.edu





155



WARRIOR BURIAL CUSTOMS IN THE LEVANT



The variety of items placed in the tombs was quite limited, as shown in table 8.3. The stan-
dard assemblage seems to have been four items that included an ax, a spearhead, a jar/juglet,
and a bowl. Sometimes one or another of the items was lacking and sometimes one was added,
but these variations do not alter the composition of the basic paraphernalia of the deceased. It
should be stated that other items which are quite common in tombs from this period, such as
toggle pins or small carinated bowls, were not found at all in the warrior tombs. Thus the
items were not placed in the tomb at random but rather according to an accepted norm linked
to the dead person’s position and social status.



Position of Offering in Relation to the Corpse



Figure 8.11 illustrates the position of the offerings in the tombs in relation to the skeleton.
The following patterns can be discerned:



Position of Ax



: Out of thirteen axes discovered in the various tombs, twelve (92%) were
placed close to the head of the deceased, and one (8%) was found near the pelvis.



Position of Spearhead:



Out of thirteen spearheads discovered in ten different tombs,
eight (61%) were found at the feet of the corpse, four (31%) close to the head, and one (8%)
close to the hands.



Position of Dagger:



Out of two daggers found in two tombs, one (50%) was placed near
the pelvis and the other (50%) near the legs.



Position of Jar/Jug:



Out of fourteen jars or jugs discovered in thirteen tombs, nine (64%)
were placed near the legs and five (36%) near the head.



Position of Bowl:



Out of thirteen bowls discovered in ten tombs, eight (61%) were found
near the legs and five (39%) near the head.



Position of Juglets:



Two juglets were discovered in one tomb, one (50%) near the head
and the other (50%) near the pelvis.

The axes were almost always found near the head, and it may be assumed that the corpse
was grasping the ax handle. As for the other items, there seems to have been no definite rule.
In four graves, all the offerings were placed near the head (top row of fig. 8.11). In four other
cases the offerings, particularly the vessels, were scattered around the head and legs (middle
line of fig. 8.11). In the other five cases, only weapons were found near the head, while ves-
sels, and occasionally some more weapons, were found near the legs.



Percentage of Warriors among the Interments



The grave sample is undoubtedly too small to support any definite conclusions, but it may
nevertheless be noted that the percentage of warriors among the total interments in Mamelon
II in Baghouz is 22.5% and is 26.2% in Gesher. These are rather similar figures, suggesting
that warriors constituted a quarter of the population buried in these cemeteries. What does this
figure imply? It is known that children were buried in storage jars under the floors of houses
in this period. Children’s graves have not been found in Gesher, Rehov, or Baghouz. The cem-
eteries thus reflect the composition of the adult community. Half of the interments are female,
and thus a quarter of the population indicates that every second male was a warrior. This per-
centage is rather high and could not represent a segregated social class of warriors; it rather
suggests that most of the adult population carried arms.

oi.uchicago.edu





156



YOSEF GARFINKEL



Origins of the Custom of Warrior Burial



It should be noted that burials associated with weapons as grave goods were common in the
southern Levant in an earlier period—the Middle Bronze Age I, ca. 2300–2000



b.c.



(Philip
1995). Such burials have been reported in almost every cemetery of the period. The following
examples are listed from north to south: ºEnan (Eisenberg 1985), Beit Shean (Oren 1973: 170–
81), Dhahr Mirzbaneh (Lapp 1966: fig. 24), Jericho (Kenyon 1960: 188–90), Lachish (Tufnell
et al. 1958: pls. 14, 21), and Tell el-Ajjul (Petrie 1932: pls. IX–XIII).



FIGURE 8.11. Position of the offerings at the various tombs.

oi.uchicago.edu

[image: image53.tif]



157



WARRIOR BURIAL CUSTOMS IN THE LEVANT



Many Middle Bronze I graves are composed of the following elements:

a. Individual burial.
b. Flexed burial; however, in many cases the bones of the decayed corpse were later

reinterred.
c. Offerings, including a few clay vessels.
d. In a few graves the offerings also include weapons, usually daggers and javelins. In rare

cases, axes of the fenestrated “eye” type were also included, as reported from Neve
Eytan, Megiddo, and Maºabarot (Miron 1992: 53). It thus appears that the custom of
burying warriors individually with weapons crystallized in the last third of the third
millennium



b.c.



in the Levant.

CONCLUSIONS

Despite the fact that bronze objects in general and bronze weapons in particular were pre-
cious items, they were nevertheless buried with their owners. This suggests that the weapons
were the personal possessions of the warrior and were not controlled by a central authority or
stored together in a community arsenal. Weapons were considered personal belongings. At the
warrior’s death, they were not bequeathed but were buried with him. Costly weapons thus
went out of circulation. The society was nevertheless able to produce new weapons, since the
copper and tin required for the production of bronze were available.

REFERENCES



Assaf, A. A.
1967 Der Friedhof von Yabrud.



Annales Archéologiques Arabes Syriennes



17: 55–68.

Bietak, M.
1991 Der Friedhof in einem Palastgarten aus der Zeit des späten Mittleren Reiches und andere For-

schungsergebnisse aus dem östlichen Nildelta (Tell el-Dabºa 1984–1987).



Ägypten und Levante



2: 47–110.

Biran, A.
1994



Biblical Dan.



Jerusalem: Israel Exploration Society and Hebrew Union College.

Buhl, M.-L.
1983



Sukas VII.



Copenhagen: Munksgaard.

Chehab, M.
1939 Tombe phénicienne de Sin el Fill.



Mélanges Syriens R. Dussaud



2: 803–10.

Damati, E., and Stepanski, Y.
1987–88 Safed, Burial Cave.



Excavations and Surveys in Israel



6: 93–94.

Dever, W. G.
1975 MB IIA Cemeteries at Ain es-Samiyeh and Sinjil.



Bulletin of the American Schools of Oriental
Research



217: 23–36.

Du Mesnil du Buisson, R.
1948



Baghouz, l’ancienne corsote.



Leiden: E. J. Brill.

Dunand, M.
1950



Fouilles de Byblos 1933–38, Vol. 2



. Paris: Librairie d’Amerique.

Eisenberg, E.
1985 A Burial Cave of the Early Bronze Age IV (MB I) near ºEnan.



ºAtiqot



(English Series) 17: 59–74.

Epstein, C.
1974 Middle Bronze Age Tombs at Kefar Szold and Ginosar.



ºAtiqot



(Hebrew Series) 7: 13–39
(Hebrew).

oi.uchicago.edu





158



YOSEF GARFINKEL



Epstein, C., and Gutman, S.
1972 The Golan. Pp. 243–98 in



Judaea, Samaria and the Golan. Archaeological Survey 1967–68,



ed. M. Kochavi. Jerusalem: Archaeological Survey of Israel (Hebrew).

Erkanal, H.
1977



Die Äxte und Beile des 2. Jahrtausends in Zentralanatolien.



Prähistorische Bronzefunde 9: 8.
Munich: C. H. Beck.

Fugmann, E.
1958



Hama III: Fouilles et Recherches de la Fondation Carlsberg 1931–38.



Copenhagen: National-
museet.

Garfinkel, Y.
1993 Gesher. Pp. 492–93 in



The New Encyclopedia of Archaeological Excavations in the Holy Land,



Vol. 2, ed. E. Stern. Jerusalem: Israel Exploration Society & Carta and New York: Simon &
Schuster.

Garfinkel, Y., and Bonfil, R.
1990 Graves and Burial Customs of the Middle Bronze IIA Period in Gesher.



Eretz-Israel



21: 132–47
(Hebrew).

Garfinkel, Y., and Nadel, D.
1989 The Sultanian Flint Assemblage from Gesher and Its Implications for Recognizing Early Neo-

lithic Entities in the Levant.



Paléorient



15: 139–51.

Gates, M.-H.
1988 Dialogues between Ancient Near Eastern Texts and the Archaeological Record: Test Cases from

Bronze Age Syria.



Bulletin of the American Schools of Oriental Research



270: 63–91.

Gautier, J. E.
1895 Note sur les fouilles entreprises dans la haute Vallée de l’Oronte.



Comptes Rendus des Séances
de l’Academie des Inscriptions et Belles Lettres



23: 435–40.

Gershuny, L.
1989 Tomb 990. Pp. 14–19 in



Excavations at Kabri, Preliminary Report of the 1988 Season,



ed. A.
Kempinski. Tel Aviv: Tel Aviv University (Hebrew).

Gerstenblith, P.
1983



The Levant at the Beginning of the Middle Bronze Age.



American Schools of Oriental
Research Research Dissertation Series 5. Philadelphia: American Schools of Oriental
Research.

Guigues, P. E.
1937 Lebea, Kafer Garra, Qraye, nécropoles de la region sidonienne.



Bulletin du Musée de Beyrouth



1



:



35–76.
1938 Lebea, Kafer Garra, Qraye, nécropoles de la region sidonienne.



Bulletin du Musée de Beyrouth



2: 27–72.

Hrouda, B.
1990 Die altbabylonischen Tumuli von Bagus bei Mari: Begräbnisse der Hanaer? Pp. 103–14 in



De la Babylonie a la syrie, en passant par Mari. Mélanges offerts à Monsieur J.-R. Kupper,



ed. O. Tunca. Liège: Université de Liège.

Ilan, D.
1992 A Middle Bronze Age Cache from Tel Dan.



Eretz-Israel



23: 9–20 (Hebrew).

Kaplan, J.
1955 Cemetery Near Tel Aviv Harbour.



ºAtiqot



(English Series) 1: 1–12.

Kenyon, K. M.
1960



Excavations at Jericho, Vol. One: The Tombs excavated in 1952–4



. London: British School of
Archaeology in Jerusalem.

1965



Excavations at Jericho, Vol. Two: The Tombs excavated in 1952–4



. London: British School of
Archaeology in Jerusalem.

Lapp, P. W.
1966



The Dhahr Mirzbaneh Tombs Vol. Two: The Tombs excavated in 1955–8



.



Three Intermediate
Bronze Age Cemeteries in Jordan.



New Haven: American Schools of Oriental Research.

oi.uchicago.edu





159



WARRIOR BURIAL CUSTOMS IN THE LEVANT



Maeir, A. M., and Garfinkel, Y.
1992 Bone and Metal Straw Tip Beer Strainers from the Ancient Near East.



Levant



24: 218–23.

Maisler, B.
1939 Tell Kurdane (Aphek?) in the Plain of Acre.



Bulletin of the Jewish Palestine Exploration Society



6: 151–58 (Hebrew).

Matthiae, P.
1980 Sulle asce fenestrate del “Signore dei Capridi.”



Studi Eblaiti



3: 53–62.

Maxwell-Hyslop, R.
1949 Western Asiatic Shaft-hole Axes.



Iraq



11: 90–30.

Miron, E.
1992



Axes and Adzes from Canaan.



Prähistorische Bronzefunde 9: 8. Stuttgart: Franz Steiner.

Montet, P.
1928



Byblos et l’Égypte.



Paris: Paul Geuthner.

Oren, E. D.
1971 A Middle Bronze Age I Warrior Tomb at Beth-shan.



Zeitschrift des Deutschen Palästina-Vereins



87: 109–39.
1973



The Northern Cemetery of Beth Shan.



Leiden: E. J. Brill.

Özgüç, T.
1986



Kültepe-Kanis: New Researches at the Center of the Assyrian Colonies, Vol. 2



. Ankara: Türk
Tarih Kurumu Yayınları.

Parrot, A.
1959



Le palais.



Mission archéologique de Mari, Vol. 2. Paris: Librairie Orientaliste, Paul Geuthner.

Petrie, W. M. F.
1917



Tools and Weapons.



London: British School of Archaeology in Egypt.
1932



Ancient Gaza II: Tell el Ajjul.



London: British School of Archaeology in Egypt.

Philip, G.
1995 Warrior Burials in the Ancient Near Eastern Bronze Age: The Evidence from Mesopotamia,

Western Iran and Syria Palestine. Pp. 140–54 in



The Archaeology of Death in the Ancient Near
East,



eds. S. Campbell and A. Green. Oxbow Monograph 51. Oxford: Oxbow.

Schaeffer, C. F. A.
1949



Ugaritica II.



Paris: Paul Geuthner.
1962



Ugantica IV.



Paris: Paul Geuthner.
1978



Ugaritica VII.



Paris: Paul Geuthner.

Seeden, H.
1980



The Standing Armed Figurines in the Levant.



Prähistorische Bronzefundel 1. Munich:







C. H.
Beck.

Thrane, H.
1978



Sukas IV.



Copenhagen: Munksgaard.

Tpilinsky, N.
1962







Meron.



Hadashot Arkheªologiyot



2: 25







(Hebrew).

Tufnell, O., et al.
1958



Lachish IV, The Bronze Age.



London: Oxford University Press.

Tufnell, O.; Baden-Powell, D. F. W.; Bate, D. M. A.;



C



ern



y



, J.; Diringer, D.; Giles, M.; Helbaek, H.; Isserlin,
B. S. J.; Murray, M. A.; Parker, B.; Porada, E.; Thompson, F. C.; Todd, E.; and Waecher, J.

1958



Lachish IV: The Bronze Age



. London: Oxford University Press.

Yadin, Y.
1963



The Art of Warfare in Biblical Lands in the Light of Archaeological Study.



New York: McGraw
Hill.

Yogev, O.
1985 A Middle Bronze Age Cemetery South of Tel Rehov.



ºAtiqot



(English Series) 17: 91–113.

oi.uchicago.edu



160 YOSEF GARFINKEL

PLATE 8.1b. Fenestrated duck-bill ax from Grave 12 at Gesher.

PLATE 8.1a. Fenestrated duck-bill ax from Grave 2 at Gesher.

oi.uchicago.edu



161WARRIOR BURIAL CUSTOMS IN THE LEVANT

PLATE 8.2b, c. Narrow elongated socket ax from Grave 14 at Gesher.

PLATE 8.2a. Fenestrated duck-bill ax from Grave 13 at Gesher.

oi.uchicago.edu



162 YOSEF GARFINKEL

oi.uchicago.edu



163



THE SIGNIFICANCE OF IRON AGE
“WAVY-BAND” PITHOI ALONG THE

SYRO-PALESTINIAN LITTORAL, WITH
REFERENCE TO THE TEL DOR PITHOI



Ayelet Gilboa



INTRODUCTION

The conspicuous pithoi with wavy and horizontal band decoration (fig. 9.1) have been a
matter of much discussion; for recently published examples, and overviews and references to
the principal literature on their mainland distribution, see Cohen-Weinberger and Goren 1996:
77; Golani and Yogev 1996: 51–54, fig. 7; and Stepansky, Segal, and Carmi 1996: fig. 9. The
accumulating evidence that has emerged from these studies (complemented here by some addi-
tional references) may be summarized as follows:

a. Chronology: On present evidence, “wavy-band” pithoi occur on the mainland from
Late Bronze Age II (LB) to ca. the mid-eleventh century



b.c.



b. Mainland distribution: The pithoi are found at Eastern Mediterranean coastal and
hinterland sites from Ugarit to Ashdod—at Ugarit; Sarepta in Area II, Sounding Y,
mainly in Strata G–E Anderson 1988: 323, fig. 6a, pl. 30:18; Tyre, at least in Strata XV



and XIV;



1



Akko, Tell Keisan 9c, Dor (see below), and Ashdod XII–XI; and in the
Upper Galilee (Tel Dan V, Hazor XII [but see also below], Ayelet Ha-Shahar, Tel Sasa,
H. ºAvot, Har Adir, H. Jelil, a site near Kibbutz Eilon, Tel ºAvdon, and Tell el-
Ghaiyada); for the four latter sites, see respectively Frankel (1994: 27) and Frankel and
Getzov (1997: figs. 2.142.10:19; 2.194:10).

c. Use: Most scholars agree on the obvious—the storage purpose of these containers—
though in most cases, what exactly they contained, or were meant to contain is still
elusive. In a few cases (e.g., at Tyre and Tell Keisan) they were shown to be used
(probably secondarily) for industrial purposes.



2







1. The example in Anderson (1988: l. 30:18) is from Stratum F; no provenance is indicated for the example in
fig. 6a. It is not clear which of the pieces recorded in table 21 (spanning Strata J/H-D1) belong to the types
examined here. Anderson (1988: 324) suggested a possible correlation between fragments with this type of
decoration and the pithos rims, designated by him RR-2, occurring throughout Strata J/H–D1 (see Anderson
1988: table 8A/B) and possibly also RR-4, but this remains inconclusive. These data are omitted from our
discussion. On the other hand, at least part of the bases designated by Anderson as B-17 certainly belong to
this type of pithos. These occur chiefly in Strata G–E (Anderson 1988: 242, pl. 52). For the distribution of
pithos rims and bases at Tyre, see Bikai 1978: tables 5A, 5B. It is not clear, however, which of them belong
to the specific types under discussion.

2. For the functions of the Cypriot pithoi discussed below, see Pilides (1996: 115–19).



9



oi.uchicago.edu





164



AYELET GILBOA



d. Production areas: All examples analyzed (only Iron Age examples), both by petro-
graphic analysis (Sasa) and Instrumental Neutron Activation Analysis (Dan and Dor),
proved to have been produced on the mainland (Cohen-Weinberger and Goren 1996;
Yellin and Gunneweg 1989; Yellin forthcoming). More specific provenances that could
be pinpointed were (mainly) the southern Phoenician coast, and Upper Galilee (one
example).



3



LATE BRONZE AGE OCCURRENCES

At least five mainland sites, all major urban centers, produced “wavy-band” pithoi in this
period: Ugarit (including Minet el-Beidha, numerous), Sarepta (apparently numerous, see
n. 1), Tyre XV (one definite, but probably more, see n. 1), Akko (“considerable amounts, late
thirteenth–early twelfth centuries”; see Raban 1988: 290), and at least one has lately been
uncovered at Hazor, in the LB IIB conflagration deposits of the Area A palace (for possible
examples in LB contexts at Ashdod and Beth Shemesh, see below). To date no other clear
examples are known, although it should be emphasized that in both coastal and mountain-
ous Upper Galilee, a crucial area in this puzzle, this could be due to lack of data. More-
over, the only LB site extensively excavated there—Hazor—seems to have been destroyed
early in the thirteenth century



b.c.



, followed by an occupational gap that lasted for at least



3. But it should be borne in mind that these are also the areas in which the analyzed pithoi were uncovered. As
yet, other provenances (northern Phoenicia and Philistia, for instance) cannot be ruled out. The results of the
petrographic examination of the Tell Keisan pithos were inconclusive (Courtois 1980: 356).

FIGURE 9.1. “Wavy-band” pithoi from Iron Age I contexts: 1: Dor; 2: Tel Sasa; 3: Tyre.

oi.uchicago.edu

[image: image54.tif]



165



IRON AGE “WAVY-BAND” PITHOI ALONG THE SYRO-PALESTINIAN LITTORAL



the second half of this century, and most of the following one (see summary in Finkelstein
1988: 100–01). This is one of the main lacunae in our data. The fabric of none of these pithoi
has been analyzed.

The Ugarit pithoi were considered by Monchambert (1983: 32) to be local products, due to
their abundance, homogeneity, and size. In contrast, because of their resemblance in shape and
fabric to the Cypriot ones and her view that it is unlikely that such large vessels were trans-
ported across long distances, Schuster-Keswani (1989: 17–18) interpreted them as products of
itinerant potters (presumably she meant Cypriot potters?). It should be emphasized that these
LB pithoi not only closely resemble the contemporaneous Cypriot ones, but that the repertoire
as a whole (evident mainly at Ugarit and Minet el-Beidha, see, e.g., Schaeffer 1949: fig. 86)
seems to echo a full range of Island types—of various fabrics, sizes, shapes, and decora-
tions—unlike the Iron Age ones (see below), indicating two different phenomena.

The locale of their production still remains to be determined. But considering the above
observations, and the fact that it is now obvious that Cypriot-made pithoi reached even more
distant destinations in the Mediterranean—such as Crete, mainland Greece, Sicily, and Sar-
dinia (see, e.g., Vichos and Lolos 1997: fig. 5,



4



and summary and references in Pilides 1996:
113; Knapp and Cherry 1994: 127)—it would be a fair guess that they were exported to Syria-
Palestine as well,



5



though a mainland origin, for at least part of the LB pithoi, is definitely
possible.

THE IRON AGE



Tel Dor Examples



6



The one pithos analyzed by INAA (Yellin forthcoming; reg. no. 77598, L.7926; fig. 1:3)
originates in the Iron Age I destruction layer in Area B1 (Phase B1–12), that is dated by E.
Stern, the site’s excavator, to the mid-eleventh century



b.c.



; see Stern 1994: 94.



7



Two similar
pithoi were crushed in situ







under a contemporaneous destruction level in Area G (Phase G-9),
but no sizable profiles of these can yet be reconstructed or illustrated. Fragments of at least three
other such pithoi were found under this destruction layer. Only a few fragments of similar ves-
sels, with horizontal and “wavy-band” decoration, were uncovered in contexts postdating this
destruction—dated to the mid-eleventh to early tenth century



b.c.



—but none in primary depo-
sition. Theoretically they could be redeposited pieces, but the pottery accompanying them
belongs to the very typical post–destruction assemblage, and thus it seems that to a certain
extent similar pithoi were at least used here until about the year 1000



b.c.



4. I thank S. Wolff for this reference.

5. I am quite confident that careful scrutiny of LB site reports (and future excavations) will indeed reveal evi-
dence of such pithoi; see possibly at Beth Shemesh (Grant 1931: pl. XL:18), and additional pieces in Stra-
tum IV. At Tell Abu Hawam a rather small pithos that seems to be Cypriot (but undecorated) was uncovered
in Building 52, probably from the end of LB (Balensi 1980: pl. 12:285).

6. See also Cohen-Weinberger and Wolff, chap. 32, this volume, which came to my attention after this paper
was submitted for publication.

7. For general discussions and interpretations of the Iron Age I stratigraphic and artifactual sequence at Dor,
see Stern (1990; 1994: 90–104) and Gilboa (1998).

oi.uchicago.edu





166



AYELET GILBOA



The fabric of most (but definitely not all) fragments appears, to the eye, strikingly similar
(however, only the composition of one has been analyzed). On the other hand, all of them are
very different, even to the naked eye, from that of the Upper Galilee pithoi I examined (from
Sasa, Har Adir, Dan, and H. ºAvot).

Of the three







in situ







pithoi, two were found in clear storage assemblages, in association with
other jars (the example presented here, and others, were also accompanied by “collared-rim”
jars). The third pithos was uncovered near an industrial installation of as yet unclear function
(Stern et al. 1997: fig. 11).



Use Contexts



The occurrence and distribution of “wavy-band” pithoi on the mainland in the Iron Age
should be considered within the framework of the general “pithoi phenomenon” of the late 13th
to early 10th century (indeed, they are frequently found in the same systemic contexts with
other pithos types, as at Dor). Like their “collared-rim” contemporaries, their distribution
encompasses settlements of different hierarchies (from hamlets/villages to major urban cen-
ters), and like them they are found in association with the three main “pottery cultures” of the
era: the highland culture, that of Philistia, and the Canaanite/coastal one (for a schematic char-
acterization of the latter, see Gilboa 1998). But unlike the “collared-rim” jars, “wavy-band”
pithoi have not been encountered in any of the numerous sites excavated in the central hill
country and as yet are not attested to in the northern valleys.

The ethnic (Israelite) affiliation of the users of the “collared-rim” jars has long been
rejected by most scholars and explanations of socioeconomic nature for their distribution
offered in its stead (e.g., London 1989b; Esse 1992). As far as consumption is concerned, the
“wavy-band” pithoi should be similarly interpreted; thus in Dan and Upper Galilee, they do
not



per force



indicate a Phoenician population, and the terms “Phoenician” and “Tyrian”
should be abandoned, and likewise they are not evidence of “Sea Peoples” (as suggested by
Raban and Stieglitz 1991: 41–42). Their distribution results from the same (as yet not entirely
comprehended or agreed upon) demand for large containers, though the larger capacity of
most of them and their wider aperture may have served different commodities than those of
the “collared-rim” jars.



Production Contexts



The most conspicuous attributes of these pithoi are their Cypriot-inspired shape and deco-
ration. Although applied clay bands and ridges are perhaps the most common decoration on
pithoi in different cultures and periods, wavy bands on pithoi, in the mainland regions under
discussion, are confined to this relatively short period. On the other hand, pithoi with this type
of decoration have been produced in Cyprus for centuries and pots similarly decorated are
indeed still being produced there. However, as already mentioned, all pithoi analyzed proved
to be of mainland manufacture.

All the mainland Iron I pithoi known to date, that are either complete or where a substan-
tial part of the vessel could be reconstructed are fairly uniform in shape and proportions (see
examples in fig. 9.1) and differ in many morphological respects from the Cypriot ones (see

spread one line short

oi.uchicago.edu





167



IRON AGE “WAVY-BAND” PITHOI ALONG THE SYRO-PALESTINIAN LITTORAL



below), and likewise from all complete or near complete LB examples known from the main-
land, and thus in this respect also seem to be “local.” They are much less elegantly shaped (or
“stately,” to use Pilides’ definition) than the Cypriot ones, and as a rule lack handles.

Except the Akko examples, whose contexts are still unpublished and cannot be evaluated,
only one pithos fragment—the one from Ashdod XII—was dated to ca. the early or mid-
twelfth century



b.c.



It originates from a fill of Stratum XII that contained redeposited material,
according to the excavators, of Strata XIIIB and XIIIA (Dothan and Porath 1993: 70–71, fig.
34:3). It differs from all other Iron Age mainland examples in that it is provided with handles
and may well be a redeposited LB piece.

Other than this sherd, all other pithoi (Tyre XIV, Dan V, Hazor XII,



8



Tell Keisan 9c, Dor
and Sasa) date between the late twelfth and late eleventh, possibly early tenth century



b.c.



(an
exact date for the Har Adir and H. ºAvot examples has not yet been established). Examples
from Sarepta, Tyre, and Akko (see above and n. 1) may prove to be earlier within Iron I.

DISCUSSION

My point of departure is based on a few comments offered by Cohen-Weinberger and
Goren (1996: 81). Considering the fact that out of the four “wavy-band” (so-called “Tyr-
ian”) Tel Sasa pithoi analyzed by them, three formed a separate petrographic family (Family
B, produced in northern Israel or Lebanon) and one was grouped with “Galilean” pithoi



9



(Family A, produced in Upper Galilee), the authors considered two possibilities. One possibil-
ity—that both types were produced by the same potters, intended to “satisfy demand related
to differing traditions, functions, or ethnic identities” (that is, the difference was determined
by the clients). The second possibility—that there were two “groups” of local potters, one pro-
ducing “Galilean” and the other “Tyrian” (i.e., “wavy-band”) pithoi (the difference was deter-
mined by the manufacturers).

To evaluate these alternatives, one may try to consider the meaning of the most conspicu-
ous formal attribute of the pithoi—the wavy band decoration. As indicated above, it does not
derive from any local tradition. The suggestion of a possible different function for the “wavy-
band” pithoi (vs. “Galilean” of both types, and for that matter also “collared-rim” jars) cannot
explain their conspicuous Cypriot character. (Also, there were other pithoi on the scene—of the
“Galilean 1” type—of similar sizes and proportions, that could satisfy the same functional
demands).

Thus if we assume that “local” potters were producing vessels with Cypriot attributes, this
would require another explanation. Although, as indicated above, Cypriot pithoi with these



8. Hazor Stratum XII, dated by the excavators to the twelfth century



b.c.



, probably does not antedate the very
end of this century; see summary in Finkelstein 1988: 100–01.

9. There seems to be some confusion regarding the use of the term “Galilean pithos.” Some scholars (e.g., Biran
1989: figs. 4.7:8, 4.16:9; 1994: 129; Stepansky, Segal and Carmi 1996: fig. 7:1) use this term to define the wide
pithoi with large aperture, which resemble in shape the LB ones, e.g., the Hazor pithoi (henceforward desig-
nated “Galilean 1”); while others refer to pithoi that closely resemble the “collared-rim” jars of the central hill
country (henceforward “Galilean 2”). See also Finkelstein 1988: fig. 30b vs. 104, fig. 32.

oi.uchicago.edu





168



AYELET GILBOA



attributes undoubtedly circulated in the vicinity at least until the end of the Bronze Age, no
evidence of imported pithoi later than that has been forthcoming.



10



It is therefore hard to
assume emulation, for whatever reasons, of Cypriot originals.

Another possibility would be to assume that potters of Cypriot origin were operating
on the mainland. In this case, three main explanations may be offered for the conspicuous
decoration:

1. The decoration may have just been the potters’ “way of doing things” (e.g., Sackett’s
[1990] “isochrestic choice”).

2. The decoration may have functioned as some conscious mode of message emission
(“stylistic behavior” in the sense promoted by Wobst 1977: 320–23), possibly per-
taining to the vessels’ quality, contents, origin, or affiliation of producers.

3. The decoration negotiated some sort of group identity (“stylistic behavior” in Wies-
sner’s 1984 terminology, only one of Wobst’s [1977: 328] “stylistic behavior” modes.)

To my mind explanation 1 should be rejected because of the very restricted range of
shapes, sizes, and decorative motifs used here, out of the much larger “original” Cypriot rep-
ertoire. The selection seems to have been deliberate.

Explanation 2 would require a specific clientele, to whom the information encoded in
shape and design was addressed. I would consider the local population out of the question for
the reasons stated above. The existence of a Cypriot clientele on the mainland, to whom the
vessels may have conveyed some sort of information, is possible but as yet lacks corroborat-
ing evidence.

On present evidence I would thus favor explanation 3—that the vessels were produced on
the mainland by Cypriot potters who deliberately adorned their products with the most con-
spicuous of traditional motifs, as a cognitive expression of group awareness and distinctive-
ness, with no other specific target population. This interpretation does not lack difficulties, the
principal one of which, to my mind, lies in the morphological differences between the main-
land pithoi and the Cypriot originals,



11



indicating a somewhat remote kind of familiarity. Also,
as already indicated, those pithoi somewhat resemble, in size and proportions, the “Galilean 1”
pithoi (cf. Biran 1989: fig. 4.7:8–9; see also Pilides 1996: 119). Still, this seems to me the most
plausible reconstruction.



10. The distinction alluded to above, between LB and Iron Age examples, is largely based on the morphology
of complete or near complete vessels. Thus fragments that may have bridged the chronological gap between
the known LB II and Iron I examples (especially at Tyre and Sarepta, and the unpublished examples from
Akko) cannot be evaluated. It is not clear whether they (or part of them) are imported or not. On present evi-
dence there was no chronological overlap between imports and mainland products, but I would not rule out
this possibility: at Dor there are two pithos fragments, in Iron I contexts, which to the naked eye look Cyp-
riot-made. This of course will have to be confirmed by analysis. At Tell Kazel too, a possibly Cypriot-made
pithos was uncovered in an early Iron Age context (Badre et al. 1990, fig. 39:f). These instances, though, are
very few.

11. Bearing in mind that there is as yet very little information concerning the contemporary (mainly Late Cyp-
riot IIIB and Cypro-Geometric G IA) repertoire in Cyprus and its shapes, and that comparison is based on
the earlier (mainly late Late Cypriot IIC) examples, see below.

oi.uchicago.edu





169



IRON AGE “WAVY-BAND” PITHOI ALONG THE SYRO-PALESTINIAN LITTORAL



Let us now examine this proposed phenomenon in its regional historical context:



Cyprus



In Cyprus, the pithoi that most resemble the mainland ones (but see more on this below),
constitute the hallmark of numerous Late Cypriot IIC sites (see Pilides 1996: 110–13, fig. 2).
They are of various sizes and fabrics; some (mainly the larger ones) bear the horizontal and
wavy decoration, some have other designs, and some bear none. Some have handles and some
not, while most have flat bases (see, e.g., Schuster-Keswani 1989: 13–17, tables 1, 2; Webb
and Frankel 1994: 10 and references therein, p.12).

Cypriot society at that time was stratified and complex, with ever increasing levels of spe-
cialized labor, production and exchange (e.g., Knapp 1993: 97, 100–101; Schuster-Keswani
1993; Webb and Frankel 1994). As suggested by Schuster-Keswani (1989: 17–18) and others
(Vermuele and Wolsky 1990: 378, 380; London 1991: 222, 231), the pithoi could only have
been manufactured by highly skilled specialists (by an “esoteric guild,” to use Åström’s defi-
nition quoted in Pilides 1996: 109), though it is not clear whether these were itinerant potters
or else to what extent production was centralized.

No doubt, the lion’s share of this pithos production served the widespread storage and
redistribution facilities of the time, most vividly demonstrated at sites like Kalavassos-



Ayios
Dhimitrios



(South 1995: fig. 3), Maroni (Cadogan 1992: 54), and Analiondas-



Palioklichia



(Webb and Frankel 1994 and extensive references on p.12; see Webb and Frankel 1994: 16).
Nearly all these centers ceased to exist either by the end of Late Cypriot IIC or during Late

Cypriot IIIA. Pithoi, including the decorated types, are still known from Late Cypriot IIIB and
early Cypro-Geometric contexts (e.g., Enkomi, Level IIIa, see Dikaios 1969: pl. 68:190;
Kition, Area II, Floor III, see Karageorghis and Demas 1985: pl. CXXVII), but the quantities
and variety were dramatically reduced, and they soon disappeared altogether (see overview in
Pilides 1996: 110–11, 115).



12



These were turbulent times in Cyprus that witnessed dramatic demographic shifts, island-
wide destructions, and possibly also population dislocations (for recent overviews of late
Late Cypriot IIIA and Late Cypriot IIIB, see Catling 1994: 136–37; Iacovou 1994), but during
which nonetheless ongoing traffic was maintained between the island and the mainland,
though perhaps reduced in scale (Negbi 1992: 611, n. 83; Mazar 1994: 51; Sherratt 1994: 70;
Gilboa 1998: 423).

CONCLUSIONS

I would thus argue that the lasting (or growing) demand for pithoi in many close-by
regions on the mainland, coupled with the drastically diminishing demand on the island, and
the collapse of the Late Cypriot IIC organizational production structure (see Knapp and
Cherry 1994: 163) attracted these specialist potters to the mainland (cf. Anthony’s [1997]
“Career Migration”).

Did they reside there permanently, or are we to envisage itinerant potters—somewhat
similar to the model suggested by London (1989a) but involving overseas venturing and



12. The paucity of excavated Late Cypriot IIIB and Cypro-Geometric IA non-funerary sites should be borne in
mind (this is obviously not a common grave good), but the process, by and large, is clear.

oi.uchicago.edu





170



AYELET GILBOA



assuming that production was performed close to the







clientele? Or perhaps were there sea-
sonal workshops operating on the mainland (see Vermuele and Wolsky’s suggestion [1990:
398–99] concerning Toumba Tou Skourou)? At present I see no way of resolving this matter,
though the clustered distribution on the mainland (excluding the naturally wider distribution
in harbor towns) may indicate a permanent operation milieu. Whether this restricted distribu-
tion was merely an outcome of marketing strategies or technical constraints, or whether it also
had political reasons (see Frankel 1994: 33; Kochavi 1984: 67) still remains to be determined.
Although Cypriot presence in northern Canaan/Phoenicia around 1150–1050



b.c.



has not yet
been demonstrated, this seems to me the more likely scenario, and would constitute a clue to
the fate of some Island



pitharades



during the great Late Cypriot III upheavals.

ACKNOWLEDGMENTS

This paper is part of a research program concerned with Iron Age Tel Dor, supported by
the Israel Science Foundation of the Israel Academy of Sciences and Humanities. It will be
incorporated in my Ph.D. dissertation that is currently being written under the tutorship of
Prof. E. Stern of the Hebrew University. The preparation and publication of this paper also
was made possible by a grant from the Memorial Foundation of Jewish Culture.

I thank the following individuals and institutions that kindly permitted me to examine
pithoi, and to mention those that are not yet published: Prof. A. Biran and Dalia Pakman (Tel
Dan expedition); Prof. A. Ben Tor and Anabel Zarzecki-Peleg (Hazor); Prof. E. Stern (Dor);
The Israel Antiquities Authority, Karen Covello-Paran, and Eliot Braun (Sasa, Har Adir, and
H. ºAvot).

REFERENCES



Anderson, W. P.
1988



Sarepta I. The Late Bronze and Iron Age Strata of Area II, Y



. Beirut: Université Libanaise.

Anthony, D.
1997 Prehistoric Migration as Social Process. Pp. 21–32 in



Migrations and Invasions in Archaeologi-
cal Explanation



, eds. J. Chapman and H. Hamerow. British Archaeological Reports, Interna-
tional Series 664. Oxford: British Archaeological Reports.

Badre, L.; Gubel, E.; Al-Maqadissi, M.; and Sader, H.
1990 Tell Kazel–Syria: Excavations of the AUB Museum 1985–1987, Preliminary Reports.



Berytus



38: 11–124.

Balensi, J.
1980 Les fouilles de R. W. Hamilton à Tell Abu Hawam, niveaux IV et V. Ph.D. dissertation, Univer-

sité des Sciences Humaines, Strasbourg.

Bikai, P. M.
1978



The Pottery of Tyre.



Warminster: Aris & Phillips.

Biran, A.
1989 The Collared-Rim Jars and the Settlement of the Tribe of Dan. Pp. 71–96 in



Recent







Excavations
in Israel: Studies in Iron Age Archaeology



, eds. S. Gitin and W. G. Dever



.



Annual of the Ameri-
can Schools of Oriental Research 49. Winona Lake, IN: Eisenbrauns.

1994



Biblical Dan



. Jerusalem: Israel Exploration Society and Hebrew Union College.

Cadogan, G.
1992 Maroni VI.



Report of the Department of Antiquities of Cyprus



: 51–58.

oi.uchicago.edu





171



IRON AGE “WAVY-BAND” PITHOI ALONG THE SYRO-PALESTINIAN LITTORAL



Catling, H. W.
1994 Cyprus in the 11th Century



b.c.



: An End or a Beginning? Pp. 133–41 in



Proceedings of







the
International Symposium: Cyprus in the 11th Century



b.c.



, ed. V. Karageorghis. Nicosia: Uni-
versity of Cyprus and Leventis Foundation.

Cohen-Weinberger, A., and Goren, Y.
1996 Petrographic Analysis of Iron Age I Pithoi from Tel Sasa.



ºAtiqot



28: 77–83.

Courtois, L.
1980 Examen pétrographique et caractérisation matérielle de quelques céramiques de Tell Keisan.

Pp. 353–60 in



Tell Keisan (1971–1976): Une cité phénicienne en Galilée



, eds. J. Briend and
J-B. Humbert. Fribourg: Editions universitaires.

Dikaios, P.
1969



Enkomi Excavations 1948–1958



. Mainz am Rhein: von Zabern.
Dothan, M., and Porath, Y.

1993 Ashdod V: Excavations of Area G.



ºAtiqot



23.
Esse, D. G.

1992 The Collared Pithos at Megiddo: Ceramic Distribution and Ethnicity.



Journal of Near Eastern
Studies



51: 81–103.
Finkelstein, I.

1988



The Archaeology of the Israelite Settlement



. Jerusalem: Israel Exploration Society.
Frankel, R.

1994 Upper Galilee in the Late Bronze–Iron I Transition. Pp. 18–34 in



From Nomadism to Monarchy



,
eds. I. Finkelstein and N. Naªaman. Washington, DC: Biblical Archaeology Society.

Frankel, R., and Getzov, N.
1997



Archaeological Survey of Israel, Map of Akhziv (1), Map of Hanita (2)



.







Jerusalem: Israel Antiq-
uities Authority.

Gilboa, A.
1998 Iron Age I Pottery Evolution at Dor: Regional Contexts and the Cypriot Connection. Pp. 413–

25 in



Mediterranean Peoples in Transition: Thirteenth to Early Tenth Centuries



b.c.e.



, eds.
S. Gitin, A. Mazar, and E. Stern. Jerusalem: Israel Exploration Society.

Golani, A., and Yogev, O.
1996 The 1980 Excavations at Tel Sasa.



ºAtiqot



28: 41–58.
Grant, E.

1931



Ain Shems Excavations (Palestine) 1928–1929–1930–1931,



Part I. Biblical and Kindred
Studies 4. Haverford, PA: Haverford College.

Iacovou, M.
1994 The Topography of 11th Century



b.c.



Cyprus. Pp. 149–65 in



Proceedings of the







International
Symposium: Cyprus in the 11th Century



b.c.



, ed. V. Karageorghis. Nicosia: University of Cyprus
and Leventis Foundation.

Karageorghis, V., and Demas, M.
1985



Excavations at Kition V: The Pre-Phoenician Levels



. Nicosia: Department of Antiquities, Cyprus.
Knapp, A. B.

1993 Social Complexity: Incipience, Emergence, and Development on Prehistoric Cyprus.



Bulletin of
the American Schools of Oriental Research



292: 85–106.
Knapp, A. B., and Cherry, J. F.

1994



Provenience Studies and Bronze Age Cyprus



. Monographs in World Archaeology 21. Madison,
WI: Prehistory Press.

Kochavi, M.
1984 The Period of the Israelite Settlement. Pp. 19–97 in



The History of Eretz-Israel,



Vol. 2:



Israel and
Judah in the Biblical Period,



ed. I. Ephºal. Jerusalem: Keter and Yad Yitzhak Ben-Zvi (Hebrew).
London, G. A.

1989a On Fig Leaves, Itinerant Potters, and Pottery Production Locations in Cyprus. Pp. 65–80 in



Cross-Craft and Cross-Cultural Interactions in Ceramics.



Ceramics and Civilization 4, eds.
P. E. McGovern and M. D. Notis. Westerville, OH: American Ceramic Society.

oi.uchicago.edu





172



AYELET GILBOA



1989b A Comparison of Two Contemporaneous Lifestyles in the Late Second Millennium



b.c.



Bulletin
of the American Schools of Oriental Research



273: 37–55.
1991 Ethnoarchaeological Evidence of Variation in Cypriote Ceramics and Its Implications for the

Taxonomy of Ancient Pottery. Pp. 221–35 in



Cypriote Ceramics



:



Reading the Prehistoric
Record



, eds. J. A. Barlow, D. L. Bolger, and B. Kling. University Museum Monograph 74, Sym-
posium Series Volume 2. Pennsylvania: University of Pennsylvania and Leventis Foundation.

Mazar, A.
1994 The 11th Century



b.c.



in the Land of Israel. Pp. 40–57 in



Proceedings of the International Sym-
posium: Cyprus in the 11th Century



b.c



., ed. V. Karageorghis. Nicosia: University of Cyprus
and Leventis Foundation.

Monchambert, J. -Y.
1983 La céramique de fabrication locale à Ougarit à la fin du Bronze Récent: quelques examples.



Syria



60: 25–45.

Negbi, O.
1992 Early Phoenician Presence in the Mediterranean Islands: A Reappraisal.



American Journal of
Archaeology



96: 599–615.

Pilides, D.
1996 Storage Jars as Evidence of the Economy of Cyprus in the Late Bronze Age. Pp.107–24 in



The
Development of the Cypriote Economy from the Prehistoric Period to the Present Day



. Nicosia:
University of Cyprus.

Raban, A.
1988 The Constructive Maritime Role of the Sea Peoples in the Levant. Pp. 261–94 in



Society and
Economy in the Eastern Mediterranean (c. 1500–1000



b.c.



)



, eds. M. Heltzer and E. Lipi



n



ski.
Orientalia Lovaniensia Analecta 23. Leuven: Uitgeverij Peeters.

Raban, A., and Stieglitz, R. A.
1991 The Sea Peoples and Their Contribution to Civilization.



Biblical Archaeology Review



17(6):
34–42.

Sackett, J. R.
1990 Style and Ethnicity in Archaeology: The Case for Isochrestism. Pp. 32–43 in



The Uses







of Style
in Archaeology,



eds. M. W. Conkey and C. A. Hastorf. Cambridge: Cambridge University Press.

Schaeffer, C. F.-A.
1949



Ugaritica II



. Paris: Librairie Orientaliste, Paul Geuthner.

Schuster-Keswani, P.
1989 The Pithoi and Other Plain Ware Vessels. Pp. 12–21 in



Vasilikos Valley Project 3: Kalavassos-
Ayios Dhimitrios II.



Studies in Mediterranean Archaeology 71, no. 3, by A. South,







P. Russel and
P. Schuster-Keswani, ed. I. Todd. Göteborg: Paul Åströms.

1993 Models of Local Exchange in Bronze Age Cyprus.



Bulletin of the American Schools of Oriental
Research



292: 73–83.

Sherratt, S.
1994 Commerce, Iron and Ideology: Metallurgical Innovation in 12th-11th Century Cyprus. Pp. 59–107

in



Proceedings of the International Symposium: Cyprus in







the 11th Century



b.c



.



, ed. V. Kara-
georghis. Nicosia: University of Cyprus and Leventis Foundation.

South, A. K.
1995 Urbanism and Trade in the Vasilikos Valley in the Late Bronze Age. Pp. 187–97 in



Trade, Con-
tact and the Movement of Peoples in the Eastern Mediterranean:







Studies in Honour of J. Basil
Hennessy,



eds. S. Bourke and J.-P. Descoeurdes. Mediterranean Archaeology Supplement 3.
Sydney: Medarch.

Stepansky Y.; Segal, D.; and Carmi, I.
1996 The 1993 Sounding at Tel Sasa: Excavation Report and Radiometric Dating.



ºAtiqot



28: 63–76.

Stern, E.
1990 New Evidence from Dor for the First Appearance of the Phoenicians along the Northern Coast

of Israel.



Bulletin of the American Schools of Oriental Research



279: 27–33.

oi.uchicago.edu





173



IRON AGE “WAVY-BAND” PITHOI ALONG THE SYRO-PALESTINIAN LITTORAL



1991 Phoenicians, Sikils, and Israelites in the Light of Recent Excavations at Tel Dor. Pp. 85–94 in



Studia Phoenicia XI: Phoenicia and the Bible



, ed. E. Lipinski



.



Orientalia Lovaniensia Analecta
44. Leuven: Uitgeverij Peeters.

1994



Dor: Ruler of the Seas



. Jerusalem: Israel Exploration Society and Bialik Institute.

Stern E.; Berg, J.; Gilboa, A.; Sharon, I.; and Zorn, J.
1997 Tel Dor, 1994–1995: Preliminary Stratigraphic Report.



Israel Exploration Journal



47: 29–56.

Vermuele, E. D. T., and Wolsky, F. Z.
1990



Toumba Tou Skourou: A Bronze Age Potters’ Quarter on Morphou Bay in Cyprus



. Cambridge,
MA: Harvard University, Museum of Fine Arts.

Vichos, Y., and Lolos, Y.
1997 The Cypro-Mycenaean Wreck at Point Iria in the Argolic Gulf: First Thoughts on the Origin of

the Vessel. Pp. 321–37 in



Res Maritimae: Cyprus and the Eastern Mediterranean from Prehis-
tory to Late Antiquity,



eds. S. Swiny, R. L. Hohlfelder, and H. W. Swiny. Atlanta: Scholars Press.

Webb, J. M., and Frankel, D.
1994 Making an Impression: Storage and Surplus Finance in Late Bronze Age Cyprus.



Journal of
Mediterranean Archaeology



7: 5–26.

Wiessner, P.
1984 Reconsidering the Behavioral Basis for Style: A Case Study among the Kalahari San.



Journal of
Anthropological Archaeology



3: 190–234.

Wobst, H. M.
1977 Stylistic Behavior and Information Exchange. Pp. 317–42 in



For the Director: Essays in Honor
of James W. Griffin, ed. C. E. Cleland. Anthropological Papers, Museum of Anthropology,
University of Michigan 61. Ann Arbor: University of Michigan Press.

Yellin, J.
Forthcoming The Provenience of an Iron Age I “Wavy-Band” Pithos from Dor.

Yellin, J., and Gunneweg, J.
1989 Instrumental Neutron Activation Analysis and the Origin of Iron Age I Collared-rim Jars and

Pithoi from Tel Dan. Pp. 133–41 in Recent Excavations in Israel: Studies in Iron Age Archae-
ology, eds. S. Gitin and W. G. Dever. Annual of the American Schools of Oriental Research 49.
Winona Lake, IN: Eisenbrauns.

oi.uchicago.edu



174 AYELET GILBOA

oi.uchicago.edu



175



10



AN UNPUBLISHED RELIEF SHERD FROM
ALI



V



AR HÖYÜK



Ronald L. Gorny



Douglas Esse was an extremely supportive and inspirational friend. He was also a gifted
scholar who possessed an investigative spirit that was thorough and relentless in the pursuit of
excellence. Though working primarily in the Levant, he had a deep and abiding interest in
Turkey, as well as other areas of the Middle East. Esse had, in fact, just before his death,
received a permit to reopen the Oriental Institute excavations in the Amuq Plain. The fact that
Doug and I were both attempting to resurrect old excavations provided us with a strong com-



mon bond. Esse was among the few people who truly believed that the excavations at Ali



v



ar,
like those in the Amuq, could be invigorated with new life. The fact that we are excavating at
Ali



v



ar today is due, in part, to Esse’s encouragement and is a testament to his belief that
dreams can become reality.



1



Although Doug’s archaeological interests were as varied as his talents, he always had a
deep appreciation for ceramics, and we spent many hours in the basement of the Oriental
Institute comparing notes on pottery from sites of mutual interest. The realm of ceramics was
one in which Doug excelled, and his research resulted in many fascinating insights. His inter-
est in Khirbet Kerak ware even brought him into contact with Anatolia where this particular
ware is known in the local jargon as “Karaz” or “Early Transcaucasian”(ETC) ware. In honor
of his abiding fascination with ceramics, I would like to dedicate this investigation of the Ali-



v



ar relief sherds to Esse, hoping that he would have approved of the topic and knowing how
pleased he would have been to see the long-held dreams I shared with him about Ali



v



ar now
being realized.

Ali



v



ar Höyük is an 18-hectare site located in the Kanak Su basin, midway between the
important ancient political centers at



H



attu



s



a-Bo



g



azköy and the Kültepe-Kane



s



(fig. 10.1).
The site is composed of a high central mound which is surrounded by a terrace that takes its
shape from the imposing fortification wall that ringed the settlement in the early second
millennium. Excavations were originally conducted at Ali



v



ar by the Oriental Institute of the
University of Chicago from 1927 to 1932, and although the identity of the site during the sec-
ond millennium has still not been positively affixed, there are good reasons for equating it
with the Hittite city of Ankuwa, an equation I use as a working assumption throughout this
paper (Gorny 1990: 395–437; 1994; 1997).



2




One of the fascinating discoveries made by the original Ali



v



ar excavation team was the
large number of sherds with relief decoration. Although some of these fragments are now



1. Excavations were resumed at Ali



v



ar Höyük in 1993 after a hiatus of sixty-one years. See Gorny (1994) and
Gorny et al. (1995).

2. Note, however, that Popko (1994), based on his identification of Zippalanda with Alaca Höyük, places
Ankuwa at Eskiyapar (cf. Gorny 1997).



oi.uchicago.edu





176



RONALD L. GORNY



housed in the Anadolu Medeniyetler Müzesi in Ankara,



3



many can be viewed in the Oriental
Institute Museum at the University of Chicago.



4



In all, about twenty-three pieces are known
to have been found at Ali



v



ar,



5



though only seventeen can be given any kind of provenience.
Five relief sherds, not including the present piece, are problematic and remain unpublished.



6



There are two distinctive types of relief sherds at Ali



v



ar that vary somewhat in style, motif,
and manner of execution (T. Özgüç 1957: 8; 1988: 105). On the first, both the background and



3. Schmidt 1932: 132, fig. 162; Osten 1937b: 112, fig. 153; p. 113, fig. 154; p. 115, fig. 155; and p. 116, fig.
156; Osten 1937a: 73, fig. 80:1.

4. Schmidt 1933: fig. 85, b 166; Osten 1937b: fig. 153 (= no. d 2100); fig. 154, nos. d 1622, d 2100, d 2516a,
and d 2937; fig. 155, nos. d 1620, d 2997a, d 2997b, and d 2999; fig. 156, nos. d 2518 and 2996; and fig.
157, no. d 1896.

5. There is some confusion over the exact number of pieces involved here as it appears that some pieces have
been counted twice or renumbered at a later time. If we look at the piece numbered in (Osten 1937a) as
d 2997a, for instance, we find it numbered on the sherd itself as d 1628. Similarly, d 2440 is mistakenly cited
as d 2648, which is actually a different sherd (cf. Schmidt 1932: 132–33, figs. 162–63; Osten 1937a: 111–
14, figs. 153–57).

6. These are to be published by the author in a future article.

FIGURE 10.1. Map of Hittite Anatolia (Courtesy of



Biblical Archaeologist



).

oi.uchicago.edu





177



the relief are finished with a heavy red monochrome slip. Details of body and dress are indi-
cated by stamped or incised designs. The second is characterized by a highly polished poly-
chrome finish that makes use of red, dark red, brown, white, and black paint to emphasize
artistic details. Human figures are the most common form of representation within this cate-
gory. The flesh on the polychrome examples remains red slipped, but accoutrements and
details are indicated by the application of additional colors of paint or slip. Relief decoration
is lower on the polychrome pieces than on the monochrome examples, perhaps because the
painted details make raised plastic ornamentation less necessary. In style, subject, and decora-
tion the polychrome examples from Ali



v



ar are clearly related to those on the large Old Hittite
sherd from Bitik



7



and are especially close to the reliefs from the ritual procession and cult
scenes depicted on the Inandık vase.



8



Although the monochrome figures from Ali



v



ar come
from the neck of a vessel differing in shape from the Bitik and Inandık examples, the close
resemblance of the various pieces seems to betray a common origin.

As noted above, several relief sherds found during the original Ali



v



ar excavations were
never published. Of this group, one example is of more than passing interest and deserves spe-
cial note.



9



The sherd in question comes from a wheel-turned vessel made of finely worked
clay. It depicts what appears to be a wooden framework similar in design to a picket fence
which is set atop a wheeled platform. This superstructure is painted white and juxtaposed
against a red background. The surface is burnished to a lustrous finish. Interestingly enough,
this sherd was found to join a piece previously published by Osten (d 2997a /d 1626).



10



Together, the two pieces provide curious new evidence that may help to illuminate the his-
tory and function of relief art in Anatolia as a whole, and at Ali



v



ar in particular.
After joining the two relief sherds, it becomes clear that, whereas the previously published

piece had depicted part of a wheeled cart and what appeared to be the feet of a person or deity
mounted at the rear, the new fragment reveals much of the vehicle’s superstructure (fig. 10.2)
and portrays a rather typical four-wheeled wagon with slatted or wicker sides that is similar
to other vehicles depicted in early Hittite art (figs. 10.3, 10.4; cf. Moorey 1986: 201 and figs.
2–3). The combination of these two relief fragments from Ali



v



ar is interesting, not only from
an artistic point of view, but from a cultural standpoint as well. Since the representation of this
presumably ox-drawn wagon is a theme known from other ancient Near Eastern sources,



11



the two sherds may offer us a glimpse into one aspect of Hittite culture. Although the Ali



v



ar



7. Bitik is a small site 42 km northwest of Ankara. It was excavated in 1942 by Remzi Ö



g



uz Arık. For the
results of the Bitik excavations, see Arık’s report (1944: 342 ff.) and for descriptions of the vase, see
T. Özgüç (1957: 57 ff.).

8. Inandık is located near Çankırı and is clearly an important Hittite site during the Old Hittite Kingdom. Alp,
relying on the cuneiform tablet found in association with the relief vase, has identified the site as the cult city
of



H



an



h



ana, but note the objections by Yakar (1980: 75–95). The results of excavations at the site, along with
a description of the vase, have been recently published by T. Özgüç (1988). A line drawing of the vase can
also be found in Boehmer (1983: p. 21, fig. 7) and a photograph appears in Macqueen (1986: 103, fig. 86).

9. Ali



v



ar Study Collection (ASC) 201.

10. In Osten 1937a: 115, fig. 155; this piece is published as d 2997a, but the actual piece is marked d 1626.
Excavation records fail to shed any light on this discrepancy or on the findspot.

11. For a discussion of wagons, see the reviews in Littauer and Crouwel (1979; 1986: 395 ff.); Strommenger
(1990: 267–306); Boehmer (1983: 36–42); and Leinwand (1984; 1992).



AN UNPUBLISHED RELIEF SHERD FROM ALI



V



AR HÖYÜK

oi.uchicago.edu





178



RONALD L. GORNY



relief sherd represents only a small portion of the original vessel, there is sound basis in pre-
Hittite and Hittite iconography for suggesting a connection between what is depicted on the
sherds and the Hittite cult. Tahsin Özgüç, for instance, places the Ali



v



ar sherds in the same
category as those from Inandık and Bitik and clearly states his belief that vessels of this type
are to be associated with the Hittite cult (T. Özgüç 1988: 100–06).

This being the case, the Ali



v



ar sherds then may reveal some interesting insights into reli-
gious life at Ali



v



ar. Nancy Leinwand provides some additional food for thought by pointing
out several metal examples of bull-drawn four-wheeled wagons dating to the Early Bronze
Age that seem to provide the earliest evidence for the Weathergod’s cult in ancient Anatolia
(Leinwand 1984: 173–77; 1992: 165–70)



12



and which may prefigure the two-wheeled bull-
drawn vehicles associated with the Weathergod in later Hittite iconography (Leinwand 1984:
174, n. 9, p. 192; 1992: 167, figs. 23, 25; also see Moorey 1986: 201, fig. 4).



13



The close asso-
ciation of this particular motif with Syria (cf. also an example from Tell Biºa, Strommenger
1990: pl. 101a) and southeast Anatolia suggests that its origins lie in the mixed Hurro-Luwian
culture of that region where the god in the wagon is associated with the Hurrian deity Te



s



up
(Leinwand 1992: 165–68).

Contextually, the Ali



v



ar relief sherd fills several gaps in our knowledge about life in Hittite
Anatolia. As Leinwand pointed out, not only is there a surprising absence of this motif in the



12. Of special note here is the rock-relief at Immamkulu that depicts the Stormgod striding over the personified
mountains and onto an ox-drawn cart (cf. Kohlmeyer 1983: 80–86, esp. 83, fig. 33, pls. 29–30).

13. Technologically, the Ali



v



ar relief sherd illustrates the continuation of the four-wheeled vehicle into the Old
Hittite period where it precedes the more maneuverable two-wheeled chariot for which the Hittites were
later to be known (cf. Moorey 1986).

FIGURE 10.2 Joined relief sherds ASC 201 (top) and d 2997a (bottom). Scale: 2:5.



spread one pica long

oi.uchicago.edu

[image: image55.tif]



179



glyptic of the Old Assyrian Colony period, or roughly between the third millennium and the
late second millennium (1992: 166), but it also remains largely unattested in the iconography
of the geographic area between southeastern Anatolia and the Hittite capital at Bo



g



azköy
(1992: 167–68). The Ali



v



ar relief sherd becomes significant, therefore, in that it not only
bridges the chronological gap between the third and the late second millennium manifesta-
tions of this motif, but it also serves geographically to link the Bo



g



azköy representations with
those of the Hurro-Luwian populations of southeast Anatolia where it seems to be at home.



14



As such, these links may well portend the subsequent reconfiguation of Hittite religion wit-



14. In light of recent discoveries, the increasing influence of Hurrian culture on Hittite Anatolia should come as
no surprise; see, for instance, the discovery of nearly 2,000 cuneiform tablets written primarily in Hurrian
at the site of Ortaköy (Hittite Sapinuwa) near Çorum (Gates 1996).

FIGURE 10.3. Enlarged drawing of four-wheeled cart of Ali



v



ar type from Karum Kane



s



II cylinder seal.
(After Littauer and Crouwel 1979; Moorey 1986.)



AN UNPUBLISHED RELIEF SHERD FROM ALI



V



AR HÖYÜK

oi.uchicago.edu

[image: image56.tif]



180



RONALD L. GORNY



nessed in the rock-cut reliefs of Yazılıkaya where the Anatolian Weathergod is represented in
the iconography by the Hurrian deity Te



s



up.



15



Bearing all this in mind, we should recall that the Ali



v



ar relief sherd is still incomplete.
However, despite the fact that no bulls or oxen are visibly linked to the wagon on the sherd,
the presence of several other fragments with oxen in relief provide good reason to believe they
once existed as part of the overall motif.



16



One relief sherd (Schmidt 1933: fig. 85, b166), in
fact, depicts a pair of yoked oxen that nicely fit the motif suggested by Leinwand (1988).
While this is the one sherd that comes from a different portion of the mound (square G8), the
presence of yoked oxen and a mounted wagon at Ali



v



ar adds further support to her suggestion,



15. We are also reminded that



S



arruma, who is often syncretized with the Stormgod of Nerik (KUB 21.27 + rev.
iv 42



u



; see Haas 1970: 107 ff.; but cf. Deighton 1982: 103, n. 4), as well as with the Stormgod of Zippalanda
(Popko 1994: 32–33), is represented at Yazılıkaya as the son of Te



s



up (the Stormgod of



H



atti) and



H



epat (the
Sungoddess of Arinna). For more on the idea of reconfiguration in Hittite religion, see Gorny 1996: 70–71.

16. Further evidence for this motif comes from a colossal statue of the Stormgod recently discovered in a field
near Adana. The white limestone statue is nearly 3.0 m in height and portrays the god on a black basalt base
in the form of a chariot drawn by bulls. While the chariot depiction represents a technological and chrono-
logical evolution from the use of a wagon, the basic motif remains the same. The statue is apparently late as
the base has a hieroglyphic and a Phoenician inscription, presumably bilingual, which is similar to Karatepe
inscriptions, perhaps the work of Urikki-Awariku (cf. Ipek, Tosun, and Toko



g



lu 1999).

FIGURE 10.4. Enlarged drawing of four-wheeled cart of Ali



v



ar type from Karum Kane



s



II cylinder seal. (After
Littauer and Crouwel 1979; Moorey 1986.)



spread one pica long

oi.uchicago.edu

[image: image57.tif]



181



and fits in well with textual materials that describe similar scenes. Of special note in this con-
nection are the festivities associated with the Weathergod at Zippalanda.



17



The types of vessels on which relief representation appears can be very different. Several
of the Ali



v



ar specimens, for example, come from the neck of a bottle-shaped jug while another
comes from a bowl.



18



The best example of a relief vessel, however, is from a large complete
vase that was found at the Old Hittite period site of Inandık.



19



The ornamentation on this
extraordinary vessel was composed of four registers of human figures, each depicting some



17. See, for instance, KBo 13.214 rev. 4 10 + 1 ff. where, after riding in a wagon to Zippalanda and performing
offerings in the temple of the Weathergod, the Hittite king mounts a chariot and rides off to either Katapa or
Ankuwa. The representation on the sherd may then be that of the Hittite king as he rides in a wagon to one
of the cities on the cult itinerary.

18. Three pieces labeled d 2100 come from the neck of a bottle-shaped jug (Osten 1937a: figs. 153, 154)
while d 2996 (Osten 1937a: fig. 157) comes from a large bowl.

19. See footnote 8, above.

FIGURE 10.5. Distribution of relief sherds at Ali



v



ar Höyük (after Gorny 1995c: fig. 6).



AN UNPUBLISHED RELIEF SHERD FROM ALI



V



AR HÖYÜK

oi.uchicago.edu

[image: image58.tif]



182



RONALD L. GORNY



aspect of the cult.



20



The representation of music, processions, and worship on the sherds is a
strong indication of the relief sherds’ cultic character. Stylistically, the vessels from all three
sites represent a similar style and must have come from the same period.

While the Inandık vase was important in and of itself, what gave its discovery special sig-
nificance was the discovery of an Old Hittite clay cuneiform tablet from the period of



H



attu-



s



ili I in the northern part of the temple (Balkan 1973; T. Özgüç 1988: 71, 110–11). Since the
vessel would then have to be contemporary with the tablet or of a slightly later date, it is pos-
sible to supply an Old Hittite kingdom date in the sixteenth century not only for the Inandık
vase, but also for the whole group of similarly executed vessels.

It should be noted, however, that numerous examples of relief ware were found in the
excavation of Bo



g



azköy’s Upper City (Neve 1992: fig. 78; also Parzinger and Sanz 1992: 62–
63, pls. 66–68). Since the Upper City is essentially an Empire period construction, the discov-
ery of so many pieces would seem to create a problem for our date in the Old Hittite Kingdom
period. Parzinger and Sanz note, however, that these pieces were fashioned in the Old Hittite
and Early Empire period (Parzinger and Sanz 1992: 62–63) but had a long life-span with
many of the vessels apparently continuing in use all the way into Upper City Period 3
(between ca. 1240 and 1220



b.c.



; for dates see Parzinger and Sanz 1992: 72–73). The pres-
ence of Old Hittite ceramic vessels of this type is also consistent with the discovery of other
Old Hittite materials in the area such as tablets and seals (Neve 1992: 30, also fig. 83). The
longevity of such materials undoubtedly results from a purposeful preservation based on the
value accorded the materials for historical or cultic reasons. This archaic style of relief ware
is no longer attested, however, in Upper City Period 2 (Parzinger and Sanz 1992: 63), perhaps
as a result of the destruction that befell the city at the hands of the usurper Kurunta in the reign
of Tud



h



aliya IV (Neve 1992: 19).
The date of the Ali



v



ar relief fragments must then, based on analogy, be set somewhere in
the Old Hittite Kingdom, a date already proposed by Bittel (1955: 32). The same conclusion
was later reached by Boehmer whose recent publication on the relief ceramics uses the same
logic for placing the Ali



v



ar relief ware in the Hittite Old Kingdom.



21



An exception is a single
piece that differs significantly enough from the other Ali



v



ar examples that it is thought to be
contemporary with the Early Hittite Empire sherds from Bo



g



azköy (above).



22



20. It may be objected at this point that, based on the two clear depictions of sexual intercourse on the vessel,
the Inandık vase is more characteristic of a (possibly royal) wedding scene. From what we can tell, such
scenes are not known in the Hittite cult. T. Özgüç, however, calls it a “sacred marriage” (1988: 100 ff.) and
states that “all of these cannot represent anything but cult scenes.” In this respect, such activities are proba-
bly tied to the divinities’ role as the guarantor of health and prosperity throughout the Hittite lands.

21. Boehmer (1983)







places sherds c 2623 (musician with cymbals, p. 128), d 2517 (offering bearer?, pp. 22–
24), d 2997b (“alter ?” p. 31), d 2100, fig. 36 (god, p. 34), d 2997a (wagon, p. 37), b 166 (bull, p. 40), d 1896
(two horse heads, p. 45), and d 2518 (cow, p. 47) in the Old Hittite period. In the course of the project, piece
d 2997a was found to join ASC 201. The resulting combination shows a wheeled wagon with vertically
slatted sides.

22. The one example that seems to fall outside of this early period is a sherd depicting a deer with back-turned
head and bulging pupils which dates to around 1400 (Osten 1937a: 114, 116, fig. 156, d 2996; cf. Boehmer
1983: 59). A similarly styled deer is also known from a complete vase now in the Cleveland Museum of
Art (see Turner 1986: 36, no. 24; also see photograph of the same piece in the



Bulletin of the Cleveland

oi.uchicago.edu





183



One point of note in the manufacture of the Inandık vase is an enclosed tubular channel
around the wide everted rim of the vessel. The channel connects spouts that are molded in the
shape of a bull’s head and allows fluids to empty through the bull’s mouth into the vessel’s
cavity.



23



Such vessels seem to have been relatively common at Anatolian sites and may have
been used for mixing liquids in the various types of rituals commonly found in the Hittite
cult (Gorny 1995a). While no complete vessels of this type were found at Ali



v



ar, several
fragments of similarly constructed vessels were unearthed there.



24



Additional vessels of this
type are also known from Alaca Höyük (Ko



s



ay 1951: pl. 70, 1a–b, 2a–b); Ma



v



at (T. Özgüç
1982: 152, pl. 87, 1a–b); and Eskiyapar (T. Özgüç 1988: 117, 145, pl. 29, D3); and Yörüklü.



25




The distribution pattern of the relief sherds at Ali



v



ar may also be of some significance
(fig. 10.5). With the exception of the relief fragment found in square G8 (b166), the location
of findspots for the other relief sherds attributed to the Old Hittite period (at least sixteen
pieces) are all clustered in the vicinity of the so-called Mansion, a multi-roomed building on
the terrace which seems to have been of some importance during the Old Hittite Kingdom and
early Hittite Empire periods (fig. 10.6; cf. Gorny 1990: 366–69; 1996). The largest group (nine
sherds in all) came from squares S-U 28 and R-U 27–31 just east of the Mansion and north
of Building B.

The concentration of Old Hittite relief sherds in this limited space may provide us with
some clues about the area’s functional character. In particular, if one accepts the premise cited
above that the relief sherds have special relevance to the Hittite cult, the density of sherds in
the area between the Mansion and Building B may well suggest the presence of a religious
sanctuary somewhere in the vicinity, perhaps associated with a much larger complex of build-
ings to which both the Mansion and Building B apparently belonged.



26



Unfortunately, a



23. For further discussion concerning this type of vase, see T. Özgüç 1988: 84–85, color pl. H, nos. 3–4, pl. 41;
Boehmer 1983: 45–54, esp. figs. 40–44.

24. Osten 1937a: 121, fig. 163, e 1721 (note that e330 is also of a similar style).

25. Two large relief vases were recently uncovered at a spot known as Huseyindede in north-central Turkey. The
site is near Yörüklü in the northern part of the Sungurlu district (Çorum province). Most of the fragments
of these two relief vases were recovered, making it possible to restore the vessels nearly completely. The
two vases resemble the Inandik vase (above) with scenes displayed in decorated bands characterized by
brightly colored figures in high relief. Based on analogy to other such pieces, the Yörüklü vases probably
date to the reign of



H



attu



s



ili I (1650–1620



b.c.



). Of note is the fact that the smaller of the two vases depicts
two somersaulting acrobats leaping over a bull, which is reminiscent of the famous scenes from the Minoan
Palace at Knossos. However, since the vases apparently predate the Minoan Palace frescoes, they may pro-
vide important new evidence for cross-cultural links, perhaps even Anatolian influences on Minoan civili-
zation (cf. Ediz et al. 1999).

26. The Mansion and Building B seem to have been the westernmost part of a larger complex of buildings that
probably focused on an interior courtyard. This can best be seen in Osten’s plan of the Mansion (Osten
1937a, pl. 18). The remainder of the building was apparently destroyed during the Phrygian renovations of
the seventh century



b.c. More of the complex may still be located in the vicinity as a large structure with a
stone foundation can still be seen in the remains of a tunnel von der Osten dug directly below square O 31
(Osten 1937b: 12, fig. 81).

Museum of Art 72: 248). Note, however, that Muscarella questions the authenticity of the Cleveland vase
(pers. comm.).

AN UNPUBLISHED RELIEF SHERD FROM ALIVAR HÖYÜK

oi.uchicago.edu



184 RONALD L. GORNY

FIGURE 10.6. Plan of the so-called Mansion at Alivar Höyük (after Osten 1937a: pl. 17).

oi.uchicago.edu

[image: image59.tif]

185

massive disturbance of the site by later Iron Age settlers has, so far, made this impossible to
substantiate in archaeological terms. If, however, we are to take this line of reasoning a little
further, and we accept the linkage of the ox-dawn cart motif on specialized ceramic vessels
with the cult of the Stormgod, it would suggest the possibility of a sanctuary located in this
area of Alivar that might have been dedicated to such a deity.

One might argue that if we accept the equation of Alivar Höyük with the ancient Hittite
city of Ankuwa, we might posit a connection between the Mansion and the goddess Katahha
who is known to have been the patron deity of that town (Ünal 1980: 477–78; 1984: 87–107).
The prominence of her cult in the city of Ankuwa might lead us to believe that any temple or
cultic area found at Alivar would most likely have been dedicated to her.27 Hittite texts from
Bogazköy show, however, that several Weathergods were also worshipped in Ankuwa, among
which was the Weathergod of Zippalanda and the Weathergod of the Rain.28 Thus, based on
our previous discussion, it is plausible to understand the presence of the relief sherds on the
Alivar terrace as indicating the approximate location of an artifice dedicated to some manifes-
tation of the Weathergod.

In the end, if Alivar can be equated conclusively with Ankuwa, the Mansion may well turn
out to have housed the worship of other deities,29 although the Weathergod remains the most
likely candidate for its primary occupant. A temple for the Weathergod in the vicinity of the
Mansion would not, however, exclude the possibility of a temple dedicated to a patron deity
such as Katahha on another part of the terrace or, more likely, on the ravaged citadel mound.30

It is also conceivable, on the other hand, that Katahha could have shared a temple with this

27. A temple is attested for Katahha at Ankuwa in an AN.TAH.SUM festival text (KUB 11.27 vi 3u).
28. A Rain Festival was known at Ankuwa to which this particular deity may have been connected. See KUB 30

73 (NH colophon for AN.TAH.SUM Festival) and KBo 22.214 (colophon). For more on the Rain Festival,
see Ünal (1984: 102–3), and more recently Jakob-Rost (1990: 35–39). Other deities known to have been
worshipped at Ankuwa include Istar (Sausga) of the fields (KUB 27.1 ii 49), the Stormgod of the Heavens
(dU ANE KUB 11.27), Stormgod of the houses (dU ÉTIM), the Sungod of the Earth of Ankuwa (KBo 34.203
iii 11u, Zawalli, divinity of Ankuwa (KUB 5.6 ii 65–57), and a variety of other general gods (for a complete
list, see del Monte and Tischler 1978: 21–22; 1992: 6–7).

29. The halentuwa, a building now understood to be a “palace” or “residence,” is known to have housed mul-
tiple temples. Puhvel notes the appearance of halentuwa in ritual text KUB 32.13 ii 3 (1992: 15–19) as the
equivalent (ibid., I 2) of the Hurrian word hikkali (Ugr. hkl , and Heb. Hekal) which translates as “palace”;
also see Popko’s description of the halentuwa as a residence (1994: pp 18 ff. and 24 ff.). For examples, see
KUB 11.34 vi 51–52, INA É DIM URU halen[tu]was É.DINGIR.MES-ya humandas, in the temple of the
Stormgod and all the temples of the halentuwa; KUB 30.34 iv 4–5, kinuna Éhalinduwas É.DINGIR.MES-ya
parkunut, and now he has cleansed also the temples of the halentuwa: KUB 30.34 rev. 7–8, nu eshananza
linkiyaz Éhalinduwa É.DINGIR.MES le epzi, may bloodshed [and] perjury not seize the temples of the hal-
entuwa. Also see Puhvel (1992: 19) where he notes that in this context the habitual plural is significant.
Also see Singer (1975: 65–84; 1983:111–12).

30. Once again, this can not be substantiated by archaeological finds because the citadel at Alivar was much dis-
turbed by the later Iron Age occupants who leveled the top of the mound for their new constructions. Very
few remains dating to either the Hittite Old Kingdom or the Hittite Empire have been preserved from the cit-
adel. The discovery of the Old Hittite relief sherd mentioned above in G 8, however, provides evidence of
at least a Hittite Old Kingdom settlement on the citadel mound. Empire period biconvex seals found in asso-
ciation with the citadel walls point to an even later thirteenth century settlement (Gorny 1993: 163–91).

AN UNPUBLISHED RELIEF SHERD FROM ALIVAR HÖYÜK

oi.uchicago.edu



186 RONALD L. GORNY

Weathergod and that relief-ware vessels come from activities held in a building with a double
adyton such as is suggested for Temples I and V at Bogazköy (Bittel 1970: 57). We hope that
future excavations will help to clarify this situation.

1 It has been suggested, in fact, that Katahha was the consort of the Weathergod of Zippalanda, making the
suggestion of a multiple adyton more appealing (see Popko 1994: 30, 35). It should be noted however, that
KUB 11.27 vi 1u–7U seems to indicate that the temple of Katahha was separate from the halentuwa and
apparently existed as an entity unto itself.

ACKNOWLEDGMENT

I would like to thank Dr. Harry Hoffner who graciously allowed me to use the files of the
Hittite Dictionary for this paper.

REFERENCES

Arık, R.
1944 Bitik Kazısı ve Hatay, Tetkikleri Hakkında Kisa Rapor. Belleten 8: 342 ff.

Balkan, K.
1973 Eine Schenkungungsurkunde aus der althethitischen Zeit, gefunden in Inandik 1966. Ankara:

Türk Tarih Kurumu Basımevi.

Bittel, K.
1955 Eine hethitische relief Vase aus Kappadokien. Pp. 29-32 in Festschrift fur Carl Weikert, ed. G.

Bruns. Verlag Gebr. Mann.
1970 Hattusha: The Capital of the Hittites. New York: Oxford University Press.
1978 Fragment einer Hethischen Reliefscherbe mit Wagendarstellung. Pp. 178–82 in Studien zur Reli-

gion und Kultur Kleinasiens (festschrift Dorner),

Boehmer, M. R.
1983 Die Reliefkeramik von Bogazköy (= Hattusa-Bogazköy 13). Berlin: Gebr. Mann Verlag.

Deighton, H. J.
1982 The “Weather-God” in Hittite Anatolia: An Examination of the Archaeological and Textual

Sources. British Archaeological Review, International Series 143. Oxford: British Archaeologi-
cal Reports.

Ediz, I.; Ipek, Ö.; Sipahi, T.; and Yildirim, T.
1999 Yörüklü/Huseyindede Kirtarma Kazısı. Pp. 189–98 in IX Mueze Kurtarma Kazıiları Semineri

(27–29 Nisan 1998). Ankara: Üniversitesi Basımevi.

Gates, M.- H.
1996 Ortaköy—Çorum. Archaeology in Turkey. American Journal of Archaeology 100: 297–98.

Gorny, R.
1990 Alivar Höyük in the Second Millennium b.c. Ph.D dissertation, University of Chicago.
1993 The Biconvex Seals of Alivar Höyük. Anatolian Studies 43: 163–91.
1994 The 1993 Season at Alivar Höyük in Central Turkey. Anatolica 20: 191–202.
1995 Viticulture in Ancient Anatolia. Pp. 133–75 in Origins and Ancient History of Wine, eds.

P. McGovern, S. Fleming, and S. Katz. Newark: Gordon and Breach.
1996 Hittite Imperialism and Anti-Imperial Resistance as Viewed from Alivar Höyük. Bulletin of the

American Schools of Oriental Research 299/300: 65–89.
1997 Zippalanda and Ankuwa: The Geography of Central Anatolia in the Second Millennium b.c.

Review article of Zippalanda: Ein Kultzentrum im hethitischen Kleinasien (Heidelberg: Uni-
versitätsverlag C. Winter). Journal of the American Oriental Society 117.

Weathergod31 and that relief-ware vessels come from activities held in a building with a double
adyton such as is suggested for Temples I and V at Bogazköy (Bittel 1970: 57). We hope that
future excavations will help to clarify this situation.

31. It has been suggested, in fact, that Katahha was the consort of the Weathergod of Zippalanda, making the
suggestion of a multiple adyton more appealing (see Popko 1994: 30, 35). It should be noted however, that
KUB 11.27 vi 1u–7U seems to indicate that the temple of Katahha was separate from the halentuwa and
apparently existed as an entity unto itself.

oi.uchicago.edu



187

Gorny, R.; McMahon, G.; Paley, S.; and Kealhofer, L.
1995 The Alivar Regional Project 1994. Anatolica 21: 65–100.

Haas, V.
1970 Der Kult von Nerik: Ein Beitrag zur hethitischen Religionsgeschichte. Studia Pohl 4. Rome:

Papistliches Bibelinstitut.

Ipek, I.; Tosun, A. K.; and Tekoglu, R.
1999 Adana Geç Hitit Heykeli Kurtarma Kazisı 1997 Yili Çalismasi Sonuçlari. Pp. 173–88 in IX

Mueze Kurtarma Kazıları Semineri (27–29 Nisan 1998). Ankara: Üniversitesi Basımevi.

Jakob-Rost, L.
1990 Bemerkungen zum hethitischen Regenfest: Ein Beitrag zum hethitische Festkalender. For-

schungen und Bericht 29–30: 35–39.

Kohlmeyer, K.
1983 Felsbilder der hethitischen großreichszeit. Acta praehistorica et archaeologica 15.

Kosay, H. Z.
1951 Türk Taruh Kurumu tarafindan yapilan Alaca Höyük Kazısı. 1937–1939 daki çalivmalara ve

keviflere ait ilk rapor. Ankara: Türk Tarih Kurumu Yayınları series 5, no. 5.
Leinwand, N.

1984 A Study of Anatolian Weathergods of the Old Assyrian Colony Period. Ph.D. dissertation, Bryn
Mawr College.

1988 A Clue to Interconnections in Cult and Art in Early Bronze Age Anatolia. Paper delivered at
College at the Rencontre Assyriologique Internationale, Bryn Mawr, Philadelphia.

1992 Regional Characteristics in the Style and Iconography of the Seal Impressions of Level II at
Kültepe. Journal of Ancient Near Eastern Studies 21: 141–72.

Littauer, M., and Crouwel, J.
1979 Wheeled Vehicles and Ridden Animals in the Ancient Near East. Leiden: E. J. Brill.
1986 The Earliest Known Three-dimensional Evidence for Spoked Wheels. American Journal of

Archaeology 90: 395 ff.

Macqueen, J.
1986 The Hittites and Their Contemporaries in Asia Minor. Revised ed. London: Thames and Hudson.

Meluzin, S. A.
1969 A Re-examination of Alivar Höyük in the Light of Recent Evidence. Masters thesis, University

of Chicago.

Monte, G. F. del, and Tischler, J.
1978 Die Orts- und Gewässernamen der hethitischen Texte. Répertoire Géographique des Textes

Cunéiformes 6. Wiesbaden: Ludwig Reichert.
1992 Die Orts- und Gewässernamen der hethitischen Texte, Supplement. Répertoire Géographique des

Textes, Cunéiformes, vol. 6, no. 2. Wiesbaden: Dr. Ludwig Reichert Verlag.

Moorey, P. R. S.
1986 The Emergence of the Light Horse-Drawn Chariot in the Near East c. 2000–1500 b.c. World

Archaeology 18: 196–215.

Neve, P.
1992 Hattusa—Stadt der Götter und Tempel. Antike Welt (special issue).

Osten, H. H. von der
1937a The Alishar Hüyük: Seasons of 1930–32, Part 2. Oriental Institute Publications 29. Chicago:

Oriental Institute, University of Chicago Press.
1937b The Alishar Hüyük: Seasons of 1930–32, Part 3. Oriental Institute Publications 30. Chicago:

Oriental Institute, University of Chicago Press.

Özgüç, T.
1957 The Bitik Vase. Anatolia 2: 57 ff.
1988 Inandıktepe: An Important Cult Center in the Old Hittite Period. Türk Tarıh Kurumu Yayınları,

series 5, no. 43. Ankara: Türk Tarih Kurumu Yayınları.

AN UNPUBLISHED RELIEF SHERD FROM ALIVAR HÖYÜK

oi.uchicago.edu



188 RONALD L. GORNY

Parzinger, H., and Sanz, R.
1992 Die Oberstadt von Hattusa: Hethitische Keramik aus dem zentralen Tempelviertel. Berlin: Gebr.

Mann.

Popko, M.
1994 Zippalanda: ein Kultzentrum im hethitischen Kleinasien. Heidelberg: Üniversitätsverlag Carl

Winter.

Puhvel, J.
1992 Halentuwa. Pp. 15–19 in Hittite Etymological Dictionary, vol. 3. Amsterdam: Mouton.

Schmidt, E.
1932 The Alishar Hüyük Seasons of 1928 and 1929, Part 1. Oriental Institute Publications 19. Chi-

cago: Oriental Institute, University of Chicago Press.
1933 The Alishar Hüyük Seasons of 1928 and 1929, Part 2. Oriental Institute Publications 20. Chi-

cago: Oriental Institute, University of Chicago Press.

Singer, I.
1975 Hittite hilammar and Hieroglyphic Luwian hilana. Zeitschrift für Assyriologie 65: 69–103.
1983 The Hittite KI.LAM Festival, Part One. Studien zu den Bogazköy Texten 27. Wiesbaden: Har-

rassowitz.

Strommenger, E.
1990 Planwagen aus dem mittleren Euphrattal. Pp. 297–306 in Resurrecting the Past: A Joint Tribute

to Adnan Bounni, ed. P. Matthiae, M. van Loon, and H. Weiss. Istanbul: Nederlands Historisch-
Archaeologisch Instituut te Istanbul.

Turner, E. H.
1985 Annual Report: Ancient Art. Bulletin of the Cleveland Museum of Art 72.
1986 The Year in Review for 1985. Bulletin of the Cleveland Museum of Art 73: 27–71.

Ünal, A.
1980 Kattaha. Reallexikon der Assyriologie 5: 477–78.
1984 Nochmals zur Geschichte und Lage der hethitischen Stadt Amkuwa. Studi micenei ed egeo-

anatolici 24: 87–107.

Yakar, J.
1980 Recent Contributions to the Historical Geography of the Hittite Empire. Mitteilungen der deut-

schen Orient-Gesellschaft 112: 75–95.

oi.uchicago.edu



189



EARLY BRONZE AGE II–III PALESTINIAN
CYLINDER SEAL IMPRESSIONS AND THE

NORTH CANAANITE METALLIC WARE JAR



Raphael Greenberg



More than two decades have passed since Ben-Tor’s pioneering compilation and discussion
of Early Bronze Age (EB) Palestinian cylinder seals and sealings (Ben-Tor 1978). During this
time, the number of glyptic finds has more than doubled, and the number of flat-carved geo-
metric and cultic seal impressions of the EB II–III—included in Ben-Tor’s Classes I and III—
has increased fivefold, approaching a total of 250 (e.g., Lapp 1989; Esse 1990; Ben-Tor 1992,
1994; Mittmann 1994; Greenberg 1996, forthcoming). Nonetheless, the following fundamental
issues left unresolved by Ben-Tor have not been systematically addressed: (a) The chronol-
ogy of the EB II–III seal impressions has not been refined; (b) their function has not been
ascertained; and (c) their specific provenance has not been established.

New evidence presented since 1974, when Ben-Tor’s corpus was compiled, casts important



light on issues



a



and



c



. As for issue



b



, the question of function, the solution seems tantalizingly
close at hand. The key to all three issues is the understanding that seal impressions are not arti-
facts in themselves; each different seal-impressed sherd represents the jar or pithos to which it
was affixed. It is these vessels, nearly always identifiable as northern Canaanite



1



metallic ware
jars or pithoi, that comprise the subject of the following pages.

CHRONOLOGY

It is a curious and significant fact that, to date, not one cylinder seal impression of the
EB II–III (Classes I, III) can be associated with a complete vessel, nor even with a rim. Thus,
ceramic morphology offers little in the way of chronology, and dating relies largely on other
considerations, such as ceramic fabric and stratigraphic context (insofar as stray sherds may
be dated by their context). Recent finds appear to provide ample evidence for the dating of the
sealings to both EB II and EB III. Finds from well-defined EB II contexts come from Beit
Yera



˙



(Esse 1990: 30–31), Tel Dan (Greenberg 1996: 149), and Qiryat Ata (Golani 1996;
Greenberg forthcoming) as well as from sites that were abandoned before the onset of EB III
(e.g., Tel Kinnerot—Fritz 1990: 23–24; Beit Ha ºEmeq—Beck 1976; Givon 1993). The very
large collection of seal impressions from Khirbet ez-Zeraqun, representing more than one
half of the entire corpus (Mittmann 1994), seems mostly to originate in EB III contexts
and can be associated with pithoi common in Zeraqun EB III levels (although at least some



1. The term northern Canaan is used to denote the area currently comprising northern Israel, northwest Jordan,
southeast Syria, and southern Lebanon. There are occasional finds from more southerly sites (e.g., Lapp
1989), but these should be seen as offshoots of what is fundamentally a northern phenomenon.



11



oi.uchicago.edu





190



RAPHAEL GREENBERG



seal-impressed sherds could be residual, from EB II strata thus far inferred only on the basis
of finds in fills—H. Genz, pers. comm.).

The attribution of the seal impressions to both EB II and EB III conforms with what is known
of the chronology of the vessels to which they were affixed. Metallic ware jars, thin-walled and
no more than 60 cm in height, were produced mainly in EB II. Pithoi, thick-walled and 80–
105 cm high, were produced both in EB II and EB III (Greenberg and Porat 1996); thus it seems
quite likely that the production of the seals spanned the transition between the two periods.

Circumstantial evidence can be adduced to narrow the chronological range within EB II–
III: numerous restorable pithoi representing the final phase of occupation at Khirbet ez-Zeraqun
are consistently unsealed (with the exception of one jar, its rim removed, impressed with a
unique variant of the cultic-type seals); of the 126 vessels represented by the Zeraqun sealings,
only the one mentioned above could be even partly restored (H. Genz, pers. comm.). Two
restorable late EB III pithoi, one at Hazor (Area A, L. 636) and one at Tel Dan (Area A, L.
18), are not sealed (Greenberg 1996: fig. 3.32:10; 1997: fig. II.4); the same is true of pithoi
found at Beit Yera



˙



(E. Eisenberg, pers. comm.). It would thus seem safe to conclude that the
seal-impressed vessels do not belong to the latest phases of settlement at these sites, but to a
somewhat earlier phase. The lack of restorable sealed vessels in EB II strata at the above-men-
tioned and other sites could be attributed to (a) the generally small exposures excavated in this
period; (b) the absence of EB II destruction layers at some sites (e.g., Tel Qashish); and (c) the
apparently smaller number of seals produced in this period (assuming EB III Zeraqun repre-
sents the high-water mark of seal production and distribution).

Thus a gradual increase in the application of cylinder seals to metallic ware jars and pithoi
may be posited, following their introduction in EB II. The use of sealings peaks in early
EB III, then rapidly declines before the end of the period.

PROVENANCE

As cylinder seals were applied to the ceramic vessels before firing (but after the comple-
tion of all other aspects of vessel-forming, including pattern-combing), the provenance of the
seals may be assumed to be identical to that of the jars themselves; and as the vessels to which
virtually all the geometric and cultic impressions from northern sites were affixed were of
metallic ware, studies of the provenance of such ware are pertinent to this discussion.

Petrographic studies of metallic ware, conducted by Naomi Porat and Yuval Goren, have
shown that all analyzed metallic ware vessels from EB II–III contexts in northern Israel are
comprised of a similar association of raw materials derived from Lower Cretaceaous formations
which crop out from the Hermon foothills and the flanks of the Naphtali hills and northward,
into Lebanon (Greenberg and Porat 1996). This would suggest that the seal-impressed vessels
originated from ceramic workshops situated on the northern margins of the Hula Valley or
northward (but not too far north, in view of the absence, so far, of Class I and III seals on
metallic ware in Lebanon).

In order to confirm this attribution, Porat examined seal-impressed sherds with both geo-
metric and cultic motifs from six widely separated sites in northern Israel, using petrographic
and trace element analyses (these will be presented in a forthcoming report). The analyses
demonstrated that sealed sherds from Tel Dan, Beit Yera



˙



, H. ºEn



Ó



or (near



Ó



anita), Beit Ha
ºEmeq, Qiryat ºAta, and Tel Qashish (figs. 11.1, 11.2) were all of similar Lower Cretaceous



spread one pica long

oi.uchicago.edu





191



EARLY BRONZE AGE II–III PALESTINIAN CYLINDER SEAL IMPRESSIONS



FIGURE 11.1. Selection of analyzed cylinder seal impressions of Classes I (1–3, 5, 7, 8) and III (4, 6, 9)
from Beit Ha ºEmeq (1–2), Tel Qashish (3–4), Beit Yera



˙



(5–6), and Tel Dan (7–9).

oi.uchicago.edu

[image: image60.tif]



192



RAPHAEL GREENBERG



clays and tempers. This again suggests a single zone of production for the sealed vessels. In
view of the distance to the nearest plausible clay sources from sites such as Beit Yera



˙



(50 km,
as the crow flies) and Tel Qashish (80 km), the suggestions regarding local production of seals
in the Jezreel Valley (Ben-Tor 1992: 160) or by itinerant potters at Beit Yera



˙



and other sites
(Esse 1990: 32*) no longer seem tenable. Figure 11.2 illustrates the congruence in the distri-
bution maps of metallic ware and of Class I and III flat-carved cylinder seal impressions.

FUNCTION

The fragmentary nature of the evidence bars confident assertions about the function of cyl-
inder seals. Nonetheless, with a corpus of more than two hundred sealings amassed over several
decades, the question of function cannot be avoided.

At the purely technical level, the seal impressions usually served to mask and strengthen
the join between the wheelmade neck and handmade body of the jars—a function often filled,
in unsealed jars, by applied rope-decoration (e.g., Ben-Tor 1975: fig. 10:4, 5, 7, 8; Givon



FIGURE 11.2. Map of northern Canaan showing area of intensive distribution of metallic ware and sites yield-
ing Classes I, III cylinder seal impressions of EB II–III. [Hatching indicates the approximate extent of mass dis-
tribution of metallic ware; dots represent known findspots of EB II, III cylinder seal impressions (Classes I, III)].



long

oi.uchicago.edu

[image: image61.png]



193



EARLY BRONZE AGE II–III PALESTINIAN CYLINDER SEAL IMPRESSIONS



1993: fig. 15:2). In this, the sealed metallic ware jars differ significantly from their EB I fore-
runners (e.g., ºEn Shadud and Megiddo—Braun 1985: figs. 34, 35) and their EB IV Syrian
successors (Mazzoni 1984: 20; 1993: figs. 8–10), on which seals were applied either randomly
or on the rim itself.

Beyond this technical aspect, the cylinder seal impressions may be assumed to have been
related to some aspect of the economic role of the vessels. Past interpretations have attempted
to associate seals either with the content of the vessels or with the circumstances of their manu-
facture (e.g., Ben-Tor 1978: 101–4; Mazzoni 1984, 1993; Esse 1990; for discussions of similar
problems in Cyprus, the Aegean, and the Near East, see Webb and Frankel 1994; Aruz 1994;
Pittman 1994). These approaches must be reexamined in light of the identification, presented
above, of a single zone of production for the seal-bearing jars. Fundamental to whatever
approach is taken is the recognition that no regular pattern has yet emerged in the distribution
of either geometric or cultic motifs, nor any predictable association of specific motifs with spe-
cific sites. Even among the 126 different sealings at Zeraqun, the vast majority are types well
known from other sites in northern Canaan, and only a small number of new types have been
identified (Mittmann 1994:15; H. Genz pers. comm.). This can only mean that all the motifs
conveyed a limited, fixed set of meanings, universally understood at all sites, and lacked con-
notations specific to the sites and contexts in which they were recovered.

Insofar as the issue of content is concerned, the sealing of the pot could (a) indicate stan-
dardization of volume, (b) comprise an a priori dedication of specific pots for a specific type
of content (best wine, temple oil, priest’s tithe, and the like), or (c) have a general protective
intent. The central production of the seal-bearing jars would indicate the existence of either a
central authority commissioning the seals and controlling weights and measures (for possibil-
ity



a



, above), or—what is more likely—a high degree of cultural uniformity and integration,
permitting the widespread use of similar symbols to denote similar functions at each site.
There is no evidence yet for a central store for sealed jars, or for the ingathering of vessels to
a specific site.

lf considered as a potter’s device, the seals may be understood as trademarks of different
ateliers in the zone of production. This may well explain their wide and apparently random
distribution and accords with their consistent functional application at the neck-rim join; yet
it leaves unanswered questions such as why most vessels (including all restorable jars found
to date) were not decorated, and what would have been the relation between potter’s seals and
the potter’s marks found on many metallic ware vessels, including jars (e.g., Givon 1993:
fig. 15:1; Greenberg 1996: fig. 3.27:12).

To complicate the issue further, the function and significance of the seal impressions may
have changed over time, in tandem with shifts in the contexts of pottery production between
EB II and EB III. In EB II, northern Canaan manifests features of a centralized scale economy:
within a cultural



oikumene



that embraced settlement patterns, architecture, and various aspects
of material culture, the pottery industry looms large, with great quantities of metallic ware
bowls, platters, vats, jugs, jars, and pithoi being exported from workshops, possibly in the
vicinity of Mt. Hermon, to points as far as Beit Ha ºEmeq and Tel Qashish (see Greenberg and
Porat 1996). The affixing of seals to EB II metallic ware



jars



, which could be transported with
their contents, might have been commissioned by an agent of authority with the purpose of
marking the content of the jar. Under this configuration, the significance of the sealing could
be understood in a manner similar to that suggested with regard to central stores, such as those

oi.uchicago.edu





194



RAPHAEL GREENBERG



found at the Ebla palace (Mazzoni 1984), or, to cite a later example, the



lmlk



jars of Iron Age
Judaea (Naªaman 1986).

Metallic ware pithoi, however, had no role in interregional or international trade in perish-
ables. As far as we know, these cumbersome handleless containers were intended mainly as
household installations: ubiquitous, yet rarely found in groups. Their function within the
household is illustrated by part of an EB II kitchen excavated at Tel Dan (fig. 11.3). The sym-
bolic import of seals applied to pithoi has to be understood in the context of the specific sites
where each pithos was located (cf. the seal-impressed cooking pots found at Ebla—Mazzoni
1993: 407). In such contexts, the sealings either conveyed a message regarding the content,
setting it apart from that of other containers, or served to identify the manufacturer of the
vessel itself. The difference may be visualized in modern terms as that between the label on a
mineral water dispenser and the label on a refrigerator.

The onset of EB III marked the end of metallic ware’s virtual monopoly of the ceramic
repertoire in northern Canaan. Sites characterized earlier by a uniform ceramic assemblage
began to show greater diversity. At Hazor, for example, several independent and nearly con-
temporaneous ceramic traditions are attested: (a) common ware bowls, platters, jugs and jars,



FIGURE 11.3. Plan and contents of presumed kitchen at Tel Dan, Phase B7 (EB II); No. 5 is the lower portion
of a pithos.

spread one pica long

oi.uchicago.edu

[image: image62.tif]



195



EARLY BRONZE AGE II–III PALESTINIAN CYLINDER SEAL IMPRESSIONS



produced in techniques similar to those of the earlier period, but in a coarse, soft-fired fabric
often covered with a peeling red slip (Yadin et al. 1961: pl. 154:11–16); (b) Khirbet Kerak
ware, clearly introduced by potters of Syro-Anatolian descent, used for small bowls, kraters or
large jugs, and stands (Yadin et al. 1961: pl. 155:1–3; Greenberg 1997: figs. III.2:9–12;
III.4:12–22); (c) metallic ware, limited to jars and pithoi (Yadin et al. 1961: pl. 154:17; Green-
berg 1997: figs. II.3:15; II.4); (d) wheelmade ware: bowls, jugs, and bottles reminiscent of
Syrian EB IV forms; (e) imports, apparently from the Lebanese coast (the latter two groups
are described in Greenberg 1997: 21–24).

This diversity appears to be linked to a more sophisticated, cosmopolitan urban phase,
marked by more pronounced social ranking. The metallic ware workshops, forced to compete
for their share in the market, abandoned the broad range of ceramic production and concen-
trated on the single item in which their advantage was preeminent—the meter-high pithos.
From the manufacturer’s point of view, the decoration of these items with cylinder seals was
intended either to enhance their value or to set apart the various metallic ware workshops.
From the consumers’ point of view, the marked pithoi may have been acquired in increasing
quantities to answer the needs of an incipient bureaucracy, as the storage and redistributive
functions of the urban centers gained in importance (cf. the pithos storerooms recently exca-
vated in the Tel Yarmuth palace: Miroschedji 1994: 148–51).

However, the conditions that led at first to the increased use of metallic ware pithoi might
have been the ultimate cause of the demise of cylinder seal production. The symbolic “lan-
guage” of the cylinder seals, universally appreciated in EB II, would not have had the same
broad appeal in the socially and ethnically fragmented EB III. As the production and distribu-
tion of the broad range of metallic ware declined and, with it, the prestige of the metallic ware
ateliers, the cost of supporting both specialist potters and seal engravers might have become
prohibitive. Local workshops began to produce pithoi of a lesser quality on their own, and the
metallic ware market slowly contracted to the immediate environs of the original workshops—
hence the discovery of restorable undecorated metallic ware pithoi in upper Jordan Valley
sites, and of common ware pithoi at sites farther to the south (Megiddo, in which no Type I or
III seal impressions have been found, being a case in point). This ended the remarkable con-
junction of the two highly specialized crafts of seal carving and pithos production.

CONCLUSION

To sum up, the use of geometric and cultic cylinder seal impressions on Canaanite metallic
ware jars was initiated in EB II by the metallic ware ateliers in the region of Mt. Hermon, per-
haps as a means of controlling the movement of specific goods (on small jars), or as a symbol
referring to content or locus of manufacture (on jars and pithoi). Their broad and apparently
random distribution, closely matching that of the metallic ware repertoire in general, is illus-
trative of the broad appeal and universal appreciation of their symbolism and could be con-
strued as a metaphor for a high degree of social and ideological solidarity in northern Canaan
at this time. By EB III the principal use of the seals appears to have been as a label on widely
distributed metallic ware pithoi. As the demand for these vessels dropped toward the end of
EB III, the production of seals of the geometric and cultic type ceased.

By way of an epilogue it is noted that both Canaanite cylinder seal production and ap-
plication to jars, as well as the metallic ware technique, survived into EB IV, taking varying

oi.uchicago.edu





196



RAPHAEL GREENBERG



trajectories. Among the former, the Class II animal-file seals continued to be used, appearing
on EB III vessels at Numeira (Lapp 1989: fig. 7), at Byblos, and on a probable Byblite import
at Giza (Ben-Tor 1978: 69–79), as well as on an EB IV jar at Beit Yera



˙



(Bar-Adon 1973);
metallic ware metamorphosed into the ubiquitous “combed ware” found in southern Canaan in
late EB III and all along the Syro-Lebanese coast throughout EB III and IV (Mazzoni 1986:
152). Their brief joining of ways in EB II–III northern Canaan remains a fascinating episode,
whose full significance continues to elude us.

REFERENCES



Aruz, J.
1994 Seal Imagery and Sealing Practices in the Early Aegean World. Pp. 211–35 in



Archives Before
Writing, Proceedings of the International Colloquium, Oriolo Romano, October 23–25, 1991



,
eds. P. Ferioli, E. Fiandra, G. G. Fissore, and M. Frangipane. Turin: Scriptorium.

Bar-Adon, P.
1973 Rare Cylinder-Seal Impressions from Beth Yera



˙



.



Eretz-Israel



11 (Dunayevsky volume):
99–100.

Beck, P.
1976 The Cylinder Seal Impressions from Beth Ha ºEmeq.



Tel Aviv



3: 120–26.

Ben-Tor, A.
1975 The First Season of Excavations at Tell-Yarmuth: August 1970.



Qedem



1: 55–87. Jerusalem:
Institute of Archaeology, Hebrew University.

1978



Cylinder Seals of Third Millennium Palestine



. Bulletin of the American Schools of Oriental
Research Supplement 22. Cambridge, MA: American Schools of Oriental Research.

1992 New Light on Cylinder Seal Impressions Showing Cult Scenes from Early Bronze Age Pales-
tine.



Israel Exploration Journal



42: 153–64.
1994 Early Bronze Age Cylinder Seal Impressions and a Stamp Seal from Tel Qashish.



Bulletin of
the American Schools of Oriental Research



295: 15–29.

Braun, E.
1985



En Shadud. Salvage Excavations at a Farming Community in the Jezreel Valley, Israel.



British
Archaeological Reports, International Series 249. Oxford: British Archaeological Reports.

Esse, D.
1990 Early Bronze Age Cylinder Seal Impressions from Beth-Yera



˙



.



Eretz-Israel



21 (Ruth Amiran vol-
ume): 27*–34*.

Fritz, V.
1990



Kinneret. Ergebnisse der Ausgrabungen auf dem Tell el ºOreme am See Gennesaret, 1982–1985.



Wiesbaden: Harrassowitz.

Givon, S.
1993



The Excavation at Bet Ha-ºEmeq, 1973.



Tel Aviv: Institute of Archaeology and Regional Coun-
cil Mateh Asher.

Golani, A.
1996 Qiryat Ata.



Excavations and Surveys in Israel



15: 31–33.

Greenberg, R.
1996 The Early Bronze Age Levels. Pp. 83–160 in



Dan



I



: A Chronicle of the Excavations, the Pottery
Neolithic, the Early Bronze Age, and the Middle Bronze Age Tombs,



eds. A. Biran, D. Ilan,
and R. Greenberg. Jerusalem: Hebrew Union College.

1997 Area A: The Early Bronze Age; Area L: The Early Bronze Age. Pp. 17–24 and 183–93 in



Hazor
V,



eds. A. Ben-Tor and R. Bonfil. Jerusalem: Israel Exploration Society.
Forthcoming The Cylinder Seal Impressions. In



Excavations at Qiryat Ata,



ed. A. Golani. Israel Antiquities
Authority Reports.

oi.uchicago.edu





197



EARLY BRONZE AGE II–III PALESTINIAN CYLINDER SEAL IMPRESSIONS



Greenberg, R., and Porat, N.
1996 A Third Millennium Levantine Pottery Production Center: Typology, Petrography, and Prove-

nance of the metallic ware of Northern Israel and Adjacent Regions.



Bulletin of the American
Schools of Oriental Research



301: 5–24.

Lapp, N. L.
1989 Cylinder Seals and Impressions of the Third Millennium



b.c.



from the Dead Sea Plain.



Bulletin
of the American Schools of Oriental Research



273: 1–15.

Mazzoni, S.
1984 Seal Impressions on Jars from Ebla in EB IVA–B.



Akkadica



37: 18–40.
1986 The Diffusion of the Palestinian Combed Jars. Pp. 145–58 in



Studies in the History and
Archaeology of Palestine,



Proceedings of the First International Symposium on Palestine
Antiquities, vol. 2. Aleppo: Aleppo University.

1993 Cylinder Seal Impressions on Jars at Ebla: New Evidence. Pp. 399–414 in



Aspects of Art and
Iconography: Anatolia and its Neighbors. Studies in Honor of Nimet Özgüç



, eds. M. J. Mellink,
E. Porada, and T. Özgüç. Ankara: Türk Tarih Kurumu.

Miroschedji, P. de
1994 Tel Yarmut. 1993 (Notes and News).



Israel Exploration Journal



44: 145–51.

Mittmann, S.
1994 Hirbet ez-Zeraqon, Eine Stadt der fruhen Bronzezeit in Nordjordanien.



Archäologie in Deutsch-
land



2 (1994): 10–15.

Naªaman, N.
1986 Hezekiah’s Fortified Cities and the LMLK Stamps.



Bulletin of the American Schools of Oriental
Research



261: 5–21.

Pittman, H.
1994 Towards an Understanding of the Role of Glyptic Imagery in the Administrative Systems of

Proto-Literate Greater Mesopotamia. Pp. 177–203 in



Archives Before Writing, Proceedings of
the Intemational Colloquium, Oriolo Romano, October 23–25, 1991,



eds. P. Ferioli, E. Fiandra,
G. G. Fissore, and M. Frangipane. Turin: Scriptorium.

Webb, J. M., and Frankel, D.
1994 Making an Impression: Storage and Surplus Finance in Late Bronze Age Cyprus.



Journal of
Mediterranean Archaeology



7: 5–26.

Yadin, Y.; Aharoni, Y.; Amiran, R.; Dothan, M.; Dothan, T.; Dunayevsky, I.; and Perrot, J.
1961



Hazor III–IV (Plates).



Jerusalem: Magnes.

oi.uchicago.edu





198



RAPHAEL GREENBERG

oi.uchicago.edu



199



TOMBS, CULT, AND CHRONOLOGY:
A REEXAMINATION OF THE MIDDLE BRONZE

AGE STRATA OF MEGIDDO



Rachel S. Hallote



INTRODUCTION

While the ceramics of the Middle Bronze Age (MB) tombs of Megiddo have often been
used to date the architectural strata of the site, the tombs themselves, and their placement
under floors of houses, have seldom been scrutinized. A careful examination of the tombs and
the architectural strata with which they were associated will lead to several conclusions about
the dating of the Megiddo strata and about cult in the MB.

Megiddo has always stood out in the archaeology of the southern Levant as a site of great
importance. Because of its location, historic significance, impressive cultural remains and
complex history of excavation, modern scholarship has had cause to carefully examine the site.
Megiddo has been excavated several times, from the German work of the early twentieth cen-
tury (Schumacher 1908; Watzinger 1929), to the current joint Tel Aviv and Pennsylvania State
University excavations, with the University of Chicago’s formative work in between (Fisher
1929; Guy 1931; Engberg and Shipton 1934; Lamon 1935; May and Engberg 1935; Guy and
Engberg 1938; Lamon and Shipton 1939; Loud 1939, 1948), as well as Yadin’s smaller-scale
explorations (Yadin 1970). The result of so many different expeditions, each with its own
methodology, has produced complicated reconstructions of the ceramic and architectural
sequences.

Kenyon’s attempts to clarify the ceramic sequence of the Bronze Age strata (1958, 1969)
succeeded in fine-tuning some aspects of these periods that were left cloudy by the main
Chicago publication (Loud 1948), but ultimately her work led to many more questions and
conflicts. The approach suggested here may help in sorting out both the chronology and the
content of the strata. Instead of tackling the issues through traditional ceramic and architec-
tural studies once again, the approach to be taken here is more oblique and only involves one
period, MB, ca. 2000–1550



b.c.



This approach reexamines the contexts of the tombs excavated in the MB levels of the
Chicago excavations. The presence of the tombs, coupled with the lack of any temple struc-
ture in these levels, suggest that private religious rituals, including those relating to death and
burial, temporarily superseded public temple worship. We begin by stating some questions
that have been raised, but only partially solved, by previous scholarship on the site. For
instance, what was the chronological extent of the temple sequences of Area BB? Was the
transition from the MB to the Late Bronze Age (LB) abrupt or smooth at Megiddo? Why did
Loud treat the tombs found on the tell as if they were simple stratigraphic remains, and
why, in refining the pottery sequence from the site, did Kenyon specifically utilize the tomb



12



oi.uchicago.edu





200



RACHEL S. HALLOTE



materials, even while acknowledging that their dates relative to the strata needed rethinking?
The answers to these related questions suggest a connection between the gap in the temple
sequence and the abundance of tombs in domestic areas.

THE LACK OF A MIDDLE BRONZE AGE TEMPLE IN AREA BB AND THE SMOOTH

TRANSITION FROM THE MIDDLE BRONZE AGE TO THE LATE BRONZE AGE

The first question involves the temples of Area BB, Temple 4040 of the Early Bronze Age
(EB) period, and Temple 2048 of the LB. Loud was the first to assume that a now lost MB
temple once existed, based on the evidence of the earlier and later temples, and also based on
the presence of a group of standing irregular monoliths in Stratum XII (Loud 1948: 92). The
possibility of a missing MB religious structure was taken up in various forms by Kenyon
(1969) and Thompson (1970). Yet when the preconception that an MB temple should exist is
laid aside, one can recognize that a public cultic structure was not as necessary as it had been
in the preceding periods and would be again in the following periods. In fact, the religious
practices of MB demonstrate a cultural discontinuity that is also visible in many aspects of the
material culture of the southern Levant ca. 2000



b.c.



(see Gerstenblith 1980, 1983). At
Megiddo as well as at other sites, public religious architecture was temporarily replaced due
to a privatization of religion which was physically represented by the placement of tombs
within domestic structures.

Although EB Temple 4040 and associated Altar 4017 appear in the Chicago plan of Stra-
tum XIV (Loud 1948: fig. 395), they had already gone out of use. The date of this complex has
been argued over repeatedly. Kenyon suggested that it belonged to the EB IV/MB I period, her
architectural phase G, and that it continued in use through Phase K and ceramic group B of the
MB IIB period and was abolished by Phase L (see table 12.1). Others have dated it to the
MB IIA period. However, in a discussion of some small-scale excavations at Megiddo,
Dunayevski and Kempinski (1973) demonstrated that this temple fits firmly into the Early
Bronze Age, as did Esse (1991: 87–89). Contrary to Kenyon’s suggestions, the structure was
probably no longer standing in the Middle Bronze Age.

Because it was such a substantial complex, some of its ruins were almost certainly still
visible during the next period of settlement, either as an abandoned mound or possibly as an
organic place of worship.



1



The plan of Stratum XIV (MB IIA) depicts the interior of the temple
as partially filled in (Loud 1948: pl. 395). Rather than implying a separate phase of use, this
stage could represent the preparations for rebuilding the area, which was becoming predomi-
nantly domestic. By Stratum XIIIA even the southern part of square N13, where the temple had
been located, contained walls of domestic units. Thus, while 4040 was probably recognized as
a former temple in MB IIA, it was in a ruined state and almost certainly was not used as an
official place of worship.

Some past studies of the next temple, LB Temple 2048, have attempted to push its date
into the MB (see below), so that there would not be a chronological gap between the last phase
of 4040 and the first of 2048. This would mean that there was no break in Area BB as a temple
area. Similarly, the presence of the irregular monoliths of Stratum XII has been used as an



1. Compare Jewish worship at the western wall of an ancient temple mount today.

oi.uchicago.edu





201



TOMBS, CULT, AND CHRONOLOGY



argument for a lost temple structure. However, a reexamination of the evidence does not sup-
port these interpretations.

Kenyon was the first to propose the now accepted idea that the “ghost walls” of 2048,
which appear in Stratum IX (Loud 1948: fig. 401) were only foundations dug into this earlier
stratum, which suggested that 2048 was not actually constructed until the LB period. She fur-
ther suggested that there was a gap in the occupation of Megiddo due to the campaigns of
Thutmosis III in the early LB II, and that 2048 was built only after occupation had resumed
(Kenyon 1969: 49–53). Her classification of the Stratum IX walls as foundations dug down
from a later level is accurate, but the notion of a gap must be questioned since the MB–LB
transition is a rather smooth one, especially at Megiddo.



2



Epstein’s study of Area BB is helpful for understanding this transition (1965). Although
she argued for an early dating of 2048, beginning in MB Stratum XII and continuing down
through LB Stratum VIIB, her analysis still contains some important clues to the MB–LB
transition. A problem in Epstein’s discussion is that she relied heavily on the unpublished field



2. See Bunimovitz (1995: 322) on the continuity from MB to LB.



Table 12.1.



Description of Kenyon’s Megiddo Stratigraphy (Adapted from Kenyon 1969)



Date Stratum Cultic features Architectural phase Ceramic group



Post-ca. 1480
(Thutmosis III)

VIIA Temple 2048 S —

Gap Gap Gap Gap Gap

Late 16th century VIIB — R —

(Early LB) VIII — Q —

IX — — —

X — — H (including bichrome)

— — G

MB IIB X — P F

XI Monoliths O E

XIIA Monoliths N D

XII Monoliths M C

XIIIA — L B

MB IIA–MB IIB XIIIA Temple 4040 K A

MB IIA Later than XIIIB Temple 4040 J —

EB IV/MB I XIIIB Temple 4040 — —

XIV Temple 4040 H —

XV Temple 4040 G —

oi.uchicago.edu





202



RACHEL S. HALLOTE



diaries of the excavators, which she took at face value, even while showing that the published
material should not be wholly trusted. In fact, the field diaries are somewhat anecdotal and
asystematic for these periods and should not be used as reliably as they have been for later
periods (cf. Esse 1992).

References in these diaries to what seemed to be dumps led Epstein to postulate an earlier
temple, or an early phase of Temple 2048 in MB (Epstein 1965: 204–5, 208–9). However, she
did not take into account the mixture of levels that took place due to the excavation techniques
of the Chicago team. It is well known that the excavators were attempting to peel layers off a
large area that was not built up in a manner suitable for such a strategy. The scale and speed
of the project prevented a more detailed approach and led to some mixture of materials from
different strata (see below). Furthermore, the ceramics of the loci that Epstein calls into ques-
tion as possibly MB could instead comfortably belong in the MB IIC/LB I and LB I periods
(Epstein 1965: 210–13; Loud 1948: pl. 235:20). Because she still followed the dating of the
strata published by Loud, she did not recognize that Stratum X represents an MB–LB transi-
tional phase, or that Stratum IX is solidly LB (table 12.2).

Within her discussion, Epstein points to the core of the issue for these strata. She acknowl-
edges the continuity from Stratum XII all the way down to Stratum VIIA in Area BB but
assumes that the change in plan that is first recognizable in Stratum XII is due to the establish-
ment of Temple 2048 (Epstein 1965: 213). However, this new plan is actually due to the
change from MB IIA (Stratum XIII) to MB IIB (Stratum XII), as seen in table 12.2. Following
this architectural shift, there is a great degree of continuity visible in the plans of Strata XII,
XI, and X, again suggesting that the transition from MB to LB is a smooth one. Similar obser-
vations can be made for these strata in Area AA.



3



This MB–LB continuity has been discussed before for Megiddo as well as for other sites.
When differences between MB and LB are suggested, they often have historical bases, not



3. The new excavations at Megiddo conducted by Tel Aviv University and Pennsylvania State University will
most likely confirm the hypothesis of a smooth transition from MB to LB at Megiddo, particularly through
the evidence uncovered so far from the northern lower terrace, Area F. This area, which had not been previ-
ously excavated, contains an apparently unbroken sequence of MB and LB remains.



Table 12.2.



Chronologies of Megiddo



Stratum Features Loud Kenyon Current proposal



VII Temple 2048 LB I LB II LB IIA

IX — MB IIB MB IIB LB IB

X — MB IIB MB IIB MB IIC/LB IA

XI — MB IIB MB IIB MB IIB/MB IIC

XII Change in plan; standing monoliths MB IIB MB IIB MB IIB

XIII — MB IIA MB IIB MB IIA

XIV Absolute end of Temple 4040 MB IIA MB IIA MB IIA

oi.uchicago.edu





203



TOMBS, CULT, AND CHRONOLOGY



archaeological ones.



4



In fact, the only significant new feature found in Megiddo’s Area BB dur-
ing the MB and LB strata is Temple 2048, in Stratum VIII. At that point the western house com-
plex had mainly gone out of use, although the structures to the east continued with little change.
Thus MB and LB can best be distinguished from each other not by new settlement patterns, new
houses, or new technologies, but by the reintroduction of public religious architecture.

THE MB TOMBS OF MEGIDDO AND THEIR ARCHITECTURAL ASSOCIATIONS

Since no large temple structure existed at Megiddo in the MB, it is logical to investigate
whether or not religious rituals of the period can be discerned archaeologically in any other
way. In fact, rituals were carried out in private, domestic contexts, not public, formal ones.
The main physical manifestation of MB religion at Megiddo and elsewhere is funerary
remains. A discussion of the tombs and their associated strata will underscore the importance
of mortuary ritual in MB, which in turn will clarify the private nature of religious ritual.

There have been many discussions of Megiddo’s MB strata and associated tombs. Kenyon
(1958; 1969) formed a new interpretation of the published data by reconstructing the contents
of all the tombs of the EB, MB, and LB periods and dividing their ceramics into her lettered
groups (see table 12.1). She chose this particular approach to the site because she believed that
there was more clarity in the ceramic assemblages of the tombs than in the ceramic assem-
blages of the settlement levels (Kenyon 1969: 25). This argument has been both criticized and
defended in several debates which only further complicated the understanding of the site (see,
for instance, Thompson 1970; Dunayevski and Kempinski 1973; Williams 1975: 906–51;
Cole 1984; Bienkowski 1989; Kempinski 1989). One of the main contributions of Kenyon’s
work was the division of the MB tombs into the eight ceramic groups, the first of which rep-
resents the MB IIA period, and the following seven, the MB IIB period. Kenyon’s results have
been indispensable to ceramic studies; however, her overseparation of the material has not
assisted in understanding the MB of Megiddo as a whole.



5



Kenyon tended to view the ceram-
ics in isolation and only looked at their architectural contexts after completing her typology.
This omission of contexts from the most important sections of her two studies led to errors in
the reconstructions of architectural phases, which she based on the dates of the tomb ceramics
alone (Kenyon 1969: 43–60). We now attempt to reunite the MB tombs with their contexts.

Megiddo is one of many examples of a southern Levantine site where burials are com-
monly found under the floors or courtyards of domestic structures. The tombs of the tell (as
opposed to the shaft-and-chamber tombs on its slopes) include masonry constructed tombs,



4. See for instance Weinstein 1981. Weinstein examines sites where there are no destructions dating to the
reign of Ahmose, only destructions dating to Thutmosis III. These sites, as well as some others that he did
not list, not only do not have a gap in occupation between MB and LB, but also demonstrate the direct con-
tinuity described for Megiddo. Additionally, he lists a separate group of sites with destructions and gaps at
the time of Ahmose—the MB–LB transition. Thus the destructions and changes of culture that took place
did not occur all at once, but rather spanned a long period of time. For further discussion of the end of the
MB in archaeological and historical terms, see Bienkowski 1989; Dever 1990; Hoffmeier 1989, 1990; Bun-
imovitz 1995.

5. Kenyon divided the MB pottery from the Jericho tombs into similar groups (see Kenyon 1960, 1965).

oi.uchicago.edu





204



RACHEL S. HALLOTE



simple pits, stone-lined pits, and infant burials in store jars (see Hallote 1994: 191, 240–54;
Ilan 1995: 122). There are a total of 150 burials published in Loud 1948







that date to MB, span-
ning Strata XIV through X. Of these, fifty-four date to MB IIA, eighty-one to MB IIB, and
only fifteen date to MB IIC. In many cases, the information on both burial type and number of
individuals interred in each tomb was not properly recorded and can only be reconstructed
through the photographic and negative records, published and unpublished, which themselves
are not complete. Often no information is available at all. Therefore, burials of unknown type
account for 56%, 48%, and 33% of the burials of MB IIA, MB IIB, and MB IIC respectively.
It is likely that most of these were simple pits. Figure 12.1 compares the frequency of tombs
of each type by period. Figure 12.2 suggests that the majority of the sixty burials for which
this information was available were single inhumations, while fig. 12.3 represents the break-
down of burial by age for the same sixty burials.

A key characteristic of the MB burials at Megiddo and contemporary sites is their place-
ment under the floors of houses. Although this crucial fact was even recognized by Kenyon
(1958: 59), she downplayed it in her chronological reconstruction. The fact that these tombs
were under floors was never given sufficient attention before now, partly because of the lack
of precision in the excavation mentioned above (Loud 1948: 1). The series of floor levels usu-
ally found in domestic areas in use for long periods was built up very slowly, with minimal
architectural changes of significance. Houses probably stayed in use for generations, with
walls removed, replaced, or added gradually during the centuries of occupation. Therefore,
when the Chicago excavators of Megiddo attempted to “peel” the strata off the various areas
of the mound one at a time (Fisher 1929: xi), they removed what they considered to be the
entirety of the latest MB IIA phases. In doing so, they inadvertently removed walls that were
in use for many consecutive phases. After the removal of most of the walls, all that remained
of the lower, earliest MB IIA phases were scraps, such as those that appear in the plans of



FIGURE 12.1. Frequency of tomb types by period.

oi.uchicago.edu

[image: image63.tif]



205



TOMBS, CULT, AND CHRONOLOGY



Strata XIV and XIII B (Loud: pls. 395, 396). This often makes associating individual tombs
with specific structures and floors impossible.

However, it is clear from sites throughout the Levant that tombs were commonly dug into
floors of the domestic structures that were concurrently in use (see Hallote 1994: 51, 61; see
also Ilan 1995: 124; 1996). Like the houses, these tombs were probably reused for several
generations. It is likely that the floor levels of the rooms would rise, but the openings to the
tomb shafts would remain at least partially visible (see fig. 12.4). Therefore, in archaeological
terms, these tombs are contemporary with the floors above them, the very floors that appear to
be cut by their shafts.



FIGURE 12.2. Proportion of single vs. multiple inhumations.

FIGURE 12.3. Age of inhumations.

oi.uchicago.edu

[image: image64.tif][image: image65.tif]



206



RACHEL S. HALLOTE



Kenyon understood that the Megiddo excavators should not have associated tombs with
the floors on which they sometimes rested. But while she acknowledged that tombs were
placed beneath houses, she did not recognize the probability that they were sealed by floors of
their own period. She therefore used burials only as termini post quem, and because of this
concluded that the architecture of Stratum XIV was pre–MB IIA (EB IV/MB I) and that Stra-
tum XIIIA included early MB IIB material (Kenyon 1958: 60; 1965: 49).

She even suggested that in MB IIA, Area BB consisted of a cemetery surrounding an MB
IIA version of Temple 4040 (Kenyon 1969: 43). She based this on a locational analysis of the
MB IIA tombs relative to each other and to the temple mound. This analysis ignored the rela-
tionships between tomb locations and the architecture of the strata to which the tombs had
been assigned since her assumption was that the assignations were incorrect. Kenyon was mis-
led by the absence of the walls that had been peeled off the earlier MB IIA levels, which
caused the loss of their associations with several different phases.

The best surviving features of these first phases of MB IIA are the burials themselves,
which were placed under the floors of the houses and therefore remained untouched and unaf-
fected by renovations. Since they were beneath the floors, they also were not displaced by the



FIGURE 12.4. Plan of Megiddo Tomb 3070. (Adapted from Loud 1948: fig. 400.)

oi.uchicago.edu

[image: image66.tif]



207



TOMBS, CULT, AND CHRONOLOGY



excavators’ removal of the associated walls and upper layers of floor plaster. The number of
undisturbed burials, combined with the amount of disturbed architecture, is what led Kenyon
to postulate a cemetery in the area in Stratum XIV. A closer look at the situation reveals that
the MB IIA tombs of Stratum XIV are all near walls or wall fragments, that is, under floors of
rooms, near the walls. In Strata XIIIB and XIIIA, where the architecture is less scrappy, many
of the tombs are clearly within rooms. Stratum XIV includes only MB IIA burials, while Stra-
tum XIII includes a few burials of the early parts of the MB IIB as well (see fig. 12.5; also see
Loud 1948: figs. 395–97). Since it is clear that the custom was to place tombs in houses, under
floors that were concurrently in use, it is likely that Strata XIV and XIII were MB IIA strata
and that they consisted largely of domestic areas in Area BB. The MB IIB strata can be viewed
through the means of a similar analysis and include Stratum XII and part of Stratum XI.

THE CULT OF THE DEAD IN THE MB SOUTHERN LEVANT

In order to understand the connection between the two topics discussed so far— the lack
of an MB religious structure at Megiddo and the presence of tombs in domestic contexts—it
is necessary to view the tombs of Megiddo as a subset of the larger topic of funerary practices
of the MB southern Levant, which in turn is part of the even larger issue of MB religious activi-
ties. Instead of concentrating on the numerical and statistical aspects of tombs, they can be
viewed more productively as evidence of culturally specific symbols, parts of rituals relating
to death (Metcalf and Huntington 1991: 62–75). It has been pointed out that the burial repre-
sents only the middle section of a funeral since rituals relating to death and mourning may well
have preceded and followed the interment (cf. Bartel 1982). Yet the burial is the only feature
of these death rituals that is visible archaeologically. Specific applications of mortuary archae-
ology and theory to the MB southern Levant have been dealt with elsewhere (Hallote 1994,
1995). What is important here is the recognition that ceremonies and rituals relating to death,



FIGURE 12.5. Numbers of MB IIA, MB IIB, and MB IIC tombs in each Megiddo stratum.

oi.uchicago.edu

[image: image67.tif]



208



RACHEL S. HALLOTE



including the burial itself, were significant cultic activities, not only in the Middle Bronze
Age, but in the preceding and following periods as well.

By the Iron Age, death related rituals had been solidified into a “cult of the dead” that is
known from both biblical and Ugaritic sources. The manifestations of this cult have been
described most recently by Bloch-Smith (1992; see also Brichto 1973). Most of its features
can be traced to the Middle Bronze Age if not to earlier periods. These features include the
offering of sacrifices to the dead, the related practice of feeding the dead, burial in one’s home-
town or city, and burial on land associated with family or ancestors, often in family tombs
(Bloch-Smith 1992: 119–27, 148–50). Bloch-Smith has suggested (1992: 131) that family and
clan loyalties were strong enough to lessen the influence of the central government, and that
in response to these loyalties, the eighth and seventh century reformers of the Monarchy
attempted to abolish the cultic practices that emphasized clan, including the cult of the dead.

It is quite possible that this cult originated in the Middle Bronze Age. For example, by the
MB, the practice of feeding the dead was a regular feature of burials. Almost every burial
included food and drink, as well as non-edible offerings. Similar offerings have also been
found in the EB and even earlier tombs, but another, more striking feature of the MB tombs
in light of the Iron Age textual material is the probability that they contained family groups.
A typical MB burial, in either a cemetery or a domestic area, consisted of one to three individ-
uals interred together. Occasionally a tomb would be reopened and another burial added.
These burials do not represent nuclear families but rather several members of an extended kin
group or clan who were interred together. Tombs, especially shaft tombs in nondomestic con-
texts, occasionally would be reopened and burials would be added, presumably alongside
those of ancestors or family members (Hallote 1994: 69–74).

Because of these similarities to the Iron Age cult, the treatment of the dead in the Middle
Bronze Age can be understood as the beginning of the family-oriented worship that character-
ized that later cult. Although ancestor worship clearly existed in the southern Levant in differ-
ent forms in earlier periods (cf. the Neolithic [Banning 1995: 4; Watkins 1992] and EB IV/
MB I [Dever 1987]), the cult of the dead was able to flourish in a more significant manner in
the MB because forms of individual clan worship temporarily overwhelmed public temple
worship at this time.

We have already seen that no temple was present at Megiddo in MB Strata XIV through IX.
During this same period, many tombs were placed under houses, and it is likely that they often
visibly protruded above the floors into which they were set (see fig. 12.4). This day to day vis-
ibility suggests a daily consciousness of a specific religious ritual. The immediate presence of
deceased family members, constant reminders of both beliefs and ceremonies concerning after-
life and death, had an active part in daily household life which was important enough to dimin-
ish the need for a large central temple. The frequent choice of domestic contexts for burial
suggests that communal religion had been superseded by the cult of the dead and probably by
other private rituals as well (Hallote 1994: 224–39).

This underdevelopment of public temple worship in the MB is not unique to Megiddo. In
MB IIB southern Levant it appears that only local shrines existed, far away from the larger,
important cities, and that true temples were not reinstated until the MB IIC–LB IA horizon.
Worship in the MB seems to have been a family or household affair, and the cult of the dead
was one of its most visible physical manifestations. It is likely that this cult propelled the

oi.uchicago.edu





209



TOMBS, CULT, AND CHRONOLOGY



household worship of the period and helped encourage familial and clan, rather than public
and community, responsibilities.

When subjected to close scrutiny, most temples that have been classed as MB actually
date to the MB IIC period rather than MB IIB proper and continue into the LB with only mini-
mal changes. For instance, the temples at Hazor mainly date to MB IIC. The Double Temple
of Area F dates to MB IIC, as well as the earliest Orthostat Temple, which begins in the MB
IIC and continues with only slight changes into the LB I. The Long Temple of Area A contin-
ues with an unchanged plan from MB II into LB I, again demonstrating the continuity from
the end of MB to the beginning of LB (Yadin 1972: 75–76, 96–98, 102–03). It appears that
for the greater part of MB there were few if any temples at Hazor, which suggests that non-
monumental expressions of religion existed. Similarly, the High Place at Gezer dates to MB
IIC. The temples at Shechem, the Migdal temple, and the Gerazim sanctuary nearby are also
MB IIC structures (Dever 1971: 94–132; G. E. Wright 1965; Boling 1969: 82–103).

Shechem’s casemate courtyard temple complex dates to MB IIB. However, this so-called
temple is built on a domestic plan (G. E. Wright 1965: figs. 58, 64, and esp. 65). Although the
scale is somewhat larger than traditional domestic architecture, the plan is certainly not one of
a temple but better resembles an elite residence. In fact, it has previously been suggested that
part of the reason that the structure was recognized as a temple was because of the later temple
in the area (G. R. H. Wright 1968). This argument is augmented by features within the complex
that belong to domestic, rather than religious structures, such as ovens and tombs (G. E. Wright
1965: fig. 64). In short, the area was not purely a religious precinct but was at least partially
domestic, again suggesting that in the MB, domestic and private life briefly took precedence
over religious and public life.

Among the few temples from the southern Levant that do clearly fall into MB IIB are the
temple at Tel Kitan, and its counterpart across the Jordan River at Tell el-Hayyat (Eisenberg
1977: 77–81; Falconer and Magness-Gardiner 1984: 49–74; Falconer 1987: 251–59). These
small temples, which are stylistically similar to each other, may well represent a specific
regional phenomenon of the Jordan Valley, not the culture of the southern Levant as a whole.

The coastal sanctuary at Nahariyah is problematic. Ben-Dor dated the temple he excavated
there to MB IIB (Ben-Dor 1950). Dothan suggested that a nearby structure slightly to the south
predated the temple. Dothan dated the entire site to the “Hyksos” period, and its destruction to
MB IIC–LB I, following Egyptian chronology (Dothan 1956). However, some of the ceramics
Ben-Dor published are clearly MB IIA in character. It is therefore likely that the use of these
structures spanned the Middle Bronze Age. Yet this sanctuary was small and had specific func-
tions. Presumably it was dedicated to the worship of the goddess Asherah of the sea (Ben-Dor
1950: 1–41; Dothan 1956: 14–25). It was not a large public temple and does not indicate that
there was a large-scale public forum for worship in this region.

Tel Kitan, Tell el-Hayyat, and Nahariyah can be understood as anomalous small sanctu-
aries representing individual regional variations. Similarly, the much discussed MB temple at
Tell el-Dabºa is located in Egypt’s Nile Delta, not in the southern Levant, suggesting that,
like the other regional temples, it does not represent the norm (Bietak 1981: 250–53). Except
for these, all the known MB sanctuaries date only to the very end of the period and continue
into the Late Bronze Age.

oi.uchicago.edu





210



RACHEL S. HALLOTE



A few other structures identified as small sanctuaries have been excavated recently. These
include an MB IIA shrine at Kfar Shemaryahu (Gophna and Ayalon 1980: 149; Kaplan 1971:
fig. 11), as well as structures at Tell Haror, Givat Sharet, Na



˙



al Rephaim, and Kfar Rupin
(e.g., Oren 1993: 580;, Edelstein and Greenhut 1990: 120; Edelstein 1993: 1282). These have
yet to be published sufficiently, and the specific dates within MB of many of them are still
unclear. Additionally, these appear to be small, local shrines, often within domestic areas,
implying household use rather than public use. This is clearly the case at Na



˙



al Rephaim
(Edelstein and Greenhut 1990; Edelstein 1993). The presence of such small, private cultic
structures does not undermine the current argument, but in fact demonstrates a second
approach to family-oriented, rather than communal, worship.

The lack of emphasis on public worship described here seems not to be unique to the
southern Levant. A similar phenomenon has been detected by Dabney and Wright (1990) for
contemporary Middle Helladic (MH) Mycenaean society. Within the larger context of palatial
society and state formation, they discussed how burial practices can reinforce allegiance to
ancestors and can demonstrate a focus on individual power. In MH Mycenaean society, as in
the MB southern Levant, cult centers and cult symbolism appear to be underdeveloped and
seem not to be used as a means to control or unify the populace in relation to king and palace,
as they were to be in the Late Helladic.

Although the temporal convergence of this phenomenon in the Aegean and the southern
Levant is probably little more than coincidence, the same principle may well be at work. In the
southern Levant mortuary practices that emphasize lineage and ancestors become prominent
at precisely the moment when communal cult activities are somewhat curtailed. When the
family-oriented cult of the dead began in the MB, it either became popular to the point of tem-
porarily eclipsing temple worship, or alternatively, it grew in importance as a response to an
already dwindling interest in public worship. In either case, the cult continued through the
Late Bronze Age, when temple worship resumed, and into the Iron Age, when the clan ties
that sustained the cult were not only perceived as a threat to communal worship but also as a
threat to the central organization of the state.

SUMMARY AND CONCLUSIONS

Because Megiddo’s role was that of an important Bronze Age center, analysis of its MB
levels is essential for unraveling the social systems of the period. The architectural and ceramic
sequences at Megiddo have helped to establish the smooth nature of the transition from MB
to LB. As the material similarities between these periods become apparent, the fact that the site
did not have an MB temple seems rather incongruous. Since the EB and LB temples were
always located in the same area of the site throughout the Bronze Age, the gap in the temple
sequence during the MB is particularly noticeable and therefore warranted investigation.

A close reexamination of the tombs in the contexts of their strata has suggested that at the
time when no temple existed at Megiddo, tombs were commonly placed under floors of
houses still in use. Furthermore, the entry shafts of these tombs most probably projected part-
way up into the living areas. This constant presence of deceased individuals, almost certainly
members of extended kin groups, reinforces an aspect of the most prominent type of the MB

oi.uchicago.edu





211



TOMBS, CULT, AND CHRONOLOGY



worship—small-scale clan or family rituals, specifically the cult of the dead, taking the place
of temple worship.

Although other sites may not demonstrate this phenomenon in as sharp relief as Megiddo
does, it is clear that intramural burial begins on a large scale throughout the southern Levant
in the Middle Bronze Age. Furthermore, no site of central importance in the MB has yet been
found with a significant cultic structure that predates MB IIC. Although future excavations at
Megiddo as well as at other MB sites may lead to refinements of this hypothesis, it is likely
that such work will help confirm the relationship between private cultic rituals and the tempo-
rary deemphasis of public temples.

ACKNOWLEDGMENTS

I would like to thank Dr. Samuel Wolff for inviting me to contribute to this volume. Some
of the ideas outlined in this manuscript were first presented in November 1993 in Washington,
DC, at the annual meetings of American Schools of Oriental Research, and later appeared as
a section of my dissertation. Discussions with Baruch Halpern, Israel Finkelstein, David
Ussishkin, J. P. Dessel, David Ilan, and Alex Joffe have helped me greatly in clarifying some
of my arguments, although none of them are responsible for any errors or inaccuracies in this
work. John Larson of the Oriental Institute kindly gave me access to the unpublished photo-
graphs, negatives, field diaries, and plans of the Chicago excavations at Megiddo. I owe my
greatest debt to Douglas Esse, who was my mentor and advisor. It was under his guidance that
I learned how to approach archaeological issues, and I will always feel privileged to have
studied with him.

REFERENCES



Banning, E. B.
1995 Herders or Homesteaders? A Neolithic Farm in Wadi Ziqlab, Jordan.



Biblical Archaeologist



58:
2–13.

Bartel, B.
1982 A Historical Review of Ethnological and Archaeological Analyses of Mortuary Practice.



Journal
of Anthropological Archaeology



1: 32–58.

Ben-Dor, I.
1950 A Middle Bronze-Age Temple at Nahariya.



Quarterly of the Department of Antiquities in Pales-
tine



17: 1–41.

Bienkowski, P.
1989 The Division of Middle Bronze IIB–C in Palestine.



Levant



21: 169–79.

Bietak, M.
1981



Avaris and Piramesse



:



Mortimer Wheeler Archaeological Lecture, 1979



. Proceedings of the
British Academy 65. Oxford: Oxford University Press.

Bloch-Smith, E.
1992



Judahite Burial Practices and Beliefs about the Dead



. Journal for the Study of the Old Testament,
Supplement Series 123; Journal for the Study of the Old Testament/American Schools of Oriental
Research Monographs 7. Sheffield: Sheffield Academic Press.

Boling, R. G.
1969 Bronze Age Buildings at the Shechem High Place: American Schools of Oriental Research Exca-

vations at Tananir.



Biblical Archaeologist



32: 82–103.

oi.uchicago.edu





212



RACHEL S. HALLOTE



Brichto, H. C.
1973 Kin, Cult, Land and Afterlife—A Biblical Complex.



Hebrew Union College Annual



44: 1–54.

Bunimovitz, S.
1995 On the Edge of Empires—Late Bronze Age 1500–1200



b.c.e.



Pp. 320–31 in



The Archaeology
of Society in the Holy Land



, ed. T. E. Levy. New York: Facts-on-File.

Cole, D. P.
1984



Shechem I: The Middle Bronze IIB Pottery



. Winona Lake, IN: Eisenbrauns.

Dabney, M. K., and Wright, J. C.
1990 Mortuary Customs, Palatial Society and State Formation in the Aegean Area: A Compara-

tive Study. Pp. 45–53 in



Celebrations of Death and Divinity in the Bronze Age Argolid



, eds.
R. Hägg and G. C. Nordquist. Stockholm: Paul Åstroms.

Dever, W. G.
1971 Further Excavations at Gezer, 1967–1971.



Biblical Archaeologist



34: 94–132.
1987 Funerary Practices in EBIV (MBI) Palestine: A Study in Cultural Discontinuity. Pp. 9–19 in



Love and Death in the Ancient Near East



:



Essays in Honor of Marvin Pope



, eds. J. H. Marks and
R. M. Good. Guilford: Four Quarters.

1990 “Hyksos,” Egyptian Destructions, and the End of the Palestinian Middle Bronze Age.



Levant



22:
75–81.

Dothan, M.
1956 The Excavations at Nahariya, Preliminary Report (Seasons 1954/55).



Israel Exploration Journal



6: 14–25.

Dunayevski, I., and Kempinski, A.
1973 The Megiddo Temples.



Zeitschrift des Deutschen Palästina-Vereins



89: 161–87.

Edelstein, G.
1993 The Mana



˙



at Excavations. Pp. 1281–82 in



The New Encyclopedia of Archaeological Excava-
tions in the Holy Land



, vol. 4. Jerusalem: Israel Exploration Society & Carta and New York:
Simon & Schuster.

Edelstein, G., and Greenhut, Z.
1990 Emek Rephaim.



Excavations and Surveys in Israel



7–8: 117–21.

Eisenberg, E.
1977 The Temples at Tel Kitan.



Biblical Archaeologist



40: 77–81.

Engberg, R. M., and Shipton, G. M.
1934 Another Sumerian Seal Impression from Megiddo.



Palestine Exploration Fund Quarterly State-
ment



90–93.

Epstein, C.
1965 An Interpretation of the Megiddo Sacred Area During the Middle Bronze II.



Israel Exploration
Archaeologist



15: 204–221.

Esse, D. L.
1991



Subsistence, Trade, and Social Change in Early Bronze Age Palestine



. Studies in Ancient
Oriental Civilization 50. Chicago: Oriental Institute Press.

1992 The Collared Pithoi at Megiddo: Ceramic Distribution and Ethnicity.



Journal of Near Eastern
Studies



51: 81–103.

Falconer, S. E.
1987 Heartland of Villages: Reconstructing Urbanism in the Southern Levant. Ph.D. dissertation,

University of Arizona.

Falconer, S. E., and Magness-Gardiner, B.
1984 Preliminary Report of the First Season of the Tell el-Hayyat Project.



Bulletin of the American
Schools of Oriental Research



255: 49–74.

Fisher, C. S.
1929



The Excavation of Armageddon



. Oriental Institute Communication 4. Chicago: University of
Chicago Press.



spread one pica short

oi.uchicago.edu





213



TOMBS, CULT, AND CHRONOLOGY



Gerstenblith, P.
1980 A Reassessment of the Beginning of the Middle Bronze Age in Syria-Palestine.



Bulletin of the
American Schools of Oriental Research



237: 65–84.
1983



The Levant at the Beginning of the Middle Bronze Age



. American Schools of Oriental Research
Dissertation Series 5. Winona Lake, IN: Eisenbrauns.

Gophna, R., and Ayalon, E.
1980 Survey of the Central Coastal Plain, 1978–1979: Settlement Pattern of the Middle Bronze Age

IIA.



Tel Aviv



7: 147–51.
Guy, P. L. O.

1931



New Light from Armageddon: Second Provisional Report (1927–1929) on the Excavations at
Megiddo in Palestine



. Oriental Institute Communication 9. Chicago: University of Chicago
Press.

Guy, P. L. O., and Engberg, R. M.
1938



Megiddo Tombs



. Oriental Institute Publications 33. Chicago: University of Chicago Press.

Hallote, R. S.
1994 Mortuary Practices and their Implications for Social Organization in the Middle Bronze South-

ern Levant. Ph.D. dissertation, University of Chicago.
1995 Mortuary Archaeology and the Middle Bronze Age Southern Levant.



Journal of Mediterranean
Archaeology



8: 93–122.

Hoffmeier, J. K.
1989 Reconsidering Egypt’s Part in the Termination of the Middle Bronze Age in Palestine.



Levant



21: 181–93.
1990 Some Thoughts on William G. Dever’s “Hyksos,” Egyptian Destructions, and the End of the

Palestinian Middle Bronze Age.



Levant



22: 83–89.

Ilan, D.
1995 Mortuary Practices at Tel Dan in the Middle Bronze Age: A Reflection of Canaanite Society and

Ideology. Pp. 117–39 in



The Archaeology of Death in the Ancient Near East



, eds. S. Campbell
and A. Green. Oxford: Oxbow Books.

1996 The Middle Bronze Age Tombs. Pp. 161–267 in



Tel Dan,



Vol. I:



History of Excavations, the Pot-
tery Neolithic, the Early Bronze Age Levels and the Middle Bronze Age Tombs



, eds. A. Biran, D.
Ilan, and R. Greenberg. Annual of the Nelson Glueck School of Biblical Archaeology. Jerusa-
lem: Keter.

Kaplan, J.
1971 Mesopotamian Elements in the Middle Bronze II Culture of Palestine.



Journal of Near Eastern
Studies



30: 293–307.

Kempinski, A.
1989



Megiddo—A City State and Royal Centre in North Israel



. Munich: C. H. Beck.

Kenyon, K.
1958 Some Notes on the Early and Middle Bronze Age Strata of Megiddo.



Eretz-Israel



5 (B. Mazar
volume): 51*–60*.

1960



Excavations at Jericho, Vol. One: The Tombs Excavated in 1952–4



. London: British School of
Archaeology in Jerusalem.

1965



Excavations at Jericho, Vol. Two:







The Tombs Excavated in 1955–8



. London: British School of
Archaeology in Jerusalem.

1969 The Middle and Late Bronze Age Strata at Megiddo.



Levant



1: 25–60.

Lamon, R. S.
1935



The Megiddo Water System



. Oriental Institute Publications 32. Chicago: University of Chicago
Press.

Lamon, R. S., and Shipton, G. M.
1939



Megiddo I: Seasons of 1925–1934, Strata I–V



. Oriental Institute Publications 42. Chicago: Uni-
versity of Chicago Press.

oi.uchicago.edu





214



RACHEL S. HALLOTE



Loud, G.
1939



The Megiddo Ivories



. Oriental Institute Publications 52. Chicago: University of Chicago Press.
1948



Megiddo II: Seasons of 1935–1939



. Oriental Institute Publications 62. Chicago: University of
Chicago Press.

May, H. G., and Engberg, R. M.
1935



The Material Remains of the Megiddo Cult



. Oriental Institute Publications 26. Chicago: Univer-
sity of Chicago Press.

Metcalf, P., and Huntington, R.
1991



Celebrations of Death: The Anthropology of Mortuary Ritual



. New York: Cambridge University
Press.

Oren, E.
1993 Tel Haror. Pp. 580–84 in



The New Encyclopedia of Archaeological Excavations in the Holy
Land



, Vol 2. Jerusalem: Israel Exploration Society & Carta and New York: Simon & Schuster.

Schumacher, G.
1908



Tell el-Mutesellim I



. Leipzig: Haupt.

Thompson, T.
1970 The Dating of the Megiddo Temples in Strata XV–XIV.



Zeitschrift des Deutschen Palästina-
Vereins



86: 38–49.

Watkins, T.
1992 The Beginning of the Neolithic: Searching for Meaning in Material Culture Change.



Paléorient



18: 63–76.

Watzinger, C.
1929



Tell el-Mutesellim II



. Leipzig: Haupt.

Weinstein, J. M.
1981 The Egyptian Empire in Palestine: A Reassessment.



Bulletin of the American Schools of Orien-
tal Research



241: 1–28.

Williams, B.
1975 Archaeological and Historical Problems of the Second Intermediate Period. Ph.D. dissertation,

University of Chicago.

Wright, G. E.
1965



Shechem, The Biography of a Biblical City



. New York: McGraw-Hill.

Wright, G. R. H.
1968 Temples at Shechem.



Zeitschrift für die Alttestamentliche Wissenschaft



80: 1–35.

Yadin, Y.
1970 Megiddo of the Kings of Israel.



Biblical Archaeologist



33: 66–96.
1972



Hazor, the Schweich Lectures of the British Academy



. London: Oxford University Press.

oi.uchicago.edu



215



EARLY BRONZE SOCIAL ORGANIZATION AS
REFLECTED IN BURIAL PATTERNS FROM

THE SOUTHERN LEVANT



Timothy P. Harrison



Ethnographic studies have long drawn attention to the behavioral significance embedded in
the symbolism and rituals expressed in the burial of the dead (Bartel 1982; Metcalf and Hun-
tington 1991). With the calls for a sounder theoretical base in archaeology ringing out in the
1960s, it was not long before attempts to submit archaeological mortuary data to more rigorous
theoretical scrutiny utilizing ethnographic case studies began to appear (see particularly Saxe
1970; and the assembled papers in Brown [ed.] 1971). This early euphoria was soon tempered
by a cautious skepticism, particularly regarding the reliability of ethnographic analogy. In an
important study, P. Ucko (1969) demonstrated the diverse functional and ideological consid-
erations that can accompany a particular method of burial and all too easily obscure the archae-
ological record.

Despite this, and other raised concerns (see particularly Hodder 1980; 1982: 195–201;
Pader 1982; Parker Pearson 1982; and also Braun 1981; with response by Tainter 1981),







the
archaeological analysis of mortuary ritual has forged ahead,



1



generating numerous operational
hypotheses. Central to these has been the assumption that the



social persona



, the composite of
the social identities maintained by the deceased in life, will be recognized symbolically at
death. Furthermore, the relative rank of the social position held by the deceased will reflect the
composition and size of the social group recognizing status responsibilities to the deceased
(Binford 1971: 17). Stated simply, the effort expended to bury an individual will be propor-
tional to the ascribed social status of that individual.

The implications for archaeological analysis logically follow. Principally, we might expect
the ceremonial complexity exhibited in a given burial to reflect, at least indirectly, the level of
social complexity experienced by the responsible social group. In a ranked (or nonegalitarian)
society, certain archaeological correlates should be discernible. At least two dimensions of
social personae have been associated with the burial practices of ranked societies:



The first,



superordinate dimension



, must be a partial ordering that is based on







symbols,
energy expenditure, and other variables of mortuary ritual, and which is







not



simul-
taneously ordered on the basis of age and sex. That is, membership in the







class and some
variability within the class are based on the ascriptive qualities of an







individual’s
genealogy. . . . The second,



subordinate dimension



, will be a partial







order based on
symbols, energy expenditure and other variables, which generally







will be ordered on the



1. The literature is immense. By way of example, see Tainter 1975, 1978, 1982; Chapman, Kinnes, and Rands-
borg 1981; O’Shea 1984; Chapman 1987; Beck 1995; Carr 1995; and Campbell and Green 1995.



13



oi.uchicago.edu





216



TIMOTHY P. HARRISON



basis of age and sex. That is, beyond the “given” features of







age and sex, variability in
this dimension will reflect achievement through life







histories of individuals. The older an
individual, the greater the opportunity for







accomplishment, therefore, on the average, the
higher the rank. (Peebles and Kus 1977: 431)



The degree to which the superordinate dimension becomes less evident and the subordinate
dimension more prominent, therefore, should be a reliable index of the increasing social com-
plexity of the particular social group in question.

Mindful of the complexity of mortuary ritual and the symbolic nature of the evidence, this
paper examines burial patterns in the Early Bronze Age (EB) cemetery at Bâb edh-Dhrâº,
located on the southeastern plain of the Dead Sea in Jordan, for what they broadly reveal of
EB social structure. Well known for its long history as an active burial ground, Bâb edh-Dhrâº
affords the opportunity to trace the evolution of burial customs practiced in the cemetery over
the course of the EB. The presence of an adjacent settlement permits comparisons with con-
temporary nonmortuary archaeological data and improves the possibility of distinguishing
recognizable patterns of social organization within the EB culture that buried its dead at Bâb
edh-Dhrâº. For further comparison, the mortuary data from the site of Jericho is also examined
to determine whether the burial customs practiced at Bâb edh-Dhr⺠reflect broader cultural
patterns. The chronological framework used here follows that established in the third edition
of



Chronologies in Old World Archaeology



(Stager 1992: 40–41).

BÂB EDH-DHRº

Bâb edh-Dhr⺠first received archaeological attention when W. F. Albright stumbled upon
the site in 1924 during an exploratory trip to the Dead Sea region (Albright 1924, 1926; Kyle
and Albright 1924; Albright, Kelso, and Thorley 1944). Despite Albright’s early interest, sys-
tematic excavations did not begin until almost four decades later, after pottery plundered from
the site began appearing on the Jerusalem antiquities market. In an effort to forestall further
destruction, an expedition was put together by P. Lapp in 1965 (Lapp 1966a, 1966b, 1968a; see
also Schaub and Rast 1989). In the course of his excavations, Lapp distinguished three basic
tomb types representing the three main periods of the cemetery’s use. The first phase consisted
of multichambered subterranean shaft tombs, and corresponded to the start of the EB period.
Charnel houses, the second tomb type, represented the urban phase of the EB II–III. The final
phase, assigned to the EB IV (or Lapp’s “Intermediate Bronze Age”), consisted of shallow pit
burials termed “cairns.”

Eventually, leadership of the expedition was transferred to two of Lapp’s colleagues, W. E.
Rast and R. T. Schaub. Under their direction, a multidisciplinary approach was adopted, and
the expedition took on a more regional perspective (preliminary reports have appeared in Rast
and Schaub 1974, 1978, 1980; Rast and Schaub [eds.] 1981; Fröhlich and Ortner 1982; and
Schaub and Rast 1984). While excavations continued on a large scale in the cemetery, the
expedition now also focused attention on the nearby settlement site. Specialists were engaged
to study a wide range of issues concerning the ancient economy and environment of the
settlement and its surroundings.







Over the course of five field seasons, the project successfully
elucidated the settlement history of Bâb edh-Dhrâº, exposed a large portion of the cemetery,
and conducted a regional survey and limited soundings at a number of neighboring sites that

oi.uchicago.edu





217



EARLY BRONZE SOCIAL ORGANIZATION AS REFLECTED IN BURIAL PATTERNS FROM THE



have helped to establish the larger regional context. It is from the published results of this
fieldwork that the data used in the analysis below have been drawn.

PATTERNS OF BURIAL

Surveys by the current research team indicate that when the cemetery reached its greatest
extent, during EB IA, the cemetery stretched over an estimated twenty-eight hectares (280,000
sq. m). When certain wadis were excluded, a more accurate approximation suggested 244,000
sq. m, still a sprawling area (Rast and Schaub 1980: 40). In the course of excavations, the
tombs were separated into four principal cemetery subgroups (A, C, G, H) (fig. 13.1). To date,
at least fifty-one EB IA shaft tombs, including six from 1981, have been investigated, about
one-third of the total number of estimated tombs dating to this period (Fröhlich and Ortner
1982: 256–64). Six EB IB shaft tombs and two circular charnel houses have so far been pub-
lished, while at least thirteen EB II–III charnels have been cleared. The extent of the EB IV
burial population is not clear, although it appears the cemetery was only sparsely used, because
only four stone-lined shaft tombs have been reported.



Early Bronze IA (3500–3300



b.c.



)



Shaft tombs were the primary form of burial during the EB IA period. They follow a
remarkably uniform pattern. Lapp has furnished a thorough description of Tomb A76, typical
of this tomb type (fig. 13.2) (1968b; see now also Schaub and Rast 1989: 151–56). Access to
the burial chambers was achieved through a vertical shaft approximately 2.1 m in depth. The
entrances to the chambers themselves were sealed with a combination of stone and mortar.
There were two chambers in Tomb A76, although the number varied in other shaft tombs any-
where from one to five chambers. The chambers in Tomb A76 were circular, averaging 1.8 m
in diameter and 0.87 m in height.

The burial deposits in a chamber typically consisted of a pile of disarticulated bones
placed on a mat in the center of the chamber floor. The long bones were arranged neatly in
parallel fashion, and the crania were aligned in a row to the left of the postcranial material. In
most cases, the number of long bones corresponded to the number of crania, while the catalog
of smaller bones was incomplete. The grave goods were gathered around the sides of the
chambers and typically consisted of large numbers of intact ceramic vessels, often stacked,
and smaller objects such as figurines, jewelry, weapons, and an occasional perishable item
such as a reed basket, a pair of sandals, or a wooden staff (Lapp 1968b: 19–26).

It was clear that the material in each chamber was the result of a single interment, and that
once a chamber entrance had been sealed it was not reopened for a second or third burial.
However, all of the chambers within a shaft tomb were not necessarily filled at the same time.
In fact, a significant amount of time could elapse between the occupation of two chambers of
the same shaft tomb, so that the life of the tomb could easily stretch over a considerable span
of time. This being the case, it is likely that the shafts were marked, or at least left open until
all of the chambers had been filled. In Cemetery A, the density of the EB IA tombs has raised
questions about the degree of planning involved in its creation. The arrangement of the tombs
appears to follow a regular pattern, implying that the tomb cutters were familiar with the
overall layout of the cemetery. If so, this would suggest that a distinct population group used

oi.uchicago.edu





218



TIMOTHY P. HARRISON



the site as a burial ground, rather than a dispersed or unrelated assortment of groups (Schaub
1981: 46–47, 56).

In light of this, it has been hypothesized that the consistent pattern of secondary, disartic-
ulated burials was the result of periodic returns to the site by a nomadic EB IA population. If
an individual died while the group was away from the cemetery, the body was given primary
burial treatment elsewhere. When the group returned to the burial site, the skeletal material
that remained of the deceased was placed in one of the chambers of the “family” (or “clan”)
shaft tomb (Schaub 1981: 57). It would appear that the sparse evidence of the EB IA settle-
ment at the nearby town site supports this hypothesis (Rast and Schaub 1980: 40).

It should be noted, however, that a few EB IA shaft tombs produced articulated skeletons.
Tomb F2, for example, located just to the west of the walled EB II–III settlement, contained
an articulated female adult (Rast and Schaub 1978: 5–6). Furthermore, articulated burials
were discovered in the EB IA shaft tombs excavated in 1981. Four chambers from two sepa-
rate tombs (A110NE, A110SE, A111E, and A111W) contained the remains of both male and
female adult articulated skeletons (Fröhlich and Ortner 1982: 256–61). Although articulation
does not necessarily indicate primary burial, it does permit only a limited amount of time
between death and disposal of the body. Thus, these articulated remains could represent the
burial of individuals from either a nearby settled community, or from a nomadic group that
was in the area of the cemetery, perhaps disposing of other deceased, at the time of death.
Based on the data currently available, it would appear that the pattern of secondary, disarticu-



FIGURE 13.1. Plan of the town site and cemetery at Bâb edh-Dhr⺠(adapted from Rast and Schaub 1980: 33,
fig. 8; and from Schaub and Rast 1989: 23, fig. 2).

oi.uchicago.edu

[image: image68.tif]



219



EARLY BRONZE SOCIAL ORGANIZATION AS REFLECTED IN BURIAL PATTERNS FROM THE



lated burials, together with the scant settlement evidence, point to a relatively transient EB IA
population, one that nevertheless felt the need for a formal burial ground and the importance
of maintaining it.



Early Bronze IB (3300–3100



b.c.



)



The tombs excavated with EB IB pottery indicate that this period was transitional in
nature. In addition to the continued use of shaft tombs, the period marked the introduction
of the charnel house tomb type. While the EB IA witnessed a proliferation of shaft tombs,
only six EB IB chambers have so far been published, along with two circular charnel
structures.

Three of these chambers were attached to shafts with other chambers that contained EB IA
material. Tomb Chamber A100N (fig. 13.3), for example, which consisted of a flat stone floor
with one articulated adult skeleton and other partially articulated remains, contained EB IB
pottery, while the other three chambers that shared its shaft held disarticulated burials and
EB IA ceramic material (Schaub 1981: 58–59). There are two possible interpretations of the
occupational history of this tomb. The first is that the life of Tomb A100 simply spanned the
transition from EB IA to EB IB, with Chamber A100N representing the final interment. Since
the various chambers of a shaft tomb were filled individually, a significant amount of time
could have elapsed between burials, drawing out the life of the tomb. An alternative possibility
is that the original EB IA contents of Chamber A100N were removed at a certain point and



FIGURE 13.2. Plan of Bâb edh-Dhr⺠Tomb A76 (after Lapp 1968b: 16, fig. 4).

oi.uchicago.edu

[image: image69.tif]



220



TIMOTHY P. HARRISON



then later replaced by the EB IB material. In this context, it is noteworthy that the adjacent
Chamber A100E contained two distinct bone piles and two groups of crania (fig. 13.3).
Schaub has proposed that the bone pile and crania found on the right were in fact originally
buried in Chamber A100N and then moved to Chamber A100E by the EB IB users of Cham-
ber A100N (Schaub 1981: 60). Since then, excavations have produced an additional example
of a shaft tomb with chambers containing both EB IA and IB pottery (Tomb A111) (Fröhlich
and Ortner 1982: 261–62). Thus, it seems likely that these two tombs were in use during the
transition from EB IA to EB IB.



FIGURE 13.3. Plan of Bâb edh-Dhr⺠Tomb A100, shaft, and Chambers A100N and A100E (after Schaub 1981:
59, fig. 18).



long

oi.uchicago.edu

[image: image70.tif]



221



EARLY BRONZE SOCIAL ORGANIZATION AS REFLECTED IN BURIAL PATTERNS FROM THE



The other EB IB shaft tombs had only one chamber. One of these, Tomb G2, contained
mostly EB IA ceramic forms, but also had some EB IB types, indicating its transitional nature
(Schaub 1981: 57–58).

One of the most unique tombs discovered in the Bâb edh-Dhr⺠cemetery was the EB IB
circular charnel house Tomb G1 (fig. 13.4). Although much of the southern part had been
destroyed by modern military trenching activity, careful clearance revealed the intact northern
half of the tomb. The structure measured 3.7 m in diameter, with a 0.37 m thick encircling
wall. The northern section of the wall, which curved upward in beehive-like fashion, was pre-
served to a height of 1.3 m. Entrance to the charnel was achieved through a semicircular fore-
court made of mudbrick. Two orthostats flanked the entryway, and a large stone slab served as
a threshold. There were fragments of a possible stone lintel, and the blocking stone was found
in situ. The skeletal material, though widely scattered, indicated primary articulation. There
was also evidence of extensive burning, with the few surviving ceramic forms all dating to
EB IB (Schaub 1981: 63–65).

Tomb G1 closely resembles the round mudbrick Charnel A53 excavated in 1967 by Lapp
(1968a: 90–91; see now also Schaub and Rast 1989: 209, 222–32). Tomb A53 was actually
better preserved, with evidence of a second story or wooden beam roof. Both date to EB IB
and mark a departure from the subterranean shaft tombs of the EB IA period. Nonetheless,
several architectural features can be seen as carryovers from the earlier shaft tombs, including
the circular plan, and the forecourt, which functionally paralleled the purpose of the shaft. At
the same time, the freestanding structure anticipated the charnel houses of the following EB
II–III. All of this serves to highlight further the transitional nature of the period. Not surpris-
ingly, these developments coincided with the expansion of the nearby village site. The func-
tional features of these charnel structures, with their more immediate accessibility and larger
storage capacity, certainly would have lent themselves to a larger, more permanently settled
population (Schaub 1981: 65).



Early Bronze II–III (3100–2250



b.c.



)



The tradition of circular charnel houses continued into early EB II with the construction
of Tomb A56. Slightly smaller than its EB IB predecessors, Tomb A56 experienced two dep-
ositional phases and contained skeletal remains exhibiting some articulation (Schaub 1981:
65–66).

In addition to the circular charnel, the EB II period saw the development of a new tomb
type, the rectangular charnel house. It is not clear whether the two were in use contemporane-
ously, or whether the rectangular structures emerged later out of the circular charnel house tra-
dition. In any event, the large rectangular charnels became the standard tomb type and remained
as such until the end of the EB III. Although quite a few have been excavated, space permits
the description of only one, Tomb A22.

Tomb A22 was cleared in 1979 for the sake of enlarging the skeletal sample available for
anthropological study (fig. 13.5). The tomb measured 7.8 x 15.5 m, making it the largest tomb
discovered in the cemetery. The mudbrick outer walls were preserved to a height of fourteen
courses, or about 1.39 m. The floor of the structure was paved with small pebbles, and there
were traces of reed matting. The presence of wattle and daub, and charred wooden beams,
hinted at what had once been the roof. An interior wall divided the charnel into two sections
(Rast and Schaub 1980: 34–37).

oi.uchicago.edu





222



TIMOTHY P. HARRISON



The depositional remains indicated that a portion of the building had been subjected to
severe burning. Most of the burials were in disarray, mixed together with mudbrick detritis
and other debris from the fire. Despite this disturbance, however, the tomb still yielded evi-
dence of primary articulation. In addition, some skeletons showed no signs of exposure to fire,
indicating that they might have been placed in the charnel after it had burned. Three articu-
lated burials, all heavily charred, were found in association with what might have been a pallet
made of poles and reed matting. The remaining burials (all disarticulated), including 161 cra-
nia, had been heaped in mixed piles along the walls of the charnel. The interment process



FIGURE 13.4. Plan of Bâb edh-Dhr⺠Charnel House G1 (after Schaub 1981: 64, fig. 23).

oi.uchicago.edu

[image: image71.tif]



223



EARLY BRONZE SOCIAL ORGANIZATION AS REFLECTED IN BURIAL PATTERNS FROM THE



seemed to involve a pattern of initial primary articulated burial, followed later (after decom-
position) by removal to the sides of the charnel, leaving space for the next interment (Rast and
Schaub 1980: 37–38).

Analysis of the associated pottery indicated that the earliest burials were concentrated in
the east room, and the latest in the west chamber. Initially, both sections were probably used,
with the east wing receiving the most activity. In time, the dividing wall was built to close off
the east room, and burials were placed only in the west wing. Among the rather meager grave
goods were pieces of gold leaf jewelry with incised designs. Based on the ceramic evidence,
Tomb A22 was finally destroyed (or abandoned) late in EB III, or early in EB IV (Rast and
Schaub 1980: 38–39). The burial practices evident in Tomb A22, and in the other charnels that
have been cleared, point to a sedentary EB II–III population, not surprising given the thriving
fortified settlement that existed at the nearby town site.

Finally, a series of “tholoi,” averaging 5 m in diameter, were found by the Lapp expedition
extending in a line eastward from the town site. Surface sherds tentatively suggested an EB
II–III date for these freestanding structures (Lapp 1968c: 10; Schaub and Rast 1989: 489).



Early Bronze IV (2250–2000



b.c.



)



As stated earlier, very few EB IV tombs have been excavated. Lapp believed that a scat-
tered group of “cairns” on the far eastern edge of the cemetery dated to this period (1966a:
106; 1968c: 10). Subsequent investigations, however, have failed to substantiate Lapp’s pre-
liminary findings, and it now appears that these tumuli were simply the result of agricultural



FIGURE 13.5. Plan of Bâb edh-Dhr⺠Charnel House A22 (after Rast and Schaub 1980: 36, fig. 12).

oi.uchicago.edu

[image: image72.tif]



224



TIMOTHY P. HARRISON



field clearing efforts (McCreery 1977–78; Clark 1979; Rast and Schaub 1978: 24, 29; Schaub
and Rast 1989: 483, 489).

Four single-chambered, stone-lined shaft tombs located directly within the cemetery have
been positively dated to the EB IV. The first two, Tombs A52 and A54, were discovered during
the Lapp excavations and have been reported by Schaub (1973; see now also Schaub and Rast
1989: 473–87), who described Tomb A54 in detail (fig. 13.6). The vertical shaft was circular
and lined with stones and had been filled with stone rubble once the burial had been placed in
the chamber below. At the bottom of the shaft, a large stone blocked the entrance to the cham-
ber. The entryway was flanked by two upright slabs seated on a threshold and crowned with a
lintel. There was a slight step down to the floor of the chamber (Schaub 1973: 3–4). The
chamber itself was kidney-shaped and contained four articulated burials resting on a thick
layer of broken bones and assorted debris. The skeletal remains appeared to have been depos-
ited in successive layers, with subsequent burials disturbing the articulation of earlier depos-
its. It is possible that the stone lining of the shaft was installed in an attempt to strengthen it
for this apparent repeated use. Arranged around the sides of the tomb were twenty-six pottery
vessels dating to the early EB IV period (Schaub 1973: 5–6, 17–18).

In addition to Tombs A52 and A54, two chambers, designated RTTI and RTT2, were acci-
dentally discovered in 1979. They produced a large number of EB IV pottery vessels and
some skeletal material. The range of forms proved to be similar to those published from Tomb
A54 (Rast and Schaub 1980: 40–41).

Together, these four single-chambered shaft tombs constitute the extent of our knowledge
of the EB IV burial practices at Bâb edh-Dhrâº. In keeping with the conclusions already
drawn about the EB IA shaft tombs, these EB IV burials would suggest a relatively transient
population.

SOCIAL ORGANIZATION



Early Bronze IA



A key question concerning EB IA Bâb edh-Dhr⺠is whether or not the cemetery was
associated with a sedentary population. EB IA Palestine is generally believed to have experi-
enced widespread nomadic activity (Richard 1987: 25). The excavators of the Bâb edh-Dhrâº
cemetery accept this prevailing view and have concluded, citing the scant settlement data and
evidence of disarticulated burials, that the site was used primarily by semi-nomadic pastoralists
(Rast 1981: 7). However, a recent study by G. Bentley has now challenged this semi-nomadic
hypothesis. In addition to emphasizing the need for a more precise understanding of nomadic
pastoralism, she questions the assumption that Bâb edh-Dhr⺠lacked a permanent EB IA
settlement (1987: 11–21).

The primary reason for her skepticism centers around the apparent formal nature of the
EB IA burial grounds (Bentley 1987: 21–31). Ethnographic studies have highlighted the close
relationship that often exists between settled populations in control of restricted land resources
and their cemeteries (see, e.g., Douglass 1969; Bloch 1971; Saxe and Gall 1977; Glazier
1984), supporting the general principle that “the most elaborate formalized descent and alli-
ance structures . . . only evolve when there is an increased scarcity of land to support them”
(Harner 1975: 129). Archaeological research has formulated a similar postulate. First pro-



SPREAD ONE PICA LONG

oi.uchicago.edu





225



EARLY BRONZE SOCIAL ORGANIZATION AS REFLECTED IN BURIAL PATTERNS FROM THE



posed by A. Saxe as his “Hypothesis 8” (1970: 119),



2



it states basically that a corporate group,
in order to affirm its ancestral right to a particular restricted resource, will often establish a for-
mal disposal area for its dead.

From what we have seen of the EB IA cemetery at Bâb edh-Dhrâº, its “formal” nature
seems apparent. The arrangement of the shaft tombs within the cemetery appears to follow a
coherent plan, while the burials themselves maintain a consistent pattern. In addition, there is
every indication that specific shafts were used repeatedly by the same corporate (possibly
even familial) unit. As Bentley insists (1987: 28), there can be little doubt that the cemetery
functioned as the formal disposal ground of a distinct and well-defined population group. Yet,
it is also difficult to ignore the absence of extensive settlement remains. It is conceivable,
therefore, that by maintaining a formal cemetery, the otherwise transient EB IA population
was not laying claim to restricted land resources, but to another limited resource, namely
water (for a survey of local environmental conditions, see Harlan 1981, 1982, 1985).







If indeed



2. Subsequent research has largely substantiated, although in revised form, his initial hypothesis. In particular,
see the work of L. G. Goldstein 1976, 1980, 1981; and more recent critique in Morris 1991.

FIGURE 13.6. Plan of Bâb edh-Dhr⺠Tomb A54 (after Schaub 1973: 6, fig. 4).

oi.uchicago.edu

[image: image73.tif]



226



TIMOTHY P. HARRISON



we are dealing with semi-nomadic pastoralists, control of one of the few perennial water
sources in the region would have been of vital concern. Since the cemetery is adjacent to one
of the principal wadis in the region, it is possible to see its creation as the result of the system-
atic efforts of a distinct population group, whether transient or not, seeking to legitimize its
claim to this valuable water source through the presence of a nearby ancestral burial ground
(see the paper by Rast, chap. 27, this volume, for a similar view; see also Rast 1999).

In addition to addressing the question of settlement, Bentley also tried to determine
whether kinship played a factor in the social organization of the EB IA Bâb edh-Dhr⺠ceme-
tery population. To further test the possibility that the spatial arrangement of the EB IA shaft
tombs was the result of repetitious use by familial, kin-based social units, Bentley studied the
nonmetric dental traits of approximately 300 incomplete skeletal individuals collected from
twenty-six of the tombs. This approach was chosen because of its well-established ability to
elucidate relationships, both spatially and chronologically, within a given population (Bentley
1987: 45–67). The results of her analysis favored the conclusion that family groups were bur-
ied together in the shaft tombs and reinforced the notion of corporate group structure within
the EB IA population at Bâb edh-Dhrâº. Moreover, the homogeneity exhibited in the dental
traits suggested a history of inbreeding and the possibility of endogamous marital relations
(Bentley 1987: 205–07).

To determine whether social ranking occurred within the cemetery population, Bentley
also studied the spatial dimensions of the accompanying tomb artifacts. Generally, she found
little differentiation in the artifact distribution, suggesting a relative absence of social ranking.
Indeed, the consistent spatial arrangement of the burials, with the central pile of postcranial
bones and row of crania on the left, may have been an attempt to downplay the individuality
of the deceased in favor of the collective group (Bentley 1987: 208–40). Altogether, the evi-
dence gathered from the EB IA mortuary data indicate a population with a strong corporate
structure in which social ranking played a minimal role.



Early Bronze IB



The limited EB IB mortuary evidence makes it precarious to venture a reconstruction of the
level of social organization achieved during this transitional period. Excavations on the mound
have revealed a more substantial settlement than in the preceding period, and with the appear-
ance of charnel houses in the cemetery, the indication is of an increasingly sedentary population.

The emergence of the round charnel tomb type provides some hint of social transformation
within the cemetery population. Its form clearly betrays a development from the earlier EB IA
shaft tombs and leaves little doubt that this transformation was an indigenous one. Also sig-
nificant was the marked increase in primary burials. In the circular charnel, Tomb G1, primary
articulation was the predominant pattern. Although the building had experienced extensive
damage from burning, it was still possible to distinguish the remains of some 150 individuals.
Interestingly, three of these individuals had sustained injuries that had healed prior to death,
including two that had received ax blows to the cranial region (Ortner 1982: 93–95).

It is possible that the growing number of primary burials reflects a decreasing concern with
mortuary ritual and the “symbolic utility” of the dead (Bentley 1987: 247). If true, this implies
that the population that used the cemetery no longer felt the same need to legitimize its claim
to certain resources through the presence of a formal burial ground, although the high visibil-



SPREAD ONE PICA LONG

oi.uchicago.edu





227



EARLY BRONZE SOCIAL ORGANIZATION AS REFLECTED IN BURIAL PATTERNS FROM THE



ity of the freestanding charnels also would have carried much of the same formal quality.
Either way, controlled access to violence, implied by the wound-inflicted crania, might well
have rendered useless the symbolic value of the cemetery as a legitimization of the settle-
ment’s control of a limited resource. The decline in the concern for the symbolic utility of the
dead may also suggest a shift in group affiliation. Rather than relating only along kinship lines,
there may now have been an emphasis on membership within a larger corporate body. Such a
shift likely would have coincided with the move to a more sedentary existence and its inevi-
table focus on agriculturally oriented subsistence strategies. Under such circumstances, where
labor becomes a critical resource, and the need for greater organization more immediate,
social stratification would have become an increasingly evident reality.



Early Bronze II–III



This period witnessed settlement expansion reach its peak. The town site not only grew to
its largest extent, but was also encircled with a fortification wall for protection (Schaub 1982:
71). In the cemetery, the charnel house became the established tomb type. These large rectan-
gular structures were capable of holding significant numbers of interments and might have
stood as visible unifying symbols for a sedentary population whose strong earlier kinship ties
had now been engulfed by the larger corporate identity of a more complex social order. They
might also have been designed to resemble contemporary domestic architecture, as recently
suggested for the Neolithic megaliths of western Europe (Hodder 1984:53), reinforcing the per-
manence and stability associated with sedentism. Although the almost haphazard treatment of
the burials could suggest a nonstratified, egalitarian society, the apparent absence of differential
treatment may be misleading. It is possible, for example, that the symbolism embodied by these
collective burials had the intended effect of masking internal conflicts, or the inherent social
disparities, existent within the EB II–III population that resided at Bâb edh-Dhr⺠(Bentley
1987: 249; see also Shennan 1982; and Bradley 1985).

These rectangular charnels, however, were apparently not the only tomb type used during
this period. The so-called tholoi discovered by Lapp, and suggested by him to have contained
the burials of the ruling “aristocracy” of the town (1968c: 10, pl. III:2), probably also date to
the EB II–III period. Although the dating and precise function of these structures remains
uncertain, if they were indeed the elaborate tombs of individuals buried during the EB II–III
period, it is tempting to see them as the final resting place of a ruling elite. Such clear differ-
entiation in burial treatment would indicate a more sharply ranked, hierarchical society, one in
which a predominantly sedentary population had come under the authority of a political (or
religious) elite in control of the social institutions and economic resources of the community.
Whether this level of social organization represented that of a chiefdom or a state is difficult
to determine and open to speculation.



Early Bronze IV



The town site at Bâb edh-Dhr⺠came to an abrupt and destructive end at the close of the
EB III. Nevertheless, there is good reason to believe that some of the local population sur-
vived and continued to settle in the area. The evidence comes in part from the burial practices
that reappear during the ensuing EB IV period. The return of a tomb type strikingly similar in
both structure and burial content to the EB IA shaft tombs allows for certain observations to

oi.uchicago.edu





228



TIMOTHY P. HARRISON



be drawn about the EB IV social organization. In particular, it argues for the reassertion of a
strong, kin-based social structure and raises the possibility that kinship ties had remained
intact throughout the (Bentley 1987: 250–53). Certainly, the ability of kinship networks to
survive, and even thrive, in complex social contexts (including urban environments) is well
established. Thus, despite the hierarchical pressure asserted during the urban phase of the EB
II–III period, it is likely that kinship ties remained a powerful organizing force, at least on the
familial level, until they reemerged more visibly with the decline of sedentary activity in the
EB IV.

In this context, it is possible that the cemetery came to function as a kind of “ritual center”
(cf. Zohar 1992), replacing the nearby destroyed (or abandoned) EB III settlement as a place
of gathering for the economic, social, and political activities associated with the former town
site, thereby establishing a territorial (and possibly ancestral) claim to the surrounding region,
not unlike that suggested for the EB IA period. A similar pattern has been identified at other
EB IV sites in the region (Harrison 1997: 17–19).

TELL ES-SULTAN (JERICHO)

Located near an important perennial spring and situated in the same Irano-Turanian phyto-
geographical zone as Bâb edh-Dhrâº, the cemetery and settlement site at Jericho provides rele-
vant comparative data on the burial patterns and social organization of the southern Levant.
Although the site of Tell es-Sultan had been subjected to earlier large-scale excavations, it was
not until K. Kenyon’s work in the 1950s that the adjacent cemetery received extensive attention
(see, in particular, Kenyon 1960, 1965). As a result of her efforts, an incredible 369 tombs were
recorded. Of these, the vast majority dated to her Intermediate Early Bronze–Middle Bronze
(i.e., EB IV) period, some 344 in all, while only twenty-five were assigned to earlier EB periods.
Ten of these twenty-five tombs were identified as Proto-Urban (i.e., EB I), and the remaining fif-
teen simply as EB, with an additional Proto-Urban tomb (Tomb A114) apparently being reused
later in the EB (Kenyon 1965: 546–47).

The Jericho Proto-Urban tombs typically consisted of a single chamber accessed by a shaft
and containing large numbers of burials. One such tomb, Tomb A94, held a minimum of 113
individuals, at least as evidenced by the number of crania found (fig. 13.7) (Kenyon 1960: 16–
40). The burials in the Proto-Urban tombs were usually single interments, implying that each
tomb was in use for a lengthy period of time. The burials generally followed a disarticulated
burial pattern, although the disturbed nature of many of the tomb deposits often obscured this
distinction. Frequently, the tombs contained little but the partial remains of crania and ceramic
grave goods. Although there are details that differ—for example, the possible emphasis on
secondary burial at Jericho—the general pattern of burial evident at Proto-Urban Jericho
broadly parallels that found in the EB IB cemetery at Bâb edh-Dhrâº.



3



Despite the disturbed nature of the mortuary data, there was a discernible shift in burial
methods by the EB III period. Single-chambered shaft tombs were still the preferred tomb
type, but articulation had clearly become a common burial practice. Some tombs contained



3. I wish to thank W. Rast for drawing to my attention the parallels between the Proto-Urban chamber tombs
at Jericho and the EB IB circular charnel houses at Bâb edh-Dhr⺠and, more generally, for his helpful com-
ments regarding the Bâb edh-Dhr⺠tomb material.



SPREAD ONE PICA LONG

oi.uchicago.edu





229



EARLY BRONZE SOCIAL ORGANIZATION AS REFLECTED IN BURIAL PATTERNS FROM THE



large numbers of individuals, as did Tomb A, which contained an estimated 300 individuals
(Kenyon 1960: 52–53). In addition, the shape of the tomb chambers was now more rectangu-
lar (fig. 13.8) (Kenyon 1960: 94–96), not unlike the charnel houses at Bâb edh-Dhrâº. These
developments suggest that the cemetery had become the burial ground of a more sedentary
population, and indeed, excavations at the nearby tell support this view.



4



Finally, during Kenyon’s Intermediate Early Bronze–Middle Bronze (i.e., EB IV) period,
there was a decisive shift to the practice of individual primary burial. Rather than large “com-
munal” repositories, the EB IV tombs contained only one, or occasionally two, individuals
(fig. 13.9) (Kenyon 1960: 180–81). In keeping with what we have seen at Bâb edh-Dhrâº, this
development suggests a breakdown in urban life, and a corresponding increase in semi-
nomadic activity, a pattern that has been documented elsewhere for the southern Levant (Dever



4. The German excavations between 1907 and 1909 succeeded in clearing a large portion of an EB wall, as
well as a residential section at the northern end of the settlement (Sellin and Watzinger 1913: 20–45). In the
course of his excavations, Garstang identified two distinct EB urban phases, with Tomb A dating to the sec-
ond phase (Garstang and Garstang 1948: 75–88).

FIGURE 13.7. Plan of Jericho Tomb A94 (adapted from Kenyon 1960: 20, 22–23, figs. 5–7).

oi.uchicago.edu

[image: image74.tif]



230



TIMOTHY P. HARRISON



1987). In addition, a study of the age, sex, and burial treatment of forty-six individuals from
these EB IV tombs found little evidence of differentiation and concluded that the social struc-
ture of the local population probably resembled that of an egalitarian society (Shay 1983,
1985). Ranking, in other words, did not play an important role in the social organization of the
community. However, a subsequent study utilizing much the same data, but different analytical
methods and assumptions, has challenged this reconstruction, arguing instead that the social
structure of the EB IV population was stratified and not egalitarian (Palumbo 1987; see also
Shay’s response, 1989).



5



While no society may ever be “egalitarian” in the strictest sense of
the term, it is nevertheless difficult to see the preserved EB IV mortuary evidence from Jericho
as reflective of a stratified social order. Despite the apparent emphasis on the individual, there
is very little unambiguous evidence to suggest a concern for ranking within the interred ceme-
tery population. It is possible, however, that the burials found in the cemetery represent only
one part of an EB IV community, presumably that segment ascribed high (or at least differen-
tial) status, with the remainder of the population receiving treatment insufficient to have been
preserved in the archaeological record.



5. Most recently, yet another study has taken up the argument for a sharply differentiated, or stratified, EB IV
population at Jericho (Baxevani 1995).

FIGURE 13.8. Plan of Jericho Tomb D12 (after Kenyon 1960: 95, fig. 32, layer 1).

oi.uchicago.edu

[image: image75.tif]



231



EARLY BRONZE SOCIAL ORGANIZATION AS REFLECTED IN BURIAL PATTERNS FROM THE



While differences in detail do exist between the burial practices at Bâb edh-Dhr⺠and
Jericho, the broader patterns exhibited by both cemeteries follow a remarkably similar path
of development, and together substantiate the observations drawn about the EB social orga-
nization in the region.



FIGURE 13.9. Plan of Jericho Tomb A111 (after Kenyon 1960: 188, fig. 69).

oi.uchicago.edu

[image: image76.tif]



232



TIMOTHY P. HARRISON



SYMBOLISM, RITUAL, AND EARLY BRONZE SOCIAL TRANSFORMATION

This paper has examined the burial patterns evident in the cemeteries associated with the
sites of Bâb edh-Dhr⺠and Jericho for what they reveal of the EB social organization in the
southern Levant. One of the most intriguing aspects of mortuary ritual is its potential to reveal
familial or kinship networks, a basic organizational feature in most societies. The possibility
of identifying such ties enables this fundamental level of human relationship to be incorporated
into the analysis of broader patterns of social development. In the case of the EB southern
Levant, the preservation of kinship structure during the urban phase of the EB II–III is implied
by its reappearance in EB IV, following the collapse of centralized political control. Rather
than seeing the successive phases of EB as disparate stages of development with distinct breaks
in between, the underlying kinship structure allows them to be placed on a social continuum
along which configurations shifted between the polar extremes of a highly stratified society at
one end, and a socially unranked society at the other.



6



Hence, kinship ties maintained in EB I
can be seen as “hidden,” rather than eliminated or suppressed beneath the larger corporate
identity constructed by the more centralized, “urban social order of EB II–III. This systematic
approach to the mortuary record thus permits the charting of emerging complex levels of social
organization without ignoring lower level structural relationships, like familial or kinship net-
works, that typically persist within a society despite changes to the social order. In this way,
the symbolism expressed in the EB mortuary ritual can enhance our understanding of the
dynamics involved in the social transformation of EB society.



6. R. M. Adams has constructed a similar model for Mesopotamia (1975), as has L. Marfoe (1979) for the Biqaº
Valley in Lebanon.



ACKNOWLEDGMENTS

This paper owes much to Douglas Esse, who first pointed me in the direction of mortuary
studies and convinced me of the importance of utilizing mortuary data when analyzing broad
patterns of sociopolitical development. [It represents an updated version of a paper written in
1989 for a graduate seminar on the “Archaeology of Death.” I wish to thank A. Graham and
S. Batiuk for their help reproducing the figure illustrations.]

REFERENCES



Adams, R.
1975 The Mesopotamian Social Landscape: A View from the Frontier. Pp. 1–20 in



Reconstructing
Complex Societies



, ed. C. B. Moore. Bulletin of the American Schools of Oriental Research
Supplement 20. Cambridge, MA: American Schools of Oriental Research.

Albright, W. F.
1924 The Archaeological Results of an Expedition to Moab and the Dead Sea.



Bulletin of the Ameri-
can Schools of Oriental Research



14: 2–12.
1926 The Jordan Valley in the Bronze Age. Pp. 13–74 in



Annual of the American Schools of Oriental
Research



6. New Haven: Yale University Press.

Albright, W. F.; Kelso, J. L.; and Thorley, J. P.
1944 Early-Bronze Pottery from Bâb ed-Dhr⺠in Moab.



Bulletin of the American Schools of Oriental
Research



95: 3–13.

oi.uchicago.edu





233



EARLY BRONZE SOCIAL ORGANIZATION AS REFLECTED IN BURIAL PATTERNS FROM THE



Bartel, B.
1982 A Historical Review of Ethnological and Archaeological Analyses of Mortuary Practice.



Journal
of Anthropological Archaeology



1: 32–58.
Baxevani, E.

1995 The Complex Nomads: Death and Social Stratification in EB IV Southern Levant. Pp. 85–95 in



The Archaeology of Death in the Ancient Near East



, eds. S. Campbell and A. Green. Oxford:
Oxbow Books.

Beck, L. A., ed.
1995



Regional Approaches to Mortuary Analysis



. New York: Plenum.
Bentley, G.

1987 Kinship and Social Structure at Early Bronze IA Bâb edh-Dhrâº, Jordan: A Bioarchaeological
Analysis of the Mortuary and Dental Data. Ph.D. dissertation, University of Chicago.

Binford, L.
1971 Mortuary Practices: Their Study and Their Potential. Pp. 6–29 in



Approaches to the Social
Dimensions of Mortuary Practices



, ed. J. A. Brown. Memoirs of the Society for American
Archaeology 25. Washington, DC: Society for American Archaeology.

Bloch, M.
1971



Placing the Dead: Tombs, Ancestral Villages, and Kinship Organization in Madagascar



. New
York: Seminar.

Bradley, R.
1985 The Archaeology of Monuments. Pp. 1–20 in



Consumption, Change and the Archaeology of
Deliberate Deposits



. Occasional Paper 13, University of Edinburgh, Department of Archaeology.
Edinburgh: University of Edinburgh Press.

Braun, D.
1981 A Critique of Some Recent North American Mortuary Studies.



American Antiquity



46: 398–416.
Brown, J. A., ed.

1971



Approaches to the Social Dimensions of Mortuary Practices



. Memoirs of the Society for Ameri-
can Archaeology 25. Washington, DC: Society for American Archaeology.

Campbell, S., and Green, A., eds.

1995



The Archaeology of Death in the Ancient Near East



. Oxford: Oxbow.
Carr, C.

1995 Mortuary Practices: Their Social, Philosophical-Religious, Circumstantial, and Physical Determi-
nants.



Journal of Archaeological Method and Theory



2: 105–200.
Chapman, R.

1987 Mortuary Practices: Society, Theory Building and Archaeology. Pp. 198–213 in



Death, Decay
and Reconstruction: Approaches to Archaeology and Forensic Science



, eds. A. Boddington,
A. N. Garland, and R. C. Janaway. Manchester: Manchester University Press.

Chapman, R.; Kinnes, I.; and Randsborg, K., eds.

1981



The Archaeology of Death



. Cambridge: Cambridge University Press.
Clark, V. A.

1979 Investigations in a Prehistoric Necropolis near Bâb edh-Dhrâº.



Annual of the Department of
Antiquities of Jordan



23: 57–77.
Dever, W. G.

1987 Funerary Practices in EB IV (MB I) Palestine: A Study in Cultural Discontinuity. Pp. 9–19 in



Love
and Death in the Ancient Near East: Essays in Honor of Marvin H. Pope



, eds. J. H. Marks
and R. M. Good. Guilford, CT: Four Quarters.

Douglass, W. A.

1969



Death in Murelaga: Funerary Ritual in a Spanish Basque Village



. Seattle: University of
Washington.

Fröhlich, B., and Ortner, D. J.

1982 Excavations of the Early Bronze Age Cemetery at Bab edh-Dhra, Jordan, 1981.



Annual of the
Department of Antiquities of Jordan



26: 249–67.

oi.uchicago.edu





234



TIMOTHY P. HARRISON



Garstang, J., and Garstang, J. B. E.

1948



The Story of Jericho



. Revised ed. London: Marshall, Morgan and Scott.

Glazier, J.

1984 Mbeere Ancestors and the Domestication of Death.



Journal of the Royal Anthropological Insti-
tute of Great Britain (Man)



19: 133–48.

Goldstein, L. G.

1976 Spatial Structure and Social Organization: Regional Manifestations of Mississippian Society.
Ph.D. dissertation, Northwestern University, Evanston, IL.

1980



Mississippian Mortuary Practices: A Case Study of Two Cemeteries in the Lower Illinois Valley



.
Northwestern University Archaeological Program, Scientific Papers, No. 4. Evanston, IL:
Northwestern University Press.

1981 One-dimensional Archaeology and Multi-dimensional People: Spatial Organization and Mortuary
Analysis. Pp. 53–69 in



The Archaeology of Death



, eds. R. Chapman, I. Kinnes, and K. Randsborg.
Cambridge: Cambridge University Press.

Harlan, J. R.

1981 Natural Resources of the Southern Ghor. Pp. 155–64 in



The Southeastern Dead Sea Plain
Expedition: An Interim Report of the 1977 Season



, eds. W. E. Rast and R. T. Schaub.



Annual of
the American Schools of Oriental Research



46. Cambridge, MA: American Schools of Oriental
Research.

1982 The Garden of the Lord: A Plausible Reconstruction of Natural Resources of Southern Jordan in
Early Bronze Age.



Paléorient



8:71–81.

1985 The Early Bronze Age Environment of the Southern Ghor and the Moab Plateau. Pp. 125–29 in



Studies in the History and Archaeology of Jordan



, Vol. 2, ed. A. Hadidi. Amman: Department of
Antiquities of Jordan.

Harner, Michael J.
1975 Scarcity, the Factors of Production, and Social Evolution. Pp. 123–38 in



Population, Ecology
and Social Evolution



, ed. S. Polgar. Hague: Mouton.

Harrison, Timothy P.
1997 Shifting Patterns of Settlement in the Highlands of Central Jordan During the Early Bronze Age.



Bulletin of the American Schools of Oriental Research



306: 1–37.
Hodder, I.

1980 Social Structure and Cemeteries: A Critical Appraisal. Pp. 161–69 in



Anglo-Saxon Cemeteries
1979,



eds. P. Rahtz, T. Dickinson, and L. Watts. British Archaeological Reports, International
Series 82. Oxford: British Archaeological Reports.

1982



Symbols in Action: Ethnoarchaeological Studies of Material Culture



. Cambridge: Cambridge
University Press.

1984 Burials, Houses, Women and Men in the European Neolithic. Pp. 51–68 in



Ideology, Power
and Prehistory



, eds. D. Miller and C. Tilley. Cambridge: Cambridge University Press.
Kenyon, K.

1960



Excavations at Jericho, Vol. One. The Tombs Excavated in 1952–4



. London: British School of
Archaeology in Jerusalem.

1965



Excavations at Jericho, Vol. Two The Tombs Excavated in 1955–8



. London: British School of
Archaeology in Jerusalem.

Kyle, M. G., and Albright, W. F.
1924 Results of the Archaeological Survey of the Ghor in Search for the Cities of the Plain.



Bibliotheca
Sacra



81: 276–91.
Lapp, P. W.

1966a The Cemetery at Bâb edh-Dhrâº, Jordan.



Archaeology



19: 104–11.
1966b Bâb edh-Dhrâº.



Revue Biblique



73: 556–61.
1968a Bâb edh-Dhrâº.



Revue Biblique



75: 86–93.
1968b Bâb edh-Dhr⺠Tomb A 76 and Early Bronze I Palestine.



Bulletin of the American Schools of
Oriental Research



189: 12–41.

oi.uchicago.edu





235



EARLY BRONZE SOCIAL ORGANIZATION AS REFLECTED IN BURIAL PATTERNS FROM THE



1968c Bâb edh-Dhrâº, Perizzites and Emim. Pp. 1–25 in



Jerusalem Through the Ages: The Twenty-
Fifth Archaeological Convention October 1967



, ed. J. Aviram. Jerusalem: Israel Exploration
Society.

Marfoe, L.
1979 The Integrative Transformation: Patterns of Sociopolitical Organization in Southern Syria.



Bul-
letin of the American Schools of Oriental Research



234: 1–42.
McCreery, D. W.

1977–78 Preliminary Report of the A. P. C. Township Archaeological Survey.



Annual of the Department
of Antiquities of Jordan



22: 50–62.
Metcalf, P., and Huntington R.

1991



Celebrations of Death: The Anthropology of Mortuary Ritual.



2nd ed. Cambridge: Cambridge
University Press.

Morris, I.
1991 The Archaeology of Ancestors: The Saxe/Goldstein Hypothesis Revisited. Cambridge Archaeo-

logical Journal 1: 147–69.
Ortner, D. J.

1982 The Skeletal Biology of an Early Bronze IB Charnel House at Bab edh-Dhra, Jordan. Pp. 93–95
in Studies in the History and Archaeology of Jordan, Vol. 1, ed. A. Hadidi. Amman: Department
of Antiquities of Jordan.

O’Shea, J. M.
1984 Mortuary Variability: An Archaeological Investigation. New York: Academic Press.

Pader, E.-J.
1982 Symbolism, Social Relations and the Interpretation of Mortuary Remains. British Archaeologi-

cal Reports, International Series 130. Oxford: British Archaeological Reports.
Palumbo, G.

1987 “Egalitarian” or “Stratified” Society? Some Notes on Mortuary Practices and Social Structure
at Jericho in EB IV. Bulletin of the American Schools of Oriental Research 267: 43–59.

Parker Pearson, M.
1982 Mortuary Practices, Society, and Ideology: An Ethnoarchaeological Study. Pp. 99–113 in Sym-

bolic and Structural Archaeology, ed. I. Hodder. Cambridge: Cambridge University Press.
Peebles, C. S., and Kus, S.

1977 Some Archaeological Correlates of Ranked Societies. American Antiquity 42: 421–48.
Rast, W. E.

1981 Patterns of Settlement at Bâb edh-Dhrâº. Pp. 7–34 in The Southeastern Dead Sea Plain Expedi-
tion: An Interim Report of the 1977 Season, eds. W. E. Rast and R. T. Schaub. Annual of the
American Schools of Oriental Research 46. Cambridge, MA: American Schools of Oriental
Research.

1999 Society and Mortuary Customs at Bâb edh-Dhrâº. Archaeology, History and Culture in Palestine
and the Near East: Essays in Memory of Albert E. Glock, ed. T. Kapitan. Atlanta: Scholars Press.

Rast, W. E., and Schaub, R. T.
1974 Survey of the Southeastern Plain of the Dead Sea, 1973. Annual of the Department of Antiquities

of Jordan 19: 5–53.
1978 A Preliminary Report of Excavations at Bâb edh-Dhrâº, 1975. Annual of the American Schools

of Oriental Research 43: 1–32.
1980 Preliminary Report of the 1979 Expedition to the Dead Sea Plain, Jordan, Bulletin of the Ameri-

can Schools of Oriental Research 240: 21–61.
Rast, W. E., and Schaub, R. T., eds.

1981 The Southeastern Dead Sea Plain Expedition: An Interim Report of the 1977 Season. Annual of
the American Schools of Oriental Research 46. Cambridge, MA: American Schools of Oriental
Research.

Richard, S.
1987 The Early Bronze Age: The Rise and Collapse of Urbanism. Biblical Archaeologist 50: 22–43.

Saxe, A. A.
1970 Social Dimensions of Mortuary Practices. Ph.D. dissertation, University of Michigan, Ann Arbor.

oi.uchicago.edu



236 TIMOTHY P. HARRISON

Saxe, A. A., and Gall, P. L.
1977 Ecological Determinants of Mortuary Practices: The Temuan of Malaysia. Pp. 74–82 in Cultural-

Ecological Perspectives on Southeast Asia, ed. W. Wood. Papers in International Southeast Asia
Studies 41. Athens, OH: University of Ohio Press.

Schaub, R. T.
1973 An Early Bronze IV Tomb from Bâb edh-Dhrâº. Bulletin of the American Schools of Oriental

Research 210: 2–19.
1981 Patterns of Burial at Bâb edh-Dhrâº. Pp. 45–68 in The Southeastern Dead Sea Plain Expedition:

An Interim Report of the 1977 Season, eds. W. E. Rast and R. T. Schaub. Annual of the American
Schools of Oriental Research 46.

1982 The Origins of the Early Bronze Age Walled Town Culture of Jordan. Pp. 67–75 in Studies in the
History and Archaeology of Jordan, Vol. 1, ed. A. Hadidi. Amman: Department of Antiquities
of Jordan.

Schaub, R. T., and Rast, W. E.
1984 Preliminary Report of the 1981 Expedition to the Dead Sea Plain, Jordan. Bulletin of the Ameri-

can Schools of Oriental Research 254 : 35–60.
1989 Bâb edh-Dhrâº: Excavations in the Cemetery Directed by Paul W. Lapp (1965–67). Winona

Lake, IN: Eisenbrauns.
Sellin, E., and Watzinger, C.

1913 Jericho: Die Ergebnisse der Ausgrabungen. Wissenschaftliche Veröffentlichungen der Deut-
schen Orient-Gesellschaft, Nr. 22. Leipzig: Hinrichs.

Shay, T.
1983 Burial Customs at Jericho in the Intermediate Bronze Age: A Componential Analysis. Tel Aviv

10: 26–37.
1985 Differential Treatment of Deviancy at Death as Revealed in Anthropological and Archaeological

Material. Journal of Anthropological Archaeology 4: 221–41.
1989 The Intermediate Bronze Period: A Reply to G. Palumbo. Bulletin of the American Schools of

Oriental Research 273: 84–86.
Shennan, S.

1982 Ideology, Change and the European Early Bronze Age. Pp. 155–61 in Symbolic and Structural
Archaeology, ed. I. Hodder. Cambridge: Cambridge University Press.

Stager, L. E.
1992 The Periodization of Palestine from Neolithic through Early Bronze Times. Pp. 28–41 in Chro-

nologies in Old World Archaeology, Vol. 1, 3rd ed., ed. R. W. Ehrich. Chicago: University of
Chicago Press.

Tainter, J. A.
1975 Social Inference and Mortuary Practices: An Experiment in Numerical Classification. World

Archaeology 7: 1–15.
1978 Mortuary Practices and the Study of Prehistoric Social Systems. Pp. 105–41 in Advances in

Archaeological Method and Theory, Vol. 1, ed. M. B. Schiffer. New York: Academic Press.
1981 Reply to “A Critique of Some Recent North American Mortuary Studies.” American Antiquity

46: 416–20.
1982 Energy and Symbolism in Mortuary Practices. Pp. 63–75 in New Uses of Systems Theory in

Archaeology, ed. E. G. Stickel. Ballena Press Anthropological Papers 24. Los Altos, CA: Ballena
Press.

Ucko, P. J.
1969 Ethnography and Archaeological Interpretation of Funerary Remains. World Archaeology 1:

262–80.

Zohar, M.
1992 Megalithic Cemeteries in the Levant. Pp. 43–63 in Pastoralism in the Levant: Archaeological

Materials in Anthropological Perspectives, eds. O. Bar-Yosef and A. Khazanov. Monographs in
World Prehistory 10. Madison, WI: Prehistory Press.

oi.uchicago.edu



237



THE HISTORY OF THE COLLARED PITHOS AT
TELL EL-ºUMEIRI, JORDAN



Larry G. Herr



INTRODUCTION

One of Doug Esse’s important interests was the collared pithos (Esse 1992). It is with plea-
sure, therefore, that I submit new discoveries regarding the history of this vessel from the exca-
vations at Tell el-ºUmeiri. Located south of Amman near the border of the Ammonite hills
where they meet the Madaba Plain, the site was occupied from the Early Bronze Age to the
Persian period with a few breaks from time to time. Five seasons of excavation have been con-
ducted there by the Madaba Plains Project in 1984 (Herr 1989), 1987 (Herr 1991a), 1989 (Herr
1997), 1992 (Herr 2000), and 1994 (Clark 1994). The pithoi under discussion came from three
episodes of settlement which spanned the end of the Late Bronze Age (LB) to the Persian
period. Each episode, separated from the others by periods of abandonment, included more
than one stratigraphic phase but without significant changes to the essential material culture.

We first present an overview of the archaeological context followed by a discussion of the
pithoi. It is the purpose of this paper to explore evidence that, whereas collared pithoi disap-
peared toward the end of the Iron I period in Cisjordan, they seem to have continued through-
out the Iron Age at Tell el-ºUmeiri and elsewhere on the central Transjordanian plateau.

STRATIGRAPHY

The stratigraphic chart of phasing at ºUmeiri in table 14.1 corresponds with that following
the fourth (1992) season. We present only the relevant phases from LB to the Persian period.
The stratigraphy of each field of excavation (the horizontal axis) is broken down into phases
on the vertical axis. Each field has its own series of phase numbers which are retained here so
that this study can be integrated with other published results of the excavation (for these phase



numbers, see the field reports in the



Madaba Plains Project 4



, Herr 2000). The short hori-
zontal lines indicate periods of abandonment separating the three episodes of activity. The
“episode” numbers occurring in this study are not found in any other preliminary or seasonal
reports. They are intended to facilitate references within this paper only.

Field A is located at the western rim of the site (fig. 14.1); Field B stretches down the west-
ern slope; Fields C and G were outside the main settlement on the northern slope; Field D was
on a shelf on the southern slope outside the settlements of our periods; Field E was at the
water source at the bottom of the north slope; and Field F was laid out on the eastern rim.

Table 14.1 is an attempt to establish a site-wide stratification. However, none of the con-
nections are certain between the fields, except most of those between Fields A and B, which
are adjacent to each other. We have tried to avoid phase proliferation by tentatively suggesting



14



oi.uchicago.edu





238



LARRY G. HERR



connections. A question mark beside a phase number indicates that the attribution is correct
for the time period, but we are uncertain about its precise relationship to phases in other fields.
Usually, the least certainty occurs in those fields outside the top of the mound, such as Fields
C, E, F, and G.

HISTORY OF THE SITE



Episode 1 (LB IIB–Iron IA)



Episode 1A consisted of fill debris with LB pottery immediately on top of Middle Bronze
Age IIC walls in Field F (Low 1997) and the earliest stages of a terrace wall in Field C (Bat-
tenfield 1991: 85).



FIGURE 14.1. Topographic map of Tell el-ºUmeiri with the fields of excavations.

oi.uchicago.edu

[image: image77.tif]



239



THE HISTORY OF THE COLLARED PITHOS AT TELL EL-ºUMEIRI, JORDAN



The remains from Episode 1B suggest a smooth transition from LB IIB to Iron IA. In
Field C, the terrace wall from the previous phase was expanded (Battenfield 1991: 85), while
in Field F a long terrace wall was constructed that preserved the LB fill behind it (Low 1997:
fig. F7) and allowed Iron IA debris to accumulate, as well. The Iron IA levels from Field E,
the water source, probably belong here (Fisher 1997: fig. E8), as do fragmentary walls in
soundings beneath the later administrative buildings in Field A. The presence of significant
numbers of Iron IA potsherds in the rampart of Episode 1C in Field B infers a settlement that
preceded the rampart construction (Episode 1B). An earthquake that cracked bedrock in
several places was apparently responsible for the destruction of this settlement.

The most astonishing phase at ºUmeiri is Episode 1C, dated to Iron IA. It contains the best
preserved fortification system so far discovered from this time in all of Palestine, including a
casemate wall, a sloping rampart or glacis laid against the outer casemate wall, a retaining
wall at the bottom of the rampart, and a dry moat (Clark 1994). The two casemate rooms so
far excavated have produced large numbers of collared pithoi standing against the walls.
Domestic structures were found inside the fortifications in both Fields B and A (Clark 1994;
Lawlor 1991: fig. 3.3). One of them seems to have contained a small cultic corner, while
another had two rows of pillar bases and might have been a “four-room house.” Some of the
house walls were up to 2.5 m high. This strong, fortified settlement was abruptly and violently
destroyed. The destruction debris (1.5–2.5 m deep) contained reconstructable collared pithoi
high above the floors, fallen from the upper stories.



Episode 2 (Early Iron II)



Episode 2 is made up of two more or less ephemeral phases. Above the inner wall of the
casemate fortification was a small, thin surface with transitional Iron I/Early Iron II pottery,
Episode 2A (Clark 1989: 249–50).



.



Table 14.1.



Stratigraphy at Tell el-ºUmeiri



Episode Period A B C D E F G



1A LB IIB 11 — 4 — — 10 —

1B Iron IA 10 11B 4 — 8? 9? Loci 9,15?

1C Iron IA 9 11A 4 — 8? 8? Loci 9,15?

2A Iron I/Iron II — 10 — — — — —

2B E. Iron II 8 9 — — 7 — —

3A L. Iron II 7 8 3? — 6 7 —

3B L. Iron II/Persian 6 7 3? — 5 6 —

3C Persian 5 6 2? — 4 5 —

3D Persian 4 5 2? — 4 4 —

3E Persian 3 4 ? — 4 3 —

oi.uchicago.edu





240



LARRY G. HERR



Directly on top of the surface was Episode 2B, a domestic storeroom containing two stor-
age jars and a jug, dated most likely to the ninth century (Clark 1989: 250–53, fig. 16.5–7). In
Field A, fragmentary walls, oriented similarly to the storeroom, also contained associated
early Iron II pottery and, at the water source, shallow debris layers can be attributed to this
period. It appears that the settlement was limited to a few houses scattered over the site but
most heavily concentrated in the west.



Episode 3 (Late Iron II to Persian)



Episode 3 contained five phases best differentiated in Fields A and B, which have the same
number of phases and similar relationships to earlier and later phases. Although Field F also
contained the same number of phases, no certain connection of the phasing can be made. The
attribution of phases in Field C is guesswork, based on their general ceramic date. We have
lumped all the late Iron II/Persian materials from the water source into one phase and con-
nected them with all tell phases. Episode 3A is made up of pits (Field B) and small, flimsy
installations (Field A). The date for the beginning of this episode is important. Based on
inscriptional and palaeographic evidence from ostraca and seals, activity appears to have
begun during the first half of the sixth century, probably soon after the Babylonian invasion of
582



b.c.



avenging Ammonite complicity in the murder of Gedaliah, the Babylonian-appointed
governor of Judah (Jer 40) (Herr 1993; Sanders 1997). The ephemeral remains from this ear-
liest phase seem to reflect the activities of builders who were constructing the new administra-
tive complex of Episode 3B for the Ammonite monarchy after it became subject to Babylon.

In Episode 3B at least two large buildings formed an administrative complex in Field A,
while several smaller structures to the north produced domestic material culture. Because the
Ammonite government was subject to the Babylonians, there was no fortification system. The
outer casemate wall of Iron IA in Field B might have been used to support the buildings, and
a flimsy entrance structure in Field F (Low 1991: 186; 1997: fig. F11) might have taken the
place of a gate, but without an associated wall. Small, extra-urban structures were found on
the northern slope (Battenfield and Herr 1989: 280) and the water source (Fisher 1997: fig.
E5). It is probably to this phase that we should ascribe several Ammonite seals and seal
impressions found in the topsoil, such as the impression of Milkomªur the servant of
Baºalyashaº (Herr 1985; Younker 1985). A mid-sixth-century jar sat on the floor of the south-
ernmost building (Herr 1991b: fig. 19.5:23), and an Attic sherd datable to the late sixth or fifth
century was found in the fill above the floors (Waldbaum 1991: 243). This phase thus lasted
most likely into the earliest years of the Persian period.

Episode 3C consisted of new floors in the administrative complex about 1 m above the
previous ones and included slight alterations to the rooms (Lawlor 1989: 238, cf. figs. A3 and
A7; Clark 1997: fig. B32; Low 1997). This phase must date to the beginning of the Persian
period. That the function of the administrative complex remained the same during this period
is clear from three Persian provincial ºAmmon seal impressions from the end of the sixth or
beginning of the fifth century (Herr 1992: 190–93 discusses two of them). The impressions
also suggest that the population was still Ammonite.

Episode 3D begins the abatement of the Persian settlement. At least parts of the adminis-
trative buildings went out of use and new structures with smaller rooms were built to the east
(Lawlor 1997: fig. A10).

oi.uchicago.edu





241



THE HISTORY OF THE COLLARED PITHOS AT TELL EL-ºUMEIRI, JORDAN



Small fragments of walls in Fields A, B, and F suggest a further thinning of occupational
intensity during Episode 3E (Lawlor 1997: fig. A17; Clark 1997: fig. B34; Low 1997: fig. F16).
At some point, perhaps during the late fifth or fourth century, settlement ceased.

THE PITHOI

By presenting examples of pithoi from Tell el-ºUmeiri ascribed to these three episodes of
activity, we hope to establish a typological connection between Iron I collared pithoi and the
extremely frequent holemouth pithoi from the end of the Iron Age on the Transjordanian
plateau; the latter are rare elsewhere. We present the pithoi from surfaces or destruction layers
that we can confidently ascribe on the basis of stratigraphy to the three episodes described
above. From Episode 1 the collared pithoi occur both as complete vessels standing on the sur-
faces and as reconstructable vessels in the massive destruction. The data from Episode 2 are
the most problematic because we could isolate very few surface deposits to this period, and
we have found no complete or reconstructable vessels. Moreover, pithos sherds came from
layers that feasibly could contain earlier material, as well. However, the pithoi from this hori-
zon are clearly foreign to the early Iron I assemblage in terms of ware and form and, because
there is otherwise apparently no settlement at the site between the twelfth and tenth centuries



b.c.



, it is safe to assume that the pithoi are not intrusions from the eleventh and tenth centuries.
For Episode 3, we do not have complete examples from ºUmeiri, but the sherds occur in such
great numbers and such large sizes (we have found the bottom halves of several vessels) that
they cannot be intrusions from earlier periods. They also occur in similar numbers at contem-
porary sites in the region (Hesban [Lugenbeal and Sauer 1972: pl. 7], Jawa [Daviau, pers.
comm.], Jalul [Younker, pers. comm.], and Amman [Dornemann 1983: fig. 57]). Identical
forms have been reconstructed at Tell Jawa (Daviau 1992: 151).

We also present other forms from fills or secondary layers that, regardless of the unclear
date of their deposition, can be ascribed to either the second or third episode on typological
grounds.



Episode 1 (LB IIB–Iron IA)



All illustrated pithoi from this period (figs. 14.2–14.4) were found together in one casemate
room in Field B, Square 7J89, smashed by the destruction (see Clark 1997, Field Phase 11). The
reconstructed vessels, of which only the rims and necks are shown in fig. 14.2 (for the complete
forms, see Clark 1997), were found at one end of the room, and the sherds in figs. 14.3–14.4
were found in the destruction layer that covered them and most likely came either from other
jars in the room that have so far not been reconstructed or from those stored in an upper story.

One of the most surprising aspects of our assemblage is the great variety of rim and collar
forms. All the rims are everted, many strongly. Most rims are varieties of thickened ovals or
rounded triangles (figs. 14.2; 14.3:8; 14.4:1, 3, 4, 6), but others are plain (figs. 14.3:1, 2;
14.4:2, 7), or profiled (fig. 14.3:5, 6, 9, 10). The most typical type of collar in Cisjordan is the
simple ridge below the neck on the upper shoulder; it is also found at ºUmeiri (figs. 14.2;
14.3:1, 2). But there are also examples with double ridges (figs. 14.2:2; 14.4:3, 4), thickened
collars (figs. 14.3:5, 8; 14.4:2), thickened collars (short or long) with grooves, ridges, or waves
(figs. 14.2:3, 4; 14.3:5–8, 10; 14.4:5), and a second ridge on the upper neck (fig. 14.4:1, 2, 6, 7).

oi.uchicago.edu





242



LARRY G. HERR



There is variety also in neck and shoulder relationships. Sometimes the neck and shoulder
turn outward to form a long shoulder (figs. 14.2:2–3; 14.3:2, 6), but most have long necks and
more horizontal shoulders.

Although all vessels went through the destruction fire which was so hot it turned some of
the stones in the surrounding walls to lime and vitrified much of the pottery, the wares of most
of the vessels seem similar to each other and to others outside the room (not illustrated here).

The other pottery forms (cooking pots, jugs, etc.) retrieved from the rampart in use with
the casemate room in which our pithoi were found were identical to those in the destruction
layer. Because the rampart pottery must have been made before the pottery from the destruc-
tion illustrated here, the similarity between the two assemblages suggests a short life-span for
the casemate room and its associated phase of Episode 1C. The variety of forms cannot, there-
fore, be attributed to temporal causes, but must be attributed to potting techniques or stylistic
preferences of potters or users. In any case, this assemblage provides an excellent picture of
the variety of contemporary vessels at one site, most likely made of the same clays (a techno-
logical study of the ware is being undertaken by Gloria London).



Episode 2 (Early Iron II)



Lying on a surface at the watercourse (Locus E.2:42) was a later version of the collared
pithos (fig. 14.5:1), in which the rim is barely thickened. Indeed, the neck and collar are so
short there is barely room for any distinctiveness to the rim. This collar is very different from



FIGURE 14.2. Collared pithoi from Tell el-ºUmeiri from a casemate room.

oi.uchicago.edu

[image: image78.tif]



243



THE HISTORY OF THE COLLARED PITHOS AT TELL EL-ºUMEIRI, JORDAN



the typical long-necked, long-collared forms from the early Iron I period. It appears to be the
descendent of the early Iron I form with the thickened collar (figs. 14.3:5, 8; 14.4:2). This ves-
sel appeared together with jug forms dated to Iron I and early Iron II (Herr 1997).

Another pithos with a very short neck is found in figure 14.5:2. It came from postoccupa-
tional debris immediately above one of the early Iron II surfaces in Field A (Locus 7K51:



FIGURE 14.3. Rims and collars of collared pithoi from Tell el-ºUmeiri from a casemate room, excavated within
two layers of the Iron I destruction.

oi.uchicago.edu

[image: image79.tif]



244



LARRY G. HERR



34). There is a ridge below the rim and a vestigial ridge on the upper shoulder. This seems to
be a descendent of the early Iron I type of collared pithos with both rim and shoulder ridges
(fig. 14.4:1, 2, 6, 7).

These two examples recall the early Iron I forms, but a third suggests the direction subse-
quent pithoi would take (fig. 14.5:3). It came from the floor of the storeroom in Field B (Locus



FIGURE 14.4. Rims and collars of collared pithoi from Tell el-ºUmeiri from a casemate room, excavated within
two layers of the Iron I destruction.



one line long

oi.uchicago.edu

[image: image80.tif]



245



THE HISTORY OF THE COLLARED PITHOS AT TELL EL-ºUMEIRI, JORDAN



7J89:6) built immediately above the destruction layer of the Early Iron I city, but below the
domestic buildings belonging to the Episode 3. The rest of the pottery in the storeroom clearly
dated to the ninth–eighth centuries (Herr 1989: fig. 19.4:1–16). This pithos, which can now be
called a “holemouth pithos,” retains a version of the short, thickened collar reminiscent of that



FIGURE 14.5. Rims and collars of pithoi from Tell el-ºUmeiri dating to early Iron II



(1–3)



and late Iron II/Persian



(4–11).

oi.uchicago.edu

[image: image81.tif]



246



LARRY G. HERR



in fig. 14.5:1, but the neck has completely disappeared, and the rim reduces the upturn to a
bulbous thickening (Sauer has noticed this “bulbous” rim on earlier forms of collared pithoi
from Hesban dating to the tenth century: 1994: 240). This holemouth form, which may be as
early as the late ninth century, is ubiquitous at Iron II sites in central Transjordan.

Although the archaeological deposits for Episode 2 are weaker than those of Episode 1
and there is a hiatus of 250–300 years between them, the jump from the Iron I forms to those
of early Iron II is not significant, except for the third example. However, other examples from
unstratified fill loci (below) illustrate its connection, as well. Thus, we would suggest the typo-
logical development of the collared pithos that occurred in the late Iron I, while ºUmeiri was
in hiatus, was to shorten the neck and collar and reduce the significance of the rim, moving
slowly toward the holemouth form of the later centuries of Iron II.



Episode 3 (Late Iron II/Persian)



During this period all pithoi are holemouth forms and are very frequent at every late Iron II
and early Persian site on the central Transjordanian plateau. Their rim and upper shoulder treat-
ment vary greatly, but many examples reflect typologically later developments of specific earlier
collar treatments. The first examples seem to recall the short, thickened collar from the previ-
ous episode (cf. the late examples in figure 14.5:4, 5 with the earlier one in fig. 14.5:1). The
rim is bulbous in one example and slightly turned up in the second, but the ridge below the rim
on the upper shoulder is where we would expect it if the collar continued to shorten after the
early Iron II examples were made. The sherd in figure 14.5:4 came from a fill layer (Locus
7K71:8) deposited after the Persian administrative complex went out of use; although the locus
dates to the Persian period, the vessel from which the sherd came might have been made earlier.
The example in figure 14.5:5 came from a surface (7K42:16) laid after the Persian administra-
tive complex went out of use. Embedded into the surface were the bottom thirds of two pithoi
from which this rim may have come. These pithoi thus continued well into the Persian period.

Holemouth pithoi also reflect the collar form with the simple ridge on the upper shoulder
(fig. 14.5:6–10). Compare them with the early Iron I examples (figs. 14.2:1; 14.3:1, 2). Al-
though the late forms are from holemouth vessels, they still retain a thickened or upturned rim
and a distinct collar-like ridge on the upper shoulder. Figure 14.5:6 came from a postoccupa-
tion fill (Locus 7K60:5), but its form clearly belongs to this episode of activity. The sherd in
figure 14.5:7 came from a surface in Episode 3C of the administrative complex in Field A
dated to the early Persian period (Locus 7K51:13). Figures 14.5:8–10 all came from fill layers
immediately beneath the surface of the Episode 3C administrative complex (Locus 7K51:30),
and therefore date to the very end of Iron II or the first decades of the Persian period.

Other forms seem to be variations of these late vessels rather than strict descendents from
earlier types. We would call attention to the form in figure 14.5:11 with two grooves below the
rim. This is a particularly frequent type at Tell el-ºUmeiri in several subforms: the grooves can
be sharply made, have the appearance of waves, or appear more like ridges. When they appear
as ridges they are reminiscent of the double-ridged variety of collared pithos from the Iron I
assemblage (figs. 14.2:2; 14.4:3, 4).



Typologically Intermediate Pithoi



Other examples have been found in loci belonging to unstratified deposits, primarily top-
soil, but their forms typologically appear to belong to earlier periods. Because it is very pos-

oi.uchicago.edu





247



THE HISTORY OF THE COLLARED PITHOS AT TELL EL-ºUMEIRI, JORDAN



sible for sherds from later loci to have been made earlier, we include them here for interest’s
sake and because there were only three examples from the stratigraphic layers of Episode 2. It
is likely that most of these forms represent collared pithoi intermediate between those of Iron I
and the Late Iron II/Persian periods. Certainly, they are typologically intermediate. In all cases
the ware is typical of Iron II, not Iron I, and two examples are slipped (fig. 14.6:2, 7), a trait
typical of Iron II.

The first example (fig. 14.6:1) looks similar to those of Episode 1, but the rim is upright,
the neck is short, and the weak collar is very high on the shoulder; the ware is also very dif-
ferent from the Episode 1 pithoi, more akin to those of Iron II. This form could very easily
belong to late Iron I. The following two examples (fig. 14.6:2–3) also have upright rims, but
there is almost no neck. Indeed, if the rims were not turned up quite as strongly, they would
appear to be holemouth forms. Note the memory of a ridge at the bottom of the rim similar
to examples from Episode 1; this is a trait on several other rims, as well.



FIGURE 14.6. Rims and collars of pithoi from Tell el-ºUmeiri from a variety of periods in unstratified deposits.

oi.uchicago.edu

[image: image82.tif]



248



LARRY G. HERR



With the next two examples (fig. 14.6:4–5), the rim is not turned up as strongly as the pre-
ceding examples, and we can begin categorizing them as holemouth forms. But strong simi-
larities with the necked pithoi still remain: above the collar is a semblance of a neck; the rim
is thickened, even with a ridge at the bottom of one (fig. 14.6:4); the collar is clear.

The next four examples are similar to the previous two but appear closer to the standard
holemouth forms of Episode 3 (fig. 14.6:6–9). The collars are weaker, except for one (fig.
14.6:9); the rims are becoming bulbous (fig. 14.6:7, 8) or triangular (fig. 14.6:6). Note the rim
ridge on figure 14.6: 7. Some of these forms are so similar to holemouth types they could
belong to Episode 3. The last example (fig. 14.6:10) is a full-fledged holemouth pithos typi-
cal of Episode 3. The collar is still apparent as is the ridge at the bottom of the rim thickening.

EVIDENCE FROM ELSEWHERE

Worschech has already presented hints that collared pithoi continued into Iron II at Balua
(Worschech 1992), but some of his late forms are jars with ridged necks, not true collared
pithoi. Although we still need more examples from Moab, it looks like collared pithoi contin-
ued into Iron II there, as well. The debate between Finkelstein (1992) and Bienkowski (1992)
over the presence of Iron I at various Edomite sites based on the presence of collared pithoi
actually illustrates their presence in Edom during Iron II, as well. Although some of the ves-
sels Finkelstein calls our attention to may not be collared pithoi (Finkelstein 1992: fig. 2:1,
11–12, 14–15), and although most of the others have significantly different rim forms from
ours at ºUmeiri, it is easy to place them in the middle to end of our typology, dating to the
ninth through the seventh centuries. All necks slant inward strongly, and the collars are mostly
weak waves, grooves, or ridges. Because Finkelstein used the history of collared pithoi in
Cisjordan for his paradigm, where they ended by the tenth century, he tried to place the
Edomite examples in Iron I, as well. We owe him thanks for recognizing them to be collared
pithoi. However, now that we can document similar pithoi continuing throughout Iron II at
ºUmeiri, recognizing the Edomite examples as Iron II forms fits the basic stratigraphy of their
sites of origin. There is still, therefore, no reason to see Iron I at these sites. Only one of the
forms from Edom has a good parallel at ºUmeiri (Finkelstein 1992: fig. 2:17). The unique rim
forms from Edom may thus represent specific regional variants. Note that at least one form is
reminiscent of smaller “Edomite” jars (Finkelstein 1992: fig. 2:4).

ACKNOWLEDGMENTS

Some of the work for this paper was done at the Albright Institute of Archaeological
Research in Jerusalem during the 1993/94 academic year. It was supported by Canadian Uni-
versity College which granted me a sabbatical, the trustees of the Albright Institute who
appointed me Annual Professor, and the Dorot Foundation.

The excavations at Tell el-ºUmeiri are part of the Madaba Plains Project sponsored by
Andrews University in consortium with Canadian University College, La Sierra University,
and Walla Walla College.

oi.uchicago.edu





249



THE HISTORY OF THE COLLARED PITHOS AT TELL EL-ºUMEIRI, JORDAN



REFERENCES



Battenfield, J. R.
1991 Field C: The Northern Suburb. Pp. 74–86 in



Madaba Plains Project 2



, eds. L. G. Herr, L. T.
Geraty, Ø. S. LaBianca, and R. W. Younker. Berrien Springs, MI: Andrews University Press.

Battenfield, J. R., and Herr, L. G.
1989 Field C: The Northern Suburb. Pp. 258–81 in



Madaba Plains Project 1



, eds. L. T. Geraty, L. G.
Herr, Ø. S. LaBianca, and R. W. Younker. Berrien Springs, MI: Andrews University Press.

Bienkowski, P.
1992 The Beginning of the Iron Age in Edom: A Reply to Finkelstein.



Levant



24: 167–69.
Clark, D. R.

1989 Field B: The Western Defense System. Pp. 244–57 in



Madaba Plains Project 1,



eds. L. T. Geraty,
L. G. Herr, Ø. S. LaBianca, and R. W. Younker. Berrien Springs, MI: Andrews University Press.

1994 The Iron 1 Western Defense System at Tell el-ºUmeiri, Jordan.



Biblical Archaeologist



57:
138–48.

1997 Field B: The Western Defense System. Pp. 53–98 in



Madaba Plains Project 3



: The 1989 Season
at Tell el-ºUmeiri and Vicinity and Subsequent Studies, eds. L. G. Herr, D. R. Clark, L. T. Geraty,
Ø. S. LaBianca, and R. W. Younker. Berrien Springs, MI: Andrews University Press.

2000 Investigating Ancient Ceramic Traditions on Both Sides of the Jordan. Pp. 100–110 in



The
Archaeology of Jordan and Beyond: Essays in Honor of James A. Sauer



, eds. L. E. Stager,
J. A. Greene, and M. D. Coogan. Winona Lake, IN: Eisenbrauns.

Daviau, P. M. M.
1992 Preliminary Report of the Excavations at Tell Jawa in the Madaba Plains (1991).



Annual of the
Department of Antiquities of Jordan



36: 145–62.
Dornemann, R. H.

1983



The Archaeology of the Transjordan in the Bronze and Iron Ages



. Milwaukee: Milwaukee
Public Museum.

Esse, D. L.
1992 The Collared Pithos at Megiddo: Ceramic Distribution and Ethnicity.



Journal of Near Eastern
Studies



51: 81–103.
Finkelstein, I.

1992 Edom in the Iron I.



Levant



24:159–66.
Fisher, J. R.

1997 Field E: The Water Source. Pp. 176–87 in



Madaba Plains Project 3



, eds. L. G. Herr, D. R. Clark,
L. T. Geraty, Ø. S. LaBianca, and R. W. Younker. Berrien Springs, MI: Andrews University Press.

Herr, L. G.
1985 The Servant of Baalis.



Biblical Archaologist



48: 169–72.
1989 The Pottery. Pp. 299–354 in



Madaba Plains Project 1



, eds. L. T. Geraty, L. G. Herr, Ø. S.
LaBianca, and R. W. Younker. Berrien Springs, MI: Andrews University Press.

1991a Organization of Excavation and Summary of Results on the Tell. Pp. 8–14 in



Madaba Plains
Project 2,



eds. L. G. Herr, L. T. Geraty, Ø. S. LaBianca, and R. W. Younker. Berrien Springs, MI:
Andrews University Press.

1991b The Pottery Finds. Pp. 232–45 in



Madaba Plains Project 2,



L. G. Herr, L. T. Geraty, Ø. S.
LaBianca, and R. W. Younker. Berrien Springs, MI: Andrews University Press.

1992 Epigraphic Finds from Tell el-ºUmeiri during the 1989 Season.



Andrews University Seminary
Studies



30: 187–200.
1993 What Ever Happened to the Ammonites?



Biblical Archaeology Review



19: 26–35.
1997 The Pottery. Pp. 228–49 in



Madaba Plains Project 3



, eds. L. G. Herr, D. R. Clark, L. T. Geraty,
Ø. S. LaBianca, and R. W. Younker. Berrien Springs, MI: Andrews University Press.

2000 Excavation and Cumulative Results at Tall al-ºUmayri. Pp. 7–20 in



Madaba Plains Project 4



,
L. G. Herr, D. R. Clark, L. T. Geraty, R. W. Younker, and O. S. LaBianca. Berrien Springs, MI:
Andrews University Press.

Shortened to Match

oi.uchicago.edu





250



LARRY G. HERR



Lawlor, J. I.
1989 Field A: The Ammonite Citadel. Pp. 233–43 in



Madaba Plains Project 1



, eds. L. T. Geraty,
L. G. Herr, Ø. S. LaBianca, and R. W. Younker. Berrien Springs, MI: Andrews University Press.

1991 Field A: The Ammonite Citadel. Pp. 5–52 in



Madaba Plains Project 2,



eds. L. G. Herr, L. T.
Geraty, Ø. S. LaBianca, and R. W. Younker, Berrien Springs, MI: Andrews University Press.

1997 Field A: The Ammonite Administrative Complex. Pp. 21–52 in



Madaba Plains Project 3



,
eds. L. G. Herr, D. R. Clark, L. T. Geraty, Ø. S. LaBianca, and R. W. Younker. Berrien Springs,
MI: Andrews University Press.

Low, R.
1991 Field F: The Eastern Shelf. Pp. 170–231 in



Madaba Plains Project 2



, eds. L. G. Herr, L. T. Ger-
aty, Ø. S. LaBianca, and R. W. Younker. Berrien Springs, MI: Andrews University Press.

1997 Field F: The Eastern Shelf. Pp. 188–221 in



Madaba Plains Project 3



, eds. L. G. Herr, D. R. Clark,
L. T. Geraty, Ø. S. LaBianca, and R. W. Younker. Berrien Springs, MI: Andrews University Press.

Lugenbeal, E. N., and Sauer, J. A.
1972 Pottery from Heshbon.



Andrews University Seminary Studies



10: 21–69.

Sanders, T. K.
1997 An Ammonite Ostracon from Tell el-ºUmeiri. Pp. 331–36 in



Madaba Plains Project 3



, eds.
L. G. Herr, D. R. Clark, L. T. Geraty, Ø. S. LaBianca, and R. W. Younker. Berrien Springs,
MI: Andrews University Press.

Sauer, J. A.
1994 The Pottery at Hesban and Its Relationships to the History of Jordan: An Interim Hesban Pottery

Report, 1993. Pp. 225–81 in



Hesban after 25 Years,



eds. D. Merling and L. T. Geraty. Berrien
Springs, MI: Andrews University Press.

Waldbaum, J. C.
1991 Two Attic Sherds. P. 243 in



Madaba Plains Project 2,



eds. L. G. Herr, L. T. Geraty, Ø. S. LaBianca,
and R. W. Younker. Berrien Springs, MI: Andrews University Press.

Worschech, U.
1992 Collared-Rim Jars aus Moab: Anmerkungen zur Entwicklung und Verbreitung der Kruege mit

“Halswulst.”



Zeitschrift des Deutschen Palästina-Vereins



108: 149–55.

Younker, R. W.
1985 Israel, Judah, and Amman and the Motifs on the Baalis Seal from Tell el-ºUmeiri.



Biblical
Archaeologist



48: 173–80.

oi.uchicago.edu



251



COMMODITIES AND CUISINE: ANIMALS
IN THE EARLY BRONZE AGE OF

NORTHERN PALESTINE



Brian Hesse and Paula Wapnish



INTRODUCTION

Before Douglas Esse began his excavations at the Jordan Valley site of Tel Yaqush, he
asked if we would be able to tackle the analysis of the animal remains he anticipated recovering
in the course of the project. We signed on enthusiastically, knowing from working with him at
Ashkelon that he would apply the meticulous standards of excavation that insure the collection
of a reliable sample of the bones, teeth, and shells preserved in an archaeological site. More
important to us, however, was the knowledge that this request was based on a question that lay
at the core of his program of research, one that demanded the engagement of zooarchaeological
skills. Esse wanted to know how the economy at Tel Yaqush, and in particular the agro-pastoral
sector, was integrated into and shaped by the surrounding and emergent urban landscape of the
Early Bronze Age (EB). He recognized that the bulk of our knowledge of this period in the
southern Levant was based on extensive and detailed survey work (e.g., Esse 1991; Joffe 1991,
1993; Portugali and Gophna 1993) combined with the excavation of a small number of the
largest settlements. While a few smaller communities had been sampled (e.g., Braun 1985),
relatively little was known of the systems of production and exchange that supported them,
and not much of that was based on biological evidence (for a review, see the survey of Horwitz
and Tchernov 1989).

From the discussions we had with Esse while setting up the project, we concluded that a
basis for constructing a model should include ideas drawn from locational geography, analo-
gies from ethnographic evidence, reports from those areas of the late fourth and third millen-
nia



b.c.



that have yielded textual evidence bearing on the problem, and archaeological
parallels from outside the region. All pointed to the proposition that Tel Yaqush was a node in
a pastoral production, distribution, and consumption system that involved interaction between
and exchange of animal products by nomadic groups, hamlets and villages of various sizes,
and the large walled cities that dotted the region (see also, Schwartz and Falconer 1994;
Falconer 1995: 399–401). Small though the site was, it likely contained distinct social compo-
nents that had differential access to animal products. Since Tel Yaqush was occupied through-
out most of EB, it was further thought that the animal remains could be brought to bear on
suggestions that the social and economic organization of the region underwent significant
oscillations in the degree of centralization and urbanization during this multi-century period
(Portugali and Gophna 1993; Finkelstein 1995).

The research plan that evolved required the examination of the faunal remains from Tel
Yaqush to see if evidence for a locally differentiated cycle of pastoral production, exchange,



15



oi.uchicago.edu





252



BRIAN HESSE AND PAULA WAPNISH



and consumption could be detected. Could we support the idea that the animal resource–
related activities at Tel Yaqush were different in some way from those going on in nearby
communities and, further, that the differences implied the existence of a coordinated system?
Was the familiar modern system already in place? Were animals and their products already
circulating in economic systems as commodities (Appadurai 1986), or were the technological
and sociopolitical limitations of the EB society in the southern Levant such that these
resources had only limited importance as sources of exchange value compared with their use
value as consumables (see the general suggestions of Kopytoff 1986: 87)? Could we show that
within Tel Yaqush itself, different sectors had different relations to the products of animal
management? Did some slices of the social pie construct and express their status through the
consumption of certain types of animal products—cuisines? Did consumers obtain meat, milk,
fiber, and access to draft power directly from herdsmen, or was there, as Zeder (1991) terms
it, indirect access to animal products, with some elements of the social system depending on
formalized systems of redistribution to obtain their shares? Could we show that these two
aspects of the system, one external and one internal from the perspective of a resident of Tel
Yaqush, developed over time in some manner?

As it turns out, these are extremely complicated and difficult questions. In our view, there
is a significant gap between the high level theory which historians and ethnographers offer to
describe the operation of a Near Eastern pastoral production system and the evidence zoo-
archaeologists offer to illustrate its operation in a particular historical context. The problem is
the familiar one of the lack of effective “middle range theories” that are able to overcome the
uncertainties brought on by the difficulties of equifinality (Trigger 1989: 20–22). As we dis-
cuss below, animal bone samples were produced by the reflexive interaction of a variety of
ancient actors and, setting aside the uncertainties implied by taphonomic processes (Lyman
1994), it is not at all clear a priori which element of the system—production, distribution, or
consumption—a particular zooarchaeological statistic reports.

THE ZOOARCHAEOLOGICAL COLLECTIONS

The comments we offer on these issues have as their core the collections of bones obtained
in the 1989 and 1991 excavations directed by Esse at Tel Yaqush together with a small sample
obtained in 1995 by David Schloen, who now has taken over responsibility for the excavation
and publication of the site under the continuing auspices of the Oriental Institute. Given the
lack of fully elaborated quantitative models of pastoral production, distribution, and con-
sumption, it became apparent in thinking about the project that our examination of the problem
would have to be more inductive and explorative than deductive and evaluative. We were not
likely to find ourselves able to test our suppositions about economic organization; rather we
hoped to refine our expectations about the potential of zooarchaeological evidence while inch-
ing toward a resolution of the larger issues.

Further, we felt that attempting to discern interaction between communities in pastoral
production from the perspective of a single site could prove frustrating. Therefore we expanded
our analysis to include other collections of EB faunal materials. The first possibility we con-
sidered was the sample recovered by the Oriental Institute excavations at Beit Yera



˙



. A large
site located in the immediate orbit of Tel Yaqush, it is an ideal candidate for comparison
and, since the materials were in Chicago, Doug was able to let us examine those bones. Unfor-

long

oi.uchicago.edu





253



ANIMALS IN THE EARLY BRONZE AGE OF NORTHERN PALESTINE



tunately we discovered that the recovery procedures used to obtain that collection, coupled
with difficulties in reconciling stratigraphic problems in the material, made comparison to Tel
Yaqush problematic. Therefore we turned to two other collections—those from the sites of
Tell el-Mutesellim (Megiddo) and et-Tell (Ai). While less ideal comparisons, in that each site
is located at some distance from Tel Yaqush and so was unlikely to have acted directly in the
local production systems of the Jordan Valley, each seemed to be sufficiently different in com-
munity character to provide grounds for comparison. However, before we could compare
these three sets of material, we had to establish rough expectations for our analysis.

ORGANIZATION OF THE PRODUCTION, DISTRIBUTION, AND

CONSUMPTION OF ANIMAL RESOURCES

Many of the texts that have been found by archaeologists in the lands of the Fertile Crescent
record domestic animals and the exchange of pastoral products. While brief, these mentions are
sufficient to permit the reconstruction of some of the components of the production and distri-
butions systems associated with husbandry. Further, works of art illustrate the appearance of
individual animals as well as the ubiquity of herd animals in the cultural landscapes of the past.
These ancient words and pictures combine with the images of day to day activity developed
by Near Eastern ethnographers to attest that raising stock, exchanging pastoral products, con-
suming meat and milk, and utilizing fiber, hides, and the labor of draft animals were core eco-
nomic activities in ancient societies. However, even in the heartland of Mesopotamia, the
textual and artistic record of husbandry and the movement of pastoral products is partial, both
in terms of the kinds of communities and animal production systems whose activities are
reported and the historical periods and places for which we have documents (see, e.g., Postgate
1975; Zeder 1994; and a number of articles in



Domestic Animals of Mesopotamia



, Parts I
and II, which appeared as Vols. VII [1993] and VIII [1995] of the



Bulletin on Sumerian Agri-
culture



). We know much more of the major cities and the herding operations of temples and
palaces than we do of the flocks and herdsmen of smaller communities. One effect of this
imbalance in the available information has been to shift attention to the most direct source of
evidence about past pastoral economies: the animal remains found on archaeological sites.

In reconstructing past animal production systems, zooarchaeologists have emphasized
three variables—the relative abundance of the taxa, the mortality or culling pattern associated
with each type of animal, and the spatial distribution of different kinds of bones (for an
extended example, see Wapnish and Hesse 1991 or Zeder 1991). For example, Redding (1981,
1993) observed that the ratios of sheep to goats and sheep/goats to cattle that are kept by pas-
toralists is shaped by their perception of risk and the nature of the surrounding agricultural
system (see also Rosen 1986), as well as the character of the environment in which the ani-
mals are raised. Payne (1973) provided a series of idealized sheep and goat age curves that
he links to three of the goals of pastoral production—dairy, meat, and fiber—while Cribb
(1984, 1985, 1987) illustrated how herd growth rates condition these models. Hellwing and
Gophna (1984) argued that the degree of complexity in the sociocultural system that moves
animals from pasture to table is reflected in the spatial disposition of different parts of animal
carcasses across a site, a theme expanded upon by Zeder (1991), while Grantham (1992) has
contended that these distributions are shaped by the cultural values collected under the con-
cept of cuisine. Doubtless everyone is right, the question is when to apply which index.

oi.uchicago.edu





254



BRIAN HESSE AND PAULA WAPNISH



To appreciate these and other zooarchaeological indicators, each has to be seen in the
context of a general model of animal exploitation (see fig. 15.1)—a version of which we
present schematically here. The work of exploiting animals is differentiated along three
axes—temporal, spatial, and social. The temporal axis expresses the vector of production.
Animals must be recognized as useful, then acquired, assigned relative value, selected for the
production of various resources, and managed. Once their useful lives as sources of secondary
products (milk, fiber, labor, and young) are deemed complete, animals enter their final circuit
of exchange where they are slaughtered, butchered, and prepared as meals, with skeletal refuse
discarded at each step. Depending on the animal, the products desired, and the technology
available, for a single creature this process may last from months to years. For a male goat it
may take only six months, while a female camel may be in the system for more than twenty
years. The pace is accelerated or slowed by the social and economic situation in which herds-
men and the consumers of animal products find themselves: do they need to “sell” the animal
now to meet pressing needs, or can that decision be put off to a more advantageous moment.

The spatial axis expresses the fact that production, exchange, and consumption may be
contained within the same local circle or extend to a chain of specialists and middlemen



FIGURE 15.1. Schematic model of the organization of pastoral production. Three axes of variability are high-
lighted—a temporal one that records the movement of individual animals through the system from birth to slaugh-
ter and final consumption, a spatial one that describes the range of possibilities for the distribution of animal-
related activities across the landscape, and a social one that expresses the possible relationships that might exist
between the human actors in the system.

oi.uchicago.edu

[image: image83.tif]



255



ANIMALS IN THE EARLY BRONZE AGE OF NORTHERN PALESTINE



including herdsmen, drovers, slaughterers, butchers, and cooks who are spread across commu-
nities and even over huge territories. Each link in the chain may be motivated by a wide range
of ideological, economic, or pragmatic factors. A striking example of the scale to which these
systems can grow is provided by Altorki and Cole (1989) who report that animal trading
systems established by Arabian extended families were able to shift large herds of animals
between the main cities of Syria and Arabia to take advantage of shifting market opportunities.

The social axis reflects the fact that the social and political relations between the various
human elements and tasks in the system may be egalitarian, heterarchical, or hierarchical and
involve mechanisms of kin-based obligation, tribute, or market relations. The social axis is
complexly reflexive—herdsmen are not free to raise what they want, they answer to the mid-
dlemen and consumers down the temporal line even if those individuals are members of their
own households. Similarly cooks must deal with the realities of what is possible, altering the
scheduling and form of the meals they prepare in response to the carcasses available to them.

The structure of human communities within which Near Eastern animal husbandry is prac-
ticed can be presented schematically (fig. 15.2). Three primary elements are identified here—
cities, villages, and nomadic groups—though it is obvious that these exemplars reflect points



FIGURE 15.2. Schematic model of the primary loci of pastoral production in a generalized Near Eastern system.
Dotted and broken lines represent the flow of animals and animal products between the three basic types of com-
munities in the system. Cities and towns not only receive products from nomadic groups, but they generate them
from their own tethered transhumant groups, and from close-herded animals maintained near to the settlements.
(T = temple; S = shrine)

oi.uchicago.edu

[image: image84.tif]



256



BRIAN HESSE AND PAULA WAPNISH



on continua of sedentism, size, and the significance of land or animals in energizing social
relations (Ingold 1987) rather than neatly bounded categories. Animals are raised by nomads
who use their products both for domestic consumption and for exchange with settled commu-
nities. Some authorities have argued that specialized nomadic pastoralism is a precipitate of
the emergence of large agricultural communities (Lees and Bates 1974), while others suggest
that nomadism is a lifestyle adopted during periods of agricultural collapse (LaBianca 1990).

We suspect that these two types of mobile herding should be separated analytically, given
the fundamentally different economic orientations to intercommunity dependence each model
presupposes. In the first case, pastoral nomadism is understood as a husbandry specialization
extruded from and rewarded by the complex agro-pastoral systems based on irrigation and
other intensive techniques that emerged in the fourth and third millennia



b.c.



Capital invest-
ment in the fields surrounding towns and cities pushed pasturage further and further to the
periphery in order to rationalize land use. While the social and political ties to sedentary com-
munities held by mobile groups became somewhat attenuated (though not lost), distant, and
formalized, economic interdependence remained strong. Perevelotsky (1986) has shown that
the type of animals exchanged by this type of nomadic pastoralist is strongly influenced by the
quality of the land they are able to use as pasture and their relative wealth. This is directly
expressed in the age of marketed/delivered stock recorded at urban centers. Only the most
favorably situated pastoralist can afford to take the risks associated with fattening animals,
which involves maintaining them into early adulthood, even when the “market” rewards for
this practice are high (Barfield 1981: 98). Usually, in fact, these herdsmen are those who have
not adopted the nomadic lifestyle at all but are husbanding stock as a sideline in conjunction
with agricultural activities. Marginal producers, whether mobile or sedentary, move animals to
“market” as soon as they are salable (often in their first year of life), in order to meet pressing
social and financial obligations. These findings appear to be examples of what Stein (1992),
following Sandford, describes as the contrast between opportunistic and conservative herding
strategies. The general point is that the kind of stock flowing along the dotted lines in figure
15.2 linking the nomads with the villages and cities will be constantly varying with the shift-
ing economic fortunes of the pastoralists as well as the relative value/price assigned animals
of different type in the sedentary communities.

Pristine ethnographic cases of the economic behavior of the second type of nomadic pas-
toralist, those that adopted a mobile lifestyle as a result of the collapse of sedentary systems,
are hard to find. On the economic grounds established by Cribb’s simulations (1984, 1985,
1987), however, we can predict that these communities will pursue dairy production, given the
fact that this alternative yields the highest energy return of the basic three pastoral strategies.
An additional observation that can be made, based on a near example, is one provided by
Shahrani (1979). He describes herding communities who occupy high altitude pastures in the
panhandle of Afghanistan and have only limited contact with settled communities and little
possibility of practicing agriculture in their habitat. While, of course, it is not reasonable to
generalize from a single case, it is interesting to note that here the outcome of long-term rela-
tive isolation from sedentary communities in an environment of climatic uncertainty is the con-
centration of great numbers of stock in the hands of a few extremely successful individuals,
rather than some kind of economic leveling effect that equalizes outcomes across the various
pastoral enterprises. Once amassed, this is capital available for investment in political as well

oi.uchicago.edu





257



ANIMALS IN THE EARLY BRONZE AGE OF NORTHERN PALESTINE



as economic ventures. This would imply that the retreat into a nomadic pastoral lifestyle as a
response to agricultural reverses actually sows the seeds for reentry into sedentism.

The visual drama of the nomadic pastoralist has deflected attention from the animals
raised by agricultural communities. Not only are draft animals maintained at these settlements
but close-herded stock, those animals which, as milk producers or for other reasons, require
daily attention. Domesticates, like pigs, which are not suitable for transient lifestyles, are hus-
banded there as well. Labeled “Fold” in the figure, these are immediately available sources of
animal products. The need to invest in these urban production modalities is inversely related
to the ability of a sedentary center to regulate its pastoral inputs from a network of external
suppliers. Towns and villages also may maintain transhumant pastoralists, often groups of
men and boys drawn from the community, who herd village stock in distant pastures “owned”
by the village. The need to develop such a strategy is often tied to the development of a ring
of intensive agriculture around the settlement and the resulting conflict with the needs of pas-
turing herds. The links describing the flow of animals between these satellite nodes of produc-
tion and the communities to which they are tied are described by solid lines to reflect the
tethered quality of the social relationships. A broken line is used to describe the link between
village and city. This choice of graphic symbol is meant to indicate the variable nature of the
tribute obligations or the market opportunities that may obtain between the two. As Fernea
(1972) has warned, the strength of the economic or political relationship between adjacent
communities may not be indicated by their physical proximity as models of locational geog-
raphy might suggest. In any case, the key point is that as villages are incorporated in the politi-
cal systems dominated by cities, they tend to act more as conduits than end points in the flow
of animals and pastoral products.

As can be seen from the outline provided here, the flow of animals between communities
may be extremely complex and multiply motivated. Even if we were able to provide one,
which at this point in the development of our knowledge of pastoral systems we cannot, it
does not require a computer simulation to recognize that similar distributions of animal
remains at a single site may be generated by different combinations of inputs and outputs
between the nomadic, transhumant, and sedentary communities tied into a system. As a result
the organizational information embedded in the zooarchaeological remains found at a single
site is difficult to characterize unambiguously. In cases where most products are obtained from
community-based herds and few animals are either exported or imported, it may be possible
to read the production goals of the herdsmen from the species proportions and mortality in-
formation recovered. That is why most of the quantitative zooarchaeological models (e.g.,
Redding 1981) of Near Eastern pastoral production that have been developed are couched in
the domestic mode of production. More qualitative models—which predict the direction of
change in zooarchaeological variables but do not provide quite such specific numerical targets
(e.g., Grantham 1992; Redding 1991; Zeder 1991)—have been useful in examining an emerg-
ing urban end point in a system, though it is dangerous to presume that just because a site is
large and politically significant, it does not engage in the export of animals to even more
powerful centers (Wapnish 1993, 1996).

For villages and towns, communities that might be expected to engage in relaying animals
produced by their satellites and nomadic partners to the cities to which they were tied politi-
cally, the interpretive problem is exceptionally confusing. This is true for two interlocked

oi.uchicago.edu





258



BRIAN HESSE AND PAULA WAPNISH



reasons. First, because we sample an unknown proportion of an original amount of deposited
material that has been subject to taphonomic processes, we never have measures of the actual
total amount of resources that were available to the community in some period of occupation
at a site, but only estimates of relative abundance. Second, it is not clear in most situations
whether a particularly salient datum in the list of species abundances or a particularly high fre-
quency of animals of a particular age reflects the import/choice of animals of that species and
age or the residue left when most of the other species or age categories have been exported.
For these reasons, it is probably premature at this point to try to characterize the production
goals of smaller sedentary communities from zooarchaeological remains, though we can
examine the consumption choices their residents made and perhaps the distribution systems
that supported them.

Within-community differentiation in access to animal products is represented in figure
15.2 by the circled T, for temple, in the city, and the circled S, for shrine, in the village. This
choice of symbol is tied to the observation, based on ethnohistoric documents, that animal
products, often in the form of sacrifices, circulated through religious institutions in ancient
Near Eastern communities. While temples and shrines may be over identified by archaeolo-
gists given the scanty architectural or artifactual evidence usually available, the existence of
strong contrasts in the size and quality of different buildings within a settlement can reason-
ably be taken as indicative of some degree of social and political differentiation. So here,
temple and shrine are taken as shorthand for the existence of hierarchical social complexity,
the existence of public and private space, and the possibility of differential access to pastoral
products, even if a textual record like the one that supported our analysis of the Tel Dan Iron
Age material (Wapnish and Hesse 1991) is not present.

Given this multidimensional model, the initial part of the task of analyzing zooarchaeolog-
ical remains from sites such as the three we report here is coming to some conclusion about
which portion or portions of the spatial-temporal-social field, within which animal exploita-
tion is accomplished, is represented by the collection at hand. Thus we turn to a description of
the character of the archaeological contexts that produced the samples.



Tel Yaqush



Tel Yaqush is located in the eastern Lower Galilee near two tributaries of the Jordan River.
During the EB it was an agricultural village on a main north-south artery connecting Canaan
with Syria. Excavations by the Oriental Institute of the University of Chicago were conducted
at the site in 1989 and 1991 under the direction of Douglas L. Esse (see Esse 1993 for a basic
description) and in 1995 by David Schloen who excavated Square H6 (fig. 15.3). Faunal
remains were recovered from most of the areas excavated.

The bulk of the animal bone fragments was assigned by the excavators based on ceramic
associations and stratigraphic considerations to one of the three main periods of EB: I, II, or
III. Some material was assigned to indeterminate or mixed periods, such as EB II/III, but these
relatively few remains are not reported in this study. The faunal distributions by square for EB
I, II, and III are found in table 15.1. A listing of finds by species and period of occupation is
provided in table 15.2.

Based on Esse’s report (1993) we can assign some of the material presented here to phases
within each of the three main periods of occupation. EB I, the earliest occupation at the site,

oi.uchicago.edu





259



ANIMALS IN THE EARLY BRONZE AGE OF NORTHERN PALESTINE



was distinguished by (at least) four architectural levels. The largest concentration of faunal
material, 1,392 specimens, derives from the earlier phases of EB I, from a long building
(11+ m) in Square P14 at the southeastern edge of the mound. Later EB I remains were found
at the crest of the site (Square H5) and again at the southeast edge (square T14). They date to
the end of EB I, ca. 3200



b.c.



Square T14 contained only thirty-two bones, but from the
summit, Square H5 yielded a substantial number (167) of bone fragments. Both areas showed
evidence of a very destructive fire. Approximately 10% of the bones from Square H5 were
burned, a high proportion for any given subsample, and certainly the highest percentage of
burned material from any area at Tel Yaqush. Square J6, also on the summit of the mound,
yielded 196 bones, but Esse does not mention it when describing the burned areas, and indeed,
no burned bones were found in this square. A large number of bones, 697, were recovered
from Schloen’s excavation in Square H6, but since we can only date these to EB I in general,
they should not be attached to the phasing based on Esse’s report. We do use them, however,
in the total counts when comparing the three subperiods of the EB. Based on the ceramic evi-
dence Esse (1993) suggested that Tel Yaqush expanded most rapidly during the later EB I,



FIGURE 15.3. Plan of the excavations at Tel Yaqush. The areas excavated by Esse in 1989 and 1991 are indicated.
An additional square, H6, was excavated in 1995 by Schloen. In this analysis, the materials from the cluster of
squares in columns H, I, and J at the top of the mound were referred to as the Northwest Group, those in columns
P, Q, R, T at the margin of the site as the Southeast Group, and those in columns K, L, and M as the Central
Group.

oi.uchicago.edu

[image: image85.tif]



260



BRIAN HESSE AND PAULA WAPNISH



reaching its greatest extent of six acres in the block of time marked by the transition between
EB I and II.

The faunal remains dated to EB II were recovered from two adjoining squares, P14 (906
bones) and Q14 (444 bones), on the southeastern edge of the mound. A house, small forecourt,
and a street that was in use over a fairly long time suggest a domestic setting, with the bones
perhaps representing discard from household meals. Wood-charcoal samples from the fore-
court date to ca. 3100



b.c.



EB III Tel Yaqush was a small village covering one to two acres. Excavations in two areas
of the site uncovered (at least) three successive architectural phases. An area on the southern
edge of the mound was the most intensively excavated and yielded the greatest concentration
of faunal remains: squares P13 and 14 (each with 205 specimens) and Squares Q13 and 14
(116 and 103 specimens respectively). Three successive architectural levels were uncovered
in this area, the earliest consisting of a complete house and courtyard. In the same area, Square
Q12 (239 faunal specimens) appeared to be distinct from this earliest house/courtyard and
belonged to the following architectural phase. It contained portions of a large building with a
substantial foundation and wall, large paving stones, and several architectural modifications.
Ten votive juglets and two votive bowls were among the finds from the building’s later phase
of use, on the basis of which Esse (1993) suggested that it had served as a shrine. In the latest
phase of EB III, the architecture on the southern mound was again domestic in character.
Accordingly, the material from squares P13/14 and Q13/14 belong to the earliest phase, and
the Q12 fauna to the second phase; we had no bones from the third phase. Squares K9 (429
bones) and L9 (105 bones) were the main units of the second area of the site with excavated
EB III remains. The scale of the building and street and the effort expended to modify this sec-
tor suggested to Esse (1993) that the area served a public function.



Table 15.1.



Distribution of Tel Yaqush Faunal Remains by
Early Bronze Age Period and Square within the Site.



Tel
Yaqush

H I J K L M P Q R T



5 I: 167 I: 30 — — — — — — — —

6 I: 697 — I: 196 — — — — — — —

9
—
—
—

—
—
—

—
—
—

I: 3
II: 3
III: 429

—
—

III: 105

—
—

III: 15

—
—
—

—
—
—

—
—
—

—
—
—

12 — — — — — — — III: 239 — —

13
—
—
—

—
—
—

—
—
—

—
—
—

—
—
—

—
—
—

I: 14
II: 23
III: 205

—
—

III: 116

—
II: 1
III: 50

—
—
—

14
—
—
—

—
—
—

—
—
—

—
—
—

—
—
—

—
—
—

I: 1,392
II: 906
III: 205

—
II: 444
III: 103

—
II: 5
III: 20

I: 32
II: 11
III: 9

15 — — — — — — I: 1 — — —



a bit short

oi.uchicago.edu





261



ANIMALS IN THE EARLY BRONZE AGE OF NORTHERN PALESTINE



Table 15.2.



Abundance of Taxa in the Faunal Remains, 1989, 1991, and
1995 Excavations at Tel Yaqush



Tel Yaqush
EBI

No. % %
EBII

No. % %
EBIII

No. % %



Cattle 100 19
TNF 23
MNI 30
RF 23

91 23
TNF 23
MNI 30
RF

113 32
TNF 34
MNI 36
RF 36

Sheep & goats
(Including next two
categories)

343 66
TNF 77
MNI 70
RF 77

292 73
TNF 76
MNI 81
RF 76

215 62
TNF 66
MNI 64
RF 64

Sheep 17 71 18 72 14 74

Goats 7 29 7 28 5 26

Pigs 22 4 4 1 4 1

Equids 1



<



1 — — 1



<



1

Gazelles 18 3 2 1 2



<



1

Deer 17 3 2 1 5 1

Total large mammals 501 96 391 98 340 97

Dogs 1



<



1 1



<



1 5 1

Foxes 1



<



1 1



<



1 1



<



1

Birds 2



<



1 1



<



1 1



<



1

Turtles 1



<



1 — — 1



<



1

Fish 8 2 3 1 1



<



1

Crabs 7 2 — — 2



<



1

Total minor species 20 4 6 2 11 3

Total species 521







397 351

Medium mammal 918 74 370 75 441 67

Large mammal 323 26 123 25 217 33

Total Identifiable 1,762 70 890 65 1,009 74

Scrap 763 30 488 35 361 26

Grand total 2,525 1,378 1,370

oi.uchicago.edu





262



BRIAN HESSE AND PAULA WAPNISH



Ai (et-Tell)



Shortly before his death, Joseph A. Callaway gave us a small collection of faunal remains
that he had excavated from EB strata at et-Tell, a site located to the east of Bethel in the central
hill country and generally accepted as the biblical city of Ai (Callaway 1972). Et-Tell began
as a small, unfortified village in the EB IB period (Callaway [1993] dates this to 3100–3000



b.c.



; in a later reworking of the chronology, Cooley [1997] dates EB IB to 3250–3100



b.c.



).
A dramatic change in the character of Ai occurred during the EB IC period (Callaway: 3000–
2860



b.c.



; Cooley: 3100–2950



b.c.



) when the site became a planned, walled city with massive
fortifications that included four city gate complexes, an acropolis, citadel and sanctuary, and
designated market and residential areas (Esse [1984: 323–24] assigns EB IC to EB II). This
settlement ended in destruction that was evidenced by a thick ash layer. During the EB II
period (Callaway: EB IIA–B, 2860–2720



b.c.



; Cooley EB II, 2950–2775



b.c.



), the earlier city
was rebuilt and modified, albeit as an inferior version of its predecessor. The massive destruc-
tion of the EB II city may have been the result of an earthquake that spawned a widespread
fire. The city and the entire fortification system were again rebuilt in EB III (Callaway: EB
IIIA–B, 2720–2400



b.c.



; Cooley EB III, 2775–2240



b.c.



), apparently an effort undertaken
with Egyptian involvement. This city also ended in a violent destruction, and the site was
abandoned for more than a millennium.

The faunal remains we analyzed come from EB IB–C, IIA, and IIIA, according to the list-
ings provided in Callaway 1972: table 15.3. Two phases of the EB IB are represented at the
site. Phase I is a series of fill layers and erosion deposits (Callaway 1972: 59). In phase II, still
the pre-urban period, the bones come from in and around a hut-like structure (Callaway 1972:
62). Faunal remains from the first urban settlement, EB IC (Esse: EB II), derive from building
C inside the fortification wall (Callaway 1972: 104), which Callaway (1972: 105) suggested
was the living quarters for the keepers of the citadel and therefore at the “hub” of city life.
The EB IIA faunal remains were recovered from building B, home to a family or clan whose
daily activities revolved around a courtyard (Callaway 1972: 149). Building B was located
far from the bustle of city traffic (Callaway 1972: 150). Finally, our material which is dated
to EB IIIA derives from a domestic complex of two living units and may be attributed to
Egyptian rebuilding activities (Callaway 1972: 253). Although major public, official, and cul-
tic contexts characterize et-Tell’s cities, all the animal bones we studied were from domestic
deposits.



Megiddo



Tell Megiddo (Tell el-Mutesellim), a large mound in the eastern Jezreel Valley, was the
site of many large cities in antiquity. Numerous archaeologists have excavated its ruins since
work began early in the twentieth century, producing an extremely complicated set of records,
reports, and interpretations. In 1994 large-scale excavations were renewed at the site in a cam-
paign jointly directed by Israel Finkelstein and David Ussishkin of Tel Aviv University and
Baruch Halpern of Pennsylvania State University. The EB occupation they uncovered is part
of the huge cultic complex at the east and northeast sectors of the mound (referred to in pub-
lications as Area BB) that had been discovered by earlier projects. Architectural and artifac-
tual remains from the Chalcolithic period suggest that the area was used for cultic purposes in

short

oi.uchicago.edu





263



ANIMALS IN THE EARLY BRONZE AGE OF NORTHERN PALESTINE



that earlier period as well. The systematic collection of faunal remains was not a part of the
first excavation efforts, although we may surmise from a few published notes and other reports
that such remains were abundant. For instance, reports of the excavations (1925–1939) by the
Oriental Institute of the University of Chicago, an effort which unearthed the large circular



* Esse (1984) considered this period EB II. The original designation, EB IC, is retained here for consistency with
the site reports.



Table 15.3.



Distribution of Faunal Remains from Ai



Ai EB IB



No. % %



EB IC*



No. % %



EB II



No. % %



EB III



No. % %



Cattle 53 21
TNF 22
MNI 12
RF 18

28 20
TNF 20
MNI 23
RF 18

54 12
TNF 14
MNI 10
RF 13

14 12
TNF 14
MNI 14
RF 14

Sheep &
goats
(Including next
two categories)

187 75
TNF 78
MNI 88
RF 82

110 78
TNF 80
MNI 77
RF 82

397 86
TNF 86
MNI 90
RF 87

100 84
TNF 86
MNI 86
RF 86

Sheep 32 74 12 60 44 58 9 60

Goats 11 26 8 40 32 42 6 40

Pigs 1



<



1



— —



1



<



1



— —



Equids 2



<



1 2



<



1 1



<



1 1 1

Gazelles 1



<



1 1



<



1 4 1 2 2

Total large mammals 244 98 140 99 457 99 117 98

Dogs 2 1 2 1 1



<



1 1 1

Bear 1



<



1



— — — — — —



Bird 1



<



1



— —



1



<



1 1 1

Turtle 1



<



1



— — — — — —



Total minor species 5 2 2 1 2 1 2 2

Total species 249 142 459 119

Medium mammal 81 80 30 65 134 80 38 95

Large mammal 21 20 16 35 35 20 2 5

Total ID 351 97 188 95 628 96 159 92

Scrap 12 3 10 5 28 4 4 8

Grand total 363 198 656 163

oi.uchicago.edu





264



BRIAN HESSE AND PAULA WAPNISH



altar (4017), noted that the surrounding area contained much broken pottery and animal bones
(Loud 1948).

A large volume of faunal remains was recovered from the cultic area during the 1994 and
1996 field seasons, but here we consider only the 1994 material for which analyses are com-
plete (table 15.4). After the 1996 field season the excavators proposed a provisional phasing



Table 15.4.



Distribution of Faunal Remains from Megiddo



Megiddo Temple XIX
No. % %

Monumental
No. % %

Squatters
No. % %



Cattle

Articulations

6 7
TNF 8

87 15
TNF 16
MNI 26
RF 16/27

Art = 1

25 8
TNF 8
MNI 18
RF 6/20

Sheep & goats
(including next two
categories)

Articulations

71 87
TNF 92

Art = 2

475 83
TNF 84
MNI 74
RF 84/73

Art = 2

273 89
TNF 92
MNI 82
RF 94/80

Art = 9

Sheep
Articulations

2 67
Art = 1

6 75
Art = 1

29 74
Art = 5

Goats
Articulations

1 33 2 25 10 26
Art = 2

Pigs 1 1 — 3 1

Deer 1 1 1



<



1 1



<



1

Total large mammals 79 96 563 98 302 98

Rodents — — 1

Birds 2 5 3

Turtles 1 — 1

Fish — 7 1

Crabs — 1 —

Total minor species 3 4 13 2 6 2

Total species 82 576 308

Medium mammal 105 93 997 87 398 84

Large mammal 13 7 141 13 77 16

Total Identifiable 200 54 1,714 50 783 65

Scrap 165 46 1,700 50 433 35

Grand totals 365 3,415 1,216

oi.uchicago.edu





265



ANIMALS IN THE EARLY BRONZE AGE OF NORTHERN PALESTINE



of the occupations in the cultic area (see Wapnish and Hesse 2000). We have sufficient ma-
terial to warrant analyses for three of these phases, all of which date to EB I. The earliest
material comes from the Stratum XIX temple, in which an earlier and a later phase were rec-
ognized. We had to combine the material from both phases for a large enough sample to
manipulate, but in fact most of the fauna date to the earlier phase (328 and thirty-seven bones
respectively). The largest EB I faunal subsample (3,415 bones) found in 1994 is stratigraphi-
cally subsequent to the Stratum XIX temple and derives from surfaces and surface buildup
that accumulated between a set of parallel and monumental walls. This characterization of the
feature is purposely left vague since a clearly delineated “building” has yet to be exposed.
Nevertheless, the massive size and construction of the walls is incontestable and clearly
implies a context of specialized public and probably ritual function. The third EB I subsample
(1,216 bones) in our collection also comes from the area of monumental construction and is
superposed on the earlier surfaces. However, the existence of small wall stubs abutting the
walls of the massive structure and other features have led the excavators to characterize the
phase as produced by “squatters” who built a number of insubstantial enclosures between the
earlier monumental EB I construction. Presumably at some time before, and during this phase,
the area had ceased to function as a cultic center. We employ the terms “Temple XIX,” “Mon-
umental,” and “Squatters” to refer to these three deposits in the tables that accompany this
report.

The character and quality of the bone material found in the Monumental and Squatter
phases of occupation contrast in ways that suggest a microbehavioral distinction. In the Mon-
umental phase much of the recovered bone is scrap, broken bits that cannot be assigned to a
particular anatomical element (e.g., femur, rib) or to a taxon more descriptive than “animal.”
The bones are very fragmentary, a condition often characteristic of surface or floor deposits
which we call the “trampling effect.” Many of the identifiable fragments are from sheep/goat
heads, with fewer bones from meaty parts of the carcass. Only two articulations are present—
that is, the recovery of two or more bones from the same individual in correct anatomical posi-
tion. Articulations are important indicators that bones were not disturbed after initial discard.
This is significant because redeposition in tell sites is the single most significant factor in
clouding the chronological assignments given animal bones since with few exceptions the
bones found in historic period tell sites in the Middle East are not datable by intrinsic
attributes (Hesse and Rosen 1988). The recovery of only two articulations points in the
direction of increased disturbance, either from activities that took place in the area, or as an
indication that bones were redeposited there as secondary refuse from activities that occurred
elsewhere. Indeed, the latter may be inferred from the substantial number of burned bones in
this subsample because the burning does not appear to have taken place in situ. We conclude
therefore that the Monumental deposit represents an accumulation of a series of small second-
ary dumpings, perhaps removed from floors elsewhere.

On top of and to some extent intermingled with this subsample was a collection of micro-
fauna (small birds, rodent-like mammals, small frogs, and lizards) from owl pellets, the regur-
gitated remains of owl (or raptor-like bird) meals. The archaeological significance of such
deposits is that the site of collection might have been unoccupied for a time, because owls do
not usually roost in the center of human activity. The area where this collection was found was
unused or deserted for a period following its earlier active phase. Accordingly, this second
time frame of the EB I period has informally been dubbed the “owl” level.

oi.uchicago.edu





266



BRIAN HESSE AND PAULA WAPNISH



The 1,216 faunal specimens found in the Squatters phase contrast in several important
ways with those of the earlier phases when the cultic center was in operation. Primary among
them was the recovery of nine articulations and substantially less scrap bone (35% rather than
50% of the total subsample) than in the Monumental phase, both evidence that the Squatters
deposits were transformed less by postdiscard factors. A complete lower jaw (mandible) from
a cow found in the deposits of this phase provided the informal moniker “cow” level. The fact
that the jaw was intact further suggests little disturbance to this portion of the deposit because
the symphysis, or mandibular section which forms the chin, is a weak area of the bone that is
often broken off. Taken together, these features suggest a more “in situ” deposit of material
discarded near the area of primary activity.

The nature of the architecture of the earlier phases of the EB I period have led us to
assume that Megiddo functioned as a large cultic center for a considerable portion of the sur-
rounding countryside, commanding a regular input of animals to be offered up as sacrifice. We
may expect that animal remains found in temple precincts will differ from their non-cultic
counterparts in socially complex environments. The power of a ritual center would have per-
mitted the selection of particular kinds of animals based on the ideological requirements of the
cult. Thus the species proportions and mortality information may contrast with that found in
“secular” urban debris. The expectations for carcass part distributions vary in a different direc-
tion. The usual prediction for city-based animal processing activities (slaughter, butchering,
cooking, eating, and discard) is that they will be dispersed across the community, resulting in
the scattering of the various parts of a single animal carcass across a site (e.g., Hellwing and
Gophna 1984). But in those cultic complexes where the animal was usually consumed by
temple personnel in the sacred precinct and carcasses were not redistributed after sacrifice, all
these distinct processes are concentrated in one area. Cultic carcass proportions may closely
mirror “domestic” debris, but we may expect species abundances and age curves to differ from
both small-scale domestic and urban secular contexts given the widely divergent objectives of
the three systems and the constraints limiting their choices.

ANALYSIS



General Features of the Collection



Table 15.5 summarizes the main chronological contexts that we were able to evaluate. The
faunal remains recovered from all the sites were divided into three broad categories of identi-
fiability (table 15.6). In those cases where a bone fragment could be assigned both an anatom-
ical and a taxonomic label, we considered it “identifiable”; where we could only determine it
was part of a limb and either a medium-sized mammal (more or less sheep-sized) or a large
mammal (more or less cow-sized), we called it a “long-bone shaft fragment”; and where only
the fact that it was bone was known, we labeled it “scrap.” In practical terms scrap bone
contains relatively large amounts of cranial, vertebral, and costal fragments. Thus abundance of
these tiny fragments is both a measure of how strong the taphonomic forces acting on a col-
lection were and an estimate of the degree to which all parts of the carcass are present in a
deposit. The ratio of long-bone shaft fragments to identifiable specimens is a measure of the
degree of taphonomic pressure (e.g., the influence of scavenging by dogs) since the diagnostic
end of most bones is also usually less robust than the shafts and survives less readily. In

oi.uchicago.edu





267



ANIMALS IN THE EARLY BRONZE AGE OF NORTHERN PALESTINE



Note: The sample from Ai contains little of this material and might have been presorted to some
degree before it was turned over for investigation.



Table 15.5.



Chronological Contexts of Faunal
Remains in the EBA, Northern Palestine



Site Phase Context



Tel Yaqush EB I Domestic

EB II Domestic

EB III Non-Cultic—Cultic—Public

Ai EB I Pre-Urban—Urban Domestic

EB II Urban Domestic

EB III Urban Domestic

Megiddo EB I Temple—Monumental—Squatters



Table 15.6.



Distribution of Identifiable Fragments, Long Bone Shaft
Fragments, and Scrap Bone in the Subsamples from Tel Yaqush and Megiddo



Find type EB I EB II EB III



Tel Yaqush NW
H5/H6/ Scrap

I5/J6 Long Bone
Ident
No.

41%
32%
27%

1,082

Tel Yaqush CEN
K9/L9/ Scrap

M9 Long Bone
Ident
No.

Public
32%
30%
38%

548

Tel Yaqush SE
P13-15/ Scrap
Q12-14/ Long Bone
R13-14 Ident
T14 No.


27%
38%
35%

1,440


28%
37%
35%

1,390

Cult Non-Cultic
49% 40%
35% 32%
16% 28%

238 571

EB I-
Temple XIX

EB I-
Monumental

EB I-
Squatters

Megiddo
Scrap
Long Bone
Ident
No.


44%
19%
36%

371


50%
20%
30%

3,423


36%
14%
50%

1,245

oi.uchicago.edu





268



BRIAN HESSE AND PAULA WAPNISH



the case of the sample from Ai, few scrap bones or long-bone shaft fragments were found in the
preserved collection. We suspect that this is a result of some degree of preselection by the
excavators in the field since, to our knowledge, intensive recovery methods such as dry-
sieving were not routinely employed during the excavation. Thus the general data from this
site are not provided.

In the Tel Yaqush sample two contrasts deserve note. The EB I sample from the northwest
portion of the excavation and the EB III samples from the southeast have somewhat more
scrap than that recovered from EB I–II southeast. We suspect that this is a result of the former
samples being less deeply buried and so more affected by depositional conditions. The second
set of contrasts relates to the material from the three EB II contexts. The ratio of long-bone
shaft fragments to identifiable bones is much lower in the “cultic” sample. We suspect that this
is due to the fact that, as we discuss below, the animals in the “cultic” sample were slaughtered
at a much younger age and so were represented in the sample by long bones with much less
well-ossified epiphyses, specimens that did not survive well.

One contrast appears in the sample from Megiddo. The Squatters sample contains rela-
tively more identifiable fragments and less long bone and scrap material. This reinforces the
picture drawn above of a collection that had undergone less taphonomic pressure than those
obtained from the two earlier phases of occupation.



Taxa Represented



The number of taxa found in the three samples is low, and most of the remains are concen-
trated in only a few species (see tables 15.2–15.4). The samples were divided into several
broad categories. Large Mammals include hunted or husbanded herbivores gazelle-sized and
larger. Minor Species include the carnivores, rodents, birds, reptiles, fish, and mollusca. In no
case did the minor species contribute more than 4% to the specimens found in one of the ten
chronologically defined contexts at the three sites. Assuming that abundance is a rough mea-
sure of economic significance, these are trivial and so not considered further. The categories
Medium Mammal and Large Mammal are used for those specimens, mostly rib and vertebra
fragments, long-bone shaft fragments, and small cranial fragments that could not be further
identified. Most of these bones are likely to have come from the three abundant taxa in the
sample: sheep, goats, and cattle. The fact that the ratio of Medium Mammal to Large Mammal
mirrors that observed between sheep/goat (this term is used to refer to the combined sample
of bones including those that can be identified as either sheep or goat as well as those that are
indeterminate) and cattle suggests that the various parts of the carcasses of the two taxa were
accumulating in the deposits in roughly similar ways.

In nine of the ten contexts, sheep/goats and cattle provided more than 94% of the sample
of large mammals. The one exception is the EB I sample from Tel Yaqush, where 10% of the
sample comprises pig (



Sus



), gazelle (



Gazella



sp.), and deer (



Cervus/Dama



)



1



bones (see
Uerpmann 1987 for a review of the archaeological distributions of the wild forms of these ani-
mals and the taxonomic complexities involved). The very low frequency of pig is particularly
surprising given the amount of swine reported at such nearby Chalcolithic and EB I sites as
en-Shadud (located not far to the northeast of Megiddo in the Jezreel Valley, where approxi-



1. In a provisional report on the Tel Yaqush fauna cited by Esse (1993), hartebeest was identified. Further com-
parison has shown that initial impression to have been wrong.



long

oi.uchicago.edu





269



ANIMALS IN THE EARLY BRONZE AGE OF NORTHERN PALESTINE



mately 25% swine is reported [Horwitz 1985]), Tell esh-Shuna North (situated across the
Jordan River from Tel Yaqush, which produced about 34–46% in the Chalcolithic and earlier
EB I [Croft 1994]), and Tel Kinrot (a multiperiod tell site northwest of the Sea of Galilee,
where about 8% pig was found in the EB [Hellwing 1988–89]). We have suggested elsewhere
(Hesse 1990) that pigs are a measure of a community’s disengagement from large-scale trib-
utary or “market” systems. Additionally, Redding (1991) has hypothesized that pigs are
dropped from husbanding strategies as agricultural effort is intensified. On the basis of these
models we believe the data from all of our sites indicate the three communities were incorpo-
rated in larger political systems and engaged in a “market”-oriented and specialized farming
system. The drop in the number of pigs at Tel Yaqush from EB I to the later two periods of
occupation may reflect the completion of that process of development and incorporation. Sig-
nificantly, no increase in the use of swine is observed as Tel Yaqush shrinks in size during
its occupation. This (admittedly single case) may suggest caution in equating site size with
degree of political significance or agricultural intensification in a way parallel to Fernea’s
(1972) observation that propinquity is not a certain indicator of political engulfment.

We were able to separate some of the sheep/goat bones by genus following the criteria
provided by Boessneck (1969). Using the total number of fragments identified to each species
as an estimator of abundance, remarkably similar values are found across the three sites. The
bones from Ai EB IC, II, and III, those that accumulated after the city was walled, have
slightly higher numbers of goats, 40%+, compared with the rest of the samples, which have
between 25% and 33%. Redding’s (1993) summary of the evidence for Near Eastern animal
production, while a survey of regions outside the Levant, provides a model for interpreting
sheep-to-goat ratios. The values in our samples, which closely center on 3.5 sheep to 1 goat,
would, following his prescriptions, indicate a lively exchange of animals, particularly sheep,
as commodities between nomadic pastoralists (or transhumant herding camps) and the seden-
tary communities (with perhaps somewhat less of a link between Ai and its surrounding satel-
lites). If production at our three EB sites was decoupled from a larger system and targeted
household consumption, Redding’s model would project that goats would outnumber sheep.
On the other hand, the value of 3.5:1 is not high enough to suggest that the surrounding
pastoral producers were engaged in some sort of specialized wool production orchestrated by
permanent settlements, which themselves did little local husbandry. The fact that the ratios
for Megiddo and Tel Yaqush are so similar suggests that despite the considerable difference in
the scales of the three sites, the larger was not significantly more dependent on external spe-
cialized producers.

The ratio of abundance of the combined sheep/goat collections to the cattle sample was
computed in three ways (Hesse and Wapnish 1985). Total Number of Fragments (TNF) is an
estimation based on the sum of bone fragments assigned to a taxon. As a measure of abun-
dance it suffers from the usually unmeasurable effects of interdependence (when more than
one bone in the collection came from the same individual animal). Minimum Number of Indi-
viduals (MNI) is established by the most frequent element type in the collection of bones
assigned to a taxon. It establishes the smallest number of carcasses that would be required to
provide the bones recovered. It tends to produce high estimates for rare taxa and is affected by
the way the spatial area represented by the sample is established. Relative Frequency (RF) is
computed by dividing TNF by the number of bone fragment categories in the sample. It
attempts to correct for the fact that different animals have different numbers of bones in their

oi.uchicago.edu





270



BRIAN HESSE AND PAULA WAPNISH



skeletons. If the estimates produced by these three counting methods produce sharply diver-
gent results, it usually means that the carcasses of the various species were processed differ-
ently by the ancient inhabitants of the sites.

Within each set of subsamples there is close agreement between the three estimators
except in the case of the Megiddo Squatters. We suspect that the range of 8% to 20% for the
cattle in that period is an artifact of the very small sample size involved. The results for the
sheep/goat-to-cattle computations show more variability than that seen in the sheep-to-goat
ratio. At Tel Yaqush there is an increase in cattle between EB II and III, while at Ai the
number of bovines decreases between EB II and III. The number of cattle at Megiddo EB I is
generally lower than found in the other sites, although in the Monumental phase at Megiddo
EB I, it reaches the lowest values found in the other two sites.

All of the sheep/goat-to-cattle ratios are within the limits, between 1:1 and 10:1, suggested
by Redding (1993) for intensive agriculture (see also Rosen 1986, for a similar estimate
arrived at from different assumptions and data). Further, within the Megiddo samples, the con-
trast between the Monumental sample (5:1) and that from the Squatters (9:1) fits the expecta-
tions of the model as well. However, what also is striking about the data is that in general the
larger the site, the higher the ratio, the reverse of expectations. Taken at face value, this would
suggest that massive, specialized, urban sites such as Megiddo were supported by less inten-
sive agricultural systems than that which provided resources for a village like Tel Yaqush.

However, there are at least three other alternatives. First, the difference may reflect the local
environmental settings of the three sites. Sheep and goat pastoralism in the Jordan Valley in the
immediate environs of Tel Yaqush might have been a less attractive strategy than husbanding
the animals at the margin of the Jezreel near Megiddo or in the central hill country at Ai. If it
was, the number of animals produced in the Tel Yaqush community folds would be reduced.
Also, the nomadic groups or transhumant camps would be situated at a greater distance, a fac-
tor that would reduce the level of contact between mobile supplier and urban customer. Both
conditions would depress the sheep/goat-to-cattle ratio. Second, the apparently larger number
of sheep and goats in the urban centers may be related to internal differences in access to ani-
mals between sectors of those communities. That appears to be the case at Tel Yaqush in EB
III (table 15.7). Comparison of the Public, Cultic, and Non-Cultic areas shows that the ratio is
about 1.5:1 in the first two of these areas while nearly 3:1 in the Non-Cultic area. The samples
we have from those large settlements are not very diverse, and excavation in domestic housing
at Megiddo or the temple complex at Ai might produce the kind of differentiation we found at
Iron Age Dan. Here the temple debris could be distinguished from that found in contemporary
dwellings on the basis of the sheep/goat-to-cattle ratio (Wapnish and Hesse 1991), though in
this case the direction of the contrast is the opposite of that found at Tel Yaqush. If the dia-
chronic contrast between the Monumental and the Squatter phases are taken as a surrogate
model for a synchronic comparison at EB I Megiddo, it suggests that in both the Tel Yaqush
shrine and the Megiddo temple cattle were preferred. Third, the lower than expected value for
sheep and goats at Tel Yaqush may be the result of animal export. If the site was required to
deliver small bovids to a higher node in the political hierarchy or there was a market to be
tapped, this demand would depress the proportion of animals available for local consumption.
At this point it is not possible to make a clear distinction between these alternatives.

short

oi.uchicago.edu





271



ANIMALS IN THE EARLY BRONZE AGE OF NORTHERN PALESTINE



Carcass Parts



The manner in which the various parts of animal carcasses are found at archaeological
sites has been taken as evidence for the complexity of the animal production, redistribution,
and consumption system. Once slaughtered, animals are reduced to packets of meat, fat, hide,
hair, horn, and bone. Each of these resources can follow a separate pathway through the social
economy of a site, and some of these packets are associated with specific skeletal parts. Thus
the fact that various sectors of the community value them differently or have different access
to them can be recorded in the spatial distribution of bone fragments. For instance, the differ-
ent parts of any carcass have more or less meat associated with them. If the elite chooses to
express its power by claiming through economic or other means the meat-rich parts of an ani-
mal (the limbs and trunk) and leaving the spare portions (the heads and feet) to hoi polloi, then
some sectors of the community will produce garbage that is relatively rich in long bones, ribs,



Note: A possible “cultic” area was found in Square Q12. An adjacent area contained “domestic”
occupation while an area in the center of the site might have contained “public” buildings. The
cultic and public material contains more cattle and more medium mammal “meaty” carcass parts
than the domestic debris. The cultic and domestic areas contrast similarly in terms of the large
mammals. The public area contained substantially more “spare” elements of large mammal. The
small samples of ageable sheep/goat teeth present a strong contrast. The cultic area produced
largely young animals while the other two areas contained mostly—and in the case of the public
area, almost all—older animals. These contrasts suggest that even in this small 1- to 2-ha com-
munity, access to animal products was differentiated.



Table 15.7.



Comparison of EB III Tel Yaqush Areas



Tel Yaqush
EB III

Cultic?
Square Q12

Non-Cultic?
Squares P13-14/
Q13-14

Public?
Squares
K9/L9/M9



Sheep and goats 60% 70% 57%

Cattle 38% 25% 41%

Pigs/deer/gazelles 2% 5% 2%

Medium mammal “meaty”
axial + limbs

62% 43% 56%

Medium mammal “spare”
head + toes

38% 57% 44%

Large mammal “meaty”
axial + limbs

67% 56% 46%

Large mammal “spare”
head + toes

33% 44% 54%

Sheep/goat 0–1 yr. 40% 22% 0%

Mortality 1–3 yrs. 30% 28% 10%

Tooth wear 3+ yrs. 30% 50% 90%

oi.uchicago.edu





272



BRIAN HESSE AND PAULA WAPNISH



and vertebrae, while others will, by default, discard fragments of cranial fragments and toes.
While the appeal to efficiency incorporated in this model makes it an attractive approach to the
study of animal remains and we employ it below, it is dangerous to reify the categories of
meat-rich and meat-poor carcass parts. In complex societies, elites can exercise their authority
by simply claiming more of what they want, ignoring the benefits of the principles of econ-
omy and thrift in their pursuit of a distinctive “lifestyle.” As Goody (1982) points out, in the
process of civilization meals are ever more elaborately translated into the language of cuisine.
As that occurs, the significance of a general category like “meat” recedes as the salience of
“pork,” “beef,” and “mutton” grows.

Unfortunately, the zooarchaeologists of the Middle East have not achieved consensus in
the way they assign different bone fragments to various parts of the carcass. As a result it is
dangerous to compare numeric values across sites although examining the general shape of the
distributions and the way they change may be useful. In table 15.8 we present the information
for Tel Yaqush and Megiddo. Because of the problems of presorting mentioned above, data
from Ai are not reported. In our system, Axial remains are fragments of vertebrae and ribs;
Limbs include shoulder blades and hips as well as the long bones down to the proximal
metapodials (cannon bones); Toes include the distal metapodials and phalanges; and Head
includes skull and mandible fragments as well as teeth. The first two categories are the meatier
carcass elements, the last two are more spare. Two taxa are used—Medium Mammals are
mostly sheep and goats, and Large Mammals are mostly cattle.

In the Megiddo samples, there are proportionately less meaty medium and large mammal
fragments in the Monumental deposit than are found in the Squatters. Some of this difference
may be due to the greater destruction brought on by taphonomic factors in the first sample,
although it may also reflect a greater degree of spatial segregation of the activities of slaughter
and initial carcass dismemberment in the Squatters phase and a greater spatial concentration
of the same steps in transforming a carcass into a meal in the Monumental phase.

At Tel Yaqush there is little difference between the two subsamples of EB I material. How-
ever, there is an important contrast in the comparison of the material from EB I to all of the
later material. In the earlier occupation, meaty parts are slightly more abundant than in the later
phases. However, this comparison conceals the even larger contrast between the abundance of
the two kinds of meaty parts, the Axial elements and the Limbs. In the earlier material, Axial
fragments are from two to four times as abundant, while in the later material, Limbs are either
equal or more abundant. We suggest that this is evidence for the emergence of the spatial seg-
regation of animal processing activities and perhaps ones related to the preparation of different
kinds of meals from different parts of the carcass.

The remains found in the three sectors of the EB III sample also showed differences (table
15.8). The Cultic area produced the most “meaty” portions of both Large and Medium Mam-
mals. The Non-Cultic area, which had the most sheep and goats, also yielded the most “spare”
Medium Mammal. The Public area, in contrast, had the least amount of cattle and the “spar-
est” portions of that carcass type, while the Medium Mammal sample approaches the values
seen in the Cultic area. We believe that these contrasts suggest that the people occupying the
different areas of Tel Yaqush, even in its shrunken state, still were differentiated in the way
they obtained their meat resources.



short

oi.uchicago.edu



273ANIMALS IN THE EARLY BRONZE AGE OF NORTHERN PALESTINE

Mortality

From an archaeological perspective, the decision about when to slaughter animals is the
second most visible decision (after the choice of what kind of stock to husband) in pastoral
management. Efficient husbandry starts with the careful evaluation of the capacity of individ-
ual animals to yield milk, meat, fiber, labor, and young in the context of the market for those
resources. By culling those with low potential, maximum benefit is obtained from the pasture
and forage available to a community. Because of this relationship, zooarchaeologists have
developed methods to reconstruct the mortality pattern associated with each type of stock in
order to gain insight into both the production goals of herdsmen and the market demands of


Note: “C” and “N-C” in the Tel Yaqush SE EB III samples refer to the material from the “cultic” and “non-cultic”
areas respectively. Data from Ai are not included because the small number of rib fragments found in the collection
seems to be a product of presorting the sample.

Table 15.8. Distribution of Skeletal Parts in the Main Periods at Tel Yaqush and Megiddo

Skeletal Parts
EB I

Medium Large
mammal mammal

EB II
Medium Large
mammal mammal

EB III
Medium Large
mammal mammal

Tel Yaqush NW
H5/H6/ Axial
I5/J6 Limbs
Head
Toes
No.

43 54
18 20
34 16
5 10
216 63

Tel Yaqush CEN
K9/L9/ Axial
M9 Limbs
Head
Toes
No.

Public
17 32
38 14
38 49
6 4
81 71

Tel Yaqush SE
P13-15/ Axial
Q12-14/ Limbs
R13-14 Head
T14 Toes
No.


45 63
20 16
31 17
4 4
337 189


24 26
31 36
41 33
4 5
373 111

C N-C C N-C
33 20 20 30
29 23 47 26
24 49 26 37
14 8 7 7
51 107 30 43

Temple XIX Monumental Squatters

Megiddo–EB I
Axial
Limbs
Head
Toes
No.


36 —
29 —
29 —
6 Insuff.
119 Data


45 43
14 10
38 44
3 3
813 152


47 62
23 6
22 27
8 5
534 66

oi.uchicago.edu



274 BRIAN HESSE AND PAULA WAPNISH

customers. Two primary methods are used—the degree of tooth eruption and wear and the
relative proportions of completely mature bones. Deciduous teeth erupt and are replaced by
permanent dentition at predictable times in the life of an animal (Hillson 1986). The rate of
attrition as expressed by the degree of crown reduction and changing patterns of the occlusal
surfaces of cheek teeth has also been employed (Davis 1987). Here the procedures outlined by
Payne (1973) based on tooth wear patterns are used to estimate mortality in sheep and goats.
The various bones mature at different times in an animal’s life as is signaled, for instance in
the case of limb bones, by the shaft or diaphysis completely fusing to the ends or epiphyses
and the metaphyseal line disappearing. While the age at which this occurs is affected by a host
of genetic and environmental factors, useful approximations have been established for the
main types of barnyard stock (Silver 1969). By calculating the proportion of fused ends, we
can estimate the percentage of animals in a cohort that lived past the age at which fusion
occurs. By arranging the fusion percentages at each diaphysis/epiphysis location by the age at
which that fusion is expected to have occurred, a mortality curve is produced.

Table 15.9 presents the information based on tooth wear for all three sites, while tables
15.10–12 are tabulations of fusion scores for Tel Yaqush, Megiddo, and Ai, respectively. Only
information on sheep/goat mortality is presented since cattle remains are not abundant enough
to produce any estimates. Also, there were not enough sheep and goat remains in the Temple
XIX sample from Megiddo to generate an estimate by either method. Sheep and goats are
merged in these calculations despite the fact that the two animals were likely managed in dif-
ferent places for different reasons. This is done because zooarchaeologists have not developed
sufficiently reliable methods for distinguishing the two species at all stages of their develop-
ment and because our samples in all contexts are small.

The site of Ai presents the most consistent pattern. Whether estimated by tooth wear or
fusion, most of the sheep and goats consumed at the site were fully adult, more than two-thirds
died at ages greater than two years, and somewhat more than half were older than three years.
Through time the number of one-to-two or one-to-three year olds declines that may indicate a
slight selection for more well-fattened animals in the latest occupation.

The Megiddo estimates for the Monumental sample and that of the Squatters are quite dif-
ferent. In the latter case, based on tooth wear, almost two-thirds of the sheep and goats were
slaughtered in their first year, while in the former substantially more animals one to two years
of age were recovered. The fusion scores for these two samples are in rough accord with this
pattern. The large kill-off of very young animals in the Squatters phase is consistent with dairy
production or the marketing of stock as soon as it has any market value. The emphasis on
yearlings found in the Monumental phase is consistent with meat production models.

The Tel Yaqush fusion scores indicate a shift between EB I and both later periods. In the
earlier phase slightly more very young animals were being consumed and a large number of
individuals were culled at the end of their second year. In later periods, both of these slaughter
emphases are reduced. On the basis of tooth wear, the most striking feature in a remarkably
similar set of scores is the absence of animals one to two years old. Does this reflect the export
of animals of this age to larger centers? Even more striking is the contrast in the dental evi-
dence between the three contexts in Tel Yaqush EB III (table 15.7). Young animals dominated
the cull in the Cultic area but were conspicuously absent in the Public area. The Non-Cultic

short

oi.uchicago.edu



275ANIMALS IN THE EARLY BRONZE AGE OF NORTHERN PALESTINE

area was intermediate. Once again it would seem that different social sectors of even so small
a site as EB III Tel Yaqush could claim different types of animals as resources.

Taken together, the three sites present quite different pictures. The two large walled sites
contrast most strongly. Megiddo samples include large numbers of young animals, while those
from Ai are mostly mature. Part of this difference seems to be linked to the secular/cultic

Note: Mortality based on the degree of dental attrition (Payne 1973). More young and less old individuals are
found in the Megiddo Squatters sample compared to Tel Yaqush. The Ai sample exhibits the reverse, more
older animals, less younger.
* = EB II (Esse 1984)

Table 15.9. Sheep/Goat Mortality at Tel Yaqush, Megiddo, and Ai

Payne stage
A
%

B
%

C
%

D
%

E
%

F
%

G
%

H
%

I
%

Tel Yaqush
EB I


3 6 19


16


12 8 17 13 6

EB II 1 10 25 12 8 15 17 8 4

EB III 0 8 15 14 17 22 20 3 1

Megiddo EB I
Monumental 0 0 31 29 7 10 12 7 4

Squatters 0 11 56 14 3 3 7 3 3

Ai
EB IB Pre-urban


0


2


8 26 18 14


19


10


3

EB IC Urban* 0 0 10 18 22 22 18 10 0

EB II Urban 0 0 15 19 14 18 19 13 2

Age in years
0 - 1

%
0 - 2

%
1 - 2

%
1 - 3

%
2 +
%

3 +
%

Specimens
No.

Tel Yaqush
EB I


28 44


16 28 56 44


24

EB II 36 48 12 20 52 44 23

EB III 23 37 14 31 63 46 18

Megiddo
Monumental


31


60 29 36 40 33 13

Squatters 67 81 14 17 19 16 9

Ai
EB IB Pre-urban


10 36 26 44 64 46 24

EB IC Urban* 10 28 18 40 72 50 10

EB II Urban 15 34 19 33 68 52 32

oi.uchicago.edu



276 BRIAN HESSE AND PAULA WAPNISH

contrast. The Megiddo Monumental and Tel Yaqush Cultic subsamples are quite similar, sug-
gesting a tradition of young animals being selected to enter sacrificial systems. The Ai urban
domestic samples compare favorably with the Tel Yaqush Non-Cultic remains, suggesting that
the secular areas at both sites received the complement, mostly mature animals. The major
anomaly is the sample from the Megiddo Squatters phase. The large number of young animals
would be understandable if goats were more common in the sample and cattle less so since
that would suggest the community, while squatting in a huge building, was in fact largely
domestic in economic focus. At this point, however, the question is unresolved.

CONCLUSION

At the outset we indicated that this report would be inductive and explorative rather than
deductive and evaluative. Zooarchaeological research in the Near East in general and the

Note: Percentages reflect the number of animals who survived the age category listed in the
left-hand column. Fusion locations are grouped based on the estimates of fusion age given
in Silver (1969). Two linked observations about the Tel Yaqush samples are most signifi-
cant. The number of young animals going to slaughter is much greater in EB I. This is true
both of the animals who died as lambs and kids and those that died at the end of their second
year. By contrast the number of mature animals in EB II and EB III culls is greater.

Table 15.10. Tel Yaqush Sheep/Goat Mortality Based on
Percentage of Fused Long Bone Epiphyses

Age in
months

Fusion
point

EB I
M I %

EB II
M I %

EB III
M I %

> 6–10 Acetabulum
Scapula
D. Humerus
P. Radius

4 2
2 2
6 0
3 0
15 4 79

18 1
0 1
6 0
4 2
28 4 88

4 1

1 0
3 0
8 1 89

> 13–16 Phalanx I
Phalanx II

8 2
4 1
12 3 80

8 2
2 0
10 2 83

10 2
1 0
11 2 88

> 18–24 D. Tibia
Metacarpal
Metatarsal

1 2
1 4
2 4
5 10 33

8 1
2 4
2 2
12 7 63

3 2
2 3
1 2
7 8 47

> 30–36 Calcaneus
P. Ulna
P. Humerus
D. Radius
P. Femur
D. Femur
P. Tibia

2 4

0 2
0 1
1 0
0 3
1 0
4 10 28

2 2

1 2
2 2

0 3
5 9 36

3 1
1 1

1 0

1 0
6 2 78

oi.uchicago.edu



277ANIMALS IN THE EARLY BRONZE AGE OF NORTHERN PALESTINE

Levant in particular is far behind the study of survey data, pottery distributions, and the eval-
uation of architectural remains in constructing a database. Thus our ability to reconstruct
actual animal production systems is limited by the necessity of comparing sites that actually
did not act in the same economic network but are used as surrogates for those that did.

In the case of the Tel Yaqush, Megiddo, and Ai samples reported here, we believe we found
ourselves in the middle of the developmental process that was the Early Bronze Age. We have
no material that is comparable to the Chalcolithic evidence, particularly in terms of the exploi-
tation of swine that has been reported elsewhere. All three sites come onto our zooarchaeolog-
ical stage as participants in complex systems of multinoded husbandry rather than collections
of independent households focused on the domestic production of consumables. Animals were
flowing into sedentary communities from satellite producers and perhaps between settled com-
munities as well. These external tethers between towns and herdsmen do not seem to have
reached the stage where specialized wool production by urban-owned flocks had developed.
Despite the fires and other disasters clearly visible in the archaeological record that befell Tel
Yaqush and Ai as they underwent cycles of building and rebuilding, the animal production sys-
tem does not seem to have returned to simpler forms—the emphases on commodity exchange
in sheep and intensive agriculture continued throughout. Interestingly, the evidence from Tel
Yaqush indicates that despite the steady diminution of the community, access to animal prod-
ucts continued to be divided socially. Comparison of the material from cultic contexts at Tel

Table 15.11. Megiddo Sheep/Goat Mortality Based on
Percentage of Fused Long Bone Epiphyses

Age in
months

Fusion
point

Monumental
M I %

Squatters
M I %

> 6–10 Acetabulum
Scapula
D. Humerus
P. Radius

8 2
– –
2 1
5 0
15 3 83

4 1
3 1
4 1
3 1
14 4 78

> 13–16 Phalanx I
Phalanx II

2 2
3 3
5 5 50

12 5
9 2
21 7 75

> 18–24 D. Tibia
Metacarpal
Metatarsal

3 2
2 4
0 4
5 11 31

2 1
1 1
0 4
3 6 33

> 30–36 Calcaneus
P. Ulna
P. Humerus
D. Radius
P. Femur
D. Femur
P. Tibia

0 2
0 2
– –
0 3
0 4
0 1
– –
0 12 0

4 3
1 2
0 7
0 4
0 4
1 2
0 1
6 23 21

oi.uchicago.edu



278 BRIAN HESSE AND PAULA WAPNISH

Yaqush and Megiddo suggests that cattle were somewhat more common in ritual activities than
they were in the more secular parts of the communities.

The potential of zooarchaeological analysis to produce information bearing on questions of
economic organization and social differentiation seems clear. What remains to be done is to
continue the example set by Esse in the careful collection of faunal materials and in the encour-
agement he gave us to examine the material from anthropological and historical perspectives.

ACKNOWLEDGMENTS

We wish to express our appreciation to David Schloen of the Oriental Institute who pro-
vided the stratigraphic information about Tel Yaqush based on the notes from both Doug Esse’s
excavations and his own, information that permitted us to conduct this study. Billy G. Grantham
worked in the field at Tel Yaqush and did the preliminary preparation and identification of the
bone fragments. The faunal remains from Ai were given to us for study by the late Joseph A.
Calloway who provided a small subvention that covered some of the costs of the curation and
analysis. The materials from Megiddo were obtained by the ongoing excavation and publication
project headed by Israel Finkelstein and David Ussishkin of Tel Aviv University and Baruch
Halpern of Pennsylvania State University. We appreciate the timely manner in which they
provided the necessary stratigraphic information about the samples. Donna Cobb kindly drew
figure 15.1. Most especially we wish to express our thanks to the editor of this volume, Samuel
Wolff, who demonstrated enormous patience as we endeavored to conclude this paper.

* = EB II (Esse 1984)

Table 15.12. Sheep/Goat Mortality Based on Fusion Proportions in the Ai Samples

Age in
months

Fusion
point

EB IB
M I %

EB IC*

M I %
EB II

M I %
EB III

M I %

> 6–10 Acetabulum
Scapula
D. Humerus
P. Radius

5 0
8 1
12 1
5 0
30 2 94

4 0
2 0
6 1
3 0
15 1 94

9 1
17 2
21 1
5 0
52 4 93

4 2
1 2
5 1
2 0
12 5 71

> 13–24 Phalanx I
Phalanx II
D. Tibia
Metacarpal
Metatarsal

2 1
2 0
3 2
1 0
1 0
9 3 75

0 1
1 0
5 1
1 1
0 3
7 6 53

5 0
– –
16 3
3 1
6 3
31 8 79

0 1
1 0
1 1
– –
1 0
3 2 60

> 30–36 Calcaneus
P. Ulna
P. Humerus
D. Radius
P. Femur
D. Femur
P. Tibia

3 3
– –
1 0
2 0
– –
1 0
1 1
8 4 67

2 1
0 1
– –
1 0
0 1
– –
1 0
4 3 57

3 2
0 1
– –
1 1
0 1
2 1
2 0
8 6 57

1 1
1 0
– –
– –
– –
– –
– –
2 1 67

oi.uchicago.edu



279ANIMALS IN THE EARLY BRONZE AGE OF NORTHERN PALESTINE

REFERENCES

Altorki, S., and Cole, D. P.
1989 Arabian Oasis City: The Transformation of ºUnayzah. Modern Middle East Series 15. Austin:

University of Texas Press.
Appadurai, A.

1986 Introduction: Commodities and the Politics of Value. Pp. 3–63 in The Social Life of Things,
ed. A. Appadurai. Cambridge: Cambridge University Press.

Barfield, T. J.
1981 The Central Asian Arabs of Afghanistan. Austin: University of Texas Press.

Boessneck, J.
1969 Osteological Differences between Sheep (Ovis aries Linné) and Goat (Capra hircus Linné).

Pp. 331–58 in Science in Archaeology, 2nd ed., eds. D. Brothwell and E. Higgs. New York:
Praeger.

Braun, E.
1985 En Shadud. Salvage Excavations at a Farming Community in the Jezreel Valley, Israel. British

Arhcaeological Reports, International Series 249. Oxford: British Archaeological Reports.
Callaway, J. A.

1972 The EB Sanctuary at Ai (et-Tell). London: Bernard Quaritch.
1993 Ai. Pp. 39–45 in The New Encyclopedia of Archaeological Excavations in the Holy Land, Vol.

1, ed. E. Stern. Jerusalem: Israel Exploration Society & and Carta and New York: Simon and
Schuster.

Cooley, R. E.
1997 Ai. Pp. 32–33 in The Oxford Encyclopedia of Archaeology in the Near East, Vol. 1, ed. E. M.

Meyers. New York: Oxford University Press.
Cribb, R.

1984 Computer Simulation of Herding Systems as an Interpretive and Heuristic Device in the Study
of Kill-Off Strategies. Pp. 161–70 in Animals and Archaeology 3. Early Herders and Their
Flocks, eds. J. Clutton-Brock and C. Grigson. British Arhcaeological Reports, International
Series 202. Oxford: British Archaeological Reports.

1985 The Analysis of Ancient Herding Systems: An Application of Computer Simulation in Faunal
Studies. Pp. 75–106 in Beyond Domestication in Prehistoric Europe, eds. G. Barker and C. Gam-
ble. New York: Academic Press.

1987 The Logic of the Herd: A Computer Simulation of Archaeological Herd Structure. Journal of
Anthropological Archaeology 6: 376–415.

Croft, Paul
1994 Some Preliminary Comments on Animal Remains from the First Three Seasons at Shuna. Levant

24: 130–31.
Davis, S. J. M.

1987 The Archaeology of Animals. London: Batsford.
Esse, D. L.

1984 A Chronological Mirage: Reflections on Early Bronze Age IC in Palestine. Journal of Near East-
ern Studies 43: 317–30.

1991 Subsistence, Trade and Social Change in EB Palestine. Studies in Ancient Oriental Civilization
50. Chicago: Oriental Institute Press.

1993 Tel Yaqush. Pp. 1502–4 in The New Encyclopedia of Archaeological Excavations in the Holy
Land, Vol. 4, ed. E. Stern. Jerusalem: Israel Exploration Society & Carta and New York: Simon
& Schuster.

Falconer, S. E.
1995 Rural Responses to Early Urbanism: Bronze Age Household and Village Economy at Tell el-

Hayyat, Jordan. Journal of Field Archaeology 22: 399–419.
Fernea, R. A.

1972 Gaps in the Ethnographic Literature on the Middle Eastern Village: A Classificatory Explora-
tion. Pp. 75–101 in Rural Politics and Social Change in the Middle East, eds. Richard Antoun
and Iliya Harik. Bloomington: Indiana University Press.

oi.uchicago.edu



280 BRIAN HESSE AND PAULA WAPNISH

Finkelstein, I.
1995 Two Notes on Early Bronze Urbanization and Urbanism. Tel Aviv 22: 47–69.

Goody, J.
1982 Cooking, Cuisine, and Class: A Study in Comparative Sociology. Cambridge: Cambridge

University Press.
Grantham, B. G.

1992 Modern Buqata and Ancient Qasrin: The Ethnoarchaeology of Cuisine in the Golan Heights.
M.A. thesis, Dept. of Anthropology, University of Alabama.

Hellwing, S.
1988–89 Faunal Remains from the Early Bronze and Late Bronze Ages at Tel Kinrot. Tel Aviv 15/16:

212–20.
Hellwing, S., and Gophna, R.

1984 The Animal Remains from the Early and Middle Bronze Ages at Tel Aphek and Tel Dalit: A
Comparative Study. Tel Aviv 11: 48–59.

Hesse, B.
1990 Pig Lovers and Pig Haters: Patterns of Palestinian Pork Production. Journal of Ethnobiology 10:

195–225.
Hesse, B., and Rosen, A.

1988 The Detection of Chronological Mixing in Samples from Stratified Archaeological Sites.
Pp. 117–31 in Recent Developments in Environmental Analysis in Old and New World Archae-
ology, ed. R. E. Webb. British Archaeological Reports, International Series 416. Oxford: British
Archaeological Reports.

Hesse, B., and Wapnish, P.
1985 Animal Bone Archeology. Washington, DC: Taraxacum.

Hillson, S.
1986 Teeth. Cambridge: Cambridge University Press.

Horwitz, L. K.
1985 Appendix C. The En Shadud Faunal Remains. Pp. 168–77 in En Shadud. Salvage Excavations

at a Farming Community in the Jezreel Valley, Israel, ed. Eliot Braun. British Archaeological
Reports, International Series 249. Oxford: British Archaeological Reports.

Horwitz, L. K., and Tchernov, E.
1989 Animal Exploitation in the EB of the Southern Levant: An Overview. Pp. 279–95 in L’urbani-

sation, ed. Pierre de Miroschedji. British Archaeological Reports, International Series 527, pt. 2.
Oxford: British Archaeological Reports.

Ingold, T.
1987 Changing Places: Movement and Locality in Hunter-Gatherer and Pastoral Societies. Pp. 165–97

in The Appropriation of Nature: Essays on Human Ecology and Social Relations, ed. T. Ingold.
Iowa City: University of Iowa Press.

Joffe, A. H.
1991 Early Bronze I and the Evolution of Social Complexity in the Southern Levant. Journal of Med-

iterranean Archaeology 4: 35–58.
1993 Settlement and Society in the EB I & II Southern Levant. Complementarity and Contradiction in

a Small-Scale Society. Monographs in Mediterranean Archaeology 4. Sheffield: Sheffield Uni-
versity Press.

Kopytoff, I.
1986 The Cultural Biography of Things: Commoditization as Process. Pp. 64–91 in The Social Life of

Things, ed. A. Appadurai. Cambridge: Cambridge University Press.
LaBianca, S.

1990 Sedentarization and Nomadization: Food System Cycles at Hesban and Vicinity in Transjordan.
Hesban 1. Berrien Springs, MI: Andrews University Press.

Lees, S. H., and Bates, D. G.
1974 The Origins of Specialized Nomadic Pastoralism. American Antiquity 39: 187–93.

oi.uchicago.edu



281ANIMALS IN THE EARLY BRONZE AGE OF NORTHERN PALESTINE

Loud, G.
1948 Megiddo II: Seasons of 1935–1939. Oriental Institute Publications 62. Chicago: University of

Chicago Press.
Lyman, R. L.

1994 Vertebrate Taphonomy. Cambridge: Cambridge University Press.
Payne, S.

1973 Kill-Off Patterns in Sheep and Goats: The Mandibles from Asvan Kale. Anatolian Studies 23:
281–304.

Perevelotsky, A.
1986 To Sell or Not to Sell—A Pastoralist’s Dilemma: A Lesson from the Slaughterhouse. Human

Ecology 14: 287–310.
Portugali, J., and Gophna, R.

1993 Crisis, Progress and Urbanization: The Transition from Early Bronze I to Early Bronze II in
Palestine. Tel Aviv 20: 164–86.

Postgate, J. N. (with a contribution by S. Payne)
1975 Some Old Babylonian Shepherds and Their Flocks. Journal of Semitic Studies 20: 1–21.

Redding, R.
1981 Decision Making in Subsistence Herding of Sheep and Goats in the Middle East. Ph.D. disserta-

tion, University of Michigan. Ann Arbor: University Microfilms International.
1991 The Role of the Pig in the Subsistence System of Ancient Egypt: A Parable on the Potential of

Faunal Data. Pp. 21–30 in Animal Use and Culture Change, eds. P. J. Crabtree and K. Ryan.
Museum Applied Science Center for Archaeology, Supplement to Vol. 8. Philadelphia: Univer-
sity Museum.

1993 Subsistence Security as a Selective Pressure Favoring Increasing Cultural Complexity. Bulletin
on Sumerian Agriculture 7: 77–98.

Rosen, B.
1986 Subsistence Economy of Stratum II. Pp. 156–85 in ºIzbet Sartah, ed. I. Finkelstein. British

Archaeological Reports, International Series 299. Oxford: British Archaeological Reports.
Schwartz, G. M., and Falconer, S. E.

1994 Rural Approaches to Social Complexity. Pp. 1–9 in Archaeological Views from the Countryside:
Village Communities in Early Complex Societies. Washington, DC: Smithsonian Institution.

Shahrani, M. N. M.
1979 The Kirghiz and Wakhi of Afghanistan. Seattle: University of Washington Press.

Silver, I. A.
1969 The Ageing of Domestic Animals. Pp. 283–309 in Science in Archaeology, 2nd ed., eds. D. Broth-

well and E. Higgs. New York: Praeger.
Stein, G.

1992 Herding Strategies and Group Mobility in the Near East: Ethnographic and Archaeological Per-
spectives. Paper presented to the Annual Meeting of the Society for American Archaeology,
Pittsburgh, Pennsylvania.

Trigger, B. G.
1989 A History of Archaeological Thought. Cambridge: Cambridge University Press.

Uerpmann, H.-P.
1987 The Ancient Distribution of Ungulate Mammals in the Middle East. Tübinger Atlas des Vorderen

Orients, Beihefte 19, Reihe A., Nr. 27. Wiesbaden: Reichert.
Wapnish, P.

1993 Archaeozoology: The Integration of Faunal Data with Biblical Archaeology. Pp. 426–42 in
Biblical Archaeology Today 1990, Proceedings of the 2nd International Congress of Biblical
Archaeology, eds. A. Biran and J. Aviram. Jerusalem: Israel Exploration Society.

1996 Is seni ana la mani an Accurate Description or a Royal Boast? Pp. 285–96 in Retrieving the Past:
Essays in Archaeological Research and Methodology in Honor of Gus W. Van Beek, ed. J. Seger.
Winona Lake, IN: Eisenbrauns and Cobb Institute, Mississippi State University.

oi.uchicago.edu



282 BRIAN HESSE AND PAULA WAPNISH

Wapnish, P., and Hesse, B.
1991 Faunal Remains from Tel Dan: Perspectives on Animal Production at a Village, Urban and Ritual

Center. Archaeozoologia 4: 9–86.
2000 Mammal Remains from the Early Bronze Age Sacred Compound. Pp. 429–62 in Megiddo III:

The 1992–1996 Seasons, Vol. II, eds. I. Finkelstein, D. Ussishkin, and B. Halpern. Tel Aviv:
Emery and Claire Yass Publications in Archaeology, Institute of Archaeology, Tel Aviv
University.

Zeder, M.
1991 Feeding Cities. Washington, DC: Smithsonian Institution.
1994 Of Kings and Shepherds: Specialized Animal Economy in Ur III Mesopotamia. Pp. 175–91 in

Chiefdoms and Early States in the Near East: The Organizational Dynamics of Complexity, eds.
G. Stein and M. S. Rothman. Monographs in World Archaeology 18. Madison, WI: Prehistory
Press.

oi.uchicago.edu



283



THE FAUNAL EVIDENCE FOR
SOCIOECONOMIC CHANGE BETWEEN THE
MIDDLE AND LATE BRONZE AGE IN THE

SOUTHERN LEVANT



Liora Kolska-Horwitz and Ianir Milevski



INTRODUCTION

The Late Bronze Age (LB) in the southern Levant (ca. 1500–1200



b.c.



) is characterized by
a marked change in demography, settlement pattern, and material culture, relative to the pre-
ceding Middle Bronze Age (MB) IIA–C period (1900–1500



b.c.



). The number and density of
very large urban sites decreases, with a concomitant increase in the number of small and
medium-sized sites. This is indicative of a trend toward ruralism (Gonen 1984; Falconer
1994). Moreover, fortifications are absent or greatly reduced in scale relative to the Middle
Bronze Age. Population demography also changes, with a shift to the coastal plain and a con-
comitant emptying of the interior, especially the highlands. Few sites show an uninterrupted
occupation from the MB to the LB, either as a result of Egyptian military activities in the
region (Falconer 1994) or due to local sociopolitical upheavals (Finkelstein 1993). Other sites
appear to have been established for the first time in this period (Gonen 1984). In contrast to
these changes which are reminiscent of a recession or deurbanization process, there is evi-
dence for the involvement of the southern Levant in extensive political activities and cultural
commerce on an international scale (Falconer 1994).

Despite these clear patterns, there is little consensus as to what these changes signify. It
has been argued (e.g., Albright 1960: 101; Kenyon 1960: 209–14, de Vaux 1978: 122; Aha-
roni 1982: 122, 150–51; Knapp 1987, 1989; Falconer 1994) that the LB represents a period of
cultural decline and impoverishment characterized by deurbanization, as evidenced by the
poor level of the material culture, demographic changes, and sparseness of large settlements.
However, Liebowitz (1987, 1989) has argued to the contrary, stating that the Late Bronze Age
was a period of prosperity and a high point in terms of material culture as attested by the
extensive assemblages of ivory artifacts. A third view (Bienkowski 1989) argues for a mosaic
pattern, with the marginal areas of the southern Levant deteriorating in the LB in contrast to
the principal areas that prospered under Egyptian rule. Finkelstein (1993, 1994) has presented
a similar view, arguing for a highland-lowland dichotomy with the MB IIC/LB I transition in
the highlands being marked by a period of upheaval resulting from local sociopolitical con-
flicts rather than Egyptian intervention. The outcome of this transition was that the MB ended
with the abandonment of many of the highland sites, and the LB in the highlands was charac-
terized by a return to pastoralism. Bunimovitz (1993, 1994) has suggested that the level of
sociopolitical organization also shows evidence for regional diversity, with the hill country



16



oi.uchicago.edu





284



LIORA KOLSKA-HORWITZ AND IANIR MILEVSKI



representing a more integrated settlement system, while an inverse pattern is seen in the low-
land regions (such as the coastal plain and Jezreel Valley), where the semiautonomous city
states show a low degree of integration. We may conclude that despite the absence of consen-
sus on the exact nature of the LB, all researchers acknowledge that at least certain regions of
the southern Levant underwent a period of recession and deurbanization. The present study
aims at evaluating this hypothesis in the light of one aspect of the archaeological record,
namely, faunal remains.

THE MODEL

Animal remains form an integral part of the ancient record and may be used to reconstruct
the mode of animal production and consumption in the past. As such, they complement and
corroborate information obtained from the study of other archaeological remains. Due to the
close interrelationship between subsistence and other aspects of a society, it can be expected
that large-scale changes in the sociopolitical, ethnic, cultural, or economic status of a site will
result in concomitant changes in the manner in which animals are exploited (Wapnish and
Hesse 1988; Crabtree 1990; Redding 1991; Zeder 1991; Hesse and Wapnish 1997). Thus, it
should be possible to detect changes of the magnitude postulated for the MB–LB transition in
the Levant outlined above, in the faunal record of archaeological sites dating to this period.

The options open to a population in the face of economic recession are limited. Hole
(1994) has outlined two possible responses to crop failure following changing climatic condi-
tions in fourth/fifth millennium



b.c.



Mesopotamia. The first response entails the population
remaining where they are (termed by him “stand and fight”), but adopting alternative subsis-
tence strategies to agriculture such as herding, fishing, and hunting, or intensifying agriculture
accompanied by large-scale storage of surpluses. The second option (termed “capitulation and
flight”) entails wide-scale abandonment of settlements and movement to areas with more ame-
nable environmental conditions. In this instance, certain sectors of the population may shift to
pastoral nomadism in marginal zones suitable for grazing, while others may remain in small-
sized settlements living as farmers, hunters, and gatherers.

We suggest that if the southern Levantine LB populations were indeed facing a period of
marked economic and/or cultural deterioration or impoverishment, they would have faced
similar options to those outlined by Hole (1994). They could either stay where they were and
alter their subsistence strategies as outlined above, or migrate and engage in alternate modes
of production elsewhere.

In the introduction, we briefly outline some of the observed changes in population demog-
raphy and settlement pattern that have been documented for the LB. However, few attempts
(Finkelstein 1993, 1994) have been made to document changes in subsistence strategies for
this period. In order to recognize new subsistence strategies in the animal economy, two fea-
tures are required:

1. Suitable archaeozoological tools for assessing change
2. A diachronic perspective of the faunal record

In order to deal with the first requirement, we have turned to a model developed by Red-
ding (1981, 1991, 1993) for Near Eastern societies. Redding’s model predicts that the faunal
record of a society involved in intensive agriculture will contain a low frequency of caprines
and pigs; a very high frequency of cattle; more goats relative to sheep; and a high number of



long

oi.uchicago.edu





285



THE FAUNAL EVIDENCE FOR SOCIOECONOMIC CHANGE



adult cattle. In contrast, a society primarily involved in nomadic or seminomadic pastoralism,
the faunal record will be dominated by caprines (sheep and goats). The frequency of cattle and
especially pigs will be low. Moreover, sheep will constitute a greater portion of the herd than
goats. To these features we would like to add another (after Hole 1994), namely, that in socie-
ties facing economic stress and exploring alternative subsistence strategies, the frequency of
wild fauna, obtained through hunting, gathering, and fishing will increase, as may also the
number of species exploited (species richness).

To satisfy the requirement of a diachronic trajectory for assessing subsistence change, two
types of faunal data were examined. The first data set included animal remains from both MB
and LB strata from the same site, while the other included remains from LB sites unoccupied
during the preceding MB, or those established for the first time during this period. The result-
ing information was compared and contrasted in order to elucidate the nature of the subsis-
tence strategies favored by the LB populations of this region.

The rationale for examining material from two different strata at the same site was that this
enables us to examine diachronic change while minimizing the effect of external parameters
known to influence the animal economy, such as environment, topography, and geography.
Furthermore, it facilitated testing the scenario proposed by Hole (1994), in which the popula-
tion chooses to remain at the same site but has to modify existing subsistence forms.

METHODS AND MATERIALS

For the purpose of this study, six so-called core sites were chosen which contained both
MB and LB strata. The sites fall into six different geographic and phytogeographic zones (fig.
16.1) such that they facilitate assessment of changes under different environmental conditions.
They include: Tel Dan (northern Galilee), Tel Michal (central coastal plain), Tel Shiloh (the
Samarian hill country), Mana



˙



at (the Judaean hills), Lachish (the Shephelah), and Tell Jem-
meh (the southern coastal plain). At each site the same archaeozoologist studied the remains
from both of the MB and LB strata of the site, using the same analytical methods. The other
LB faunal data were derived from published reports of domestic and cultic sites in this region.

All the assemblages are discussed in terms of their total bone counts (Number of Identified
Specimens [NISP]), which were converted into relative frequencies. As most of the assem-
blages differed markedly in size (NISP), rarefaction tests were run for the MB and LB assem-
blages represented at the same site. Rarefaction is a statistical test that calculates the number
of species one would expect to find given that the assemblages studied were of the same size
(NISP) (Ludwig and Reynolds 1988). This test establishes whether the differences observed in
the number of species from the two periods were due to differences in sample size as opposed
to differences in patterns of animal exploitation.

Several problems were encountered that limited the extent of intersite comparisons that
could be made: differences in the state of bone preservation, techniques of bone recovery dur-
ing excavation, and differences in the extent and character of the areas excavated. Moreover,
the published reports differ in the types of analyses undertaken as well as in the form of data
presentation when dealing with chronology, bone element representation, as well as age and
sex breakdowns for the different species. Consequently, the emphasis in this paper is on pat-
terns in species composition. If more complete data sets are available, age, sex, and bone ele-
ment representation can be added to elucidate the patterning. With regard to chronology, for

oi.uchicago.edu





286



LIORA KOLSKA-HORWITZ AND IANIR MILEVSKI



FIGURE 16.1. Map of sites mentioned in text according to their phytogeographic regions (after Orni and Efrat
1971: 165). Drawn by Michael Smelansky.



long

oi.uchicago.edu

[image: image86.tif]



287



THE FAUNAL EVIDENCE FOR SOCIOECONOMIC CHANGE



most sites no distinction was made between the different phases of MB and LB (i.e., LBI or
LB II, MB IIA, B or C). Consequently, the different phases for each period were pooled to
form two larger categories: MB II versus LB. Due to these limitations, it has proved impossi-
ble to assess with any refinement the chronological development of the faunal changes and
only large-scale diachronic trends can be addressed.

INTRASITE COMPARISONS



Tel Dan



The site (Israel grid 2112/2949) lies some 200 m above sea level at the foot of Mount Her-
mon (fig. 16.1). It is situated in a fertile part of northern Galilee, which receives an average
annual rainfall of about 500–600 mm, and is characterized by typical Mediterranean hill zone
vegetation (Orni and Efrat 1971: 144–46). In addition, several springs flow in and around the
tell, which covers some 200 dunams (Biran 1993, 1996).

The MB settlement at the site, spanning MB IIA to MB IIC, contains a massive earthen
rampart but appears to have been unwalled. Only limited remains from the LB have been
uncovered at the site, including a built tomb and a large structure in Area K which, according
to the excavator (Biran 1993; 1996), might have served a ritual function.

The faunal data used in this study are derived from two faunal reports: that of Wapnish,
Hesse, and Ogilvy (1977) which details the collection from the 1974 season and that of Wap-
nish and Hesse (1991) dealing with material from the 1981 to 1986 excavation seasons. To
increase sample size, the medium mammal counts given in Wapnish and Hesse (1991) have
been combined with the caprine counts, while their large mammal counts were included in the
cattle category. The bulk of the LB bone assemblage (87.5%) originates from one locus in the
large structure in Area K. The remainder of this sample was recovered from a deep sounding
in a destruction level in Area T. The faunal report does not differentiate between the LB I and
LB II periods. The MB sample, dating from the end of MB IIA to the beginning of MB IIB,
is derived from fills in Areas B and Y that are associated with the MB II rampart. Those recov-
ered from Area M were recovered from inside a pit. The small size of both Bronze Age assem-
blages (NISP of



>



100 bones) limits the depth of the analysis that may be undertaken.
Although the number of species represented in both periods is similar (N = 7 in MB; N = 8

in LB), they differ in the relative proportions of the different animals. Domestic species pre-
dominate in both assemblages, with fewer caprines and more cattle in the LB relative to the
MB (fig. 16.2). Wapnish and Hesse (1991: 30) propose that the high frequency of cattle in this
latter sample may be related to the greater robustness of these bones relative to those of smaller
mammals such as sheep and goat. Since the LB sample is burnt, this is a factor that would have
made the smaller bones even more susceptible to diagenetic processes. However, as the Mini-
mum Number of Individuals (MNI) counts also reflect a high number of cattle, they conclude
that the high cattle numbers probably reflect a reliable estimate of cattle–caprine utilization at
the site. The fact that the LB sample may be derived from a cultic context raises the possibility
that cattle were specifically selected for ritual activities, and this may account for their presence
in such high numbers.

The relative frequency of the three main species, calculated from their combined NISP
counts (fig. 16.3), clearly illustrates a change in species dominance over time which follows

oi.uchicago.edu





288



LIORA KOLSKA-HORWITZ AND IANIR MILEVSKI



that observed when all species are included (fig. 16.2). Caprines decline in importance in the
LB, while concomitantly cattle increase. Pigs also increase in importance in the LB. However,
as they probably represent remains of wild boars and not domestic animals (Wapnish and
Hesse 1991: 13), they represent hunted rather than managed animals. Their high frequency in
the assemblage is probably a reflection of the favorable environment close to Tel Dan for wild
boar: abundant water sources and dense vegetation including oak forest.

In both the MB and LB more sheep were exploited than goats. However, in the LB, there
is an increase in the frequency of sheep relative to the MB (fig. 16.4). There is also an increase



FIGURE 16.2. Histograms showing the relative frequencies of all species found at the six “core” sites.

oi.uchicago.edu

[image: image87.tif][image: image88.tif][image: image89.tif][image: image90.tif][image: image91.tif][image: image92.tif]



289



THE FAUNAL EVIDENCE FOR SOCIOECONOMIC CHANGE



in the frequency of wild animals in the LB (fig. 16.5), with this assemblage particularly rich
in remains of deer, boar, and fox.



Tel Shiloh



Tel Shiloh (Israel grid 1775/1626; fig. 16.1) is situated on a fertile valley in the Samarian
hills, surrounded by ravines which merge on the west into the Wadi ºAli, east of the Jerusalem-
Nablus road. The area of the mound is about 30 dunams, rising about 710 m above sea level
(Finkelstein, Bunimovitz, and Lederman 1993). The vegetation is characteristic of the Medi-
terranean hill zone with an annual average rainfall of 500–600 mm (B. Rosen 1993: 364).

The MB was present in most of the excavated areas of the site and is derived from two strata:
VIII and VII. Stratum VIII contained mid-MB IIB–early MB IIC remains that are associated
with fills, some within structures adjoining the city wall (Areas F and H) as well as the glacis
in Area D. It has been proposed that the settlement of this period was small and unfortified. In
contrast the MB IIC site (Stratum VII) was fortified with a peripheral stone wall and glacis. The
faunal remains from this stratum are derived from the city wall and its adjoining rooms as well
as earth fills. Finkelstein (Finkelstein, Bunimovitz, and Lederman 1993: 376–77) has suggested
that during the MB IIC the site functioned as a cult place and was unoccupied as illustrated by
the absence of dwellings. The LB (Stratum VI), in Area D, contained both LB I and LB IIA
material. It appears that during this period the site served as an uninhabited cultic center.
Indeed, the bulk of the LB assemblage, including the animal remains, was recovered from a
dump of an LB I and LB IIA favissa such that these remains relate to the cultic activities carried
out at the site (Finkelstein, Bunimovitz, and Lederman 1993: 45; B. Rosen 1993).

The faunal report (Hellwing, Sadeh, and Kishon 1993) deals with the faunal remains from
both periods. However, they are not discussed in relation to their exact archaeological prove-
nance, and no separation is made between the two phases of the LB. Distinction is made, how-
ever, between the MB IIB and late MB IIC deposits. For the purposes of comparison with
other sites in this paper, the two MB strata were combined. The sizes of the bone assemblages
differ, with the combined MB samples comprising only 1,200 bones and the LB sample com-
prising close to 3,000 identifiable bones.

Both of the MB and LB assemblages contain a broad range of animal species, but the LB
has a slightly higher frequency of remains of wild animals (fig. 16.5). Both periods are domi-
nated by domestic animals. The LB assemblage is characterized by an extremely high fre-
quency of caprine remains and lower frequencies of both cattle and pigs than in the MB (fig.
16.2). As at other sites, the LB has a higher frequency of sheep relative to the MB (fig. 16.4).
The high caprine but low cattle frequencies in the LB have been interpreted as indicative of a
reversion to nomadic or seminomadic pastoralism (Finkelstein 1993; B. Rosen 1993), and the
deterioration in the economic status of the site (Hellwing, Sadeh, and Kishon 1993: 324).
Conversely, the diachronic differences may be related to the diversity of contexts from which
the remains were excavated; the LB sample represents remains from a cultic favissa, raising
the possibility of species selection for cultic purposes. In contrast, the MB sample is derived
from the city wall and associated structures and probably represents domestic debris. Conse-
quently, though the environmental and topographic location of the site in the highlands makes
it more suited to caprine herding than intensive agriculture, the high frequency of caprines
may be the result of selection for ritual practices. By comparison with fauna from another

oi.uchicago.edu





290



LIORA KOLSKA-HORWITZ AND IANIR MILEVSKI



cultic site, B. Rosen (1993) has argued that the Shiloh LB assemblage does in fact provide a
true reflection of herd composition during this period.

Examination of relative frequencies of the three most prevalent domesticates found at
Shiloh (fig. 16.3) corroborates the overall pattern observed in figure 16.2, with an increase in
caprines in the LB and a concomitant decrease in the frequency of cattle and pigs.



Tel Michal



Tel Michal (Israel grid 131/174) is located on the coastal plain, west of Herzliya within a
coastal dune vegetation (fig. 16.1) and an annual average rainfall of 500–600 mm (Orni and
Efrat 1971: 145). The site is spread over five hills covering an area of about 135 dunams (Her-
zog, Rapp, and Negbi 1989). The MB IIB assemblage (Stratum XVII) is derived from the high
tell, from fills associated with the earthen rampart and from a constructed “platform” on top
of which the settlement was built. During the MB, the settlement lacked a defensive city wall.
The MB IIB settlement was totally destroyed, probably by tectonic activity (Herzog, Rapp,
and Negbi 1989).

The site was rebuilt during the LB, and most of the remains dating to this period are
derived from the high tell (Area A) from Strata XV (LB II) and XVI (LB I). The bulk of the
LB I sample originates from a structure identified as a military fort as well as from the fill of
a rampart on top of the MB IIB earthworks. The LB II remains are derived from two structures
and parts of a rampart (Herzog, Rapp, and Negbi 1989: 40–41; fig 16.4). The presence of
many imported ceramic vessels and the absence of agricultural settlements around the tell dur-
ing both the MB and LB periods have led researchers to suggest that the agricultural potential
of its hinterland was not developed and that during both periods the settlement served as a
small (half acre) trading station (Gophna and Ayalon 1989). The excavators have noted that
unlike many of the coastal sites, there is no evidence from Tel Michal to indicate that the site
underwent a decline during the Middle and Late Bronze transition (Herzog, Rapp, and Negbi
1989: 41).

The faunal report (Hellwing and Feig 1989) describes the remains by period but does not
differentiate between the LB I and LB II strata, nor does it relate to the exact archaeological
context of these remains. As there is a great disparity in sample size between the MB (NISP =
110) and LB (NISP = 641), a rarefaction test was run. The results indicate that the difference
in the number of species represented in the two periods (MB = nine species; LB = eleven spe-
cies) may be related solely to differences in sample size and not to differences in patterns of
animal exploitation.

Domestic caprines and cattle are the most frequently represented species in both periods,
although their relative frequencies increase in the LB (fig. 16.2). It is interesting to note that in
both periods, caprines and cattle are represented in almost equal amounts. In contrast, pig
remains are significantly more common in the MB. Unfortunately no data is given on the rel-
ative numbers of sheep to goats. Comparison of the relative proportions of the three main
domestic species in the MB and LB (fig. 16.3) indicates that while the frequencies of caprines
and pigs resemble each other for both periods, there is an almost twofold increase in the rela-
tive frequency of cattle in the LB. Another marked diachronic difference is in the frequency of
wild to domestic species (fig. 16.5), with a significant increase in the LB in the frequency of
wild species including fish, gazelle, and especially deer, suggesting increased hunting.

Long

oi.uchicago.edu





291



THE FAUNAL EVIDENCE FOR SOCIOECONOMIC CHANGE



FIGURE 16.3. Relative frequencies of (a) sheep/goat, (b) cattle, and (c) and pigs in the six “core” sites. Calculated
from the total NISP counts for these three species.

oi.uchicago.edu

[image: image93.tif][image: image94.tif][image: image95.tif]



292



LIORA KOLSKA-HORWITZ AND IANIR MILEVSKI



Mana



˙



at



Mana



˙



at (Israel grid 1289/1679; fig. 16.1) is located in the Rephaim Valley (Wadi el-
Ward), 5 km southwest of the Old City of Jerusalem. The site is about 690 m above sea level
and covers an area of about 30 dunams. The vegetation around the site is typical of the Med-
iterranean hill zone with a mean annual rainfall of 550 mm (Orni and Efrat 1971: 144–46).
Close to the site is the perennial spring of Ein Yalo (Edelstein, Milevski, and Aurant, 1998).

The archaeozoological report (Horwitz 1998) discusses a small assemblage of bones asso-
ciated with an LB IIB domestic structure (Building 1028 in Area 1000). The MB IIB–C
remains are derived from Area 800 (terrace fills and house floor fills) and Area 1000 (house
floors). The character of the LB settlement at the site is unclear, but during the MB it appears
to have served as an unfortified rural village (Edelstein, Milevski, and Aurant, 1998).

The number of species represented in the two periods differ somewhat (MB = nine; LB =
thirteen) as does sample size; the MB assemblage, which has a lower number of species, is
nearly twice the size of the LB assemblage. This indicates that in the LB, the wider range of
animal species represented is not a bias resulting from sample size but reflects selection. It is
thus not surprising that the frequency of wild fauna increases in the LB.

Compared with the MB, the LB has a lower frequency of pigs but a higher frequency of
cattle and caprines, especially sheep (figs. 16.2, 16.4). However, both Bronze Age assem-
blages at Mana



˙



at contain relatively high proportions of pigs. This may be explained by the
moist, local conditions around the site as evidenced by the palaeobotanical and phytolith
remains which suggest the proximity of an oak forest as well as the presence of a marsh area



FIGURE 16.4. Relative frequency of sheep in the six “core” sites. Calculated from the relative frequency of iden-
tified sheep and goat bones from NISP counts.

oi.uchicago.edu

[image: image96.tif]



293



THE FAUNAL EVIDENCE FOR SOCIOECONOMIC CHANGE



(Edelstein and Milevski 1994). These data suggest that the water table in the Rephaim Valley
might have been higher during the Bronze Age (A. Rosen 1998).

When the relative frequency of caprines, cattle, and pig are calculated (fig. 16.3), a slightly
different picture is observed, with caprines represented in equal numbers in the two periods. In
the LB, cattle frequencies increase while pigs decrease. The faunal data from Mana



˙



at illus-
trate that there is a greater resemblance in subsistence base between the LB IIB and MB IIB
faunal assemblages, than between them and the earlier EB IV (Horwitz 1998; Edelstein and
Milevski 1994).



Lachish



Lachish (Israel grid 1357/1083; fig. 16.1) is located in the Shephelah about 250 m above
sea level. The site covers approximately 120 dunams. It is situated in the fertile region at the
southern edge of Na



˙



al Lachish (Wadi Ghafr). Wells appear to have supplied the city with
most of its water (Ussishkin 1993). The surrounding vegetation is typical of the Mediterra-
nean lowland zone giving way to semiarid scrub (Drori 1979). The annual average rainfall is
300–400 mm (Orni and Efrat 1971: 144–45).

The MB site was fortified and probably functioned as one of the major centers of southern
Canaan until its destruction at the end of this period (Ussishkin 1993). The MB IIB–C depos-
its were recovered from Level VIII and relate to a cult site in Area D and a palace in Area P.
The LB remains are derived from two levels at the site: Level VII (LB I–III) containing the
Fosse Temple, and Level VI LB III Acropolis Temple. During Level VI the city was under



FIGURE Fig. 16.5: Relative frequency of wild species in the six “core” sites. Calculated from the relative
frequency of wild versus domestic species from NISP counts.

oi.uchicago.edu

[image: image97.tif]



294



LIORA KOLSKA-HORWITZ AND IANIR MILEVSKI



Egyptian rule. The LB settlement at Lachish was unfortified, and by the end of this period the
site might have been one of the largest cities in the southern Levant (Ussishkin 1993).

The LB IIB faunal assemblage used in this study is derived from the excavations directed
by D. Ussishkin and includes material recovered during the 1973 to 1978 excavation seasons
(Drori 1979). It includes material from a public building and the Acropolis Temple in Level VI
of Areas S and P as well as remains recovered from within and underneath the temple floor
(Drori 1979: 26, plan 7). The fauna from the sanctuary in Level VI recovered during Aha-
roni’s excavations and analyzed by Lernau (1975) were not included in this study, despite the
fact that they are derived from the same excavation area since different methods were used in
their retrieval and analysis.

The Lachish LB sample contains a lower frequency of caprines and pig than the MB sam-
ple, but a higher frequency of cattle (fig. 16.2). This complements the trend reported by Ler-
nau (1975) who found lower frequencies of caprines than cattle (28.3% caprines, 66.3%
cattle) in the sanctuary in Level VI. The same pattern is observed when the relative frequen-
cies are calculated for the three main species (fig. 16.3).

Remains of sheep are more prevalent in the LB than in the MB (fig. 16.4). There is an
increase in the relative frequency of remains of wild animals in the LB relative to the MB,
especially of birds and deer (fig. 16.5). At least a portion of the MB assemblage, as well as the
entire LB sample, are derived from cultic contexts. Although it is difficult to assess the exact
influence of this factor on their composition, it is possible that the relative proportions of the
animals reflect their selection for ritual purposes. Finkelstein (1988: 343–44) has suggested
that there was a pastoral population at Lachish during the LB. However, the faunal remains
from the “Solar Shrine” (Lernau 1975) as well as those from the sanctuary (Drori 1979) con-
tain lower frequencies of caprines than cattle, which, according to the model presented here,
is characteristic of an agricultural or agro-pastoral community. Based on an analysis of the
pollen contained in mudbricks from the excavation, Drori (1979) concluded that the environ-
ment in the LB was richer in arboreal pollen than in the MB, suggesting a more humid envi-
ronment than today (contra A. Rosen 1986: 57). This factor may also explain the relatively
high cattle frequencies at the site, which today lies on the edge of a semiarid zone.



Tell Jemmeh



Tell Jemmeh is situated in the northern Negev, close to the southern coastal plain, about
10 km south of present-day Gaza (map ref. 097.088), on the southern bank of Na



˙



al Besor
(fig. 16.1). This area receives about 300 mm in annual rainfall. The surrounding vegetation is
typical of the Irano-Turanian zone, as an extension of the northern Negev to the coast (Orni
and Efrat 1971: 172). The original area of the tell was about 50 dunams, some 50 m above sea
level (van Beek 1993).

During the MB IIB, the site appears to have been a large Canaanite agricultural town with
an earthen rampart. The LB occupation, represented by both of the LB I and LB II phases,
consisted of a small settlement (about 11–50 dunams). In the northwest excavation area,
remains of an LB mudbrick fortification wall were found while remains of several domestic
structures, including a large “elite” residence, were also excavated. It is probable that during
the LB the site had an Egyptian governor (van Beek 1993).

The faunal remains have only been partially published (Wapnish and Hesse 1988; Wap-
nish 1982, 1993). Concerning their context, Wapnish (1993) notes that the majority of the LB

long

oi.uchicago.edu





295



THE FAUNAL EVIDENCE FOR SOCIOECONOMIC CHANGE



faunal assemblage is derived from the large elite building while the MB sample represents
household debris.

Both periods contain a wide range of species. There is a significant difference, however, in
the number of species represented in the LB and the MB at the site (LB = twenty-one species;
MB = eleven). Since there is a marked difference in the size of the two samples (LB = approxi-
mately 4,300 bones; MB = approximately 200 bones), a rarefaction test was run. The results
indicate that given the MB sample size, the LB sample should contain eleven species, a result
that is very close to that of the MB sample. Consequently, the larger number of species in the
LB appears to be directly correlated to the larger size of this sample. Once again wild animal
species comprise a relatively larger portion of the LB sample than in the MB sample, indicat-
ing a greater emphasis on hunting, gathering, and fishing in the LB (fig. 16.5).

Examination of the frequencies of the various species indicates an increase in the relative
proportions of caprines and cattle in the LB and a concomitant decrease in the frequency of
pigs (fig. 16.2). Wapnish (1993) has suggested that the decrease in pigs in the LB may either
reflect drier climatic conditions than in the MB or, since pigs were not considered an accept-
able dietary item for the Egyptian elite (see also Redding 1991), they are absent from the
debris of the elite LB residency. Relative frequencies of the three main species (fig. 16.3) show
an increase in cattle in the LB but a reduction in the numbers of both pigs and caprines in the
LB. In contrast to the other sites studied here, the frequency of sheep decreases in the LB rel-
ative to the preceding MB (fig. 16.4).



Diachronic Trends within Sites



The six sites that form the central data set in this study exhibit marked diachronic differ-
ences in the relative frequencies of animals exploited between the MB and the LB. These dif-
ferences are more marked in some sites than in others, a feature undoubtedly influenced by the
geographic and environmental differences between them. Similarly, the influence of differ-
ences in the function and size of the sites must be considered.

The rarefaction tests showed that interperiod differences in the number of species found at
some sites are due solely to differences in sample size and not to differences in human exploi-
tation patterns. Consequently, no significant differences were evident in the number of species
exploited between the MB and the LB. Moreover, little diachronic difference was found in the
range of animal species exploited.

Examination of the relative frequencies of species represented at the six sites (fig. 16.2)
indicates a predominance of the same domestic animals in both periods: primarily sheep and
goat (



Ovis aries



/



Capra hircus



), cattle (



Bos taurus



), and pig (



Sus scrofa



). Additional domesti-
cates found in all the sites include equids (both horse and donkey), camel, dog, and domestic
birds. Similarly, the same wild faunal elements were represented in both the MB and LB
assemblages (fig. 16.2). Their representation at a site appears to be related to the specific envi-
ronmental conditions around the site, such as the presence of deer species and boar in sites
close to woodlands and thickets. It should also be kept in mind that some of the wild species
(such as rodents and birds) may represent accidental intrusions into the deposit rather than
food items. It may be concluded, then, that the main diachronic trend between the MB and LB
periods is not expressed in differences in the range or number of species represented, but
rather in their relative frequencies. Indeed, marked diachronic differences are observed in five
such features:

oi.uchicago.edu





296



LIORA KOLSKA-HORWITZ AND IANIR MILEVSKI



(1) At four of the six sites, caprines increased in frequency in the LB relative to the MB
(Shiloh, Tel Michal, Mana



˙



at, and Tell Jemmeh). However, at Tel Dan and Lachish caprines
decrease in frequency, and at Tel Dan are the predominant species (fig. 16.2).

It is interesting to note that the Tel Dan and Lachish LB samples are from suspected cultic
deposits suggesting that human selection for ritual purposes may have influenced the species
proportions, with cattle the preferred ritual animal at some sites. In contrast, the LB levels at
Shiloh, although also derived from a cultic deposit, show a clear predominance of caprines. It
is possible that local environmental conditions might have played a role in determining the
herd composition in the hill country.

(2) There is an increase in the frequency of sheep in the LB relative to the MB at all sites
with the exception of Tell Jemmeh (fig. 16.4). At this latter site, an opposite trend is observed,
and the frequency of sheep declines in the LB, although sheep still outnumber goats in this
period.

Based on the changing frequencies of pigs and gazelle, Wapnish (1993) suggested that the
LB at Tell Jemmeh might have been drier, therefore indicating that it is possible that the
increase in goats in this period, a species known to be more heat- and water-tolerant, may also
be related to this factor. However, Wapnish (1993) also presents contradictory evidence based
on the increased frequency of water birds in the LB, which suggests that this period might in
fact have been wetter. The environmental data for this site are therefore equivocal, and the rea-
son for the decrease in sheep may simply represent a local trend. We may conclude that at the
majority of sites there is a greater emphasis on sheep herding in the LB relative to the MB.

(3) There is an increase in the frequency of cattle at five of the six LB sites (fig. 16.2). The
exception is Shiloh where cattle frequencies are lower in the LB than in the MB.

Two factors may be responsible for the inverse trend observed at Shiloh. The first is that
the specific environmental conditions of the hill country may be better suited to caprine herd-
ing than cattle keeping. The second is the possible role played by human selection due to the
cultic nature of the LB deposits at Shiloh.

(4) In all sites but one examined here, the frequency of pigs is markedly reduced in the LB
relative to the MB (fig. 16.2). The one exception is Tel Dan, where the pig frequency increases
in the LB.

This apparent anomaly may be explained by the fact that the pigs from Dan have all been
identified as wild boar rather than domestic pig. They probably inhabited the environs of the
site because of the natural springs in its proximity. Consequently, their increased frequency at
the site reflects the increased incidence of hunting rather than an increase in pig herding.

(5) The relative contribution of wild animal species to the assemblage increases in the LB
(fig. 16.5).

This includes increased frequencies of fallow and red deer, gazelle, carnivores such as fox
and wild cat, antelope species such as hartebeest, and a variety of wild birds and fish. The fact
that this trend is found at all the sites studied here suggests that it reflects a substantive change
in subsistence strategy over time.

COMPARISON WITH ADDITIONAL LATE BRONZE AGE SITES

In addition to the LB strata from the six “core” sites described above, only a few published
reports from other LB sites are available and most describe very small samples. The data from



long

oi.uchicago.edu





297



THE FAUNAL EVIDENCE FOR SOCIOECONOMIC CHANGE



these sites are briefly summarized below and discussed in relation to the LB “core” sites
described above.

A small but highly selective assemblage was recovered from the LB Egyptian Garrison at
Tel Beit Shean in the Jordan Valley (Crabtree and Kane in James and McGovern 1993: 199–
200). Species identified included caprines, cattle, equid, fallow deer, dog, wild cat, and fish,
indicating that as at the other LB sites both domestic and wild species were common.

At Tel Kinrot a site overlooking the northwestern shore of the Sea of Galilee, the faunal
assemblage (Hellwing 1988–89), contains over 60% caprines and 30% cattle but few pigs
(2%). Another small assemblage from Tel



Ó



arasim, located in the Shephelah (Horwitz 1996a),
contains a similar range of species, and the three predominant domesticates are represented in
frequencies similar to Tel Kinrot. Both assemblages reflect similar patterns of animal exploi-
tation to that found at the site of Tel Michal (fig. 16.2). They contain a broad range of wild ani-
mals: gazelle, deer, hare, fox, birds, and fish. Although Tel Dan broadly shares a similar
macro-phytogeographic zone (Mediterranean vegetation) with these sites (fig. 16.1), the
extremely high cattle frequencies found in the LB layer at this site deviate markedly from the
pattern found elsewhere, suggesting a high degree of species selection as may be expected at
ritual sites (Horwitz 1999).

Some thirty bones were recovered from an LB locus at the City of David, Jerusalem (Horwitz
1996b). Only caprines and cattle were represented, the former comprising 74% of the total sam-
ple. Although little can be said about such a small sample, the relative frequencies of these two
species are similar to those found at Mana



˙



at (fig. 16.2) which lies only a few kilometers away.
The LB assemblage from Tell es-Sharia in the northern Negev (Davis 1982) contained

over 94% caprines, of which goats comprised the majority (75%) relative to sheep. Only 3%
of the assemblage was made up of cattle, while the other remaining 3% comprised equids, pig,
roe deer, and gazelle. Unfortunately the report does not deal with remains of carnivores, birds,
fish, or small mammals which were probably represented at the site. The very southern, and
hence very arid, geographic position of Tell es-Sharia probably accounts for the extremely
high caprine frequencies, especially goat. When this assemblage is compared with two other
north Negev sites, Tell



Ó



alif and Tell Jemmeh, notable differences are observed between
them. At Tell



Ó



alif (Seger et al. 1990: 18–21) both LB I and LB II phases are represented. For
the combined LB periods, caprines constitute some 78% of the assemblage, compared with
94% at Tell es-Sharia. Initially it was thought that the caprine frequencies at Tell es-Sharia
might be inflated as the other minor faunal elements were not included in the calculations.
However, when the relative frequencies of only the three main domesticates were compared,
the caprine frequency remained markedly higher than that of the other Negev sites, suggesting
a real difference in subsistence base. At Tell



Ó



alif cattle form 15% of the assemblage and pig
only 3%. Wild species that are represented include gazelle, birds, and fish, including Mediter-
ranean Sea species. Sheep and goats are represented in equal proportions (Zeder 1990: 26–28).
With regard to the range of species represented and their relative frequencies, Tell



Ó



alif
closely resembles Tell Jemmeh (Wapnish 1982), while Tell es-Sharia represents a subsistence
base adapted to more arid conditions.



Synchronic Trends between Sites



The data presented in this report indicate some synchronic diversity in faunal patterning
between the LB sites discussed here. The relative frequency of species such as caprines or

oi.uchicago.edu





298



LIORA KOLSKA-HORWITZ AND IANIR MILEVSKI



cattle show a clear relation to the geographic and environmental position of the sites (fig. 16.1)
and reflect the nutritional and water requirements of the different animal species. Caprines
(goats especially) are the best-suited species for herding in marginal conditions, while cattle
are least adapted to such regimes and require extensive human intervention (Redding 1981).
Consequently, LB sites in phytogeographic zones not suitable for the keeping of cattle, such
as the hill and arid regions, have relatively low cattle frequencies (8% at Shiloh, 10% at
Mana



˙



at, 10% at Tel



Ó



alif, 16% at Tell Jemmeh, 3% at Tell es-Sharia) compared with sites in
prime cattle-grazing areas such as Tel Dan (about 60%), Tel Kinrot (about 30%), Tel Harassim
(about 30%), and Tel Michal (40%). In contrast, the sites in the hill country and arid regions
have higher caprine frequencies (fig. 16.2).

Where all faunal elements have been collected and studied, all the LB assemblages contain
a wide range of wild faunal species including hunted animals (deer, gazelle, antelopes), those
obtained through gathering and trapping (small mammals, carnivores, and birds) and others
obtained by fishing.



Cultic Sites



In addition to Shiloh, Tel Dan, and Lachish, faunal reports have been published for two
other LB cultic sites or temples: Timna, in the southern Negev (Lernau 1988a, 1988b), and
Nahariya, on the northern coastal plain of Israel (Ducos 1968). The Timna site contains
remains of caprines—sheep (44%) and goats (56%)—as well as remains of freshwater and
Red Sea fish. The Nahariya Temple contained remains of caprines, predominantly goat (87%),
cattle (11%), pig, and fox. When we compare these assemblages with those derived from cul-
tic contexts at Lachish, Shiloh, and Tel Dan, it is evident that there is a high level of intersite
variation, both in the species represented as well as in their relative frequencies. The unifying
feature appears to be geographic; sites in arid areas such as in the Negev or the hill country
have very high caprine and low pig and cattle frequencies. Those lying in the well-watered
Mediterranean zone show an inverse pattern. Consequently, the LB populations appear to have
exploited for ritual purposes the species most commonly represented in their herds—caprines
in the hill country (Shiloh) and in the desert (Timna), versus cattle in the Mediterranean region
(Tel Dan and Lachish). However, at Nahariya, a northern coastal site where we would have
expected a predominance of cattle, an opposite pattern is seen with a predominance of
caprines. There appears, then, to be no single pattern of animal exploitation at the LB cult sites
discussed here.

GENERAL CONCLUSIONS

The data presented here from the six “core” sites show a clear, diachronic shift in subsis-
tence strategies from the MB to the LB. The LB assemblages are characterized by an overall
increase in the frequency of cattle and an associated decrease in the relative frequencies of
pigs and caprines. Sheep are more prevalent in the LB sites than in those of the MB, while the
contribution of wild fauna in LB assemblages increases. This pattern is corroborated by the
finds from the other LB sites discussed here.

When these data are compared with Redding’s model (1991, 1993), they exhibit marked
similarities. According to his model, societies practicing intensive agriculture will be charac-

oi.uchicago.edu





299



THE FAUNAL EVIDENCE FOR SOCIOECONOMIC CHANGE



terized by a high frequency of cattle. Cattle are preferred because (1) they are needed for agri-
cultural labor; (2) the dung they produce can be used for fuel; (3) they are relatively very
efficient milk producers; (4) they require less human control than either pigs or caprines; and
(5) they require high quality forage that can be provided as a by-product of agriculture (sur-
pluses and fallow field grazing). As the intensity of agricultural production increases, and the
accompanying reliance on cattle, so pig and caprine frequencies will decline (Redding 1993).

The main reason for the decline in pig frequencies is that they can cause enormous damage
to the crops. Consequently, raising pigs together with intensive agriculture is only cost effec-
tive if enough surplus crops can be grown to feed the pigs so that they do not compete with
people over food, and they can be efficiently kept away from the fields (Redding 1991). As
pigs are unsuited to long distance transhumance (Zeder 1996), this last criteria may be difficult
to meet. An additional feature that makes pig raising problematic is their dependence on a
daily intake of water. In areas such as the southern Levant where water sources are frequently
limited, this may result in pig raising and intensive agriculture being mutually exclusive. Fur-
thermore, watering the animals on a daily basis involves a high cost—human labor—which
may be more profitably invested in agriculture that is seasonal. Pigs produce fewer products
than cattle or caprines (only meat and skins) and have to be slaughtered in order to obtain
these items, making them relatively high in cost relative to the other two species which pro-
vide, in addition to meat and skin, milk, wool/hair, and labor, the last three as secondary prod-
ucts. In this regard, Redding (1991) notes that relative to their intake of feed protein, cattle
milk provides more animal protein than pork.

Finally, if Redding (1991) is correct and the Egyptian elite refrained from consuming pig,
then this might have supplied both a sociopolitical as well as market-oriented motive for
decreasing the extent of swine herding. Zeder (1996) has proposed that under conditions of
agricultural intensification, pig raising would still have proved profitable on the small-scale
level of individual households through the use of sty-raised pigs. Indeed, this factor may
account for the continued presence of pigs at all LB sites and their relatively higher frequency
at the village site of Mana



˙



at.
As cattle become more important (i.e., as agricultural intensification increases), so the

frequency of caprines in the assemblage should decrease since caprine herding requires a
nomadic or seminomadic lifestyle, which is in direct contradiction to the sedentary require-
ments of intensive agriculture. The latter entails almost year-round crop activities as well as
the protection of boundaries and water rights, both features that are less stringently enforced
in pastoral regimes. Redding (1991) has noted that as cattle and sheep compete for similar
resources, as the one species increases in frequency, so the other will be reduced. However,
the LB data presented here show a positive correlation between the two, with both sheep and
cattle frequencies increasing in this period.

This apparent contradiction of the model can perhaps be explained by several factors.
First, Redding’s initial model of increased goat herding relative to sheep comprised not only
an increase in agricultural intensity, but also an increase in settlement density (Redding 1981:
260). This feature was not discussed in the later publication (Redding 1991) and is in contrast
to the settlement data for the LB. Second, sufficient agricultural surplus was produced in the
LB to cope with the interspecies competition for food. Third, with the exception of iron con-
tent, sheep milk is nutritionally richer than that of goats, although the latter produce more milk

oi.uchicago.edu





300



LIORA KOLSKA-HORWITZ AND IANIR MILEVSKI



(Redding 1981). In addition, in the Near East mutton rather than goat flesh is generally pre-
ferred for consumption, probably because it yields more calories per kilogram due to its higher
fat content (Redding 1981). Consequently, it is possible that if caprines were being raised for
markets, then mutton production was more economically viable than that of goats.

Lastly, it is possible that the demand for wool increased such that the keeping of sheep,
rather than goats, proved economically beneficial. It should be borne in mind that by keeping
sheep, the producer has access to all three usable and marketable caprine products: meat, milk,
and wool. In Crete, from ca. 1400



b.c.



onward, during the Mycenaean occupation of the
island, we have evidence for a flourishing wool industry (Killen 1964), while the earliest con-
vincing evidence for wool sheep in Egypt dates to the New Kingdom (Ryder 1983). These
data suggest that by 1400



b.c.



, there was an interest in wool production both in Egypt and in
the Aegean. Taking into account the political and commercial contacts between these regions
and the Levant during the LB, it is possible that they influenced the orientation of Levantine
caprine production toward one aimed at wool production.

An example of this may be found in the Egyptian tribute and booty lists of Levantine live-
stock during the New Kingdom. In the booty taken from Megiddo, 20,500 sheep and 2,000
goats are mentioned (Pritchard 1954: 237), while in the Thutmosis III tribute list of years 29–
40, about 21,670 sheep and goats are collected (Breasted 1962: §§447, 462, 471, 482). The
ratios of goats to sheep is 1:10 in the booty of Megiddo and 1:7 in the lists of the tribute and
booty. The last ratio seems to be closer to the archaeological data of the sites.

As a society increases its involvement in pastoralism, an inverse trend to that outlined
above occurs, with an increase in the frequency of caprines, and a decrease in the frequencies
of both pig and cattle (Redding 1993). As can be understood, pig frequencies decrease in both
instances, except in situations where the economy is essentially deurbanized and made up of
private households involved in subsistence farming (Zeder 1996).

We would argue that the pattern of animal exploitation observed in the majority of the LB
sites closely follows that outlined in Redding’s model, being indicative of a population
involved in intensive agriculture on a greater scale than previously practiced in the MB. Popu-
lation demography for this period supports this claim, as we see a general movement away
from areas least suited for agriculture, in the interior and hill country. Further corroboration is
provided by the available palaeobotanical information which indicates that during the LB cul-
tivated areas probably existed at Tel Michal (Lipschitz and Waisel 1989; Thompson and Rapp
1989), Mana



˙



at (Edelstein and Milevski 1994), and Lachish (Ussishkin 1978; Drori 1979).
Data for Shiloh are limited for LB, but Lipschitz (1993) has indicated that the abundance of
olive trees suggests that they were cultivated. At Tell



Ó



alif (Seger et al. 1990: 26) and Tel



Ó



arasim (Givon 1993), large silos were found in the LB II strata. Their size and number sug-
gests not only the importance of agriculture in these societies, but also the importance of stor-
age of agricultural produce, perhaps surpluses. It seems feasible to suggest that the need for
agricultural surpluses was a direct result of the economic demands exerted by the Egyptians.
Perhaps, as proposed by Falconer (1994), it was this very factor that played a critical role in
the decline of the LB urban communities.

The six “core” sites studied here all show continuity of occupation from the MB to the LB,
but discontinuity in subsistence strategies. The innovation of intensive agriculture observed in
these assemblages correlates well with what Hole (1994) sees as the main option open to a
society that chooses to remain in the same location during a period of economic instability or

Long

oi.uchicago.edu





301



THE FAUNAL EVIDENCE FOR SOCIOECONOMIC CHANGE



stress. Additional evidence for economic stress and instability at these sites is seen in their
increased reliance on wild animals: hunted, gathered, or fished. This too supports Hole’s
(1994) claim that people undergoing a period of economic stress turn to alternative subsis-
tence forms. There is, however, little faunal evidence to support a widespread return to pasto-
ralism. The animal remains from sites in the hill country (Shiloh and Mana



˙



at), with the
exception of the extremely high frequency of caprines at Shiloh, do not deviate from the gen-
eral pattern of animal exploitation observed at the other LB sites. As discussed above, this
may be due to factors other than a reversion to pastoralism. These conclusions must remain
tentative until further sites are excavated in the hill region.

This paper has offered a theoretical framework against which to assess the nature and
extent of changes in the transition from the MB to the LB. We feel confident that our data sup-
port the archaeological scenario of the LB as a period of impoverishment and recession. The
seemingly anomalous presence of luxury items at many of the LB sites, cited by Liebowitz
(1987, 1989) as evidence for prosperity, may be accounted for with reference to an analogous
situation: the collapse of Early Bronze Age sociopolitical systems in southeast Turkey (Watten-
maker 1994). In her case study, Wattenmaker (1994) illustrates that the system of specialist
(elite) economies, i.e., specialist-produced goods and the extraction of surpluses, continued
unaffected despite the disintegration of the political system in the region, which shifted from
a centralized hierarchical one to that based on increased local autonomy. It is possible that in
the LB of the southern Levant a similar situation occurred, with small households continuing
to be involved in the production of specialist goods as long as there was a demand, possibly
from the resident Egyptian population or for international markets. Consequently, these data
need not stand in contradiction to the interpretation of the LB in the southern Levant as repre-
senting a period of socioeconomic decline and impoverishment.

REFERENCES



Aharoni, Y.

1982



The Archaeology of the Land of Israel.



Trans. A. F. Rainey. Philadelphia: Westminster.
Albright, W. F.

1960



The Archaeology of Palestine.



Harmondsworth: Penguin.
Bienkowski, P.

1989 Prosperity and Decline in LBA Canaan: A Reply to Liebowitz and Knapp.



Bulletin of the Ameri-
can Schools of Oriental Research



275: 59–63.
Biran, A.

1993 Dan. Pp. 323–32 in



The New Encyclopedia of Archaeological Excavations in the Holy Land



, ed.
E. Stern. Jerusalem: Israel Exploration Society & Carta and New York: Simon & Schuster.

1996 Part I: A Chronicle of the Excavations 1966–1992. Pp. 7–63 in



Dan







I



.



A Chronicle of the Ex-
cavations, the Pottery Neolithic, the Early Bronze Age and the Middle Bronze Age Tombs



, eds.
A. Biran, D. Ilan, and R. Greenberg. Jerusalem: Nelson Glueck School of Biblical Archaeology,
Hebrew Union College.

Breasted, J. H.

1962



Ancient Records of Egypt



, vol. 2. Chicago: University of Chicago Press.
Bunimovitz, S.

1993 The Study of Complex Societies: The Material Culture of Late Bronze Age Canaan as a Case
Study. Pp. 443–51 in



Biblical Archaeology Today



, 1990,



Proceedings of the 2nd International
Congress of Biblical Archaeology



, eds. A. Biran and J. Aviram. Jerusalem: Israel Exploration
Society.

oi.uchicago.edu





302



LIORA KOLSKA-HORWITZ AND IANIR MILEVSKI



1994 Socio-Political Transformations in the Central Hill Country in the Late Bronze-Iron I Transition.
Pp. 179–202 in



From Nomadism to Monarchy. Archaeological and Historical Aspects of Early
Israel



, eds. I. Finkelstein and N. Naªaman. Jerusalem: Yad Itzhak Ben-Zvi and Israel Explora-
tion Society.

Crabtree, P. J.
1990 Zooarchaeology and Complex Societies: Some Uses of Faunal Analysis for the Study of Trade,

Social Status and Ethnicity. Pp. 155–206 in



Archaeological Method and Theory



, ed. M. B.
Schiffer. Tucson: University of Arizona Press.

Davis, S. J.
1982 Climatic Change and the Advent of Domestication: The Succession of Ruminant Artiodactyls in

the Late Pleistocene-Holocene in the Israel Region.



Paléorient



8: 5–15

de Vaux, R.
1978



The Early History of Israel



. Trans. David Smith. Philadelphia:Westminster.

Drori, I.
1979 Tel Lachish: Subsistence and Natural Environment during the Middle Bronze, the Late Bronze

and Iron Age Periods. M.A. thesis, Tel Aviv University (Hebrew).

Ducos, P.
1968



L’origine des animaux domestiques en Palestine.



Publications de l’Institut de Préhistoire de
l’Université de Bordeaux Memoire 6.

Edelstein, G., and Milevski, I.
1994 The Rural Settlement of Jerusalem Re-evaluated: Surveys and Excavations in the Rephaim Val-

ley and Mevasseret Yerushalayim.



Palestine Exploration Quarterly



126: 1–25.

Edelstein, G.; Milevski, I.; and Aurant, S., eds.
1998



Villages, Terraces and Stone Mounds: Excavations at Mana



˙



at, Jerusalem, 1987–1989



. Pp.
104–12 in Israel Antiquities Authority Reports 3. Jerusalem: Israel Antiquities Authority.

Falconer, S. E.
1994 The Development and Decline of Bronze Age Civilization in the Southern Levant: A Reassess-

ment of Urbanism and Ruralism. Pp. 305–33 in



Development and Decline in the Mediterranean
Bronze Age



, eds. C. Mathers and S. Stoddart. Sheffield Archaeological Monograph 8. Sheffield:
Sheffield Academic Press.

Finkelstein, I.
1988



The Archaeology of the Israelite Settlement



. Jerusalem: Israel Exploration Society.
1993 The Sociopolitical Organization of The Central Hill Country in the Second Millennium



b.c.e.



Pp. 110–31 in



Biblical Archaeology Today, 1990 Supplement.



Jerusalem: Israel Exploration
Society.

1994 The Emergence of Israel: A Phase in the Cyclic History of Canaan in the Third and Second Mil-
lennia



b.c.e.



Pp. 150–78 in



From Nomadism to Monarchy: Archaeological and Historical
Aspects of Early Israel



, eds. I. Finkelstein and N. Naªaman. Jerusalem: Yad Itzhak Ben-Zvi and
Israel Exploration Society.

Finkelstein, I.; Bunimovitz, S.; and Lederman, Z.
1993



Shiloh. The Archaeology of a Biblical Site



. Monograph 10. Tel Aviv: Tel Aviv University, Insti-
tute of Archaeology.

Givon, S., ed.
1993



The Third Season of Excavation at “Tel Harasim” 1992. Preliminary Report 3



. Tel Aviv: Tel
Aviv University, Institute of Archaeology.

Gonen, R.
1984 Urban Canaan in the Late Bronze Period.



Bulletin of the American Schools of Oriental Research



253: 61–73.

Gophna, R., and Ayalon, E.
1989 History of Settlement in the Tel Michal Region. Pp.16–28 in



Excavations at Tel Michal, Israel



,
eds. Z. Herzog, G. Rapp, Jr., and O. Negbi. Tel Aviv: University of Minnesota Press and Tel
Aviv University.

oi.uchicago.edu





303



THE FAUNAL EVIDENCE FOR SOCIOECONOMIC CHANGE



Hellwing, S.
1988–89 Faunal Remains from the Early Bronze and Late Bronze Ages at Tel Kinrot.



Tel Aviv



15–16:
212–20.

Hellwing, S., and Feig, N.
1989 Animal Bones. Pp. 236–47 in



Excavations at Tel Michal, Israel



, eds. Ze. Herzog, G. Rapp Jr.,
and O. Negbi. Tel Aviv: University of Minnesota Press and Tel Aviv University.

Hellwing, S.; Sadeh, M.; and Kishon, V.
1993 Faunal Remains. Pp. 309–50 in



Shiloh: Archaeology of a Biblical Site



, eds. I. Finkelstein,
S. Bunimovitz, and Z. Lederman. Monograph 10. Tel Aviv: Tel Aviv University, Institute of
Archaeology.

Herzog, Z.; Rapp, G., Jr.; and Negbi, O., eds.
1989



Excavations at Tel Michal, Israel



. Tel Aviv: University of Minnesota Press and Tel Aviv
University.

Hesse, B., and Wapnish, P.
1997 Can Pigs be Used for Ethnic Diagnosis in the Ancient Near East? Pp. 238–70 in



The Archaeol-
ogy of Israel: Constructing the Past, Interpreting the Present



, eds. N. A. Siberman and D. B.
Small. Journal for the Study of the Old Testament. Supplement Series 237. Sheffield: Sheffield
Academic Press.

Hole, F.
1994 Environmental Instabilities and Urban Origins. Pp. 121–51 in



Chiefdoms and Early States in the
Near East: The Organizational Dynamics of Complexity



, eds. G. Stein and M. S. Rothman.
Monographs in World Prehistory 18. Madison, WI: Prehistory Press.

Horwitz, L. K.
1996a Late Bronze Age Fauna from the 1994 Season at Tel Harasim. Pp. 6*–13* in



The Sixth Season
of Excavation at Tel Harasim (Na



˙



al Barkai) 1995. Preliminary Report 6



, ed. S. Givon. Tel
Aviv: Bar-Ilan University.

1996b Faunal Remains from Areas A, B, D, H and K. Pp. 302–17 in



City of David



, Vol. 4, eds. D. T.
Ariel and A. de Groot. Qedem 35. Jerusalem: Hebrew University.

1998 Faunal Remains. Pp. 104–12 in



Villages, Terraces and Stone Mounds: Excavations at Mana



˙



at,
Jerusalem, 1987–1989



, eds. G. Edelstein, I. Milevski, and S. Aurant. Israel Antiquities Author-
ity Reports 3. Jerusalem: Israel Antiquities Authority.

1999 The Contribution of Archaeozoology to the Identification of Ritual Sites. Pp. 63–69 in



The Prac-
tical Impact of Science on Near Eastern and Aegean Archaeology



, eds. S. Pike and S. Gitan.
Wiener Laboratory Publication No. 3. London: Archetype Publications.

James, F. W., and McGovern, P. E., eds.
1993 Faunal Remains. Pp. 199–200 in The Late Bronze Egyptian Garrison at Beth Shan: A Study of

Levels VII and VIII. Philadelphia: University of Pennsylvania Press.
Kenyon, K.

1960 Archaeology in the Holy Land. London: Praeger.
Killen, J. T.

1964 The Wool Industry of Crete in the Late Bronze Age. Annals of the British School of Archaeology,
Athens 59: 1–15.

Knapp, A. B.
1987 Pots, PIXE, and Data Processing at Pella in Jordan. Bulletin of the American Schools of Oriental

Research 266: 1–30.
1989 Response: Independence, Imperialism, and the Egyptian Factor. Bulletin of the American

Schools of Oriental Research 275: 64–68.
Lernau, H.

1975 Animal Remains. Pp. 86–103 in Investigations at Lachish: The Sanctuary and the Residency
(Lachish V), ed. Y. Aharoni. Tel Aviv: Institute of Archaeology, Tel Aviv University.

1988a Fish Remains. Pp. 246–52 in The Egyptian Mining Temple at Timna, ed. B. Rothenberg. London:
Institute of Archaeology and University College.

1988b Mammalian Remains. Pp. 241–46 in The Egyptian Mining Temple at Timna, ed. B. Rothenberg.
London: Institute of Archaeology and University College.

oi.uchicago.edu



304 LIORA KOLSKA-HORWITZ AND IANIR MILEVSKI

Liebowitz, H.
1987 Late Bronze II Ivory Work in Palestine: Evidence of a Cultural Highpoint. Bulletin of the Ameri-

can Schools of Oriental Research 265: 3–24.
1989 Response: LB IIB Ivories and the Material Culture of the Late Bronze Age. Bulletin of the

American Schools of Oriental Research 275: 63–64.
Lipschitz, N.

1993 Paleobotanical Remains. Pp. 351–61 in Shiloh: The Archaeology of a Biblical Site, eds. I. Finkel-
stein, S. Bunimovich, and Z. Lederman. Monograph 10. Tel Aviv: Tel Aviv University, Institute
of Archaeology.

Lipschitz, N., and Waisel, Y.
1989 Botanical Remains. Pp. 219–22 in Excavations at Tel Michal, Israel, eds. Z. Herzog, G. Rapp, Jr.,

and O. Negbi. Tel Aviv: University of Minnesota Press and Tel Aviv University.
Ludwig, J., and Reynolds, J.

1988 Statistical Ecology. London: Wiley.
Orni, E., and Efrat, E.

1971 Geography of Israel. 3rd rev. ed. Jerusalem: Israel Universities.
Pritchard, J. B.

1954 The Ancient Near East in Pictures Relating to the Old Testament. Princeton: Princeton
University Press.

Redding, R.
1981 Decision Making in Subsistence Herding of Sheep and Goats in the Middle East. Ph.D. disser-

tation. University of Michigan. Ann Arbor: University Microfilms.
1991 The Role of Pig in the Subsistence System of Ancient Egypt: A Parable on the Potential of Fau-

nal Data. Pp. 20–30 in Animal Use and Culture Change, eds. P. J. Crabtree and K. Ryan.
Museum Applied Science Center for Archaeology Research Papers in Science and Archaeology,
Supplement 8. Philadelphia: University of Pennsylvania Press.

1993 Subsistence Security as a Selective Pressure Favoring Increasing Cultural Complexity. Bulletin
on Sumerian Agriculture 7: 77–98.

Rosen, A.
1986 Environmental Change and Settlement at Tel Lachish, Israel. Bulletin of the American Schools

of Oriental Research 263: 55–60.
1998 Phytolith Remains from Selected Loci. Pp. 119-21 in Villages, Terraces and Stone Mounds:

Excavations at Mana˙at, Jerusalem, 1987-1989, eds. G. Edelstein, I. Milevski, and S. Aurant.
Israel Antiquities Authority Reports 3. Jerusalem: Israel Antiquities Authority.

Rosen, B.
1993 Economy and Subsistence. Pp. 362–67 in Shiloh: The Archaeology of a Biblical Site, eds.

I. Finkelstein, S. Bunimovitz, and Z. Lederman. Monograph 10. Tel Aviv: University of Tel
Aviv, Institute of Archaeology.

Ryder, M. L.
1983 Sheep and Man. London: Duckworth.

Seger, J. D.; Brent, B.; Borowski, O.; Cole, D. P.; Forshey, H.; Futato, E.; Jacobs, P. F.; Laustrup, M.; O’Connor
Seger, P.; and Zeder, M.

1990 The Bronze Age Settlements at Tell Óalif: Phase II Excavations, 1983–1987. Preliminary
Reports of American Schools of Oriental Research Sponsored Excavations 1983–1987. Bulletin
of the American Schools of Oriental Research Supplement 26: 1–32.

Thompson, M., and Rapp, G. R.
1989 Paleobotany from Phytoliths. Pp. 223–25 in Excavations at Tel Michal, Israel, eds. Z. Herzog,

G. Rapp, Jr., and O. Negbi. Tel Aviv: University of Minnesota and Tel Aviv University.
Ussishkin, D.

1978 Excavations at Tel Lachish, 1973–1977: Preliminary Report. Tel Aviv 5: 1–97.
1993 Lachish. Pp. 897–902 in The New Encyclopedia of Archaeological Excavations in the Holy

Land, ed. E. Stern. Jerusalem: Israel Exploration Society & Carta and New York: Simon &
Schuster.

oi.uchicago.edu



305THE FAUNAL EVIDENCE FOR SOCIOECONOMIC CHANGE

van Beek, G.
1993 Tell Jemmeh. Pp. 667–67 in The New Encyclopedia of Archaeological Excavations in the Holy

Land, ed. E. Stern. Jerusalem: Israel Exploration Society & Carta and New York: Simon &
Schuster.

Wapnish, P.
1982 Archaeozoology at Tell Jemmeh: Taphonomy and Paleoeconomy in Historic Archaeology.

Unpublished Research Report, National Geographic.
1993 Archaeozoology: The Integration of Faunal Data with Biblical Archaeology. Pp. 426–42 in

Biblical Archaeology Today, 1990. Proceedings of the 2nd International Congress of Biblical
Archaeology, eds. A. Biran and J. Aviram. Jerusalem: Israel Exploration Society.

Wapnish, P., and Hesse, B.
1988 Urbanization and Organization of Animal Production at Tell Jemmeh in the Middle Bronze Age

Levant. Journal of Near Eastern Studies 47: 81–94.
1991 Faunal Remains from Tel Dan: Perspectives on Animal Production at a Village, Urban and Ritual

Center. Archaezoologia 4: 9–86.

Wapnish, P.; Hesse, B.; and Ogilvy, A.
1977 The 1974 Collection of Faunal Remains from Tel Dan. Bulletin of the American Schools of

Oriental Research 227: 35–62.

Wattenmaker, P.
1994 Political Fluctuations and Local Exchange Systems in the Ancient Near East: Evidence from the

Bronze Age Settlements at Kurban Höyük. Pp. 193–208 in Chiefdoms and Early States in the
Near East: The Organizational Dynamics of Complexity, eds. G. Stein and M. S. Rothman.
Monographs in World Prehistory 18. Madison, WI: Prehistory Press.

Zeder, M. A.
1990 Animal Exploitation at Tell Óalif. Pp. 24–30 in The Bronze Age Settlements at Tell Óalif:

Phase II Excavations, 1983–1987, J. D. Seger et al. Preliminary Reports of American Schools
of Oriental Research Sponsored Excavations 1983–1987. Bulletin of the American Schools of
Oriental Research Supplement 26: 1–32.

1991 Feeding Cities. Washington, DC: Smithsonian Institution.
1996 The Role of Pigs in Near Eastern Subsistence: A View from the Southern Levant. Pp. 297–312

in Retrieving the Past: Essays on Archaeological Research and Methodology in Honor of Gus
Van Beek, ed. J. D. Seger. Mississippi: Cobb Institute of Archaeology.

oi.uchicago.edu



306 LIORA KOLSKA-HORWITZ AND IANIR MILEVSKI

oi.uchicago.edu



307



THE RIDDLE OF STRUCTURE 5239 AT
MEGIDDO, STRATUM IX



David Ilan



As a product of the University of Chicago and a scholar of Syro-Palestinian archaeology,
Douglas Esse was keenly interested in Megiddo (Tell el-Mutesellim) and knew the site well.
His work on the Early Bronze Age (EB) led him to reanalyze the Megiddo phases and early
strata in light of the expanded horizons achieved in his doctoral work. One of his last projects
was the renewed study of the Iron I levels based on unpublished data in the Oriental Institute
archives. This latter study resulted in ideas that hold great explanatory potential for recon-
structing Iron I society and social processes (e.g., Esse 1992). Doug would certainly have been
enthusiastic about the new excavations at Megiddo directed by Israel Finkelstein and David



Ussishkin,



1



and I think his sense of adventure would have been piqued by the following inves-
tigation, if tempered by a wry, good-humored skepticism.

For all the myriad studies of Megiddo, the site remains an archaeological enigma due to
the great extent of its excavation, the wealth of its finds, and the incomplete nature of its pub-
lication. At the top of the tell there is now a small bluff that provides the best view over the
EB III double megaron temples of Stratum XV (Loud 1948: fig. 394; Kempinski 1989: fig.
14). It is located in the northwestern portion of the University of Chicago expedition’s Area
BB, mostly in their grid square M12. A few meters east of this bluff, and south of the path that
leads down into Area BB, in the northeast corner of square M12 (fig. 17.1) lie



[t]hree deep chambers of a single unit, 5239 (figs. 244–45), substantially built and carefully paved
with excellent lime floors 4.50 m below the level to which the other walls had been destroyed.
(Loud 1948: 102: fig. 401)



The southernmost of these chambers is larger than the other two. The remains are still
remarkably well preserved and easily visible from all sides (pl. 17.1). What, then, is this finely
constructed shaft, and to what, if any, structural configuration did it belong?

Strangely enough, this rather conspicuous feature has received little attention. Moreover,
in every study of its contemporary remains, it has been left an architectural orphan. Chicago’s
Megiddo expedition attributed the shaft-like structure to Stratum IX (Late Bronze Age I). The
tops of its walls (top elevation 159.85 meters above sea level) were apparently sealed under a
clear Stratum VIII floor—presumably Locus 5227 in Loud’s (1948) figure 402. The triple-
shafted configuration penetrated down to “absolute levels equivalent to or below the general
level of Stratum XV,” base elevation being 155.35 meters above sea level. It is further noted
that “with no openings in the walls, they [the shafts] must have been storage pits accessible
from above only” (Loud 1948: 102).



1. The author participated as an area supervisor in the first full-fledged season of the new Megiddo excavations
in 1994. This contribution grew out of the days spent walking about the tell with colleagues and old plans.



17



oi.uchicago.edu





308



DAVID ILAN



The register of finds (Loud 1948: 187) lists only a modest artifactual assemblage, but it
includes six complete pottery vessels, a bronze spearhead, and a bronze toggle pin (table 17.1,
fig. 17.2).



2



All the reported finds were dated to the early part of the Late Bronze Age (LB). It
is submitted here that Structure 5239 was a tomb. This idea is not surprising in itself, given the
general proclivity at Megiddo (and elsewhere) for sinking tombs under almost every Middle
Bronze Age (MB) and early LB edifice. Parallels for a tomb configuration of this kind are few,
but certain features do have counterparts. Closest to home is T.3063, located in square O14 of
Area BB and assigned to Stratum X by the Chicago expedition (Loud 1948: 170, 192, fig.
400). In the Megiddo Locus Registry it was termed “a burial in a chimney” (Kassis 1973: 14)
and, together with its surrounding features, is a virtual replica of the Structure 5239 plan. Kas-
sis recommended that it be considered as sunk into the debris of Stratum X from a surface of
Stratum IX, and Gonen (1992: 107) concurs. It too contains a typical burial offering assem-
blage, including several Bichrome Ware vessels and a silver bracelet.

Another crypt construction, less similar but more striking in scale, occurs in the Alalakh
Stratum VII “Palace of Yarim-Lim” (Woolley 1955: 95–97, fig. 35, pls. 20–22), dated to MB
II–III. The general configuration and size of Room 17 and the rooms around it resembles that
of the Megiddo unit. In contrast to Structure 5239, Room 17 itself was built of basalt ashlars
in its lower portion, with a single row of rubble masonry above, followed by brick construction
almost 3 m high—typical of the palace walls. The subterranean room was accessed via a stone
slab stairway, or dromos, leading down to a basalt door slab on hinges. The jambs and lintel
were also basalt ashlars. But the doorway of Room 17 was blocked from the inside (Woolley
1955: pl. 20b). Whoever had done so would have had to exit vertically out of the shaft. Indeed,



2. Only complete or otherwise outstanding objects were published in the Megiddo reports. It is also clear that
much of what today would have been sent for restoration was never even registered. During the 1994 season
of the new excavations, several important small finds and many large, typologically indicative sherds—
many, probably from restorable vessels—were uncovered in the University of Chicago backfills.



Table 17.1.



Register of Finds from Structure 5239, Megiddo Stratum IX
(from Loud 1948: 187; corresponds to fig. 17.2)



Object Type Field no. Plate Figure



17.2



here



Jug 354 d 745 51:9, 133:21 2

Bowl 156 d 747 54:8 6

Bowl 246 d 746 53:13 1

Bowl 260 d 707 54:21, 134:10 7

Cooking bowl 10 d 749 55:2 5

Chalice 12 d 748 55:16, 134:16 8

Arrowhead — d 781 174:6 3

Toggle pin — d 780 223:61 4

oi.uchicago.edu





309



THE RIDDLE OF STRUCTURE 5239 AT MEGIDDO, STRATUM IX



the chamber was found intentionally filled. There was no clear floor over this fill, but neither
was there any evidence for the chamber’s having been roofed. It was Woolley’s contention that
the dromos and chamber were intended as ephemeral entities, never to be accessed again (he
suggested the possibility of a foundation offering). His description of Room 17’s hard, well-
paved, lime cement floor matches that of the Structure 5239 floor.

Woolley was mystified as to the purpose of this chamber, despite the presence of four adult
skeletons—and fragments of a child’s—contained in a wooden box set against the southwest
wall. He took issue with the parallels drawn by Schaeffer between Room 17 and the chamber-
and-dromos tombs of Ugarit (Recent), citing structural differences and the dearth of burial
goods in the subterranean chamber at Alalakh, which seemed to contrast with the quality of its
construction. In this context, one must remember that Woolley had excavated the fabulously
wealthy shaft tombs of Ur. In retrospect, his skepticism seems unwarranted; the skeletons,
after all, speak for themselves. Rather, it is the differences in technique and the poverty of the
burial offerings that require explanation. Several hypotheses might be forwarded, but none
could be easily substantiated.

At Megiddo, Schumacher’s (1908) T.1 and T.2—dated to MB III—are also reminiscent of
the Alalakh Room 17 configuration in that they were constructed together with the overlying
architecture as a planned unit. Though we now have no way of knowing, Structure 5239 may



FIGURE 17.1. Stratum IX Area BB at Tell Megiddo as restored by Kempinski (1989: Plan 6; hatched lines are
his reconstruction). Structure 5239 is located in the northeastern corner of Square M12.

oi.uchicago.edu

[image: image98.tif]



310



DAVID ILAN



have been similarly integrated, especially if the superstructure walls were of brick that was not
preserved or not detected by the excavators (but see the comments below concerning Kemp-
inski’s and Baumgarten’s reconstructions).

The “royal” tombs found under the western palace of Tell Mardikh Stratum IIIB may be
seen as conceptually similar to Structure 5239 in their location and in the shaft entries of some
(e.g., Matthiae 1980). Indeed, the fact that dromos entries coexisted with shaft entries is
instructive in itself. The same concept was probably in force with regard to the contempora-
neous subfloor chamber tombs at Ugarit, and particularly those within the confines of the
palace (e.g., Salles 1987; Schaeffer 1939; Yon 1990).

The artifactual assemblage of Structure 5239 is a fairly typical one for a LB tomb at
Megiddo and elsewhere (for a compilation of Megiddo assemblages see Gonen 1992: 98–
118). The cooking pot is highly unusual in the mortuary contexts of other periods. In LB,
however, it is ubiquitous, and especially so at Megiddo (e.g., Guy 1938: 155 and various
plates; Gonen 1992: 49). The toggle pin also is particularly characteristic of the MB and LB
tomb offerings; at Megiddo 80% of the LB and MB toggle pins reported originated in tombs
(forty-two out of a total of sixty-four items illustrated in Loud 1948 [pls. 219–23], and a total
of fifty in Guy 1938). Similar impressions can be gathered from contemporaneous sites. Like-
wise, the rest of the Structure 5239 assemblage is at home in other LB I–II tomb assemblages.



FIGURE 17.2. Finds from Structure 5239 (after Loud 1948). Pottery is drawn to a scale of 1:5 and metal objects
to a scale of 1:1.

oi.uchicago.edu

[image: image99.tif]



311



THE RIDDLE OF STRUCTURE 5239 AT MEGIDDO, STRATUM IX



While its context does not facilitate making a clear connection between the two, the corpus
of finds from the Stratum VIII loci immediately above the subterranean structure is also sug-
gestive (see table 17.2, fig. 17.3). Despite the fragmentary nature of the architecture, these
artifacts seem to imply that the context was not a simple domestic one. Perhaps these were
remains of an elite Stratum VIII structure (see below). But the assemblage may also have been
related to the underlying tomb, perhaps as part of an ancestor cult (cf. Hallote 1994). Stands,
whorls, and curved knives are most prominent in the LB tombs (Guy 1938: 152, 165, 170, and
plates). In fact, Structure 5239 may have been inserted from Stratum VIII; there does not seem
to be anything in the artifact repertoire to discount such a notion.

The lack of any mention of bones in the final report is acknowledged as a weakness in this
hypothesis—one that requires an explanation. At least one tomb at Megiddo (T.2034, assigned
to Stratum X) was also identified as such though it contained no bones (Kassis 1973: 9). Tomb
4663 at Tel Dan was reported as lacking any skeletal remains, though it clearly was intended
as a tomb (Biran 1986). In preparing the material from T.4663 for final publication, the animal
bone assemblage was submitted for analysis to L. Kolska-Horwitz, who identified a few
human left metacarpal bones (Ilan 1996: 172). Biran ventured that this tomb might have been
a sort of cenotaph, following a similar proposition made by Petrie (1932: 6) for a feature at
Tell el-Ajjul. Elsewhere, it has been suggested that the lack of human remains in T.4663 could
be the work of scavenging animals (porcupines in particular), diagenesis, or ancient plunder-
ers (Ilan 1996: 176). The tombs reported and illustrated in Loud 1948







all show good skeletal
preservation. The question is, were highly fragmentary remains of human skeletal material
discerned or collected and recorded by the workmen and the Chicago Megiddo Expedition
staff? In the reports of the burial assemblages on the tell, there are few examples of such frag-
mentary material. In any event, the skeletal material from Megiddo’s intramural tombs was
not published in any detail by a specialist, unlike the assemblages of the extramural cemetery
published by Hrdli



c



ka in



Megiddo Tombs



(1938).
Finally, we must ask to what architectural and sociopolitical context this superb shaft tomb

belonged. On the one hand, its location in the northwestern corner of Area BB has always
limited the potential of plan reconstruction. Square M12 was excavated down to widely varying
levels; half of the square—the northwestern part—was taken down only to Stratum IV (Loud
1948: fig. 377). On the other hand, Temple 2048—apparently constructed in Stratum X—seems
to have focused attention away from Structure 5239 and every other fragment of architecture

.



Table 17.2.



Register of Finds from L.5227 and L.5262 (Stratum VIII) Sealing Structure 5239
(from Loud 1948: 187–88; corresponds to fig. 17.3)



Object Type Field no. Locus Plate Figure



17.3



here



Stand 12 d652 5227 62:12, 137:2 3

Scarab — d656 5227 152:164, 158:164 4

Blade (bronze) — d663 5227 179:32 5

Whorl (bone) — d796 5262 172:24 2

“Wand” (ivory) — d795 5262 203:1 1

oi.uchicago.edu





312



DAVID ILAN



in this zone. The shaft feature was ignored by Kenyon (1969) in her reevaluation of the MB and
LB stratigraphy because it did not contain a rich assemblage, because it was not thought to be
a tomb, and because it was one of the few cases in which the Chicago excavators recognized
a subterranean feature that originated in a higher level, therefore requiring no stratigraphic revi-
sion. For similar reasons, Müller (1970), Kassis (1973), and Gonen (1987, 1992), in detailed
reassessments of their own, completely passed it over.

Both Kempinski (1989: plan 6) and Baumgarten (1992: fig. 1) made rather careless
attempts to integrate Structure 5239 into some kind of larger configuration. Kempinski con-
nected it to a single wall fragment in the southwest edge of square M12, while Baumgarten
construed it as the corner of a large, palace-type structure. Evidently, both considered its sub-
terranean walls as surface level foundations or superstructure for the sake of plan reconstruc-
tion. Expanding somewhat on the excavators’ suggestion, Kempinski (1989: 133) interpreted
Structure 5239 as a public granary administered by the personnel of Temple 2048. He com-
pared it to the small square MB buildings, identified by Welter as Egyptian-style granaries,

short



FIGURE 17.3. Finds from the Stratum VIII loci immediately above Structure 5239 (Loci 5227 and 5262; from
Loud 1948). Pottery is drawn to a scale of 1:5, other objects to a scale of 1:1.

oi.uchicago.edu

[image: image100.tif]



313



THE RIDDLE OF STRUCTURE 5239 AT MEGIDDO, STRATUM IX



erected



over



(not down into) the C Embankment at Shechem, east of the “tower temple”
(Wright 1965: 120, fig. 70). Needless to say, the Megiddo report cited no remains of grain in
Structure 5239.



3



In any event, the shaft structure remained essentially stranded. Although Kempinski
(1989: 64) remarked that the architectural changes that took place between Strata X and IX
were slight, his plans and his other statements indicate that he viewed the transformation of
the area northwest of Temple 2048 as more pronounced. He asserted that the erstwhile palace
of this area “fell into disuse” after what had been a long, continuous occupation throughout
most of the MB. However, by Stratum VII the zone was once again built up as a public or
palatial complex (Kempinski 1989: plan 8). The question is, how do we interpret the empti-
ness of this zone in Strata IX and VIII. Or was it so empty?

From Stratum XII to Stratum X, the Area BB MB palace plan tended to change in incre-
ments, maintaining the same northeast-southwest orientation, gradually adding new areas or
subdividing spaces. The few identifiable remains in Strata IX and VIII preserve this orienta-
tion, as does, more prominently but less precisely, the renewed Stratum VIIB palace (Loud
1948: figs. 398–403; Kempinski 1989: plans 3–9; Kenyon 1969: fig. 20). By its very exist-
ence, the fragmentary Stratum IX and VIII architecture suggests that the area was built up and
subsequently greatly disturbed (cf. Baumgarten 1992: fig. 1, aside from the mistaken above-
surface integration of Structure 5239). The fact that so few Stratum IX loci were clean in Area
BB (Gonen 1987: 89) lends weight to this notion. Kassis (1973: 7–8) has pointed out how
similar the plans of Strata X and IX really are and how arbitrary some of the separation
between the two is. Their distinction is mostly based on a clear break observed in Area AA.

It is most probable that the large public or palatial configuration of northwest Area BB
Stratum X continued to exist here in some form in Strata IX and VIII, similar to what was
actually preserved to the east of Temple 2048.



4



This context may reveal something of the
social personae accruing to people who would have been interred in our hypothetical tomb—
the elite of the town and the territory. Gonen (1987: 89) recognized in Stratum IX the most
affluent phase of the LB at Megiddo, comparing it with the contemporaneous levels at Hazor.
The formidable construction of Structure 5239 fits such a characterization and assures that,
despite the travails of over sixty years’ exposure to the elements, the edifice remains an intact
source of wonder for the modern visitor.



3. Of course, as with the missing bones, grain may have existed and gone unreported, or it may have decayed.
The same holds true for the Shechem structures. By the same token, other hypotheses might be submitted to
define Structure 5239. It might have been a cooling cellar, or a water cistern whose wall plaster decomposed
and redeposited on the floor. Moreover, it might have been designed with one of these functions in mind and
later converted into a tomb or depository of some other kind (cf. the cisterns and tunnels of Hazor’s lower
city of the MB and LB periods [Yadin 1972: 38–47]).

4. The identification of Structure 5239 as a tomb, and the assignment of T.3063 to Stratum IX, might require
the amending of two of Gonen’s (1992: 117) assertions: (a) that constructed burials typical of the MB II
were no longer erected in that stratum and (b) that the northwestern part of Area BB was no longer used for
burial by that stratum.

oi.uchicago.edu





314



DAVID ILAN



REFERENCES



Baumgarten, J.
1992 Town Planning in the Late Bronze Age. Pp. 143–50 in



The Architecture of Ancient Israel from
the Prehistoric to the Persian Periods,



eds. A. Kempinski and R. Reich. Jerusalem: Israel Explo-
ration Society.

Biran, A.
1986 The Dancer from Dan, the Empty Tomb and the Altar Room.



Israel Exploration Journal



36:
168–87.

Esse, D.
1992 The Collared Pithos at Megiddo: Ceramic Distribution and Ethnicity.



Journal of Near Eastern
Studies



51: 81–103.

Gonen, R.
1987 Megiddo in the Late Bronze Age—Another Reassessment.



Levant



19: 83–100.
1992



Burial Patterns and Cultural Diversity in Late Bronze Age Canaan



. American Schools of Ori-
ental Research Dissertation Series No. 7. Winona Lake, IN: Eisenbrauns.

Guy, P. L. O.
1938



Megiddo Tombs



. Oriental Institute Publications 33. Chicago: University of Chicago Press.

Hallote, R.
1994 Mortuary Practices and their Implications for Social Organization in the Middle Bronze Age

Southern Levant. Ph.D. dissertation, University of Chicago.

Hrdli



c



ka, A.
1938 Skeletal Remains. Pp. 192–208 in



Megiddo Tombs



, ed. P. L. O. Guy. Oriental Institute Publica-
tions 33. Chicago: University of Chicago Press.

Ilan, D.
1996 The Middle Bronze Age Tombs. Pp. 161–329 in



Dan I: History of Excavation, the Pottery Neo-
lithic, the Early Bronze Age and the Middle Bronze Age Tombs,



eds. A. Biran, D. Ilan, and R.
Greenberg. Annual of the Hebrew Union College/Nelson Glueck School of Biblical Archaeol-
ogy. Jerusalem.

Kassis, H. E.
1973 Beginnings of the Late Bronze Age at Megiddo—A Reexamination of Stratum X.



Berytus



22:
5–22.

Kempinski, A.
1989



Megiddo: A City-State and Royal Centre in North Israel.



Materialien zur Allgemeinen und
Vergleichenden Archäologie 40. Munich: C. H. Beck.

Kenyon, K.
1969 The Middle and Late Bronze Age Strata at Megiddo.



Levant



1: 25–60.

Loud, G.
1948



Megiddo II, Seasons of 1935–39.



Oriental Institute Publications 62. Chicago: University of Chi-
cago Press.

Matthiae, P.
1980 The Princely Tombs of Middle Bronze II at Ebla and the Contemporary Syro-Palestinian Cem-

eteries.



Studi Eblaiti



2: 195–204.

Müller, U.
1970 Kritische Bemerkungen zu den Straten XIII–IX im Megiddo.



Zeitschrift des Deutschen
Palästina-Vereins



86: 50–86.

Petrie, W. F.
1932



Ancient Gaza II, Tell el Ajjul.



London: British School of Archaeology in Egypt.

Salles, J.-D.
1987 Deux nouvelles tombes de Ras-Shamra. Pp. 187–95 in



Ras-Shamra Ougarit III: Le centre de
la ville, 38



e



–44



e



campagnes (1978–1984),



ed. M. Yon



.



Paris: Éditions Recherche sur les Civil-
isations.

oi.uchicago.edu





315



THE RIDDLE OF STRUCTURE 5239 AT MEGIDDO, STRATUM IX



Schaeffer, C. F. A.
1939



Ugaritica: études relatives aux découvertes de Ras Shamra.



Première série. Mission de Ras
Shamra III. Paris: P. Geuthner.

Schumacher, G.
1908



Tell el Mutesellim I: Fundbericht.



Leipzig: Haupt.

Woolley, C. L.
1955



Alalakh: An Account of the Excavations at Tell Atchana in the Hatay, 1937–1949.



Oxford:
Oxford University Press.

Wright, G. E.
1965



Shechem: The Biography of a Biblical City.



New York: McGraw-Hill.

Yadin, Y.
1972



Hazor: The Schweich Lectures of the British Academy 1970.



London: Oxford University Press.

Yon, M.
1990 Ras Shamra-Ougarit 1988–89 (48



e



–49



e



campagnes).



Syria



67: 442–49.

oi.uchicago.edu





316



DAVID ILAN



PLATE 17.1a. Deep Chambers (5239) of [Megiddo] Stratum IX. From Northeast. Wall in Foreground Belongs
to Stratum XI, Small Walls at Corner (left center) to XIV; Porch Walls of XV Temples 5192 and 5269 Appear in
Background. (After Loud 1948: fig. 244).

PLATE 17.1b. Largest Chamber of [Megiddo] 5239. Figure Stands on Floor. From Southwest. (After Loud
1948: fig. 245).

oi.uchicago.edu



317



HOUSEHOLD ARCHAEOLOGY AT ARAD AND AI
IN THE EARLY BRONZE AGE II



Ornit Ilan



The prosaic dwelling—its plan, dimensions, method of fabrication, and its movable con-
tents—is a mirror of lifestyle and social matrix. Recognition of its archaeological potential in
this regard has ushered in a branch of our discipline called



household



archaeology (e.g., Wilk
and Rathje 1982; Levy and Holl 1987, references therein).

In this study, the principles of household archaeology are applied to domestic assemblages
of the Early Bronze Age (EB) II, a period that saw the establishment of urban settlements at
most of the large tell sites in Canaan.



1



Two sites with similar settlement histories are investi-
gated: Arad and Ai (fig. 18.1). Both were late EB I village occupations that evolved into
walled towns in the EB II.



2



However, each occupied a different ecological zone: Arad, the
semiarid Arad Basin, and Ai, the Mediterranean habitat of the central hill country. Two main
research questions were posed:

• Do incongruities exist in the material cultures of the respective settlements that can
be explained by these differing ecological circumstances?

• Do such incongruities, if they exist, reflect cultural identity (ethnicity)?

We begin by describing and discussing the domestic architecture of the two sites, touching
on compound size and complexity, room size, and architectural hallmarks. Following this, the
house’s contents are analyzed, with emphasis placed on the pottery assemblage. Methodolog-
ically, two ethnographic approaches are utilized:

• Analogy with contemporary societies living in proximity to the ancient sites that
maintain preindustrial subsistence patterns to a significant degree. This is essentially
an attempt to maximize the precision of the archaeological remains’ interpretation
(e.g., Kramer 1982; Hirschfeld 1995).

• Cross-cultural analogy with broader, even universal applications (e.g., Flannery and
Winter 1976; Wilk and Rathje 1982; Henrickson and McDonald 1983).



1. It is now clear that walled settlements first arose in the EB I in several places (Eisenberg 1986; Beck and
Kochavi 1993), but the urban impulse achieved endemic proportions only in the EB II.

2. Contra Callaway (1972: 99, 115; 1980: 63), Phase III at Ai should not be dated to EB I, but rather to early
EB II (Schaub 1982: 69; Esse 1984: 323).



one line short



18



oi.uchicago.edu





318



ORNIT ILAN



FIGURE 18.1: Map of Major EB II sites in Israel.



one line short

oi.uchicago.edu

[image: image101.tif]



319



HOUSEHOLD ARCHAEOLOGY AT ARAD AND AI IN THE EARLY BRONZE AGE II



THE PHYSICAL ATTRIBUTES OF DOMESTIC ARCHITECTURE AT ARAD

Residential architecture was uncovered in Arad Strata III–II inside and parallel to the line
of the fortification wall in the south and the west (Areas K, N, H), between the western gate
and the Temple compound (Area TN) and between the palace and the band of structures sur-
rounding the water reservoir (Area TE) (figs. 18.2, 18.3). These dwellings were organized in
“neighborhoods” not always tangent to each other. While not a “typical” neighborhood, some
of the structures surrounding the water reservoir in Area M did reveal evidence for domestic
activity and are included therefore in this analysis. The dwellings of Stratum I were con-
structed alongside the town wall at a time when the settlement’s urban fabric had unraveled.
Architecture was apparently limited to the domestic; public edifices of any kind are lacking.



The Size of the Arad House and Its Components

Stratum III (fig. 18.2)



Twelve domestic compounds of this level were excavated, seven completely, and only the
main room in the remaining five. The compounds vary greatly in size (see table 18.1), ranging
between 32.8 m



2



and 171.1 m



2



(the biggest being 5.3 times larger than the smallest). The com-
pounds can be divided into three size categories (fig. 18.4):

• small compounds, 32.8–42.4 m



2



(



N



= 2)
• medium-sized compounds, 54.5–67.5 m



2



(



N



= 3)
• large compounds, 153.0–171.1 m



2



(



N



= 2)



Stratum II (fig. 18.3)



The residential areas of Stratum III continued to function as such in Stratum II. Of the
twenty-five compounds excavated, sixteen were revealed in their entirety (see table 18.2).
Compound area in Stratum II varies between 32.8 m



2



and 126.4 m



2



and can also be divided
into three categories of size:

• small compounds, 30.0–45.0 m



2



(



N



= 5)
• medium-sized compounds, 55.0–73.0 m



2



(



N



= 4)
• large compounds, 100.0–126.0 m



2



(



N



= 7)



Stratum I



(Amiran and Ilan 1996: pls. 94, 95)
The remains of thirteen residential units were excavated, most of them poorly preserved

being so close to the surface. Only two displayed plans that could be traced completely (with
areas of 16.7 m



2



and 33.2 m



2



respectively). Of the other compounds, three individual cham-
bers remained (7.6 m



2



, 12.0 m



2



, 3.1 m



2



) and one courtyard (5.5 m



2



). The two complete com-
pounds would be classed as “small units” by the criteria applied to Strata II–III, and each
consisted of one chamber and one courtyard. The modest size and inferior construction of
these buildings contrast starkly with the urban nature of the previous two levels and appear to
indicate socioeconomic collapse. It is suggested that Tel Arad came to be occupied by pasto-
ralists and farmers on a seasonal basis, i.e., in the winter and spring, when pasture was avail-
able and dry farming expedient.

oi.uchicago.edu





320



ORNIT ILAN



N



ote



: See also fig. 18.2.



Table 18.1.



Domestic Compounds and Component Size at Arad, Stratum III



Unit no.
Excavation

field
Total area

(m



2



)
Area of main

room (m



2



)
Area of

courtyard (m



2



)
Area of subsidiary

room (m



2



) Size category



1 M ? 21.8 ? ? ?

2 M 153.8 34.0 103.0 a = 8.3 Large

b = 8.5

3 M 54.5 14.3 37.0 3.2 Medium

4 M 32.8 14.3 18.5 none Small

5 M 63.7 20.3 29.0 14.4 Medium

6 M ? 16.5 ? ? ?

7 TN 67.5 26.1 32.4 9.0 Medium

8 H ? 26.4 ? ? ?

9 H ? 27.5 ? ? ?

10 K 42.5 14.3 21.2 7.0 Small

11 K ? 44.0 ? ? ?

12 K 171.1 47.5 100.0 a = 17.6 Large

b = 6.0

oi.uchicago.edu

[image: image102.tif]



321



HOUSEHOLD ARCHAEOLOGY AT ARAD AND AI IN THE EARLY BRONZE AGE II



The following contrasts between the domestic compounds of Strata III and II can be
observed in the above data:

• In Stratum III, compounds of various size categories coexisted in the same
neighborhood, while in Stratum II, Area H contained only small units and Area K only
large ones (the other areas maintained the previous coexistence).

• Large compounds are larger in Stratum III than in Stratum II.



FIGURE 18.2: The location of the dwelling units at Arad, Stratum III.



1:



Unit 5564+5599;



2:



Unit 5347b;



3:



Unit
5006b;



4:



Unit 5071b+5080b;



5:



Unit 5513b+5514b;



6:



Unit 5053b;



7:



Unit 4496+4610;



8:



Unit 1282c+1290c;



9:



Unit 1069b;



10:



Unit 1157b;



11:



Unit 1162b+1163b, 1169b+2318b;



12:



Unit 2440+2519.

oi.uchicago.edu

[image: image103.tif]



322



ORNIT ILAN



The study of domestic architecture in ethnographic research tends to find a positive corre-
lation between dwelling size and status and wealth—wealthier people often live in larger
homes, e.g., in Iran and Turkey (Kramer 1982: 70, 136) and in Africa (Holl and Levy 1993:
175). The newly instituted separation of neighborhoods of large domiciles from neighborhoods
of small ones may be an indication of increased social stratification at Arad (Marfoe 1980: 320).



FIGURE 18.3: The location of the dwelling units at Arad, Stratum II.



1:



Unit 5555;



2:



Unit 5347a;



3:



Unit 5006a;



4:



Unit 5071a+5080a;



5:



Unit 5060a+5063a;



6:



Unit 5513a+5514a;



7:



Unit 5021a;



8:



Unit 4974a+4978;



9:



Unit
5053a;



10:



Unit 5821a;



11:



Unit 4699+4866;



12:



Unit 4266+4267;



13:



Unit 4494;



14:



Unit 4387;



15:



Unit
1282b+1290b;



16:



Unit 1065a;



17:



Unit 1076;



18:



Unit 1081;



19:



Unit 1039a;



20:



Unit 1030;



21:



Unit 1234;



22:



Unit 2326;



23:



Unit 1157a;



24:



Unit 1169a+2318a;



25:



Unit 2539.

oi.uchicago.edu

[image: image104.tif]



323



HOUSEHOLD ARCHAEOLOGY AT ARAD AND AI IN THE EARLY BRONZE AGE II



Analysis of the domestic compounds’ architectural components in Stratum II (see tables
18.2, 18.3, 18.4) bolsters the notion of a correlation between compound size and wealth. On
average, the larger the compound the greater the proportion of the courtyard. Conversely, the
larger the compound the smaller the proportion of the main room.

Table 18.2 shows that main room size in Stratum II ranges from 12.2 m



2



to 33.5 m



2



, some-
what less than a factor of three, while courtyard size ranges from 9.5 m



2



to 84.2 m



2



, approxi-
mately a factor of nine. We infer (below) that main rooms served as sleeping quarters and as
loci for domestic activities such as cooking, while courtyards functioned as livestock pens and
contained storage facilities for agricultural produce. Hence, larger compounds did not neces-
sarily house more occupants. Rather, larger courtyards (and more installations) imply more
produce requiring storage and more livestock needing to be penned. Larger compounds mean
wealthier people.

The ratio of open area to built-up area was 1:1.2 in Stratum III and 1:1.6 in Stratum II, i.e.,
the latter was more densely occupied than the former—most probably a sign of population
increase. Decrease in residential compounds’ average size also seems to reflect a rising popu-
lation density; the town’s denizens had to be content with smaller parcels of land.



Architectural Features of the Arad Domestic Compounds and the Organization of Space



The characteristics of the Stratum II and Stratum III house compounds are identical and
can be summarized as follows (see figs. 18.4, 18.5):

1. The compound is comprised of a main room, one (sometimes two) subsidiary rooms,
and a courtyard, usually surrounded by a stone fence.

2. Generally, no passage exists between the main and subsidiary rooms. Both have
openings into the courtyard.

3. Both the main and subsidiary rooms exhibit a broad room plan, i.e., the doorway is in
the long wall.

4. Walls are most frequently 50–60 cm thick and only rarely reach a thickness of 90–
100 cm.

5. Doorways, located in one of the long walls, vary in width from 40 cm to 100 cm. They
usually range 60–70 cm. Thresholds are stone paved or of tamped earth.

6. The doorpost socket is placed to the left as one enters the chamber.
7. Benches 30–40 cm high and 20–40 cm wide are often appended to the walls of the

main room. The subsidiary rooms rarely have benches.
8. Floors are made of tamped earth and are lower than the surfaces of the courtyards.
9. One, and sometimes two, stone slabs are usually found stuck in the floor near the

center of the main room. These served as socles for wooden columns that supported
the roof beams.

10. Various kinds of installations can be enumerated. In main rooms, and less frequently
in subsidiary rooms, one finds cooking stoves constructed of flint slabs. Cupmarks—
small holes paved with small pebbles within and around their immediate
circumference—are present in rooms and, rarely, in courtyards. Stone mortars were
placed in the rooms and, rarely, in courtyards. Semicircular and rectangular platforms



short

oi.uchicago.edu





324



ORNIT ILAN



N



ote



: See also fig. 18.3.



Table 18.2.



Domestic Compounds and Component Size at Arad, Stratum II



Unit no. Excavation field
Total area

(m



2



)
Area of main room

(m



2



)
Area of courtyard

(m



2



)
Area of subsidiary

chamber (m



2



) Size category



1 M ? 30.0±



? 11.0±



?

2 M 105.3±



30.0 75.3±



None Large

3 M 55.3 12.7 31.0 a = 7.3 Medium

b = 4.3

4 M 31.5 12.8 12.5 6.2 Small

5 M 44.8 13.8 31.0 None Small

6 M 73.8 23.2 42.4 8.2 Medium

7 M 33.5 21.5 9.5 2.5 Small

8 M ? 13.0 ? ? ?

9 M ? 16.8 ? 4.0 ?

10 TE 107.9 24.6 61.8 a = 12.0 Large

b = 9.5

11 TE ? 31.0 ? ? ?

12 TN 110.0 25.8 84.2 None Large

13 TN 61.5 19.7 32.4 9.4 Medium

14 TN 57.2 14.3 37.2 5.7 Medium

15 H ? 25.8 ? ? ?

16 H 41.5 17.4 19.1 5.0 Small

17 H ? 24.8 ? 3.6 ?

18 H ? 30.4 ? 14.3 ?

19 H 42.2 12.2 30.0 None Small

20 H ? 13.9 ? 3.8 ?

21 K 126.4 33.0 83.8 9.6 Large

22 K 113.5 33.5 65.0 a = 9.4 Large

b = 5.6

23 K 101.3 a = 13.8 a = 22.0 a = 7.4 Large

b = 11.7 b = 35.7 b = 5.7

c = 5.0

24 K 123.2 17.8 82.0 a = 11.4 Large

b = 12.0

25 K ? 12.6 ? ? ?

oi.uchicago.edu

[image: image105.tif]



325



HOUSEHOLD ARCHAEOLOGY AT ARAD AND AI IN THE EARLY BRONZE AGE II



of stone or brick, and small cells (silos) defined by low stone walls, were uncovered
in rooms and courtyards.

11. Walls, benches, and installations are constructed mostly of local chalk, partially
masoned. Column bases, thresholds, and doorposts are sometimes made of dolomite,
from the Hebron Hills to the north of Arad.

The broadroom plan and the above enumerated properties also appear in Stratum I. How-
ever, the Stratum I structures, besides being smaller, display asymmetry in plan and inferior
construction—walls are often crooked and of varying thickness.



Table 18.3.



The Relative Size of Architectural Components in Domestic Compounds
in Arad, Stratum II (completely excavated units only)



Unit no.
Excavation

field
Total area

(m



2



)
Area of main room

(%)
Area of courtyard

(%)
Area of subsidiary

room (%) Size category



2 M 105.3 (±



) 28.5 71.5 None Large

3 M 55.3 23.0 56.0 21.0 Medium

4 M 31.5 40.6 39.7 19.7 Small

5 M 44.8 30.8 69.2 None Small

6 M 73.8 31.4 57.5 11.1 Medium

7 M 33.5 64.2 28.3 7.5 Small

10 TE 107.9 22.7 57.3 20.0 Large

12 TN 110.0 23.5 76.5 None Large

13 TN 61.5 32.0 52.7 15.3 Medium

14 TN 57.2 25.0 65.0 10.0 Medium

16 H 41.5 42.0 46.0 12.0 Small

19 H 42.2 29.0 71.0 None Small

21 K 126.4 26.1 66.3 7.6 Large

22 K 113.5 29.5 57.3 13.2 Large

23 K 101.3 25.2 56.9 17.9 Large

24 K 123.2 14.4 66.6 19.0 Large

Averages — 76.81 30.49 58.61 13.41 —



Table 18.4.



Average Proportions of Architectural Components in
Domestic Compounds in Arad, Stratum II (%)



Size Category Courtyards Main rooms Subsidiary rooms



Large 64.6 24.3 11.1

Medium 57.8 27.8 14.4

Small 50.8 41.3 7.9

Average 57.73 31.13 11.13

oi.uchicago.edu





326



ORNIT ILAN



In all strata, only the houses’ walls and artifactual contents were preserved; all roofs col-
lapsed. Fortunately, the discovery of a mold-formed pottery casket in the form of a house allows
us to assume a flat roof bordered at the edges by a low (15–20 cm) parapet perforated at inter-
vals to permit collection of rainwater (Amiran et al. 1978: pl. 115). Similar features have been
identified in contemporary Syrian villages (Seeden 1985: fig. 1) and in Palestinian villages in
the Hebron region, where channels are fashioned to drain rainwater into courtyard cisterns
(Hirschfeld 1995: 139). No evidence for windows was discerned at Arad. Walls were plastered
(Amiran et al. 1978: 30).



FIGURE 18.4: Large



(1),



medium



(2),



and small



(3)



domestic compounds at Arad, Strata III–II. A: main room,
B: subsidiary room, C: courtyard, D: platform, E: silo.

FIGURE 18.5: A domestic compound at Arad with its various installations. Unit 1234, Stratum II.



a:



stove;



b:



clay bin;



c:



mortar;



d:



grinding stone;



e:



silo;



f:



stone base for wooden post;



g:



cupmark;



h:



bench;



j:



door socket.

oi.uchicago.edu

[image: image106.tif][image: image107.tif]



327



HOUSEHOLD ARCHAEOLOGY AT ARAD AND AI IN THE EARLY BRONZE AGE II



Utilization of Space



The contents of the main room suggest that much of the inhabitants’ daily routine took
place here—food preparation and craft activities such as weaving and sewing. Most of the
family property was stored here as well. This large room, and not the subsidiary one, was also
the living quarters. This is deduced from the following observations:

1. Many of the domestic compounds contain only one room. This room’s contents
generally evince the prosaic activities of daily life. Preindustrial societies tend to be
consistent and conservative concerning the use of space in the family home,
unconnected to economic and social status differences (e.g., Kramer 1982: 99–113;
Hirschfeld 1995: 119–31). We assume that the citizens of Arad were guided by the
same principle and that the space where most of the daily routine took place and
where most of the family wealth was stored also functioned as its sleeping quarters.

2. Cooking was carried out in the main room, resulting in the room’s being heated. This
would have been crucial in the winter, particularly for a family with infants, while in
the hot summer months the room could be aired out. Subsidiary rooms, on the other
hand, contain almost no cooking stoves and show no evidence of being heated by
other means.

3. One assumes that the most important household property would not have been left
unattended at night (cf. Hirschfeld 1995: 116).

We suggest that the subsidiary rooms served as supplementary storerooms and as hutches
for livestock, especially in the winter. In the cold winter months, the Bedouin of the Beersheva
Valley separate the young kids from their mothers and pen them in a heated hutch (Ben-David
1982: 126). In the Hebron Hills, pastoralists and their flocks cohabit in the winter (cf. Hirschfeld
1995: 116).

Domestic Installations: Spatial and Functional Analysis

By the term “installation” we refer to any permanent fixture in a room or courtyard of a
domestic compound: benches, mortars, stoves, cupmarks, platforms, and silos (fig. 18.5). A
spatial and morphological analysis of these features can illuminate a society’s daily life—how
people stored their food, prepared their meals, and went to sleep.

Benches: Mostly found in main rooms, benches are 30–40 cm high and 20–40 cm wide. We
believe that these were shelves rather than for sitting on; restorable small and medium-size
pottery vessels were often found on them or at their feet. Food-preparing installations such as
mortars, grinding stones, and stoves were always placed at some distance from the benches,
suggesting that these activities, executed in a kneeling or crouching position, were not per-
formed on the benches. It can also be proposed that sleeping mats or rugs were rolled up and
placed on the benches to clear floor space during the day, while night preparations would have
included the spreading of the mats or rugs on the floor and placing the utensils of daytime
activity on the benches. The cave-dwelling Arabs of the southern Hebron Hills have adopted
such a routine (Havakook 1985: 38–39).

Mortars: These are generally inserted into the floor, though sometimes a second, portable
mortar is also present. Mortars have a full diameter of 30–40 cm, an interior diameter of 15–

oi.uchicago.edu



328 ORNIT ILAN

20 cm, and a depth of 10–15 cm. They are most often located about 1 m from the stone
shelves. Mortars were used for the pounding and grinding of grain into a coarse flour and for
other foodstuffs such as lentils and for the preparation of raw materials like ocher. Grinding
stones, for grinding coarse flour into a fine one, were usually found next to the mortars (Seb-
bane forthcoming).

Stoves: Fashioned of flint slabs, stoves have an average diameter of 40 cm. Flint was used
because it is a better heat conductor than either limestone or chalk. Fuel was placed on the slab
surface. In one instance, two upright stones abutted the flat surface of the stove to form a sort
of cooking pot stand. In other cases, three or more freestanding stones placed on the stove sur-
face would have served the same function (Amiran and Ilan 1992: fig. 61). Stoves were also
found closer to the center of the room, usually at some distance from the shelves.

Cupmarks: Small paved surfaces, about 50 x 60 cm in diameter and densely packed with peb-
bles in a plaster matrix were often found in the center of the main room next to the column
bases. In their centers were depressions 10–12 cm in diameter and 7–10 cm deep. Such cup-
marks were sometimes found in subsidiary rooms and, very rarely, in courtyards and open
spaces. Despite their frequent occurrence at Arad, their function still eludes us; perhaps they
were stands for cooking pots just removed from the fire.

Platforms: These were found in both rooms and courtyards. Platforms display a wide variety
of size and form, but most are constructed by the same method—a low stone wall containing
a fill of small stones or packed earth and stone. They are usually 20–30 cm high. Indoors, plat-
forms are most often located in corners, having a quarter circle configuration. Outdoors, in
courtyards, they are appended to walls and have a square or semicircular outline. Ethnographic
parallels suggest that some functioned as working surfaces and others, particularly indoors, as
silo bases supporting brick superstructures. In the modern-day villages of Syria and in the
Hebron Hills of Palestine, platforms are almost always found in courtyards; in Syria small
animal cages and prepared food are placed on small platforms (Seeden 1985: 294, fig. 1), and
in the Hebron Hills they are used for sitting and entertaining, cooking, eating, and, in the sum-
mer, for sleeping (Hirschfeld 1995: 139–40).

Silos: These are constructed in rooms (mostly main rooms) and courtyards, in corners or
along walls, and vary in size. Their surface tends to be flush with that of the floor or somewhat
lower. “Silos” were used as storage cells, probably for materials kept in jars or sacks of skin
or cloth.

THE PHYSICAL ATTRIBUTES OF DOMESTIC ARCHITECTURE AT AI

In the 1930s Marquet-Krause (1949: 21) uncovered what she termed “the lower city” of
Ai—a 114 m long strip of structures south of the large Building 195b + 238, which she inter-
preted as a residential quarter, in contrast to the zone of public architecture on the “acropolis”
(figs.18.6, 18.7). Callaway (1980: figs. 45, 80) excavated a small portion adjacent to the south
end of Marquet-Krause’s exposure—Area C—adding 15 m to the residential quarter. For the
most part, excavation in this long narrow exposure revealed remains of the EB III town, and
only in a strip 49 m long and 3–9 m wide were EB II remains reached. Fragmentary portions
of other domestic structures were uncovered in Area A (under the remains of an EB III sanc-

long

oi.uchicago.edu



329HOUSEHOLD ARCHAEOLOGY AT ARAD AND AI IN THE EARLY BRONZE AGE II

tuary) and in Area G, 200 m east of the Acropolis temple. In short, the residential aspect of
the EB II remains exposed at Ai is scanty and not nearly as informative as that of Arad.

The Size of the Ai House and Its Components

Phases III–IV3

Eight residential structures were excavated in Phase III and, it would seem, ten in Phase IV
(figs. 18.6, 18.7). Most of the units were located next to or nearby the town wall. Table 18.5
gives an indication of the paucity of data available compared to the data from Arad—only
three compounds display complete plans with areas of 33.9 m2, 43.7 m2, and 115 m2 respec-
tively. This renders useless any attempt to establish a size hierarchy for domestic compounds.
With some reservation, it is possible to propose a size hierarchy for rooms (front and back):
smaller rooms 13.0–19.2 m2 (n = 7), and larger rooms 30.0–40.0 m2 (n = 4). Similar to the
main rooms at Arad, the rooms at Ai display a threefold division into small, medium, and
large sizes.

The fact that residential zones exist both along the fortification wall and east of the temple
suggests the existence of “neighborhoods” which, like those at Arad, were not necessarily

3. Phase IV includes Phase V, which is the destruction layer of Phase IV. There is no need to differentiate
between the two (Callaway 1972: 191).

Note: See also figs. 18.6, 18.7.

Table 18.5. Domestic Compounds and Component Size at Ai, Phases III–IV

Area (m2)

Unit/Building Field Phase Total Front room Back room Courtyard

C-north A III ? 13.0 13.0 Possibly none

C-south A III 33.9 14.8 19.1 Possibly none

83+78 L III ? ? ? ?

111 L III ? ? ? ?

OX C III ? ? Possibly none ?

ON C III ? ? Possibly none ?

MN G III ? ? Possibly none ?

KL G III ? ? Possibly none ?

B-north A IV ? 13.0 13.0 Possibly none

B-south A IV 43.7? 19.2 None 24.5?

195b+238 L IV 115.0± 39.6 31.0 45.0±
85+78 L IV ? ? ? ?

111 L IV ? ? ? ?

OX C IV ? ? Possibly none ?

MN G IV ? 30.8 Possibly none ?

KL G IV ? ? Possibly none ?

Average — — 64.2 23.03 19.02 34.75

oi.uchicago.edu



330 ORNIT ILAN

tangent. While the lack of data from Ai precludes the degree of socioeconomic analysis pos-
sible at Arad, a hierarchy in building size does manifest itself at Ai—a possible indicator of
wealth or status.

Architectural Features of the Ai Domestic Compounds and the Organization of Space

Although the lack of data inhibits identification of archetypal architectural features at Ai
in Phases III–IV, we can still discern some trends in the planning (figs. 18.8, 18.9):

1. The domestic unit is comprised of one or more rooms, apparently with an appended
open space. This open space can be enclosed, forming a courtyard (fig. 18.8:1–2, 5), or
not, such that it was essentially a public area, even a street (e.g., the section next to the
east side of Unit B-south [fig. 18.9:2]).

2. In units comprised of two or more rooms, the rooms are adjoined in linear fashion
(figs. 18.8:5; 18.9:1). Ben-Tor (1992a: 64–66) has called this plan “the forecourt
building.”

3. The broadroom is the essential component of both residential and public architecture at
Ai (fig. 18.8:1–2, 5). However, structures erected next to the fortification wall seem to

FIGURE 18.6: The dwelling units at Ai, Phase III. 1: Unit C-north; 2: Unit C-south; 3: Unit 83+78; 4: Building
111; 5: Unit OX; 6: Unit ON; 7: Unit MN; 8: Unit KL.

oi.uchicago.edu

[image: image108.tif]

331HOUSEHOLD ARCHAEOLOGY AT ARAD AND AI IN THE EARLY BRONZE AGE II

be influenced by topographical, and perhaps other factors that dictate longroom or
irregular plans (figs. 18.8:3–4, 18.9:1).

4. Wall thickness ranges 0.8–1.3 m, which may betoken a second story.
5. Doorways range 1.0–2.2 m wide. Thresholds are slab paved or compacted earth.
6. No doorpost sockets were recovered in situ at Ai. This may be fortuitous or it may

suggest that doorway closure was achieved by means other than a hinged door. The
excavation reports illustrate small mortars with dimensions similar to the doorpost
sockets found at Arad, but at the former site these were sometimes found in pairs
(Callaway 1980: fig. 94:10, 19), a circumstance that better recommends their
interpretation as mortars.

7. No benches were discerned except in Building 195b + 238 (fig. 18.8:5).
8. Floors are made of tamped earth and in most cases are flush with the threshold.
9. The only stone pillar bases were found in Building 195b + 238 (fig. 18.8:5). But they

must have existed in other structures as well—it is difficult to envision alternative
roofing techniques.

10. Installations: mortars were uncovered in rooms, stoves were found in rooms and
courtyards, and pebble-paved surfaces were revealed in rooms as were tabun (bread-
making) ovens (figs. 18.8, 18.9).

FIGURE 18.7: The dwelling units at Ai, Phase IV. 1: Unit B-north; 2: Unit B-south; 3: Unit E1K; 4: Building
195b + 238; 5: Unit 83+78; 6: Building 111; 7: Unit OX; 8: Unit ON; 9: Unit MN; 10: Unit KL.

oi.uchicago.edu

[image: image109.tif]

332 ORNIT ILAN

FIGURE 18.8: Domestic units at Ai Areas G (1, 2), C (3, 4), and L (5). A: back room; B: front room, C: court-
yard. a: stove, b: oven, c: mortar, d: stone base for wooden post, e: bench.

oi.uchicago.edu

[image: image110.tif]

333HOUSEHOLD ARCHAEOLOGY AT ARAD AND AI IN THE EARLY BRONZE AGE II

Utilization of Space
The fact that only three domestic compounds were excavated at Ai with any semblance of

completeness (figs. 18.8:5, 18.9) precludes definitive spatial analysis of their components. Our
initial impression is that where two chambers are adjoined (Unit C-south and Building 195b +
238; figs. 18.8:5, 18.9:1)—and we do not know whether all or most domestic architecture at
Ai was so planned—the interior, larger room was the main one.

FIGURE 18.9: Domestic units at Ai Area A, Phases III (1) and IV (2). A: back room; B: front room; C: courtyard
or open space; D: room. a: stove; b: flat stone installation.

oi.uchicago.edu

[image: image111.tif]

334 ORNIT ILAN

Domestic Installations: Spatial and Functional Analysis
Phase IV shows a steep decline in the number of installations in rooms relative to Phase

III. Stoves and ovens disappear completely, and only a few pebble surfaces and mortars occur
(figs. 18.8, 18.9; Callaway 1980: fig. 81). It is conceivable that the activities represented by the
vanishing installations were transferred to the courtyards and other open spaces in Phase IV
(Callaway 1980: 81).

Mortars: It is not clear from the reports whether or not mortars were imbedded into the floor.
Their measurements range as follows: exterior diameter 20–45 cm, diameter of bowl 10–25 cm,
depth 2–4 cm. The only mortar indicated in the plans (fig. 18.8:3) is located at some distance
from the walls; it could be used from any direction. In all rooms where mortars were found,
grinding stones were found too (e.g., Callaway 1980: fig. 67:1–2, 4).

Stoves: Stoves are large, surrounded by a low stone wall, and have a circular (fig. 18.8:1),
semicircular (fig. 18.9:2, Stove N2), or rectangular (fig. 18.8:3) outline. Some are paved with
limestone slabs and others not at all. Dimensions are (by shape): circular, 1.4 m (diameter);
semicircular, 1.5 x 1.8 m; rectangular, 1.5 x 1.9 m. Several cooking vessels could have been
heated simultaneously on stoves of this size. The usual cooking vessel at Ai was the hole-
mouth jar with the flat base, placed in the fire and not over it (explanation below). The use of
limestone (as opposed to flint) surfacing, or no surfacing at all, may be explained by the rela-
tive plethora of fuel available in this ecological zone. Stoves are located next to walls and not
in corners.

Flat Stone Installations: Square or semicircular stone surfaces were laid even with the
floor. On average they are 70 x 70 cm (fig. 18.9:1). These may have been the foundations of
grain silos with mudbrick superstructures. Such surfaces were, like the stoves, found next to
walls but not in corners.

Ovens: These were round, sunk installations with diameters of 40–50 cm. Though included in
the final reports’ plans (e.g., fig. 18.8:1), they do not appear in sections and are not discussed
in the texts. Hence, their identification as ovens cannot be substantiated, although it seems
most logical.

THE FUNCTION OF POTTERY AT ARAD AND AI

In the following section, we seek to discern the culinary habits of Arad and those of Ai and,
if possible, to differentiate between them. Subsequently, we address questions of economy,
culture, and ethnicity to which ceramic traditions are germane.

Two universal founding assumptions underlie much of the research on vessel function in
antiquity:

• Vessel shape and volume reflect vessel function (e.g., Henrickson and McDonald
1983: 630; Smith 1985: 254).

• The above morphological characteristics are cross-cultural (e.g., Flannery and Winter
1976: 36; Henrickson and McDonald 1983: 261).

SHORT

oi.uchicago.edu



335HOUSEHOLD ARCHAEOLOGY AT ARAD AND AI IN THE EARLY BRONZE AGE II

* HM = holemouth, LH = ledge handled, MS = medium-sized, PH = pillar handled, s.b. = sickel blade, t.s. = tab-
ular scraper, zoo. = zoomorphic figurine.

Table 18.6. Selected Domestic Compounds at Arad and Their Contents*

Room
1162a+
1163a 2318a 1045a

1801+
1503

1030+
1077 1076

4496+
4610

4266+
4267

4699+
4866

5060a+
5063a 2326

Total
pottery

Total %
(pottery)

Area K K H H H H TN TN TE M K

Stratum II II II II II II III II II II II II–III

Total pottery
vessels

8 6 16 22 12 8 21 53 19 34 38 237 100.0

Lamp-bowls 1 1 1 — — 1 — 2 2 2 2 12 5.1

Bowls/
platters

1 — 1 1 — — 1 2 — 3 — 9 3.8

Cups 2 — 1 1 1 — — — — 1 — 6 2.5

Juglets — 1 3 1 — — 3 3 4 2 7 24 10.1

Jugs — — 1 1 2 — 1 6 2 3 3 19 8.0

Amphoriskoi — — — — — — — — — 2 1 3 1.3

Small jars — — 2 2 2 — 2 1 — 2 2 13 5.5

Unident. jars 1 — — — — — — — 1 — — 2 .08

LH jars — — 1 2 1 — 2 8 — 4 2 20 8.4

PH jars — — — — 1 — — 3 — 1 2 7 3.0

Knobbed jars — — — — — — — — — — 1 1 0.4

Kraters — — — — 1 — — — 2 1 — 4 1.7

HM jars — — 3 7 1 3 4 15 4 11 10 58 24.5

Cooking pots 1 — 3 6 1 — 4 3 3 1 4 26 11.0

Painted jars — — — — 1 1 — — — — — 2 0.8

Pithoi 2 4 — 1 1 2 3 6 1 1 1 22 9.3

MS jars — — — — — 1 1 4 — — 3 9 3.8

Copper tools 2 awls — — 1 awl — — — — — — — 3 —

Bone tools — — — 1 awl — — — — — — — 2 —

— — — — 1 pendant — — — — — — — — —

Flint tools — — — — 1 s.b. 1 s.b. — — — — 1 t.s. 3 —

Whorls — — — — 1 — — — — — — 1 —

Mortars — 1 — 1 — — — — — — — 2 —

Varia—
ceramic

— — — — — — — — — — zoo. 1 —

Varia—stone 1 bowl — — 1 bead 1 pendant
(hematite)

1 polished ax — — — — — 8 —

— 1 digging — — 1 basin — — — — — — — — —

— stick — — 1 trough — — — — — — — — —

— ring — — 1 pendant — — — — — — — — —

Varia—
copper

— 1 bead — — — — — — — — — 1 —

oi.uchicago.edu



336 ORNIT ILAN

These assumptions, and the evidence that supports them, allow us to utilize ethnographic
data from widely divergent cultural contexts worldwide to construct a hypothetical series of
functional definitions for vessel types that can be applied cross-culturally (Henrickson and
McDonald 1983: 630).

a Abbreviations: bl. = blade, HM = holemouth, l.g.s. = lower grinding stone, LH = ledge handled, l.p.w. = lower pottery wheel, Medit. =
Mediterranean, PH = pillar handled, spat. = spatula, t.s. = tabular scraper, u.g.s. = upper grinding stone.

b Possibly an unfinished needle.
c One may be a doorpost socket.

Table 18.7. Domestic Compounds at Ai and Their Contentsa

Room
C-south,
front rm.

C-south,
back rm. C-north B-south B-north ON ON OX OX MN KL

Total
pottery

Total %
(pottery)

Area A A A A A C C C C G G

Phase III III III IV–V IV–V III IV–V III IV–V IV–V IV–V
III–

IV–V

Total pottery
vessels

11 34 32 55 78 50 20 10 26 10 4 329 100

Lamp-bowls 1 — 1 — 2 1 — — — — 1 5 1.5

Bowls/platters 3 9 8 10 12 9 5 3 6 2 1 68 20.7

Cups — — — — 2 1 1 — 1 — — 5 1.5

Juglets — — — 1 1 3 — — 1 — — 6 1.8

Jugs — 1 — 5 7 4 1 — 1 — — 19 5.8

Amphoriskoi — — 1 1 3 — — — — — — 5 1.5

Small jars — 1 2 2 — — — — — — — 5 1.5

Unident. jars 3 3 1 3 7 2 3 1 4 1 1 29 8.8

LH jars — — — — 4 1 — — — — — 5 1.5

PH jars — — — — 2 2 — — — — — 4 1.2

Kraters 3 5 — 9 5 — — 2 1 — — 25 7.6

HM jars 1 15 18 21 31 26 8 4 7 7 1 139 42.3

Pithoi — — 1 2 2 1 2 — 5 — — 13 4.0

Copper tools — — — 1 awl — — — — — — — — —

Bone tools 1 spat.b — — — — — — — 2 spat. — — — —

Flint tools — — — — 1 bl. 1 t.s. — — 3 bl. 1 bl. — — —

Spheroids — — — — — 1 — — 4 — — — —

Whorls — — — — — 1 — — 2 — — — —

Grind. stones — — — — — 1 l.g.s. — — — — — — —

— — — — — — 1 u.g.s. — — — — — — —

Mortars — — — — — 1 — 1 2c — — — —

Varia
1 Medit.
sea shell

1 hippo
tusk

—
1 pottery

goblet
1 brick
frag.

1 brick —
1
l.p.w.

— — — — —

oi.uchicago.edu



337HOUSEHOLD ARCHAEOLOGY AT ARAD AND AI IN THE EARLY BRONZE AGE II

Method

The artifactual assemblages of eleven domestic compounds at Arad (out of a total of
thirty-seven excavated) and eleven (out of a total of thirteen) at Ai were compared.4 The resi-
dences of Arad were selected in a manner intended to represent all the size categories defined
for Strata II–III (data in tables 18.6, 18.7) and all excavation fields. Calculations have taken
into account all complete vessels and fragmentary vessels and rims of defined types. Handles
and bases have not been included so as to avoid duplication.5 While we are aware of possible
problems inherent in sample sizes, counting criteria, and collection techniques,6 we are confi-
dent that the following analysis reflects the ancient reality to a reasonable extent.

Kitchenware at Arad and Ai

In most human societies kitchenware serves the following purposes (Flannery and Winter
1976: 36; Henrickson and McDonald 1983; Smith 1985: 281; Dessel 1991: 246):

• cooking, baking, and other food preparation
• eating
• storage of fluids
• storage of dry goods
• transport of water
• illumination
• dipping and pouring
• drinking

Figures 18.10–18.14 illustrate the similarity of the Ai and Arad ceramic repertoires. The
principal difference lies in three vessel types present at Arad but absent at Ai: holemouth cook-
ing pots (fig. 18.11:5–7), knobbed jars (fig. 18.12:4), and medium-sized store jars (fig.
18.10:18). This similitude permits the two repertoires to be considered together where vessel
function is concerned.

Cooking, Baking, and Other Food Preparation

This category includes three components:

Cooking: Globular holemouth vessels were found at Arad (fig. 18.11:5–7) whose exteriors
and rims were covered by soot, making their identification as cooking pots a simple matter
(Amiran et al. 1978: 48). In the ethnographic literature pertaining to preindustrial societies,

4. Building 195b + 238 has not been included in this analysis because the separation of its two living surface
levels is not sufficiently clear (Wagner 1972: 14–21, figs. 8–10).

5. A few handles and bases are included in the sample but only in cases where it was beyond doubt what type
of vessel they represent and where no other fragments of that vessel are included.

6. After many years of work with Ruth Amiran, the author is intimately familiar with her collection methods.
From Callaway’s (1972, 1980) publications of Ai it is clear that collection and publication were similarly
meticulous. Marquet-Krause’s (1949) sampling is of an unknown quality and conclusions concerning this
material should be considered provisional.

oi.uchicago.edu



338 ORNIT ILAN

FIGURE 18.10: Typical EB II pottery assemblage from Arad (from Amiran et al. 1978; plate references herein
are to Amiran et al.1978). 1: Lamp-bowl (pl. 22:42); 2: lamp-bowl (pl. 13:1); 3: lamp-bowl (pl. 13:28); 4: lamp-
bowl (pl. 22:59); 5: small bowl (pl. 22:56); 6: bowl (pl. 23:14); 7: bowl (pl. 23:22); 8: bowl (pl. 13:41); 9: platter
(pl. 23:1); 10: platter (pl. 23:7); 11: juglet (pl. 25:17); 12: juglet (pl. 25:34); 13: cup-bowl (pl. 24:20); 14: cup-
bowl (pl. 24:17); 15: jug (pl. 27:15); 16: jug (pl. 27:16); 17: jug (pl. 26:5); 18: medium-sized jar (pl. 28:4); 19:
amphoriskos (pl. 30:10); 20: amphoriskos (pl. 30:18); 21: small jar (pl. 24:9); 22: small jar (pl. 27:1).

oi.uchicago.edu

[image: image112.tif]

339HOUSEHOLD ARCHAEOLOGY AT ARAD AND AI IN THE EARLY BRONZE AGE II

FIGURE 18.11: Typical EB II pottery assemblage from Arad (continued; from Amiran et al. 1978). 1: krater (pl.
42:9); 2: painted small jar (Abydos style) (pl. 30:7); 3: painted medium-sized jar (Abydos style) (pl. 28:7); 4:
krater (pl. 42:2); 5: cooking pot (pl. 43:10); 6: cooking pot (pl. 43:11); 7: cooking pot (pl. 43:8); 8: pithos (pl.
40:5); 9: holemouth jar (pl. 46:1); 10: holemouth jar (pl. 47:3); 11: holemouth jar (pl. 46:3).

oi.uchicago.edu

[image: image113.tif]

340 ORNIT ILAN

FIGURE 18.12: Typical EB II pottery assemblage from Arad (continued; from Amiran et al. 1978). 1: Pillar-
handled jar (pl. 38:3); 2: jar with ledge handles (pl. 31:1); 3: painted jar (Abydos style) (pl. 33:2); 4: knobbed
jar (pl. 34:1).

oi.uchicago.edu

[image: image114.tif]

341HOUSEHOLD ARCHAEOLOGY AT ARAD AND AI IN THE EARLY BRONZE AGE II

FIGURE 18.13: Typical EB II pottery assemblage from Ai . 1: hemispherical bowl (Callaway 1980: pl. 61:28);
2: hemispherical bowl (Callaway 1972: pl. 43:1); 3: small bowl (Callaway 1980: pl. 61:36); 4: small bowl (Cal-
laway 1980: pl. 92:6); 5: bowl (Callaway 1980: pl. 62:4); 6: bowl (Callaway 1972: pl. 56:19); 7: bowl (Callaway
1980: pl. 62:5); 8: platter (Callaway 1980: pl. 62:9); 9: cup-bowl (Callaway 1980: pl. 92:27); 10: juglet (Calla-
way 1980: pl. 62:16); 11: juglet (Callaway 1980: pl. 62:13); 12: amphoriskos (Callaway 1972: pl. 45:9); 13:
painted juglet (Abydos style) (Callaway 1972: pl. 45:10); 14: painted juglet (Abydos style) (Callaway 1972: pl.
45:11); 15: cup-bowl (Callaway 1972: pl. 45:12); 16: jug (Callaway 1980: pl. 62:18); 17: jug (Callaway 1972:
pl. 56:24); 18: jug (Callaway 1980: pl. 62:11); 19: small jar (Callaway 1972: pl. 43:7); 20: small jar (Callaway
1980: pl. 92:28); 21: krater (Callaway 1972: pl. 29:5); 22: krater (Callaway 1972: pl. 50:25).

oi.uchicago.edu

[image: image115.tif]

342 ORNIT ILAN

FIGURE 18.14: Typical EB II pottery assemblage from Ai (continued). 1: Jar with loop handles (Callaway 1972:
pl. 57:6); 2: jar with ledge handles (Callaway 1972: pl. 46:13); 3: pithos (Callaway 1972: pl. 57:7); 4: pithos
(Callaway 1972: pl. 57:8); 5: holemouth jar (Callaway 1972: pl. 58:6); 6: pillar-handled jar (Callaway 1980: pl.
63:10); 7: holemouth jar (Callaway 1972: pl. 58:13); 8: holemouth jar (Callaway 1972: pl. 43:11).

short

oi.uchicago.edu

[image: image116.tif]

343HOUSEHOLD ARCHAEOLOGY AT ARAD AND AI IN THE EARLY BRONZE AGE II

cooking pots are reported as characteristically globular, having rounded bases intended to
maximize the distribution of heat. Angular joins—between base and wall, for example—are
undesirable since fracture tends to occur at these points of weakness with the introduction of
intense heat (Arnold 1985: 226; Dessel 1991: 33). The cooking vessel is wider than it is high,
and the opening is narrow so as to avoid rapid evaporation (Henrickson and McDonald 1983:
631). The more the vessel is to be heated, the thinner its walls (Smith 1985: 261). All these
features exist in the cooking pots of Arad.

But what vessels did the people of EB Ai cook with? At Ai, and also at Arad, a large num-
ber of flat-bottomed holemouth jars were found bearing the remains of soot on their exteriors
(figs. 18.11:9–11; 18.14:5, 7–8). Obviously, these too were cooking vessels, but perhaps for
different foodstuffs or a different mode of preparation. One assumes that the contents of the
round-based and flat-based cooking vessels were heated by different methods: the round-based
vessel was placed over the flame, on several stones, while the flat-based jar was placed in the
fire. Smith (1985: 274, 276) notes that the contents of a pot placed over a fire are heated grad-
ually from bottom to top, while those of a vessel placed in the fire are heated from the sides,
a process that speeds evaporation. The question is, why wasn’t the round-bottomed pot adopted
at Ai? Might this indicate cultural preferences in food preparation?

Baking: No portable baking vessels were identified at either site. At Ai, the above-noted tabun
ovens probably served this purpose, and at Arad, a pita-like bread might have been baked on
hot stone slabs, ceramic vessel walls, or metal sheets in the Bedouin fashion.

The spouted krater or vat, beer and oil production: The krater is the only vessel having a spout
(figs. 18.11:1, 4; 18.13:21–22). To make beer, barley grains and water were placed in the
krater to create a fermented liquid. The beer was poured out the spout, and the floating mash
of fermented grain remained on the surface inside the container (Katz and Voigt 1986: 32).

Another interpretation of this vessel’s function has been offered by Esse (1991: 123), who
discerned several of them in one of the rooms at Beit Yera˙ associated with signs of olive oil
production. Using parallels from traditional oil production in modern Greece, Esse proposed
that following their soaking in warm water and subsequent crushing, the fruit was placed in
the vats to settle. Water was then added, causing the lighter oil to rise to the surface, to be
poured out the spout (cf. Fargo 1979: 38).

Eating Vessels
Eating vessels are often divided into personal and collective categories in the ethnographic

literature (Henrickson and McDonald 1983: 632). Personal eating vessels are comprised of
small bowls 6.0–8.0 cm high (7.0 cm on average) and 10.0–23.0 cm in diameter (14.0 cm on
average). Collective eating vessels are between 4.4 and 23.4 cm high (10 cm on average) and
have a diameter that ranges from 8.4 to 95.0 cm (24.6 cm on average). The dimensions of
bowls and platters from Arad and Ai (figs. 18.10:1–10; 18.13:1–8) show similar modality and
allow us to assert a similar functional dichotomy. Dessel (1991: 32), in remarking that slipped
and burnished surfaces are easier to clean, suggests that open vessels exhibiting such treat-
ment can thus be considered eating and serving vessels. Indeed, most of the bowls and platters
at Arad and Ai were slipped and/or burnished.

oi.uchicago.edu



344 ORNIT ILAN

The Storage of Fluids

There is no essential difference between the ceramic containers for the storage of fluids
and those for dry goods; easy access to contents and imperviousness are required of both.
Hence, it is usually hard to determine the original contents of such vessels. Studies of vessel
use in preindustrial societies have shown that some containers have multiple uses; Smith
(1985: 280), for example, discerned forty-eight functions for thirty-nine vessel types.

Water Containers: These tend to be large and heavy, almost immovable. The mouth is
wide, for easy filling and access, and the rim curves out and is profiled to allow cloth or skin
to be stretched and tied over the mouth to seal it (Henrickson and McDonald 1983: 633). All
these features are present in the pithoi of Arad and Ai (figs. 18.11:8; 18.14:3–4).

Other Containers: Oil, wine, beer, and perhaps perfumes were other liquids used in the
Early Bronze Age. It is not clear whether milk was kept in its initial state or converted entirely
into secondary products. Large vessels for the storage of liquids other than water share some
of the water jars’ features: wide mouths for easy dipping and everted rims to facilitate sealing.
These include pillar-handled, knobbed, ledge-handled, medium-sized, and holemouth jars—
though the latter are difficult to seal (figs. 18.10:18; 18.11:9–11; 18.12; 18.14:1–2, 5–8). The
pillar-handled jar’s pillar was perforated and had a concave depression on top to house a dipper
juglet. After dipping, precious excess liquid would drain back into the jar via the perforated
channel (Amiran and Ilan 1992: fig. 54). Jugs, juglets, small jars, and amphoriskoi, most of
which are slipped and have handles, were also used for the storage of liquids in small quantities
(figs. 18.10:11, 15–17, 19–20, 22; 18.13:12–14, 16–19). Their narrow mouths and necks were
intended to minimize spillage if the vessel was overturned, while the slip retarded evaporation
(Henrickson and McDonald 1983: 633).

Storage of Dry Goods

Aside from the pillar-handled jar that clearly contained liquids, any jar could contain
either fluids or dry goods. Dry goods, however, could also be stored in baskets, sacks, or con-
structed bins.

Transport of Water

The shape and size of water jars vary widely and depend on distance from water source,
topography over which the water is to be carried, means of transport, and the number of
people to be supplied. For the most part, water jars have two or three handles and are carried
by one person (Henrickson and MacDonald 1983: 634). The most likely candidate for water
transport is the necked jar with ledge handles (figs. 18.12:2; 18.14:2) that could be comfort-
ably borne on the head. The less ubiquitous loop-handled jar (fig. 18.14:1) might have also
had this function.

Illumination

Many of the small bowls bear a thick strip of soot around the perimeter of their rims, testi-
fying to their use as lamps (fig. 18.10:1–4). The bowl contained oil, and the wick—probably of
flax—protruded over the rim. It is likely that two or more wicks were used simultaneously.

oi.uchicago.edu



345HOUSEHOLD ARCHAEOLOGY AT ARAD AND AI IN THE EARLY BRONZE AGE II

Dipping and Pouring
Dipper juglets and cups (figs. 18.10:12–14; 18.13:9–11, 15) were used to extract the con-

tents of large jars and cooking pots. Both dipper juglets and cups are usually slipped and bur-
nished—probably to prevent the fabric from absorbing liquids and to facilitate cleansing.

Drinking
Some of the small and medium-sized vessels described in the above section concerning

the storage of liquids may also have served as drinking vessels. The small globular jar with the
wide mouth (fig. 18.10:21) might also have been a drinking vessel. Though not especially
large, it is too big to allow dipping, yet it is generally slipped and burnished and bears two
handles.

The Arad “Kitchen” vs. the Ai “Kitchen”

Table 18.8 summarizes the differences in the frequencies of various vessel types between
Arad and Ai, based on the data presented in tables 18.6 and 18.7. The following contrasts
between the Arad and Ai assemblages are notable:

1. Although we don’t know what proportion of the holemouth storage jars served as
cooking vessels at Ai, and thus we cannot juxtapose the proportion of cooking vessels
at the respective sites, the combined figures for all holemouth vessels reveal higher
percentages at Ai (42.3%) than at Arad (35.5%). However, we cannot know whether
there were more cooking vessels or more storage vessels at Ai.7

7. At Ai, where most of the pottery assemblage is comprised of rim sherds, it is hard to differentiate between
rims of ledge-handled jars and those of pillar-handled jars. Therefore, though the two clearly had different
functions, they are grouped together in a category that includes vessels that might have had wet- or dry-
storage purposes.

Table 18.8. Relative Frequencies (%) of Vessel Categories at Arad and Ai7

Vessel types Usage Arad Ai

Holemouth cooking pots
(round base)

Cooking 11.0 0.0

Holemouth jars Cooking and/or storing 24.5 42.3

Kraters/Vats
Oil and/or beer
production

1.7 7.6

Bowls, platters Eating 3.8 20.7

Pithoi Water storage 9.3 4.0

Small jars, jugs, juglets,
amphoriskoi

Liquid storage and
perhaps drinking

24.9 10.8

Jars (ledge-handled, pillar-
handled, knobbed, painted,
medium-sized)

Liquids and dry goods
storage

17.2 11.6

Lamp-bowls Illumination 5.1 1.5

Cups Drinking, dipping 2.5 1.5

Total 100.0 100.0

oi.uchicago.edu



346 ORNIT ILAN

2. Ai has more kraters than Arad by a factor of four. This divergence can be interpreted
in two ways:

• If kraters were indeed used for the production of beer, then more beer was produced at
Ai. This would be somewhat surprising since barley, the most frequent base ingredient
of ancient beer (Zohary and Hopf 1994: 55 and references therein), was the primary
grain cultivated and consumed at arid-zone Arad (Hopf 1978: 65), while wheat was
readily cultivable in the region of Ai (Zohary and Hopf 1994: 18). A number of
explanations could be proffered: for example, that grain was too precious to waste
profligately on beer, or that all or part of the population at Arad maintained a cultural
bias against beer. Neither of these is very convincing.

• Perhaps a more plausible interpretation is that the krater’s main function, at least at Ai,
concerned the production of olive oil. Ai is located in the Mediterranean zone in the
central highlands of Canaan where, it has been suggested, the olive was firmly
established already by the Early Bronze Age (Ben-Tor 1989: 41; Miroschedji 1989:
69–70; Stager 1992: 32; Finkelstein and Gophna 1993: 13; Joffe 1993: 86). Olea would
not have flourished in the semiarid climate of EB Arad, given the agricultural
techniques of the time (Zohary and Hopf 1994: 137). Of course, the krater might well
have served both beer and olive industries, beer being dominant at Arad and oil at Ai.

3. A major variance occurs in the category of eating vessels—i.e., bowls and platters—
whose proportion at Ai is seven times greater than at Arad. Perhaps this contrast can
be accounted for by different eating habits in the two towns; at Ai food was more often
served or partaken in individual portions and at Arad more often in larger communal
vessels, or from cups dipped into the communal vessels (see below). The Bedouin of
the Negev and Sinai deserts eat off large communal vessels, a habit common in
nomadic and seminomadic societies that shun excess baggage (Ben-David, pers.
comm.).

4. Pithoi, used principally for the storage of water, are almost 2.5 times more frequent at
Arad than at Ai. Water availability is probably reflected here, with semiarid Arad (170
mm average yearly precipitation) requiring greater storage capacity than Ai with about
435 mm average yearly precipitation and a nearby perennial spring. At Ai more public
water reservoirs seem to have been constructed (Helms 1982: fig. 18), obviating the
need to collect and store large quantities in the home. Yair and Garti (1996) have
estimated the runoff and water collection potentials of EB Arad, concluding
unequivocally that collection of rainfall from the roofs of houses would have been
essential to the existence of a sedentary population.

5. Small and medium-sized vessels for the storage of liquids other than water are also 2.5
times more frequent at Arad. The explanation for this may be that Arad’s population
purchased expensive commodities such as wine and oil in small quantities from the hill
country to the north where such goods were cultivated and processed. Oil and wine
used would have been stored in smaller vessels designed to minimize spillage (cf. the
Iron Age wine decanter [Zevulun and Olenik 1978: 26]). Conversely, at Ai oil and
wine were also stored in bulk in jars.

6. Lamp bowls are three times more frequent at Arad than at Ai. We have no explanation
for this.

Footnote call
covered here

oi.uchicago.edu



347HOUSEHOLD ARCHAEOLOGY AT ARAD AND AI IN THE EARLY BRONZE AGE II

7. Cups are proportionally better represented at Arad than at Ai. Here too the reason may
have something to do with eating habits. Since small bowls are not common, the cup
at Arad might have been both a serving and an eating vessel.

Other Artifacts

Other artifacts found in the Arad and Ai houses (tables 18.6, 18.7) indicate that certain
crafts were household activities and not necessarily practiced by specialized artisans. Those
discerned include spinning (whorls), weaving (bone shuttles), sewing (bone and metal needles
and awls), and drilling (flint drill bits and stone inertia wheels). The flint assemblage from Ai
was not analyzed, but at Arad, Schick (forthcoming) has demonstrated that the flint industry
was largely a domestic one.

THE EARLY BRONZE AGE HOUSEHOLD AT ARAD AND AI: A SUMMARY

When cross-cultural principles of household archaeology are applied to the patterns ex-
hibited by domestic architecture and artifactual assemblages from well-excavated, well-
preserved, well-endowed sites, the result can be a spectacular facsimile of daily life, social
intercourse, economic relations, and political structure. This study of the detailed data from
Arad and Ai also demonstrate how quantified description can bring to the fore essential con-
trasts between assemblages which, on the face of things, look quite similar.

Close analysis and comparisons of the two sites lead us to the following conclusions:

Architecture

1. Domestic compounds display hierarchy in size, particularly at Arad, which is better
preserved and more extensively exposed than Ai, but also at Ai. At Arad, the
compounds of Strata III–II—the paramount strata—can be divided into three size
categories: small, medium, and large. The large compounds are three to five times
larger than the small ones. This is held to reflect different degrees of wealth among the
town’s inhabitants, beyond the special and more obvious status accorded the temple
and the palace units.

Since the courtyard functioned as a storage complex and animal pen, its size and
complexity can attest to household wealth. In large compounds, the courtyard com-
prises 64.6% of total compound area—in small compounds only 50.8%. Hence, we
have deduced that larger compounds are more likely to mean more goods to store and/
or more animals to pen, rather than larger families to house.

2. In Arad Stratum III, 28% of the compounds are of the large variety, while in Stratum
II the figure is 44%. This increase may testify to a strengthening of the town’s
economy and the broadening of its economic elite. Elsewhere we have examined
Arad’s importance as a focal point for the copper trade and as a gateway community
between the desert and the sown (Ilan and Sebbane 1989; Amiran and Ilan 1992). Its
primacy in these roles apparently reached its peak in Stratum II, as reflected in the
greater number of large compounds with large courtyards.

3. In both Arad and Ai, houses are grouped in neighborhoods. In Arad Stratum III, large,
medium, and small compounds existed side by side. In Stratum II, however, Area H

oi.uchicago.edu



348 ORNIT ILAN

consists only of small units and Area K only of large ones. The data at Ai are meager,
but the four units in Area A are smaller than the single ones in Areas L and G. As a
hypothesis for further testing, we can say that, for Arad at least, the wealthier society
of Stratum II was also a more stratified one.

4. Despite certain similarities, the architectural traditions of Arad and Ai differ:
• The Arad compound is usually comprised of two rooms (main and subsidiary) opening

onto a common courtyard. At Ai, the dominant plan is that of a “forecourt building”
whose components are arranged in a linear fashion—a courtyard in front followed by
one or two connected rooms.

• In this context, the plan of a single room must be distinguished from the compound’s
general layout. The single room at both sites is a broadroom. This was the predominant
plan in the Chalcolithic period from northern to southern Canaan (Porath 1992: 42;
Braun 1989: 18). It was used for both domestic and public architecture in EB I, e.g., at
Megiddo (Kempinski 1989: fig. 7), Hartuv (Mazar and Miroschedji 1993: 585), and
Small Tel Malhata (Amiran, Ilan, and Arnon 1983: fig. 3), though much domestic
architecture of EB I is curvilinear (Braun 1989: 17). The broadroom, usually equipped
with a stone column base and sometimes with benches, was again very common in
EB II architecture (Ben-Tor 1992b: 100–102). EB II Arad and Ai share this prevalent
broadroom tradition. Where they differ is in the arrangement of architectural
components.

Finkelstein (1990: 43) interprets the Arad compound as reflecting the traditions of
nomadic pastoralists who opted for a sedentary existence. He emphasizes the open
spaces between the Arad compounds, in contrast to the northern sites where houses are
built in clusters with common walls (Finkelstein 1990: 39). In contrast, the forecourt
building layout that characterizes the architecture of Ai is understood by Ben-Tor
(1992a: 64–66, fig. 7) as being rooted in the EB I. The forecourt building seems to be
confined to northern Canaan and to the central hill country.8

• Both the Arad-type compound and the forecourt-type include an enclosed courtyard
and one or two rooms. At Arad, the main room doubled as a sleeping space and as a
work area during the day, while subsidiary rooms probably served as additional storage
space. At Ai, the sample is so meager as to allow only hesitant speculation. It may be
that in the earlier phases cooking took place in the front (or only) room. In Phase IV,
installations were no longer present in the rooms. If Unit B-south is any indication,
they had moved into the courtyards. Not enough domestic units with more than one
room were excavated to differentiate between the functions of back and front rooms.
The size of the courtyards at the two sites may be at variance; Building 195b + 238 at
Ai, which falls in the category of large compounds at Arad, incorporates a courtyard
that comprises 39% of its total expanse. This, of course, is a much lower proportion
than is usual for courtyards in large compounds at Arad (50.8%–64.6%).

Since the courtyard at Arad was also used, in all probability, to hold animals, we
might infer that at Ai fewer (or no) animals were penned. Unfortunately, the data from

8. The two ostensible forecourt buildings discerned by Ben-Tor (1992b: fig. 8) at Arad do not really fall in this
category since they are modifications of broadroom units.

oi.uchicago.edu



349HOUSEHOLD ARCHAEOLOGY AT ARAD AND AI IN THE EARLY BRONZE AGE II

Ai are too lean at the moment to make a definitive statement, but it might also be
inferred that the economy of Ai placed more emphasis on horticulture and cereal agri-
culture, perhaps deriving the greater portion of its animal products from a more pasto-
rally oriented population occupying the arid margins and the uncultivated hill country.

Located in a semiarid zone, Arad could not have relied solely on agriculture. Ani-
mal husbandry would always have remained a subsistence component, one to be
expanded in times of need. Ben-David (1988: 1, 41) has recently described how,
despite the accelerated process of sedentarization experienced lately by the Negev
Bedouin, and despite the fact that true shepherds are now rare, most families still keep
a small herd of three to twenty head of sheep and goats—what Finkelstein (1990: 41)
has called “a propagation core . . . should changes in the future require a withdrawal
back to pastoralism.”

• The average wall thickness of the Ai houses is greater than at Arad—second stories
may have existed at Ai.

• At Arad, the floors are sunken, a common feature in the southern EB sites of the Negev
and Sinai (Beit-Arieh 1989: 189, 195). At Ai they are not.

• At Ai, rounded corners appear in some of the houses, as opposed to the right-angled
walls of Arad. Ben-Tor (1992a: 62) has suggested the Lebanese coast as the source of
the rounded-corner type of house, also found at Megiddo, Tel Qashish, and Tel Kitan
in the EB I.

We surmise that the disparate architectural traditions displayed by the contemporary towns
of Arad and Ai correspond to two populations having distinct cultural identities.

Installations

5. The installations of the Arad and Ai compounds are at some variance:
• There are many more stone installations at Arad than at Ai. We infer that this reflects

less access to wood and water (for making bricks) in the vicinity of Arad. At Ai similar
installations, shelves for example, were probably built of wood or mudbrick, neither of
which preserves well in a moister climate.

• At Arad stoves are made of flint; at Ai they are sometimes formed of limestone slabs
and sometimes not paved at all. Flint, being a good conductor of heat, was meant to
maximize thermal potential so as to minimize wood consumption in a locale where
wood was not easy to come by. The use of limestone, or no stone at all, in the cooking
installations at Ai, may indicate that firewood was more abundant.

• No tabun ovens were found at Arad. At Ai, it is unclear whether the sunken oval
installations defined by Callaway (1980: fig. 59) as tabuns, were indeed such. If
Callaway was correct, bread was baked differently at each site.

Ceramic Assemblage

6. Comparing the pottery corpus of the two sites, two points are apparent:
• Three vessel types occur only at Arad: globular cooking pots, knobbed jars, and

medium-sized jars. Aside from these, the assemblages are very similar to each other.
Obviously, the two towns had much in common culturally.

oi.uchicago.edu



350 ORNIT ILAN

• While the assemblages are morphologically similar, the numerical proportions of
vessel types at the two sites differ. This observation must mirror both different
ecologies and, unavoidably, some degree of cultural distinction between the two
towns.

Ecology and Economy

7. Water was scarce at Arad, inducing the inhabitants to store as much as possible in
pithoi, which are 2.5 times more abundant at Arad than at Ai. Ai enjoyed more rainfall
(channeled into several reservoirs, as opposed to the single reservoir at Arad) and had
access to a nearby perennial spring.

8. Olive groves and vineyards would have thrived in the hill country around Ai, but not
in the semiarid climate of Arad (even under a somewhat moister climate in the Early
Bronze Age). This fact resonates in the percentage of kraters—four times more
abundant at Ai than at Arad—probably used in the production of olive oil. At the same
time, small closed vessels—which probably contained olive oil and wine—are 2.5
times more frequent at Arad. These precious liquids were acquired in relatively small
quantities from the central hills and the Shephelah in exchange for goods that Arad
specialized in—copper, sheep and goat products, bitumen, and salt. At Ai, where olive
oil and wine were manufactured, the liquids were probably accumulated in the larger
storage jars.

Eating Habits

9. The denizens of Arad and Ai probably had different eating habits. Eating vessels are
seven times more frequent at Ai. At Arad, where those vessels are comparatively few,
the local population might have eaten from communal bowls, as do the Bedouin of the
Negev and Sinai, who wish to be burdened with as little as possible. Three vessel
types found at Arad are known from the Negev and Sinai (Amiran, Beit-Arieh, and
Glass 1973: pls. 50–51; Porat 1989: 172–74, 178; Amiran and Ilan 1992: fig. 38) but
not at Ai or other northern sites. Among these, the globular cooking pot is most
evocative, as it involved a method of cooking different from that applied with the flat-
based holemouth jars used at Ai—perhaps a “southern dish” not adopted by northern
populations.
To return to the two main questions posed at the beginning of this paper:

• Incongruities do exist in the material cultures of EB II Arad and Ai that can be
explained by differing ecological circumstances and the divergent economic bases that
grew out of those circumstances.

• Other incongruities exist that do not have direct ecological or economic explanations.
These surely reflect distinct cultural identities (ethnicity).

The two communities underwent a similar settlement process in which towns of the EB II
emerged from the villages established in the late EB I. Arad was settled mainly by peoples of
the desert with a tradition of pastoral nomadism, while Ai was populated initially as part of the
penetration of the hill country by lowland people and the intensification of horticulture there.

oi.uchicago.edu



351HOUSEHOLD ARCHAEOLOGY AT ARAD AND AI IN THE EARLY BRONZE AGE II

In addition, a local sedentarizing element of pastoral nomads might have been part of this pro-
cess at Ai.

At the time of their initial settlement in the EB IB, the material cultures of Arad and Ai
exhibited less similarity to each other than was the case in the EB II. The intensification of
production and exchange, and the growth of economic interdependency between the towns
and regions of Palestine, resulted in the agglomeration and standardization of material culture
discernable in the EB II. Yet a close inspection of the Arad and Ai assemblages reveals that,
alongside the commonalities of the EB II material culture (and perhaps of cultural values),
old traditions were preserved.

ACKNOWLEDGMENTS

This article is based on a portion of my Master’s thesis written under the supervision of
Prof. Amnon Ben-Tor. I am grateful to him for his exacting yet encouraging criticism. I have
worked with Prof. Ruth Amiran for many years on the material from Arad and cannot thank
her enough for her collegial approach and her enthusiasm for household archaeology. My col-
league Michael Sebbane read the paper and made valuable remarks. I also owe a special debt
to my husband, David Ilan, for editing my English and sharing his wide and discerning knowl-
edge of processual archaeology and its bibliography. Avi Hajian drew the plans and Carmen
Hersch made the pottery drawings.

REFERENCES

Amiran, R.; Beit-Arieh, I.; and Glass, J.
1973 The Interrelationship between Arad and Sites in Southern Sinai in the Early Bronze Age II (Pre-

liminary Report). Israel Exploration Journal 23: 193–97.

Amiran, R., and Ilan, O.
1992 Arad, eine 5000 Jahre alte Stadt in der Wüste Negev, Israel. Neumünster: Wachholtz.
1996 Early Arad II, The Chalcolithic and Early Bronze IB Settlements and the Early Bronze II City:

Architecture and Town Planning, Sixth to Eighteenth Seasons of Excavations, 1971–1978, 1980–
1984. Jerusalem: Israel Museum and Israel Exploration Society.

Amiran, R.; Ilan, O.; and Arnon, C.
1983 Excavations at Small Tel Malhata: Three Narmer Serekhs. Israel Museum Journal 2: 75–83.

Amiran, R.; Paran, U.; Shiloh, Y.; Brown, R.; Tsafrir, Y.; and Ben-Tor, A.
1978 Early Arad I, The Chalcolithic Settlement and Early Bronze City, First–Fifth Seasons of Exca-

vations, 1962–1966. Jerusalem: Israel Exploration Society.

Arnold, D. E.
1985 Ceramic Theory and Cultural Process. Cambridge: Cambridge University Press.

Beck, P., and Kochavi, M.
1993 Aphek (in Sharon). Pp. 62–72 in The New Encyclopedia of Archaeological Excavations in the

Holy Land, vol. 1, Jerusalem: Israel Exploration Society & Carta and New York: Simon &
Schuster.

Beit-Arieh, I.
1989 The Early Bronze Age Pattern of Settlements in the Sinai. Pp. 189–97 in L’urbanisation, ed.

P. de Miroschedji. British Archaeological Reports, International Series 527, part 1. Oxford: Brit-
ish Archaeological Reports.

Ben-David, Y.
1982 Stages in the Sedentarization of the Negev Beduin, A Transition from Former Semi-Nomadic to

Settled Population. Ph.D. dissertation, Hebrew University, Jerusalem (Hebrew).

oi.uchicago.edu



352 ORNIT ILAN

1988 Agricultural Settlement for the Beduin Population, Israel. Jerusalem: Jerusalem Institute for
Israel Studies (Hebrew).

Ben-Tor, A.
1989 Byblos and the Early Bronze Age I of Palestine. Pp. 41–52 in L’urbanisation, ed. P. de Miroschedji.

British Archaeological Reports, International Series 527, part 1. Oxford: British Archaeological
Reports.

1992a Early Bronze Age Dwellings and Installations. Pp. 60–67 in The Architecture of Ancient Israel,
eds. A. Kempinski and R. Reich. Jerusalem: Israel Exploration Society.

1992b The Early Bronze Age. Pp. 81–125 in The Archaeology of Ancient Israel, ed. A. Ben-Tor. New
Haven: Yale University and Open University of Israel.

Braun, E.
1989 The Problem of the Apsidal House: New Aspects of Early Bronze I Domestic Architecture in

Israel, Jordan and Lebanon. Palestine Exploration Quarterly 121: 1–43.

Callaway, J. A.
1972 The Early Bronze Age Sanctuary at Ai (et-Tell). London: Quaritch.
1980 The Early Bronze Age Citadel and Lower City of Ai (et-Tell). Cambridge, MA: American

Schools of Oriental Research.

Dessel, J. P.
1991 Ceramic Production and Social Complexity in Fourth Millennium Canaan: A Case Study from

the Halif Terrace. Ph.D. dissertation, University of Arizona, Tucson.

Eisenberg, E.
1986 Tel Shalem. Excavations and Surveys in Israel 5: 96–97.

Esse, D. L.
1984 A Chronological Mirage: Reflections on Early Bronze Ic in Palestine. Journal of Near Eastern

Studies 43: 317–30.
1991 Subsistence, Trade, and Social Change in Early Bronze Age Palestine. Studies in Ancient Ori-

ental Civilization 50. Chicago: Oriental Institute Press.

Fargo, V. M.
1979 Early Bronze Age Pottery at Tell el-Hesi. Bulletin of the American Schools of Oriental Research

236: 23–40.

Finkelstein, I.
1990 Early Arad—Urbanism of the Nomads. Zeitschrift des Deutschen Palästina-Vereins 106: 34–50.

Finkelstein, I., and Gophna, R.
1993 Settlement, Demographic and Economic Patterns in the Highlands of Palestine in the Chalco-

lithic and Early Bronze Periods and the Beginning of Urbanism. Bulletin of the American
Schools of Oriental Research 289: 1–22.

Flannery, K. V., and Winter, M. C.
1976 Analysing Household Activities. Pp. 34–47 in The Early Mesoamerican Village, ed. K. V.

Flannery. New York: Academic Press.

Havakook, Y.
1985 Cave-Dwellers of the Hebron Mountains. Tel-Aviv: Israel Ministry of Defence (Hebrew).

Helms, S. W.
1982 Paleo-Beduin and Transmigrant Urbanism. Pp. 97–114 in Studies in the History and Archaeol-

ogy of Jordan I, ed. Adnan Hadidi. Amman: Department of Antiquities.

Henrickson, E. F., and McDonald, M. A.
1983 Ceramic Form and Function: An Ethnographic Search and an Archaeological Application.

American Anthropologist 85: 630–43.

Hirschfeld, Y.
1995 The Palestinian Dwelling in the Roman-Byzantine Period. Jerusalem: Israel Exploration Society

and the Pontifical Biblical Institute.

oi.uchicago.edu



353HOUSEHOLD ARCHAEOLOGY AT ARAD AND AI IN THE EARLY BRONZE AGE II

Holl, A. F. C., and Levy, T. E.
1993 From the Nile Valley to the Chad Basin: Ethnoarchaeology of Shuwa Arab Settlements. Biblical

Archaeologist 56: 166–79.

Hopf, M.
1978 Plant Remains, Strata V–I. Pp. 64–82 in Early Arad I, The Chalcolithic Settlement and Early

Bronze City, First–Fifth Seasons of Excavations, 1962–1966, eds. R. Amiran et al. Jerusalem:
Israel Exploration Society.

Ilan, O., and Sebbane, M.
1989 Copper Metallurgy, Trade and the Urbanization of Southern Canaan in the Chalcolithic and

Early Bronze Age. Pp. 139–62 in L’urbanisation, ed. P. de Miroschedji. British Archaeological
Reports, International Series 527, part 1. Oxford: British Archaeological Reports.

Joffe, A. H.
1993 Settlement and Society in Early Bronze Age I and II Southern Levant, Complementary and Con-

tradiction in a Small-Scale Complex Society. Monographs in Mediterranean Archaeology 4.
Sheffield: Sheffield Academic Press.

Katz, S. H., and Voigt, M. M.
1986 Bread and Beer, The Early Use of Cereals in the Human Diet. Expedition 28: 23–34.

Kempinski, A.
1989 Megiddo, A City-State Royal Centre in North Israel. Munich: Verlag C. H. Beck.

Kramer, C.
1982 Village Ethnoarchaeology: Rural Iran in Archaeological Perspective. New York: Academic

Press.

Levy, T. E., and Holl, A. F. C.
1987 Theory and Practice in Household Archaeology: A Case Study from the Village at Shiqmim.

Pp. 373–410 in Shiqmim I: Studies Concerning Chalcolithic Societies in the Northern Negev
Desert, Israel, ed. T. E. Levy. British Archaeological Reports, International Series 356, part 1.
Oxford: British Archaeological Reports.

Marfoe, L.
1980 Review of The Rise of an Urban Culture by A. Kempinski, and Early Arad by R. Amiran et al.

Journal of Near Eastern Studies 39: 315–22.

Marquet-Krause, J.
1949 Les fouilles de ‘Ay (et-Tell) 1933–1935. 2 vols. Paris: Librairie Orientaliste Paul Geuthner.

Mazar, A., and de Miroschedji, P.
1993 Hartuv. Pp. 584–85 in The New Encyclopedia of Archaeological Excavations in the Holy Land,

vol. 2., ed. E. Stern. Jerusalem: Israel Exploration Society & Carta and New York: Simon &
Schuster.

Miroschedji, P. de
1989 Le processus d’urbanisation en Palestine au Bronze Ancien: Chronologie et rythmes. Pp. 63–79

in L’urbanisation, ed. P. de Miroschedji. British Archaeological Reports, International Series
527, pt. 1. Oxford: British Archaeological Reports.

Porat, N.
1989 Petrography of Pottery from Southern Israel and Sinai. Pp. 169–88 in L’urbanisation, ed.

P. de Miroschedji. British Archaeological Reports, International Series 527, pt .1. Oxford: Brit-
ish Archaeological Reports.

Porath, Y.
1992 Domestic Architecture of the Chalcolithic Period. Pp. 40–48 in The Architecture of Ancient

Israel, eds. A. Kempinski and R. Reich. Jerusalem: Israel Exploration Society.

Schaub, R. T.
1982 The Origin of the Early Bronze Age Walled Town Culture of Jordan. Pp. 67–75 in Studies in the

History and Archaeology of Jordan I, ed. A. Hadidi. Amman: Department of Antiquities.

oi.uchicago.edu



354 ORNIT ILAN

Schick, T.
1978 Flint Implements, Strata V–I. Pp. 58–63 in Early Arad I, The Chalcolithic Settlement and Early

Bronze City, First–Fifth Seasons of Excavations, 1962–1966, eds. R. Amiran et al. Jerusalem:
Israel Exploration Society.

Forthcoming Flint Implements, Strata V–I. In Early Arad III: Finds of the 6th–18th Seasons (1971–1984), eds.
R. Amiran, O. Ilan, and M. Sebbane. Jerusalem: Israel Museum and Israel Exploration Society.

Sebbane, M.
Forthcoming Pounding and Grinding. In Early Arad III: Finds of the 6th–18th Seasons (1971–1984), eds.

R. Amiran, O. Ilan, and M. Sebbane. Jerusalem: Israel Museum and Israel Exploration Society.

Seeden, H.
1985 Aspects of Prehistory in the Present World: Observations Gathered in Syrian Villages from

1980–1985. World Archaeology 17: 289–303.

Smith, M. F.
1985 Economic Interpretation of Ceramics: Relating Vessel Size and Shape to Use. Pp. 254–309 in

Decoding Prehistoric Ceramics, ed. B. A. Nelson. Carbondale, IL: Southern Illinois University
Press.

Stager, L. E.
1992 The Periodization of Palestine from Neolithic Through Early Bronze Times. Pp. 22–41 in Chro-

nologies in Old World Archaeology, 3rd ed., ed. R. W. Ehrich. Chicago: University of Chicago
Press.

Wagner, N. E.
1972 Early Bronze Age Houses at Ai (et-Tell). Palestine Exploration Quarterly 104: 5–25.

Wilk, R. R., and Rathje, W. L.
1982 Household Archaeology. American Behavioral Scientist 25: 617–39.

Yair, A., and Garti, R.
1996 The Water Supply at Ancient Arad. Pp. 127–38 in Early Arad II: The Chalcolithic and Early

Bronze Ib Settlements and the Early Bronze II City—Architecture and Town Planning, 6th–18th
Seasons of Excavations, 1971–1978, 1980–1984, eds. R. Amiran and O. Ilan. Jerusalem: Israel
Museum and Israel Exploration Society.

Zevulun, U., and Olenik, Y.
1978 Function and Design in the Talmudic Period. Tel Aviv: Haªaretz Museum (Hebrew).

Zohary, D., and Hopf, M.
1994 Domestication in the Old World: The Origin and Spread of Cultivated Plants in West Asia,

Europe, and the Nile Valley. 2nd ed. Oxford: Oxford University Press.

oi.uchicago.edu



355



EARLY BRONZE AGE SEAL IMPRESSIONS
FROM THE JEZREEL VALLEY AND

THE PROBLEM OF SEALING IN
THE SOUTHERN LEVANT



Alexander H. Joffe



INTRODUCTION

An Early Bronze Age (EB) seal impression was discovered during the 1996 excavation
season at Megiddo. Four EB seal impressions were also found in 1995 during a survey of
Mizpe Zevulun in the northwest Jezreel Valley. These sealings present the opportunity again
to address two of the outstanding issues of the Early Bronze Age: the nature of socioeconomic
and political organization, and the nature and extent of “foreign” contacts.

THE SEAL IMPRESSIONS



Megiddo



The seal impression (reg. 96/H/004/AR001) depicts two horned animals facing left, one
behind the other (fig. 19.1). The rightmost animal appears to be in the foreground, while that
on the left is in the background. The head of the animal on the right is indistinct, which may
be partially a function of its position in the composition and also due to wear. The horns of the
animal on the left are rounded and curve downward, while the tail on the right-hand animal
dips downward and flares up over what may be pointed ears. This may indicate that the left-
hand animal is an ibex



(Capra ibex nubiana)



while the right-hand animal could be a lion or
leopard



(Felis leo



or



Felis pardus).



The Megiddo impression was discovered in Area H, one of the major fields currently being
excavated to reexamine the Iron Age stratigraphy of the mound. This find is especially unusual
in that the vast majority of fills encountered in Iron Age contexts excavated in 1994 and 1996
contained virtually no EB pottery. The particular context of this object was Locus 004, a fill
below the courtyard of Building 1853, a structure generally assigned to Stratum III at Megiddo
and dated after the Assyrian conquest of Israel in 732



b.c.



(Joffe 2000a).
The impression is on a small sherd of pink (Munsell 5 YR 7/4) calcite tempered ware

typical for the EB. The sherd itself is 1.8 cm wide and 2.7 cm high. The impression is only
1.3 cm high. Three sides of the sherd appear to have been straightened by chipping. The
shape and condition of the sherd suggests that it had been found, modified, and preserved
some time after the original impression was made. It might have been a curiosity, trinket, toy,
or heirloom.



19



oi.uchicago.edu





356



ALEXANDER H. JOFFE



Parallels



The Megiddo seal impression has its closest parallel in an example excavated in the
“Stages” on the slope of Megiddo by the University of Chicago. Seal impression 34.2754 shows
at least two horned animals facing left with one behind the other (Ben-Tor 1978: no. IIA–4;
Engberg and Shipton 1934: 36, figs. 10:C, 11:C). The curved horns of the right-hand animal
suggest it is an ibex. The stratigraphic position of the earlier Megiddo sealing, interpreted by
Ben-Tor as a pit in Stage V (Ben-Tor 1978: 43–44), simply places it in the EB I. Other impres-
sions from the Megiddo Stages and from the tell are much less similar to our example. A frag-
mentary impression from Jericho shows a feline pursuing an ibex (Ben-Tor 1978: no. IIA1–3).
A recently published example of a feline striding left is found on a handle recovered during a
survey of Dayr Qiqub, near Pella (Vieweger 1997). Less precise parallels may be drawn with
the stamp seal impression of an ibex from Tell el-Hesi (Ben-Tor 1978: no. IIA-6), the impres-
sion of a schematic ibex from Tell el-Fârºah North (Ben-Tor 1978: no. S-3; Keel-Leu 1989: no.
35), and the stamp seal from Tell Qishyon (Ben-Tor 1978: no. S-4; Keel-Leu 1989: no. 20).
There is no indication of



tête-bêche



arrangement as found on seal impressions from Megiddo,
ºEn Shadud, and Tel Qashish, and which Ben-Tor regards as the product of a distinctive Jezreel
Valley “school” (Ben-Tor 1995: 67–68).

Overall the new Megiddo example fits well in the corpus of small naturalistic seal impres-
sions (Ben-Tor’s Class IIA: Animal File [1978: 8–9, 52–57]) and should be dated on the basis
of the ceramic fabric and the impression to the late EB I.



Mizpe Zevulun



Mizpe Zevulun (Khirbet el-Mushreifeh, 1697: 2389)



1



is a classic EB “enclosure” site,
located on a spur overlooking the western Na



˙



al Zippori. It is approximately 6 ha in size and
is defended on the north, south, and west by heavy fortification walls. Gate complexes are vis-
ible on the northwest and the southeast. There are three descending terraces which appear to be



1. Note that the grid references given in Gal 1992: 17 are incorrect.

FIGURE 19.1. Megiddo seal impression (reg. 96/H/004/AR001) depicting two horned animals. Scale 2:1.

oi.uchicago.edu

[image: image117.tif]



357



EARLY BRONZE AGE SEAL IMPRESSIONS FROM THE JEZREEL VALLEY



separated by fortification walls. The arrangement is uncannily similar to that of Tell el-Fârºah
North, Mitham Leviah in the Golan (Kochavi 1994), Givºat Rabi near Sepphoris, Mitham
Shahal on the Na



˙



al Tabor (Gal 1988), and other examples in the eastern lower Galilee and
western Samaria (Zertal 1993).

The Mizpe Zevulun sealings were discovered on the surface in 1995 during a survey of the
western Jezreel Valley, directed by Professors Israel Finkelstein and Baruch Halpern in con-
nection with the Megiddo project. The site had been previously surveyed by Raban (Raban
1982: no. 69) and Gal (Gal 1992: 17), but its true extent had not been realized. In contrast to
previous research, the 1995 survey found some EB I and possibly EB III pottery, and an over-
whelming predominance of EB II material, especially “Metallic Ware.” The four seal impres-
sions (fig. 19.2) and two figurine fragments were discovered roughly in the center of the site.



FIGURE 19.2. Seal impressions (reg. 16–23/99/01) from Mizpe Zevulun (Khirbet el-Mushreifeh). Scale 1:1.

oi.uchicago.edu

[image: image118.tif]



358



ALEXANDER H. JOFFE



Seal impression 1 (fig. 19.2:1) contains a herringbone flanked by three vertical lines, and
half of a lozenge, which cuts into the vertical lines, framed by horizontal lines. It is 4 cm wide
and 5.5 cm high. The sherd is pink (Munsell 7.5 YR 7/3) and is the shoulder of a medium stor-
age jar. Vertical combing is present above and below the impression.

Seal impression 2 (fig. 19.2:2) contains two concentric circles. It is 3.3 cm wide and 3.8
cm high. The sherd is light reddish brown (Munsell 5 YR 6/4) and is probably the shoulder of
a medium storage jar.

Seal impression 3 (fig. 19.2:3) contains two lozenges separated by a triangle above a sin-
gle horizontal line. It is 3.2 cm wide and 5.5 cm high. The sherd is pinkish gray (Munsell 7.5
YR 7/2) and is the shoulder of a medium storage jar.

Seal impression 4 (fig. 19.2:4) contains two crudely impressed ladders separated by vertical
lines. A third ladder may be partially preserved. This impression is 3.7 cm high and 3.9 cm wide.
The sherd is light red (Munsell 2.5 YR 7/6). It is probably the shoulder or body of a medium jar.

All four seal impressions are on Metallic Ware, although number 4 has a slightly different
surface appearance and contains more visible calcite grits. Numbers 1–3 contain traces of
white lime wash, while 4 does not. Combing is evident on Number 1 but might have been
present on other parts of the vessels from which Numbers 2 and 3 derive.



Parallels



The Mizpe Zevulun examples have a number of general parallels but few precise ones in
the increasingly large Southern Levantine corpus of seal impressions. Overall they fall into
Ben-Tor’s Class I of geometric motifs (Ben-Tor 1978: 4–8, 47–52).

The herringbone is paralleled on examples from a number of sites such as Tel Dan (Green-
berg 1996: fig. 3.40.88, 11), and Shamir (Ben-Tor 1978: IE–3, IE-5). Concentric circles are
found on a number of impressions but usually in association with herringbones or rhombuses
that frame or connect the circles. It is of course possible that such motifs existed but are not
preserved on our sealing. Parallels are found with impressions with full circles such as Tel Dan
no. 100091 (Greenberg 1996: fig. 3.41.15), Tel Qashish Qa 81/86 (Ben-Tor 1994: fig. 12), and
the half-circles of Bâb edh-Dhr⺠no. 2947 (Lapp 1989: 3–4, fig. 3). There are also many par-
allels with the unpublished seal impressions from Khirbet ez-Zeraqun.



2



The use of multiple
lines as a divider and the lozenges which use concentric lozenges rather than horizontal or ver-
tical lines as fillers are unusual and lack parallels at present. The ladder motif is found on many
impressions, such as IE-6 from Hazor (Greenberg 1997: fig. III.4), but is usually surrounded
by other elements. The shortness of the ladders also has no parallel.

Interestingly, the Mizpe Zevulun sealings do not contain any animal, human, or architec-
tural representation similar to examples from nearby Jezreel Valley sites such as Tel Qashish
and Givºat Rabi (Ben-Tor 1992). The Mizpe Zevulun sealings should be dated to EB II–III on
the basis of their style and ware. The question of whether they can be dated more precisely is
addressed below.



2. I am grateful to Matthias Flender for sharing copies of the unpublished Khirbet ez-Zeraqun seal impressions
with me.

oi.uchicago.edu





359



EARLY BRONZE AGE SEAL IMPRESSIONS FROM THE JEZREEL VALLEY



THE FUNCTIONS OF SEALS AND SEALINGS IN THE EARLY BRONZE AGE



Approaching Seals and Sealing



What were the functions of seals and sealings in the southern Levant? Most scholars either
state or imply that they are administrative devices used by complex institutions and/or in com-
plex economic interactions taking place over long distances (Ben-Tor 1995: 76). Motifs that
include groups of people and structures in particular have been interpreted as representing or
belonging to religious institutions (Beck 1976; Ben-Tor 1992; Epstein 1972).

In order to discuss how seals and sealings might have worked in the Early Bronze Age
southern Levant, it is useful to review the features of sealing systems. As elements of a visual
communication system, seals must contain a limited number of motifs which, while usable in
a variety of combinations, still retain sufficient consistency to make them recognizable. Formal
variability in size, shape, treatment, and execution may also vary, but not so much as to com-
promise recognition. The number of repeatable elements may be large, so as to be presentable
as a formal syntax, but not so large as to exceed the capacity of individuals or groups working,
in the southern Levantine case, without mnemonic devices. As administrative devices sealings
serve as “witness to a fact of a legal or administrative nature, elaborated by observing certain
set forms which are destined to guarantee its reliability and to give it the force of proof” (Fissore
1994: 340).

What features associated with seals and sealing practices elsewhere are not present in the
southern Levant? These can be easily enumerated:

1. There is no writing system, and no other evidence of literacy or numeracy in the form
of other mnemonic or accounting systems, such as tokens.

2. There are no bullae in the form of door locks, indicative of sealed installations.
3. There are no bullae that sealed baskets, bags, or other portable items (with the

exception of the locally made EB I Egyptian examples from ºEn Besor, Tel Erani, and
the Tel



Ó



alif Terrace [van den Brink 1995] indicative of a variety of containers and
goods being recorded for transport or storage.

4. There are no caches of bullae or vessel sealings, indicative of temporary or permanent
archival activities.

5. There is no practice of countersealing the same object, indicative of information being
processed by more than one office or administrator.

6. There is almost no other imagery that extends or complements seal iconography in
EB I and none at all in EB II–III.

7. Large EB II and EB III architectural complexes such as Building 3177 at Megiddo, the
“Building with Circles” at Beit Yera



˙



, and Area G at Tel Yarmuth, all suggested to be
elite residences (“palaces”), storage facilities, or both, have no published seals or
sealings at all.



3



8. There are extremely few excavated examples of seals and sealings from even broad
exposures at major EB II and EB III sites, such as at Arad, Tell el-Fârºah North, Beit
Yera



˙



, Tel Yarmuth, and Megiddo.



3. The assemblage from Khirbet ez-Zeraqun and its context remain unpublished.

oi.uchicago.edu





360



ALEXANDER H. JOFFE



9. EB II–III sealings have an extremely limited distribution, primarily the Jezreel, Huleh,
and Northern Jordan Valleys, the Galilee, and Northern Transjordan.

These lacunae suggest that southern Levantine sealings have a fundamentally different
structure and function from those in the Syro-Mesopotamian, Anatolian, and Iranian worlds
(e.g., Alizadeh 1988; Ferioli and Fiandra 1983; Mazzoni 1992; Ferioli and Fiandra 1994;
Rothman 1994). Unfortunately, the vast majority of seals and sealings from the southern
Levant, the present examples included, were found in either secondary contexts or on the sur-
face. Our ability to reconstruct the use of seals is therefore severely limited. An examination
of the motifs of EB glyptic, however, in conjunction with other data gives some indication as
to the larger purpose of sealings.



Early Bronze I Seals and Sealings



There are few securely dated EB I seals and seal impressions but many that display a vari-
ety of motifs largely oriented toward wild animals and agro-pastoral activities. The splendid
stamp seal from Tel Kitan showing two cattle pulling a plow is the most elaborate example of
an agro-pastoral scene (Keel-Leu 1989: no. 42). A seal impression apparently depicting a cow
suckling a calf was found at Beit Yera



˙



in an EB IV level but is consistent with late EB I styles
and ceramics (Bar-Adon 1973; Ben-Tor 1978: no. IIC-3). The “Animal File” and



tête-bêche



arrangement may also depict ovicaprids. These objects find conceptual parallels in other EB I
material culture, such as the bowl from Tell el-Fârºah North containing two yoked oxen (Ami-
ran 1986; de Vaux 1952: 577, pl. LXXXb) and the “laden animal” figurine from Azor (Amiran
1985).

Wild animals, primarily lions and ibexes but also snakes, fish, and a crab, are also common
on EB I seals. The most elaborate examples are the impressions from Jericho (Ben-Tor 1978:
IIA-1, 2, 3) and a purchased seal in the École Biblique, probably from Byblos (Ben-Tor 1978:
IIBb-2). Wild animals, including lions, cattle, and ibex, are also found on the famous Picture
Pavement in the courtyard of the Stratum XIX temple at Megiddo (Loud 1948: pls. 275–79),
now securely dated to the late EB I period (Joffe 2000b). The long-standing cultic significance
of lions and felines is also evident from the Late Neolithic mosaic pavement at the open sanc-
tuary Uvda 6 and other sites in the southern Negev and eastern Sinai (Avner 1984), all the way
to the Iron Age, where lions are found on cult stands and an actual skull was found in the
temple at Jaffa (Kaplan and Ritter-Kaplan 1993: 656). The carved basalt (or phosphorite) ibex
bowl from Cave 134 at ºEn HaNasiv is an especially beautiful example of EB I art (Amiran
1989; see also Braun 1990; Philip and Williams-Thorpe 1993).

Several seals depict wild and domestic animals together. Seal impression c538 found in
Megiddo Stratum XI, but certainly originating in EB I, depicts a horned ox and may include a
lion (Ben-Tor 1978: no. IIC-2; Loud 1948: pl. 160:4). Numeira impression 1301 shows a
horned animal followed by a lion (Lapp 1989: fig. 7).



4



In contrast, a series of geometric
impressions also date to EB I and are localized to the central Jordan Valley, especially Tell
Umm Hamad, Jawa, and nearby sites (Helms 1987; 1991b: 110–29, figs. 160–69; 1992:



4. The appearance of animal procession motifs on sherds dated to EB III from Tell el-Handaquq South sug-
gests the survival or imitation of an older style (Chesson 1998: fig. 12).

oi.uchicago.edu





361



EARLY BRONZE AGE SEAL IMPRESSIONS FROM THE JEZREEL VALLEY



fig. 272; Leonard 1992: fig. 22). A possible seal with cross-hatched design carved from a small
wadi cobble was found on the surface of Tell el-Handaquq (Mabry 1989: fig. 14.5). The geo-
metric impressions would appear to represent the continuation of Neolithic-Chalcolithic stamp
seal traditions (Keel-Leu 1989).

The contrasting animal motifs represent a fundamental tension in the EB I society,
between lifeways and perhaps ideologies oriented toward “proto-urban” agro-pastoral activi-
ties and those that were more wild. EB I society found itself divided between the lion and the
oxen, between subsistence and prestige activities which were rapidly shifting from village to
small-scale urban socioeconomic bases. That this tension is captured, however fleetingly, in
the glyptic and other art of the EB I reflects the transformational nature of the period. Ami-
ran’s persuasive analysis of the ºEn HaNasiv ibex bowl as an EB I variant of a Chalcolithic
shape also supports the idea of significant continuity between the two periods, a conclusion
indicated by settlement pattern and other data (Joffe 1993: 36–37, 41–48).

Only the broad outlines of seal distribution in the EB I can be detected at present. Both
agro-pastoral and wild animal motifs have been found at the same sites, such as Megiddo, and
agro-pastoral motifs have been found at extremely small sites, such as Tel Kitan. The exca-
vated Megiddo examples derive from the domestic area on the slope of the site, while the tell
itself was occupied contemporaneously by an extremely large religious complex that has pro-
duced no seals or sealings whatsoever. That geometric motifs predominate in the central Jordan
Valley may indicate a greater degree of continuity with the Chalcolithic period than elsewhere
(Joffe 1993: 37; Joffe and Dessel 1995). But sealings appear on a variety of shapes and sizes
of vessels, suggesting that they are not being used as part of a system of administration.

At best we may say that the EB I seals and seal impressions were emblems that might have
symbolized emic concepts and visually facilitated communication, not economic contacts.
Like other types of the EB I art, they symbolize a selective and idealized view of society,
which though increasingly urbanized, still had very little need or capacity for administrative
devices.



Early Bronze II–III Seals and Sealings



The practice of sealing in the EB II–III periods is qualitatively and quantitatively different
from that of the EB I. First there is a major dating problem to be addressed. While the Metallic
Ware tradition clearly begins at the very end of the EB I and overlaps, at least regionally, with
the Grain Wash tradition (Esse 1993), the bulk must now be dated to the EB II (Greenberg and
Porat 1996). Production of a limited variety of larger storage jars continues, however, into the
early part of the EB III prior to the appearance of the Khirbet Kerak tradition. Given that the
Mizpe Zevulun and Tel Dan seal impressions come from sites with no appreciable Khirbet
Kerak component, a date in late EB II or early EB III seems most likely. The position of Khir-
bet ez-Zeraqun and its extensive assemblage of sealings in this scheme must await fuller pub-
lication. These issues will only be resolved with further radiocarbon assays.

How did seals and sealings function in the EB II–III periods? Once again it must be noted
that there are no caches of seals or sealings, although sites such as Tel Dan and Khirbet ez-
Zeraqun have produced numerous examples. Vessels were impressed with seals before firing,
indicating that the information encoded permanently was known or available at that stage of
production. There is a greater degree of repetition and consistency in the geometric and

oi.uchicago.edu





362



ALEXANDER H. JOFFE



representational motifs employed. The human, animal, and architectural motifs have been
found at a number of sites ranging from the very small, such as Rosh ha-Niqra, to the very
large, such as Beit Yera



˙



.
The strong connection of the sealing with Metallic Ware, and the restricted distribution of

the latter, indicate that encoded information was highly localized and highly specific. While
Metallic Ware is distributed into the central Levant (Marfoe 1995: 96; Mazzoni 1985, 1987),
sealings appear to be restricted to the Jezreel, Huleh, and northern Jordan valleys and imme-
diately surrounding regions. And while a number of sites display similar motifs, such as the
“ritual dance,” all the examples are made by separate seals. Indeed, virtually all the EB II–III
impressions now known, over 200, were made from different seals. Even at sites with large
corpora of seals show almost no duplication. Furthermore, no geometric seal impression has
been found on any post-EB I southern Levantine vessel in Egypt. This fact also argues for dat-
ing sealings later in the EB II after the cessation of Egypto-Canaanite trade. In short, while
there was a significant amount of sealing being done, there were not many vessels impressed
with the same seal, and sealed vessels were not carried especially far or in significant numbers
from one site to another. The geographic extent of known EB II–III sealings is perhaps 80 by
80 km. That they are found almost exclusively on storage jars and pithoi indicates that, if they
moved at all, vessels and the commodities they contained traveled only short and medium dis-
tances. Petrographic and neutron activation analyses of seal impressions are necessary to
determine more precisely the production and distribution patterns of vessels.

If sealings did not play a role in administering long distance trade, what did they do? As
noted above, the institutional basis of sealings has been much discussed. Major architectural
complexes are becoming better known from all periods of the EB, with “temples” in particular
being the focus of much attention (Miroschedji 1993). Were religious institutions separate
from political or economic institutions, or were they all one and the same? It has been argued
elsewhere that EB II organization was largely communal, with urban elites sharing political,
economic, and religious roles (Joffe 1993: 84–87). But by the late EB II and certainly EB III,
changes in settlement patterns, the layout of sites, and intrasite architecture all indicate a more
centralized and competitive environment. Huge, unequivocal temples appear at Megiddo and
Khirbet ez-Zeraqun during the EB III, along with apparently more specialized facilities, such
as the Building with Circles at Beit Yera



˙



, and the large “palatial” complex at Tel Yarmuth.
In short, late EB II and early EB III see the transformation of the social and economic land-
scapes of the southern Levant creating new requirements for control of commodities and
attendant information.

In the second half of the EB II as Egyptian contacts evaporated, intra- and interregional
exchange of commodities, especially Mediterranean crop products, expanded. Local elites,
probably urban based, exchanged goods with one another and expanded local control over the
rural landscape through control of production, consumption, and exchange of Mediterranean
crops. These practices, probably by a variety of competing and increasingly specialized eco-
nomic, political, and religious elites, culminated in the fully specialized landscape of the
EB III (Esse 1991: 98–124; Joffe 1993: 82–86; Rosen 1995). Seals and sealings were another
method for local elites, including religious elites with aspirations to become major economic
institutions, to control the flow of goods around specific regions and between regions. The
proliferation of seals indicates that there were a great number of aspirants, or a great number
of local participants, in this system. The reorganization of the landscape coincided with the

oi.uchicago.edu





363



EARLY BRONZE AGE SEAL IMPRESSIONS FROM THE JEZREEL VALLEY



tremendous expansion of Metallic Ware, providing in a sense an increasingly standardized
medium for communication and means of exchange. But the system remained highly local-
ized, and institutions never felt the need or achieved the capability to go beyond literally
superficial communication. Seals were symbols that marked either producer, distributor, or
consumer of the storage vessel and/or its contents and had no further use beyond a single
event or transaction. They were in a sense advertisements rather than records of transactions.
And once the institutional landscape was fully articulated and centralization intensified around
sites, even this amount of information was no longer necessary.

The lack of geometric and representational sealings at Arad, the largest and most impor-
tant southern site during the EB II, may be accounted for in several ways. First, it was well
outside the area of Metallic Ware where most impressions are found. Second, its primary
socioeconomic interactions were not with other urban sites but with sedentarizing pastoral
nomads in the Negev and Sinai, perhaps even as “urbanism of the nomads” (Finkelstein
1990). Third, Arad might well have been in rapid decline by the time the northern sealing
tradition began. Overall, Arad, and the Dead Sea Plain sites which continued into the EB III,
had different administrative, organizational, and ideological needs. While the latter sites did
employ a few geometric motifs similar to those on Metallic Ware, most were more idiosyn-
cratic (Lapp 1989, 1995).



Seals in Cross-Cultural Perspective



Lerna may provide a hint about the use of seals by highly localized intra-urban institu-
tions, such as the House of Tiles, with specific functions not necessarily related to long-distance
trade. Aegean sealing practices provide important similarities and contrasts with those of the
Levant. First, only stamp seals are known. Their distribution is also limited, with most seal-
ings deriving from the excavations at Lerna, although several actual seals have been found
elsewhere (Heath [Wiencke] 1958; Wiencke 1969; Pullen 1994; Weingarten 1997). But it is in
contexts and function that Aegean seals diverge most profoundly from their Levantine coun-
terparts. Stamp seal impressions are used to decorate ceramic hearth rims and vessels (Wiencke
1970; Caskey 1990; Pullen 1994). But at Lerna they are also found on clay bullae uncovered
within a corridor room of the House of the Tiles with sealing boxes, jars, and baskets
(Wiencke 1969; Aruz 1987: 192–96; Pullen 1994: 43–46). This usage parallels that of earlier
and contemporary Mesopotamia, southwestern Iran, Anatolia, and Egypt and indicates an
inchoate record-keeping system that is conspicuously absent from the Levant.

Lerna is not the only site with seals and sealings, but it does have the largest corpus
(Pullen 1994: 49–50, table 2; Aruz 1994). What was in the various containers being sealed at
Lerna? What did the system record: the origins of tribute or offerings received, the ownership
of goods being stored, or the destination of disbursements? What is the time span represented?
And what is the nature of regional organization? Pullen suggests that the system recorded the
contributions of followers to the chief or paramount at Lerna but dismisses the possibility that
the seventy separate seals represent items being stored by or for individual groups (Pullen
1994: 45–47, n. 47). In contrast, Wiencke notes that the Lerna sealings were not necessarily
administrative or indicative of contents or quality. She believes that the “simplest conclusion
is that the different seal designs indicated the providers of goods” (Wiencke 1989: 505).
Finally, Weingarten suggests that Lerna was an Anatolian trading outpost employing eastern
sealing practices (Weingarten 1997).

oi.uchicago.edu





364



ALEXANDER H. JOFFE



It should be noted that Lerna is only 1.5 hectares in size (Konsola 1990: 469), which makes
it an unlikely candidate for an urban site with wide-ranging contacts. It also casts suspicion on
the site as the center of a powerful chief, since the resident population was also extremely small.
In this respect it resembles Late Neolithic Sabi Abyad (Akkermans and Duistermaat 1997) or
Ubaid period Tepe Gawra, both similarly endowed with rich corpora of sealings but with small
resident populations (Rothman 1994). Some urban-rural hierarchies existed in Greece, but these
were highly localized. No pan-regional pattern of hierarchy is evident. More notable are a hand-
ful of large “corridor houses” (Shaw 1990), lacking any storage facilities, mostly at coastal sites
with little other evidence of significant social differentiation (Wiencke 1989). The conclusions
here can only be that intraregional integration was uneven, and interregional integration spo-
radic, at best. The Early Helladic sealings should therefore be interpreted in a minimalist fash-
ion: at best as precocious but highly localized efforts at administration, token gifts given in
patron-client relationships, or more likely as lower order signs of personal identification within
a framework of communal storage and ritual activities.



5




The lack of bullae or record-keeping indicates that southern Levantine sites did not move

goods, secure facilities, and process information at the same level as sites in regions to the
north. Whether or not the sealed vessels represent gifts to institutions or from institutions is
one major outstanding question, which could be partially addressed by analyzing the ceramics
and their sources. The individuals and institutions sealing vessels apparently had no need for
more complex information, such as quantitative or volumetric data, probably lacked the abil-
ity to process such information, and rarely executed transactions over distances of more than
a few dozen kilometers. But the apparent success of sealing on pottery at EB IV Ebla (Maz-
zoni 1992) indicates that the method itself was useful, provided that the institution needed and
was capable of processing information encoded in such a manner.

The EB II–III sealings in the southern Levant should be interpreted in a minimalist fash-
ion, as efforts to encode limited quantities of information on highly specific types of commod-
ities that traveled only short and medium distances. Sealing was a system that served as an
ideological statement as much as administrative information. When the ideology and organi-
zations that sealings supported was fixed and the environment transformed, other means of
ideological maintenance and administration became necessary.

SEALS AND FOREIGN CONTACTS DURING THE EARLY BRONZE AGE

Foreign iconography in EB glyptic has been extensively debated, but the significance of
motifs alleged to be foreign, the conditions of their adoption, and possible routes and mecha-
nisms of transmission have not. This review proceeds chronologically.



Early Bronze I



The question of southern Levantine contacts with the north during the EB I has been
discussed by several authors, most notably Hennessy, Amiran, and Ben-Tor (Amiran 1970;
Ben-Tor 1985, 1986, 1989; Hennessy 1967). Relations with Egypt are of course well docu-
mented, but their precise nature remains a topic of debate. Ben-Tor has discussed in a number



5. See generally Rutter 1993.

oi.uchicago.edu





365



EARLY BRONZE AGE SEAL IMPRESSIONS FROM THE JEZREEL VALLEY



of publications what he sees as widespread Mesopotamian and Iranian influences on the EB I
southern Levantine seal iconography and practices. Other scholars such as Beck and Lapp
have similarly stressed these influences (Beck 1975, 1976, 1984; Lapp 1995). Interestingly,
the discussion by a Mesopotamian glyptic specialist, Beatrice Teissier, concludes that Meso-
potamian and Iranian motifs were used haphazardly in southern and northern Levantine glyp-
tic, suggesting little or no understanding of the iconographic vocabulary (Teissier 1987).
Several of her points may be repeated and amplified.

First, it must be recognized that the cylinder seal is a Mesopotamian innovation of the
(Middle?) Uruk period which increased the ability of an impression to contain information
and complemented existing administrative systems based on stamp seals and tokens (Nissen
1977: 16; Pittman 1994: 25). The adoption of the cylinder seal in Syria and the Levant built
on preexisting carved bead and stamp seal traditions and was strongly related to the contacts
between Uruk settlements along the Euphrates River and in Egypt in the mid to late fourth
millennium. These contacts, although probably small-scale, were of enormous significance for
Egyptian political iconography.



6



But the evidence for Uruk material culture in the southern
Levant is virtually nil. Locally made bent-spout vessels (Amiran 1970, 1992) and some Uruk-
like pottery at the Eastern Desert site of Jawa (Helms 1991a)



7



are the only tenuous evidence
for northern contacts during the EB I. With one possible exception (ºEn Besor [Ben-Tor
1976]), there are no southern Levantine seals or impressions that can be argued to have been
imported from Mesopotamia or Iran, while there are a number of examples at Byblos and in
north Syria that may be imports from the Uruk settlement sites along the Euphrates (Teissier
1987: 40–41). Thus, while the concept of the cylinder seal seems to have arrived in the south-
ern Levant either directly from the north or via Egypt, there are few other data to provide a
context for this “influence.”

Second, there is the matter of glyptic styles. Animal motifs similar to Iranian examples
prevail in southern Levantine glyptic, but as noted above these build on both the local stamp
seal tradition and particular local iconographic and ideological conditions. The most distinc-
tive Iranian motifs—such as animal protomes, griffins, animals on snakes, rosettes entwined
by serpents, lions attacking quadrupeds, and birds perched on animals—are very rare in the
southern Levant, especially during the EB I. Far more common are



tête-bêche,



a device rather
than a motif, animals with extra-long goatees, three pronged tails, five legged animals, and
exaggerated curved horns (Teissier 1987: 43–45). Given the agro-pastoral economy of the
Early Bronze Age, we may ask whether these motifs are simply local innovations rather than
foreign imports. Most of these motifs are also common to earlier stamp seal traditions of both
Iran and Syria (see generally von Wickede 1990) and the stamp seal traditions of Anatolia
(Aruz 1992) and the Levant. The most distinctive Uruk, Jemdet Nasr, and Early Dynastic
scenes are conspicuously absent, including scenes of battle, manufacturing, such as weaving
and potting, contests, and banquets (see generally Brandes 1979).



6. See Boehmer 1974a, 1974b, 1991a, 1991b; Boehmer, Dreyer, and Kromer 1993; Moorey 1990, 1995;
Sievertsen 1992; Smith 1992; von der Way 1992; 1993: 67–75; Mark 1998; Hendrickx and Bavay forth-
coming; Joffe 2000c.

7. The dating of Jawa and its pottery is problematic. The “Uruk” elements may, in fact, be slightly later than even
the second phase of expansion proposed in the new chronology. See McClellan and Porter 1995; Philip 1995.

oi.uchicago.edu





366



ALEXANDER H. JOFFE



Teissier concludes that “traits characteristic of Iranian glyptic are less common in Pales-
tinian glyptic than those found in sites in the Lebanon and Syria. What can be detected is more
likely to have been derived from the north than received second hand from Egypt” (Teissier
1987: 47). She goes on to state that in Syro-Palestine “motifs were absorbed into glyptic art
only . . . stripped of their context and used singly” (Teissier 1987: 49). The route of transmis-
sion for these motifs to the Levant and Egypt appears to have been via the Uruk period settle-
ments on the Euphrates.



8




That complex Uruk institutional imagery with a rich iconographic vocabulary is paralleled

by only limited animal motifs in the contemporary EB I southern Levant indicates that the
southern Levant received nothing directly from the Syro-Mesopotamian world and compre-
hended less.



Early Bronze II–III



The question of southern Levantine contacts during the Jemdet Nasr and the Early Dynas-
tic periods raises other problems. Does any other evidence exist to suggest late fourth and
early–mid-third millennium contacts between the southern Levant and Syro-Mesopotamia?
Despite extensive excavation and study, little can be cited besides model beds (Beck 1993; De
Cree 1987–88), decorated bone tubes (Zarzecki-Peleg 1993), and most recently a burial at Tell
Assawir containing north Syrian pottery (Yannai 2000). Glyptic style and practices comprise
the only other evidence. The matter of Khirbet Kerak ware and related material dating to EB
III requires separate discussion and has no (clear) bearing on the questions raised here.

The relationship of geometric and representational motifs on southern Levantine and
Greater Mesopotamian glyptic has been approached in a way that emphasizes specific motifs
rather than the larger iconographic vocabulary and contexts of use in either region. The Early
Dynastic period “piedmont” or “geometric International style” (Collon 1987: 20) has recently
been discussed by Pittman (1994), who refers to it as the “glazed steatite style.” This style is
distributed from western Iran to Syria, primarily in peripheral regions surrounding the Meso-
potamian basin. The similarities between this widely distributed style and the glyptic of the
southern Levant have been long pointed to as evidence for early third-millennium contacts with
Mesopotamia. The sealings on vessels from Tell Gubba in the Hamrin Basin have recently been
cited as showing practices similar to those of the southern Levant and by implication some
organic relationship between the two regions (Ben-Tor 1995: 71; Esse 1990: 31–32; Ji 1988).

The southern Levantine use of several motifs such as concentric circles, lozenges, and
chevrons do indeed echo the “piedmont” or “glazed steatite” style (Collon 1987: 23). Without
any evidence of imports or other contacts, however, it is difficult to reconstruct a mechanism
for transmission. A closer examination of the style in Greater Mesopotamia may yield some
insights on its use and possible conditions of transmission. Pittman has suggested that the
“glazed geometric style” motifs are at least partially related to signs in the proto-Elamite script,
which was emerging at precisely this time. The presence of these motifs on seals in the pied-
mont suggests proto-Elamite activities in central Mesopotamia (Pittman 1994: 260–61). Mat-
thews disagrees, noting that geometric motifs are found on earlier Uruk sealings (D. Matthews



8. Teissier’s conclusions on the Iranian elements of Levantine glyptic also agree with those of Algaze and
Moorey on the role of Susa in the Uruk expansion (Algaze 1993; Moorey 1990, 1995).

oi.uchicago.edu





367



EARLY BRONZE AGE SEAL IMPRESSIONS FROM THE JEZREEL VALLEY



1997: 79–86). Both scholars agree that the style and motifs must be seen in the context of
Uruk–proto-Elamite interaction.

Rather than seeing the seals as elements of a commercial network, however, Pittman
believes that the style might have emerged as a way of visually mediating between two or
more language groups (Pittman 1994: 261–62). She also draws parallels with the Jemdet Nasr
and Early Dynastic I city seals that apparently relate to specific items exchanged between
Mesopotamian cities named on the seals (R. Matthews 1993; Pittman 1994: 262). In both
cases geometric motifs complemented other forms of visual communication to facilitate con-
tacts between diverse units. D. Matthews (1997: 84–85) notes that the function of the seals
differed in the core areas of glazed (or, as he prefers, fired) steatite glyptic of southwestern
Iran, southern Mesopotamia, the Hamrin, and at Nineveh; central administration, ritual con-
texts, on vessels, and on containers, respectively. The degradation of meaning is apparent
over space, as well as over time.



9




In the absence of long-distance trade networks, local exchange of commodities served as

means for down-the-line transmission of motifs. The presence of geometric motifs on sealings
served as a form of visual communication, originally based on proto-Elamite forms, which
facilitated local interaction along the eastern and northern peripheries of Mesopotamia. From
there the motifs, increasingly few in number and without their original symbolism and steatite
technique, eventually reached the Levant some centuries after their original diffusion. Whether
or not the original intent of the motifs was intact is unknowable, but the simplicity and local
effectiveness of sealing pottery with geometric motifs could be relearned or reinvented at each
step along the way. In the southern Levant the style, now reduced to a bare handful of motifs,
fell on fertile ground (cf. Mazzoni 1993; Flender in Finkbeiner 1995).

The use of architectural facades, animals, masked humans, and other devices raises other
problems since these are sometimes suggested to be in direct imitation of Greater Mesopota-
mian practice (cf. Ben-Tor 1992: 164). These raise again the problem of Early Bronze Age
religion and cult. Two problems should be separated. The first is whether religious institutions
existed and were socioeconomic entities during the EB II and III periods. The evidence of
temples during this period is clear, although it must be noted that no significant associated
storage installations have been detected at Megiddo, Khirbet ez-Zeraqun, Ai, Arad, or Tell
Yarmuth. This may simply be an accident of excavation. Suggestions as to how such institu-
tions might have used sealings as visual advertisement is offered above. A second and critical
question is whether southern Levantine religion and cult were at all related to or modeled after
Mesopotamian practice. Amiran (1962, 1972a, 1972b), for example, has long suggested that
specific Mesopotamian deities such as Dumuzi were worshipped in the southern Levant, and
that this unity of beliefs goes back into the Neolithic.



9. A parallel may be drawn with the third millennium “Intercultural Style” of carved steatite bowls, distributed
widely from the Indus Valley and eastern Iran to Greater Mesopotamia and containing complex represen-
tational iconography. Most studies have stressed the role of chlorite vessels as commodities in complex
interregional and international exchange mechanisms (Kohl 1978, 1982). Recently, however, Lamberg-
Karlovsky has reexamined the data and suggested instead that the vessels were highly symbolic objects
relating to death and burial and whose motifs united widely divergent groups that shared similar beliefs and
ideologies (Lamberg-Karlovsky 1988, 1993).

oi.uchicago.edu





368



ALEXANDER H. JOFFE



If it was indeed the case that a set of beliefs and practices was common across western
Asia from prehistory onward, we are presented with possible explanations for the motifs on
cylinder seals that appear to depict Mesopotamian cultic scenes, such as masked individuals,
temple facades, and processions of animals. One possibility is that specific motifs and their
meanings were indeed transmitted to the southern Levant, although as noted above, comple-
mentary evidence is lacking. A second possibility is that an underlying matrix of pan-western
Asian beliefs provided a similar background for religious iconography in Mesopotamia and the
southern Levant. These concepts, including the same basic pantheon, cosmology, and cosmog-
ony, manifest themselves in parallel iconography when similar social conditions developed in
each area. Given that religious and cultic scenes in the southern Levant never duplicate entirely
Mesopotamian iconography but appear to use only isolated or singular motifs, this latter sug-
gestion may be more reasonable.

Motifs such as the “ritual dance” (groups of people holding hands in front of buildings)
have only conceptual parallels on Mesopotamian glyptic. Other iconographic elements sug-
gested as importation—such as the use of masks—are well attested in the Levantine Neolithic,
while specific motifs such as masked individuals seated before structures echo Chalcolithic
iconography—evident, for example, on Ghassul wall paintings. In other words, southern
Levantine iconography could be a local and parallel development. In the absence of clear and
demonstrable archaeological evidence for early third-millennium contacts and cultural trans-
mission between Mesopotamia and the southern Levant, the deep common roots of the two
areas may suffice to explain most iconographic parallels.

The similarities between southern Levantine sealing practices and those at Lerna have also
been frequently cited (Ben-Tor 1995: 74–75; Wiencke 1970; see generally Aruz 1994). The
Early Helladic IIA period, or Korakou culture, and Lerna III specifically should now be dated
to ca. 2500–2400



b.c.



(Manning 1994: 186), some two centuries after the beginning of the
Levantine EB III. If southern Levantine motifs and techniques current in late EB II and early
EB III indeed reached the Argolid, several possibilities are raised. The first is that they arrived
in Greece ca. 2700



b.c.



but could not be effectively employed until much later. The second is
that contacts were sporadic and that information regarding seals and sealings did not arrive
until Early Helladic II. A final suggestion is to see the Lerna practices indirectly inspired by
areas increasingly remote from the southern Levant: EB Byblos, EB IVA Ebla (Mazzoni
1992), the Syro-Anatolian traditions of sealing bullae, and the increasingly attested practice of
sealing vessels, such as at Late Uruk Hassek Höyük (Behm-Blancke et al. 1981: 24–28; Behm-
Blancke 1993) and EB Jerablus-Tahtani (Peltenburg et al. 1996: 5–6, fig. 4; Peltenburg et al.
1997: 4, fig. 3). Overall, the internationalism of the mid-third millennium provides a better
context for Aegean-Levantine contacts than the more insular world of EB II (Joffe n.d.).



10



CONCLUSIONS

The Megiddo and Mizpe Zevulun sealings are, respectively, typical and atypical examples
of Southern Levantine glyptic in the EB. While glyptic art was, in a most attenuated fashion,
related to larger Western Asian styles, the motifs and practices were particularly attuned to



10. See also Broodbank 1993; Agouridis 1997.

oi.uchicago.edu





369



EARLY BRONZE AGE SEAL IMPRESSIONS FROM THE JEZREEL VALLEY



local ideological and organizational needs. That these features in the southern Levant differed
dramatically from those in contemporary large-scale societies points again to the importance
of understanding local, small-scale complexity first its own terms and then in comparative
perspective.

ACKNOWLEDGMENTS

I wish to thank the directors of the Megiddo project, Israel Finkelstein, David Ussishkin,
and Baruch Halpern, for their permission to publish these objects. The figures were drawn by
Rodica Penhas. I also thank Rachel Hallote, J. P. Dessel, Linda Bregstein, Stan Hendrickx, and
Adi Kafri for their help with this research. Finally, my thanks to Benjamin Sass for graciously
making the Megiddo sealing available for publication and for discussing the object with me.

REFERENCES



Agouridis, C.
1997 Sea Routes and Navigation in the Third Millennium Aegean.



Oxford Journal of Archaeology



16:
1–24.

Akkermans, P. M. M. G., and Duistermaat, K.
1997 Of Storage and Nomads: The Late Sealings from Late Neolithic Sabi Abyad, Syria.



Paléorient



22: 17–44.

Algaze, G.
1993



The Uruk World System: The Dynamics of Expansion of Early Mesopotamian Civilization.



Chicago: University of Chicago Press.

Alizadeh, A.
1988 Socioeconomic Complexity in Southwest Iran during the Fifth and Fourth Millennia



b.c.



: The
Evidence from Tall-i Bakun.



Iran



26: 17–34.

Amiran, R. B. K.
1962 Myths of Creation of Man and the Jericho Statues.



Bulletin of the American Schools of Oriental
Research



167: 23–25.
1970 The Beginnings of Urbanization in Canaan. Pp. 83–100 in



Near Eastern Archaeology in the Twen-
tieth Century: Essays in Honor of Nelson Glueck,



ed. J. A. Sanders. Garden City: Doubleday.
1972a A Cult Stele from Arad.



Israel Exploration Journal



22: 86–88.
1972b Reflections on the Identification of the Deity at the EB II and EB III Temples at ºAi.



Bulletin of
the American Schools of Oriental Research



208: 9–13.
1985 Canaanite Merchants in Tombs of the Early Bronze I at Azor.



ºAtiqot



17: 190–92.
1986 Some Art-and-Cult Objects of the EB I period. Pp. 7–13 in



Insight through Images, Studies in
Honor of Edith Porada,



ed. M. Kelly-Buccellati. Malibu: Undena.
1989 An Early Bronze I Basalt Cult-bowl with Ibex Reliefs.



Israel Museum Journal



8: 17–23.
1992 Petrie’s F-Ware. Pp. 427–32 in



The Nile Delta in Transition: 4th–3rd Millennium B.C.,



ed.
Edwin C. M. van den Brink. Tel Aviv: Edwin C. M. van den Brink.

Aruz, J.
1987 The Aegean and the Orient in the Neolithic and Bronze Ages: The Evidence of Stamp and Cyl-

inder Seals. Ph.D. dissertation, New York University.
1992 The Stamp Seals from Tell Esh Sheikh.



Anatolian Studies



42: 15–28.
1994 Seal Imagery and Sealing Practices in the Early Aegean World. Pp. 211–35 in



Archives before
Writing, Proceedings of the International Colloquium, Oriolo Romano, October 23–25, 1991



,
eds. P. Ferioli, E. Fiandra, G. G. Fissore, and M. Frangipane. Torino: Scriptorium.

Avner, U.
1984 Ancient Cult Sites in the Negev and Sinai Deserts.



Tel Aviv



11: 115–31.

oi.uchicago.edu





370



ALEXANDER H. JOFFE



Bar-Adon, P.
1973 Rare Cylinder Seal Impressions from Beth Yera



˙



.



Eretz-Israel



11: 99–100.

Beck, P.
1975 The Cylinder-Seal Impression from Megiddo, Stage V, and Related Problems.



Opuscula Athe-
niensia



11: 1–16.
1976 The Cylinder Seal Impressions from Beth HaºEmeq



. Tel Aviv



3: 120–26.
1984 The Seals and Stamps of Early Arad.



Tel Aviv



11: 97–114.
1993 Early Bronze Age “Bed-Models” Reconsidered.



Tel Aviv



20: 33–40.

Behm-Blancke, M. R.
1993 Glyptische Traditionen Beiderseits des Ost-Taurus im Ausgehenden 4. und Frühen 3. Jahrtausen

V. Chr. Pp. 247–59



in Between the Rivers and the Mountains,



eds. M. Frangipane, H. Hauptmann,
M. Liverani, P. Matthiae, and M. Mellink. Archaeologica Anatolica et mesopotamica Alba
Palmieri Dedicata. Rome: Università Degli Studi Di Roma La Sapienza.

Behm-Blancke, M. R., et al.
1981 Hassek Höyük, vorlaüfiger Bericht über die Grabungen in den Jahren 1981–1983.



Istanbuler
Mitteilungen



34: 31–149.

Ben-Tor, A.
1976 A Cylinder Seal from ºEn Besor.



ºAtiqot



11: 13–15.
1978



Cylinder Seals of Third-Millennium Palestine.



American Schools of Oriental Research Supple-
ment Series 22. Cambridge, MA: American Schools of Oriental Research.

1985 Glyptic Art of Third Millennium Palestine and Its Foreign Relations. Pp. 1–25 in



The Land of
Israel: Cross-Roads of Civilization,



ed. E. Lipi



n



ski. Leuven: Peeters.
1986 The Trade Relations of Palestine in the Early Bronze Age.



Journal of the Economic and Social
History of the Orient



29: 1–27.
1989 Byblos and the Early Bronze Age I of Palestine. Pp. 41–52 in



L’urbanisation,



ed. P. de Miroschedji.
British Archaeological Reports, International Series 527. Oxford: British Archaeological Reports.

1992 New Light on Cylinder Seal Impressions Showing Cult Scenes from Early Bronze Age Palestine.



Israel Exploration Journal



42: 153–64.
1994 Early Bronze Age Cylinder Seal Impressions and a Stamp Seal from Tel Qashish.



Bulletin of
the American Schools of Oriental Research



295: 15–29.
1995 Cylinder Seal Impressions of Early Bronze Age Israel—The Present State of Research. Pp. 65–

80 in



Seals and Sealing in the Ancient Near East, Proceedings of the Symposium Held on
September 2, 1993, Jerusalem, Israel,



ed. J. G. Westenholz. Jerusalem: Bible Lands Museum.

Boehmer, R. M.
1974a Das Rollsiegel im prädynastichen Ägypten.



Archäologischer Anzeiger



4: 495–514.
1974b Orientalische Einfluss auf verzeiten Messergriffen aus dem prädynastiche Ägypten.



Archäolo-
gische Mitteilungen aus Iran



7: 15–40.
1991a 14C-daten aus Uruk und Abydos—Ägyptisches (?) im frühen nordsyrien, Sumer und Elam.



Baghdader Mitteilungen



22: 223–30.
1991b Gebel-el-Arak- und Gebel-el-Tariff-Griff: Keine Fälschungen.



Mitteilungen des Deutschen
Archäologischen Institut, Abteilung Kairo



47: 51–60.

Boehmer, R. M.; Dreyer, G.; and Kromer, B.
1993 Einige frühzeitliche 14C-Datierungen aus Abydos und Uruk.



Mitteilungen des Deutschen
Archäologischen Instituts



,



Abteilung Kairo



49: 63–68.

Brandes, M. A.
1979



Siegelabrollungen aus den archaischen Bauschichten in Uruk-Warka.



Freiburger Altorientalische
Studien Band 3. Wiesbaden: Franz Steiner Verlag.

Braun, E.
1990 Basalt Bowls of the EB I Horizon in the Southern Levant.



Paléorient



16: 87–96.

Broodbank, C.
1993 Ulysses without Sails. Trade, Distance, Knowledge and Power in the Early Cyclades.



World
Archaeology



24: 315–31.

oi.uchicago.edu





371



EARLY BRONZE AGE SEAL IMPRESSIONS FROM THE JEZREEL VALLEY



Caskey, M.
1990 Thoughts on Early Bronze Age Hearths. Pp. 13–21 in



Celebrations of Death and Divinity in the
Bronze Age Argolid,



eds. R. Hägg and G. C. Nordquist. Stockholm: Svenska Institutet i Athen.
Chesson, M. S.

1998 Preliminary Results of the Excavations at Tell el-Handaquq South.



Palestine Exploration Quar-
terly



130: 20–34.
Collon, D.

1987



First Impressions, Cylinder Seals in the Ancient Near East.



Chicago: University of Chicago
Press.

De Cree, F.
1987–88 Votive Beds in the Rockefeller Museum, Jerusalem. A Case of Mesopotamian Influence in IIIrd

Millennium Palestine?



Acta Archaeologica Lovaniensia



26–27: 109–12.
de Vaux, R.

1952 La quatrième campagne de fouilles à Tell el-Fârºah, près Naplouse.



Revue Biblique



59: 551–83.
Engberg, R. M., and Shipton, G. M.

1934



Notes on the Chalcolithic and Early Bronze Age Pottery of Megiddo



. Studies in Ancient Oriental
Civilization 10. Chicago: University of Chicago Press.

Epstein, C.
1972 Early Bronze Age Seal Impressions from the Golan.



Israel Exploration Journal



22: 209–17.
Esse, D. L.

1990 Early Bronze Age Cylinder Seal Impressions from Beth Yera



˙



.



Eretz-Israel



21: 27*–34*.
1991



Subsistence, Trade, and Social Organization in Early Bronze Age Palestine. Studies in Ancient
Oriental Civilization 50. Chicago: Oriental Institute, University of Chicago.

1993 Tel Yaqush. Pp. 1502–4 in The New Encyclopedia of Archaeological Excavations in the Holy
Land, vol. 4, ed. E. Stern. Jerusalem: Israel Exploration Society & Carta and New York: Simon
& Schuster.

Ferioli, P., and Fiandra, E.
1983 Clay Sealings from Arslantepe VIA: Administration and Bureaucracy. Pp. 455–509 in Perspec-

tives on Protourbanization in Eastern Anatolia, Arslantepe (Malatya). An Interim Report on the
1975–1983 Campaigns, eds. M. Frangipane and A. Palmieri. Origini 12.

1994 Archive Techniques and Methods at Arslantepe. Pp. 149–61 in Archives before Writing, Pro-
ceedings of the International Colloquium, Oriolo Romano, October 23–25, 1991, eds. P. Ferioli,
E. Fiandra, G. G. Fissore, and M. Frangipane. Turin: Scriptorium.

Finkbeiner, U.
1995 Tell el-ºAbd. Vorläufiger Bericht über die Ausgrabungen 1992 und 1993. Damascener Mitteilun-

gen 8: 51–83.
Finkelstein, I.

1990 Early Arad—Urbanism of the Nomads. Zeitschrift de Deutschen Palästina-Vereins 106: 34–50.
Fissore, G. G.

1994 Conceptual Development and Organization Techniques in the Documents and Archives of the
Earliest Near Eastern Civilizations. Pp. 339–54 in Archives before Writing, Proceedings of the
International Colloquium, Oriolo Romano, October 23–25, 1991, eds. P. Ferioli, E. Fiandra,
G. G. Fissore, and M. Frangipane. Turin: Scriptorium.

Gal, Z.
1988 H. Shahal Tahtit and the Early Enclosures. Israel Exploration Journal 38: 1–5.
1992 Lower Galilee during the Iron Age. American Schools of Oriental Research Dissertation Series

8. Winona Lake, IN: Eisenbrauns.
Greenberg, R.

1996 The Early Bronze Age Levels. Pp. 83–160 in Dan I: A Chronicle of the Excavations, the Pottery
Neolithic, the Early Bronze Age and the Middle Bronze Age Tombs, eds. A. Biran, D. Ilan, and
R. Greenberg. Jerusalem: Nelson Glueck School of Biblical Archaeology, Hebrew Union
College-Jewish Institute of Religion.

oi.uchicago.edu



372 ALEXANDER H. JOFFE

1997 The Early Bronze Age Phase in Area L. Pp. 183–93 in Hazor V: An Account of the Fifth Season
of Excavation, 1968, eds. A. Ben-Tor and R. Bonfil. Jerusalem: Israel Exploration Society.

Greenberg, R., and Porat, N.
1996 A Third Millennium Levantine Pottery Production Center: Typology, Petrography and Provenance

of the Metallic Ware of Northern Israel and Adjacent Regions. Bulletin of the American Schools
of Oriental Research 301: 5–24.

Heath (Wiencke), M. C.
1958 Clay Sealings from the House of the Tiles at Lerna. Hesperia 27: 81–120.

Helms, S. W.
1987 A Question of Economic Control during the Proto-historical Era of Palestine/Transjordan. Pp. 41–

51 in Studies in the History and Archaeology of Jordan III, ed. A. Hadidi. Amman: Department
of Antiquities.

1991a The Pottery. Pp. 55–109 in Excavations at Jawa 1972–86: Stratigraphy, Pottery and Other
Finds, ed. A. V. G. Betts. Edinburgh: Edinburgh University Press.

1991b Stamped, Incised, and Painted Designs on Pottery. Pp. 110–39 in Excavations at Jawa 1972–86:
Stratigraphy, Pottery and Other Finds, ed. A. V. G. Betts. Edinburgh: Edinburgh University
Press.

1992 The Pottery Typology. Pp. 39–121 in Excavations at Tell Um Hammad 1982–1984: The Early
Assemblages (EB I–II), ed. A. V. G. Betts. Edinburgh: Edinburgh University Press.

Hendrickx, S., and Bavay, L.
Forthcoming The Relative Chronological Position of Egyptian Predynastic and Early Dynastic Tombs with

Imported Objects from the Near East and the Nature of Interregional Contacts.

Hennessy, J. B.
1967 The Foreign Relations of Palestine During the Early Bronze Age. London: Quaritch.

Ji, H.
1988 Seals and Seal Impressions from Tell Gubba. Al-Rafidan 9: 98–133. (Japanese).

Joffe, A. H.
1993 Settlement and Society in the Early Bronze I and II of the Southern Levant: Complementarity and

Contradiction in a Small-Scale Complex Society. Monographs in Mediterranean Archaeology 4.
Sheffield: Sheffield Academic Press.

2000a The Stratigraphy and Architecture of Area H. Pp. 140–60 in Megiddo III, eds. I. Finkelstein,
D. Ussishkin, and B. Halpern. Tel Aviv: Ramot.

2000b The Early Bronze Age Pottery. Pp. 161–85 in Megiddo III, eds. I. Finkelstein, D. Ussishkin, and
B. Halpern. Tel Aviv: Ramot.

2000c Egypt and Syro-Mesopotamia in the Fourth Millennium: Implications of the New Chronology.
Current Anthropology 41: 113–23.

n.d. The Levant in the Mid-Third Millennium. Unpublished manuscript.

Joffe, A. H., and Dessel, J. P.
1995 Redefining Chronology and Terminology for the Chalcolithic of the Southern Levant. Current

Anthropology 36: 507–18.

Kaplan, J., and Ritter-Kaplan, H.
1993 Jaffa. Pp. 655–59 in The New Encyclopedia of Archaeological Excavations in the Holy Land,

vol. 2, ed. E. Stern. Jerusalem: Israel Exploration Society & Carta and New York: Simon &
Schuster.

Keel-Leu, H.
1989 Die frühesten Stempelsiegel Palästinas. Von den Anfängen bis zum Ende des 3. Jahrtausends.

Pp. 1–38 in Studien zu den Stempelsiegeln aus Palästina/Israel, eds. O. Keel, H. Keel-Leu, and
S. Schroer. Orbis Biblicus et Orientalis 88. Freiburg: Universitätsverlag.

Kochavi, M.
1994 The Land of Geshur Project, 1993. Israel Exploration Journal 44: 136–41.

oi.uchicago.edu



373EARLY BRONZE AGE SEAL IMPRESSIONS FROM THE JEZREEL VALLEY

Kohl, P. L.
1978 The Balance of Trade in Southwestern Asia in the Mid-Third Millennium b.c. Current Anthro-

pology 19: 463–83.
1982 The First World Economy: External Relations and Trade in West and Central Asia in the Third

Millennium b.c. Pp. 23–32 in Mesopotamien und Seine Nachbarn, eds. H. Nissen and J. Renger.
Berlin: Dietrich Reimer.

Konsola, D.
1990 Settlement Size and the Beginning of Urbanization. Pp. 463–71 in L’Habitat Égéen Préhis-

torique. Actes de la Table ronde internationale organisée par le Centre national de la recherche
scientifique, l’Université de Paris I et l’École français d’Athenes (Athenes, 23–25 juin 1987),
eds. P. Darcque and R. Treuil. École Française d’Athenes. Paris: Diffusion de Boccard.

Lamberg-Karlovsky, C. C.
1988 The Intercultural Style Carved Vessels. Iranica Antiqua 23: 45–95.
1993 The Intercultural Style Vessels of the Third Millennium b.c. Pp. 270–92 in History from Things:

Essays on Material Culture, eds. S. Lubar and W. D. Kingery. Washington, DC: Smithsonian
Institution.

Lapp, N. L.
1989 Cylinder Seals and Impressions of the Third Millennium b.c. from the Dead Sea Plain. Bulletin

of the American Schools of Oriental Research 273: 1–15.
1995 Some Early Bronze Age Seal Impressions from the Dead Sea Plain and Their Implications for

Contacts in the Eastern Mediterranean. Pp. 43–51 in Trade, Contact, and the Movement of Peo-
ples in the Eastern Mediterranean: Studies in Honour of J. Basil Hennessy, eds. S. Bourke and
J.-P. Descoeudres. Mediterranean Archaeology Supplement 3. Sydney: Meditarch.

Leonard, A., Jr.
1992 The 1953 Jordan Valley Survey: Some Unpublished Soundings Conducted by James Mellaart.

Annual of the American Schools of Oriental Research 50. Winona Lake, IN: Eisenbrauns.

Loud, G.
1948 Megiddo II, Seasons of 1935–1939. Oriental Institute Publications 62. Chicago: University of

Chicago Press.

Mabry, J.
1989 Investigations at Tell el-Handaquq, Jordan. Annual of the Department of Antiquities of Jordan

33: 137–44.

Manning, S. W.
1994 The Absolute Chronology of the Aegean Early Bronze Age, Archaeology, Radiocarbon and

History. Monographs in Mediterranean Archaeology 1. Sheffield: Sheffield Academic Press.

Marfoe, L.
1995 Kamid el-Loz 13. The Prehistoric and Early Historic Context of the Site. Saarbrücker Beiträge

zur Altertumskunde Band 41. Bonn: Rudolph Habelt.

Mark, S.
1998 From Egypt to Mesopotamia: A Study of Predynastic Trade Routes. College Station: Texas

A&M University Press.

Matthews, D. M.
1997 The Early Glyptic of Tell Brak. Cylinder Seals of Third Millennium Syria. Orbis Biblicus et

Orientalis, Series Archaeologica 15. Fribourg and Göttingen: University Press and Vandenhoeck
& Ruprecht.

Matthews, R. J.
1993 Cities, Seals and Writing: Archaic Seal Impressions from Jemdet Nasr and Ur. Materialien zu

den frühen Schriftzeugnissen des Vorderen Orients 2. Berlin: Gebr. Mann.

Mazzoni, S.
1985 Giza ed Una Produzione Vascolare Di Biblo. Pp. 317–38 in Studi in Onore di Edda Bresciani,

eds. S. F. Bondì, S. Pernigotti, F. Serra, and A. Vivian. Pisa: Giardini.

oi.uchicago.edu



374 ALEXANDER H. JOFFE

1987 The Diffusion of the Palestinian Combed Jars. Pp. 145–58 in Studies in the History and Archae-
ology of Palestine II. Aleppo: Aleppo University.

1992 Le impronte su giara eblaite e siriane nel Bronzo Antico. Materiali e Studi Archaeologici de
Ebla 1. Rome: Università degli Studi di Roma la Sapienza.

1993 Cylinder Seal Impressions on Jars at Ebla: New Evidence. Pp. 399–414 in Aspects of Art and
Iconography: Anatolia and Its Neighbors. Studies in Honor of Nimet Özgüç, eds. M. Mellink,
E. Porada and T. Özgüç. Ankara: Türk Tarih Kurumu Basımevi.

McClellan, T. L., and Porter, A.
1995 Jawa and North Syria. Pp. 49–65 in Studies in the History and Archaeology of Jordan V: Art and

Technology throughout the Ages, eds. K. ºAmr, F. Zayadine, and M. Zaghloul. Amman: Depart-
ment of Antiquities.

Miroschedji, P. de
1993 Chalcolithic and Early Bronze Age Religion. Pp. 208–20 in Biblical Archaeology Today 1990:

Proceedings of the 2nd International Congress of Biblical Archaeology, eds. A. Biran and
J. Aviram. Jerusalem: Israel Exploration Society.

Moorey, P. R. S.
1990 From Gulf to Delta in the Fourth Millennium: The Syrian Connection. Eretz-Israel 21: 62*–69*.
1995 From Gulf to Delta and Beyond. Beer-Sheva 8. Beer-Sheva: Ben-Gurion University.

Nissen, H. J.
1977 Aspects of the Development of Early Cylinder Seals. Pp. 15–23 in Seals and Sealing in the

Ancient Near East, eds. M. Gibson and R. D. Biggs. Malibu: Undena.

Peltenburg, E.; Bolger, D.; Campbell, S.; Murray, M. A.; and Tipping, R.
1996 Jerablus-Tahtani, Syria, 1995: Preliminary Report. Levant 28: 1–26.

Peltenburg, E.; Campbell, S.; Carter, S.; Stephen, F. M. K.; and Tipping, R.
1997 Jerablus-Tahtani, Syria, 1996: Preliminary Report. Levant 29: 1–18.

Philip, G.
1995 Jawa and Tell Um Hammad: Two Early Bronze Age Sites in Jordan: Review Article. Palestine

Exploration Quarterly 127: 161–70.

Philip, G., and Williams-Thorpe, O.
1993 A Provenance Study of Jordanian Basalt Vessels of the Chalcolithic and Early Bronze Age I

Periods. Paléorient 19: 51–63.

Pittman, H.
1994 The Glazed Steatite Glyptic Style: The Structure and Function of an Image System in the Admin-

stration of Protoliterate Mesopotamia. Berliner Beiträge Vorderen Orient 14. Berlin: Dietrich
Reimer.

Pullen, D. J.
1994 A Lead Seal from Tsoungiza, Ancient Nemea, and Early Bronze Age Aegean Sealing Systems.

American Journal of Archaeology 98: 35–52.

Raban, A.
1982 Nahalal Map (28) 16–23. Jerusalem: Archaeological Survey of Israel.

Rosen, A. M.
1995 The Social Response to Environmental Change in Early Bronze Age Canaan. Journal of Anthro-

pological Archaeology 14: 26–44.

Rothman, M.
1994 Sealing Use and Changes in Administrative Oversight and Structure at Tepe Gawra during the

Fourth Millennium b.c. Pp. 97–119 in Archives before Writing, Proceedings of the International
Colloquium, Oriolo Romano, October 23–25, 1991, eds. P. Ferioli, E. Fiandra, G. G. Fissore, and
M. Frangipane. Turin: Scriptorium.

Rutter, J. B.
1993 The Prepalatial Bronze Age of the Southern and Central Greek Mainland. American Journal of

Archaeology 97: 745–97.

oi.uchicago.edu



375EARLY BRONZE AGE SEAL IMPRESSIONS FROM THE JEZREEL VALLEY

Shaw, J. W.
1990 The Early Helladic II Corridor House: Problems and Possibilities. Pp. 183–94 in L’Habitat

Égéen Préhistorique. Actes de la Table ronde internationale organisée par le Centre national de
la recherche scientifique, l’Université de Paris I et l’École français, eds. P. Darcque and R.
Treuil. École Française d’Athenes. Paris: Diffusion de Boccard.

Sievertsen, U.
1992 Das Messer vom Gebel el-Arak. Baghdader Mitteilungen 23: 1–75.

Smith, H. S.
1992 The Making of Egypt: A Review of the Influence of Susa and Sumer on Upper Egypt and Lower

Nubia in the Fourth Millennium b.c. Pp. 235–46 in The Followers of Horus: Studies Dedicated
to Michael Allen Hoffman, eds. R. Friedman and B. Adams. Oxford: Oxbow Books.

Teissier, B.
1987 Glyptic Evidence for a Connection between Iran, Syro-Palestine and Egypt in the Fourth and

Third Millennia. Iran 25: 27–53.

Vieweger, D.
1997 An Early Bronze Age Handle with Cylinder Seal Impression from Dayr Qiqub, Jordan. Levant

29: 147–52.

Weingarten, J.
1997 Another Look at Lerna: An EH IIB Trading Post? Oxford Journal of Archaeology 16: 147–66.

van den Brink, E. C. M.
1995 The ºEn Besor Cylinder Seal Impressions in Retrospect. Pp. 201–14 in Excavations at ºEn

Besor, ed. R. Gophna. Tel Aviv: Ramot Publishing.

von der Way, T.
1992 Excavations at Tell el-Faraªin/Buto in 1987–1989. Pp. 1–10 in The Nile Delta in Transition: 4th–

3rd Millennium B.C., ed. Edwin C. M. van den Brink. Tel Aviv: Edwin C. M. van den Brink.
1993 Untersuchungen zur Spätvor- und Frühgeschichte Unterägyptens. Studien zur Archäologie und

Geschichte Altägyptens Band 8. Heidelberg: Heidelberger Orientverlag.

von Wickede, A.
1990 Prähistorische Stempelglyptik in Vorderasien. Münchner Vorderasiatische Studien Band 6.

Munich: Profil.

Wiencke, M. H.
1969 Further Seals and Sealings from Lerna. Hesperia 38: 500–21.
1970 Banded Pithoi of Lerna III. Hesperia 39: 93–110.
1989 Change in Early Helladic II. American Journal of Archaeology 93: 495–509.

Yannai, E.
2000 Tell Assawir. Presentation at the Conference on Early Bronze Age Pottery in the Southern

Levant, Durham, July, 1995.

Zarzecki-Peleg, A.
1993 Decorated Animal Bones of the Third Millennium b.c. from Palestine-Syria: Stylistic Analysis.

Israel Exploration Journal 43: 1–22.

Zertal, A.
1993 Fortified Enclosures of the Early Bronze Age in the Samaria Region and the Beginning of

Urbanization. Levant 25: 113–25.

oi.uchicago.edu



oi.uchicago.edu



377



THE COLLARED PITHOS IN CONTEXT: A
TYPOLOGICAL, TECHNOLOGICAL, AND

FUNCTIONAL REASSESSMENT



Ann E. Killebrew



The collared pithos has attracted the attention of many scholars and archaeologists for
nearly seventy years, largely owing to its reputation as a container associated with the emer-
gence of Israel during the Iron I period (e.g., Finkelstein 1988: 275–85; Esse 1991, 1992;
Wengrow 1996). However, though it is one of the most discussed vessels appearing in the
ceramic assemblage of ancient Israel, its significance is still poorly understood. This paper,
dedicated to the memory of Douglas Esse and in appreciation of his ground-breaking articles
(Esse 1991, 1992) on the collared pithos, attempts to reexamine its significance in the archae-
ological record. Two sites where collared pithoi have appeared, Beit Shean and Giloh, are
showcased here as representing two different phenomena associated with the appearance of
this vessel. A typology is proposed for the collared pithos, and an examination of its context
and provenience is summarized with the goal of reaching several conclusions about its devel-
opment, modes of production, and function from its earliest appearance at the close of the Late
Bronze Age (LB) through its floruit in the Iron I period.

The collared pithos was first defined as a specific ceramic type by Kjaer (1930), who
described it as a storage jar whose rim was folded over with a ridge in low relief at the base
of the neck. This ridge is the characteristic feature that later inspired the term “collared store-
jar” applied to this pithos.



1



Most of the discussion surrounding the collared pithos has focused
on whether these jars can be used as “ethnic” indicators or “type fossils” of premonarchic
Israelites (see, e.g., Albright 1940: 548; London 1989a; Esse 1992). Several studies have
traced the geographic distribution of the collared store jars in general (Ibrahim 1978; Mazar
1981: 27–31; Finkelstein 1988: 281–82; Ji 1995: 136–38; Wengrow 1996: 312–16). They
have been found in large quantities during the Iron I period in the central hill country;



2



in



1. W. F. Albright (1934: 12) was the first to coin this term. The name,



collared store jar



or



pithos



, is an unfor-
tunate term resulting in a great deal of confusion regarding this vessel type. Several scholars have mis-
understood this term as applying to the “collar-shaped rim” (e.g., Callaway 1969: 8–9; Wengrow 1996: 307),
rather than the ridge at the base of the neck. It should also be noted that most pithoi of LB and Iron I periods
have a “ridge” at the base of the neck. Thus, this feature is not distinctive enough to be used to characterize
these pithoi. In order not to further confuse the issue, I have followed the conventional designation for this
vessel; however, I prefer the term



collared pithos



rather than



collared rim pithos



since the former more
accurately describes the feature after which these pithoi were named.

2. See, e.g., Manasseh: Zertal 1991: 35–36; Mount Ebal: Zertal 1986–87: 134–36; Shiloh: Kjaer 1930: 97–
104; Buhl and Holm-Nielsen 1969: pls. 15, 16: 190, 191, 192; Finkelstein 1988: 220–27; Finkelstein, Bun-
imovitz, and Lederman. 1993: figs. 6.48:1, 2, 4; 6.49:3, 4; Northern Ramat Manasseh ºEn



Ó



agit: Wolff
1998: fig. 3; Bethel: Kelso 1968: pls. 56, 57:1–5; Tell en-Nasbeh: Wampler 1947: 5; pl. 2:16–23, 26–28; Tell



20



oi.uchicago.edu





378



ANN E. KILLEBREW



FIGURE. 20.1. Collared Pithos Type A from Beit Shean (scale: 1:5).

oi.uchicago.edu

[image: image119.tif]



379



THE COLLARED PITHOS IN CONTEXT



large to moderate numbers in Transjordan;



3



in smaller quantities in the Upper Galilee



4



and the
Jezreel Valley;



5



and only occasionally on the coastal plain



6



and Shephelah.



7



Recently, it has
become clear that this jar made its debut in modest numbers already at the end of the LB IIB
period at several sites along the coast



8



and in the Jezreel and Jordan Valleys.



9





TYPOLOGY

The large storage container, or pithos, is a shape known from the LB II period in the region
of Canaan north of the Jezreel Valley; until the end of the thirteenth century, pithoi are virtu-
ally unknown south of the Jezreel Valley. In southern Canaan, only the smaller storage jar—
the so-called Canaanite storage jar—measuring some 50 cm in height, was used as a storage
container. The collared pithos makes its initial appearance at the close of the LB. It is twice as
tall as contemporary Canaanite storage jars, measuring 1 m or greater in height. The defining



3. See, e.g., Ibrahim 1978, Ji 1995: 122–26, Wengrow 1996: 315–16 and Herr, this volume, chap. 14 for a gen-
eral overview of the distribution of the collared pithos in Transjordan. Numerous fortified settlements in the
Transjordanian highlands dated from the thirteenth/twelfth to twelfth centuries



b.c.



contained notable quan-
tities of collared pithoi; see Herr 1991, e.g., fig. 8.9:1; Herr et al. 1991: 159 [Field B, Phase 11], figs. 5, 6,
which discuss in detail collared pithoi at Tell el-Umeiri. Large pithoi similar to our Pithos Type A also
appear as far south as Edom: Weippert 1982: figs. 6:9, 8:9 for two pithoi fragments that are not clearly “col-
lared”; Bienkowski 1992: 167, who dates the pithoi from Buseirah and Tawilan to the seventh century, while
Finkelstein 1992a; 1992b: 171 suggests an Iron I date.

4. See, e.g., Dan: Biran 1989: Stratum VI: figs. 4.16:8, 4.18:6–8, 4.22, 4.23; 1994: Stratum VI: Ill. 92; Stratum
V: Ill. 95; Hazor, Stratum XII: Yadin et al. 1961: pl. CLXVII:1–7, 9–11 for rims that are similar in shape to
the rims on the collared jars; Yadin et al. 1989: 29 note that these jars “are also related to Albright’s collared
type”; see Yadin et al. 1961: pl. CLXVIII: esp. 9–11 for rim profiles similar to our collared pithos Type A
from Beit Shean. Sasa: Bahat 1986: 100: no. 1; Golani and Yogev 1996: 51; fig. 6:4; Stepansky, Segan, and
Carmi 1996: 68; fig. 7:2.

5. See, e.g., ºAfula: Dothan 1955: Stratum IIIA: fig. 11:25; Stratum IIIB: fig. 16:4; Megiddo, mainly Stratum
VIA: see, e.g., Loud 1948: pl. 83:4; and possibly a few from Stratum VII; see Esse 1992 for a summary of
the evidence; Qiri: Ben-Tor and Portugali 1987: Stratum VIII: Photos 38–41; Hunt 1987: 200; fig. 36;
southern edge of the Jezreel Valley: Tanaach: Rast 1978: 9–10; figs. 4:1, 9:1, 10:1, 3, 4.

6. See, e.g., Tell Keisan: Niveau 9c: Briend and Humbert 1980: pl. 68:1, 2, 2a; Tel Mevorakh: Stern 1978: 68–
69 (tenth century date); fig. 19:4; Qasile: Mazar 1985: 57–58.

7. Beth Shemesh: Grant and Wright 1939: 129; Tell Beit Mirsim: Albright 1932: 59; pl. 26:18.

8. Nahariya: Yogev 1993: 1088–89; Tel Nami: Artzy 1993; 1994: 127–28; in press; Tel Zeror: Ohata 1970: pl.
56; Aphek: Beck and Kochavi 1985: fig. 5:1.

9. Beit Shean, Stratum VII: James and McGovern 1993: figs. 23:2, 4; 32:4; Megiddo: see Esse 1992: n. 47 who
states there is at least one pithos Type A attributed to Stratum VIIB; ºAfula: Stratum IIIB: see Dothan 1955:
38–41 for a description of pottery found with the collared store jar rim including Cypriot imports; Tell es-
Saºidiyeh: Tubb et al. 1996: Stratum XV: fig. 20.



el-Ful: Sinclair 1960: 16–18; pl. 20:2–17; Khirbet Raddana: Callaway and Cooley 1971: 11; Ai: Callaway
1969: 8–9; fig. 8; 1980: fig. 154; 155:1–13; Giloh: Mazar 1981: 27–31; fig. 9; pl. 5A, B; 1990: fig. 5; Beth
Zur: Funk 1968: 44–47; figs. 7, 8:21–29; pls. 21, 22a. The collared pithos appears in much smaller quantities
at Khirbet Rabud: Kochavi 1974: fig. 5:12 and Izbet



Í



ar



†



ah (on the western fringe of the hill country):
Finkelstein 1986: 77–84, Type 21.

oi.uchicago.edu





380



ANN E. KILLEBREW



characteristic features of this pithos with two handles include the raised ridge at the base of the
neck and the wide variety of rim and body profiles.

Several attempts have been made to trace a typological and chronological development.
Scholars such as Kelso (1968: 63) and Rast (1978: 9) divide the collared pithoi at Bethel and
Taanach into two main types—tall-necked jars and short-necked jars. Other proposals for a
typological division of collared pithoi include Callaway (1969: 8–9), who posits that the
length of the folded rim has a chronological significance, concluding that the longer fold of the
rim is earlier in date while the shorter folded rim is later, unrelated to the length of the neck.



10



A different approach is suggested by Finkelstein (1986: 77–78; Finkelstein, Bunimovitz, Led-
erman 1993: 159), who distinguishes between “thick” and “thin” rims and concludes that the
thickened rim type seems to be earlier in date. However, due to the great variety of rim forms
that appear contemporaneously (as noted by Mazar [1981: 28–29] at Giloh and by M. Artzy
[pers. comm.] at Tel Nami), it is doubtful that a typology can be developed based on rim pro-
files (also see Wampler 1947: 4, who also notes the great variety of rims and the difficulty in
establishing a typological sequence for these jars).

I follow the typological differentiation based on neck height, as first suggested by Kelso
and Rast. I divide the collared pithoi into Types A and B, based on general morphological fea-
tures of the entire rim and neck and on an analysis of their chronological and stratigraphic
appearance in the archaeological evidence. These two main groups are based on a survey of
published collared pithoi and use of one complete example from Beit Shean and one from
Giloh to illustrate the two main types. Both collared pithoi types share basic morphological
characteristics, including a piriform-shaped body profile, with two handles on the shoulder,
and a raised “collar” at the base of the neck. The distinguishing feature in this typology is the
length of the neck measured from the collar to the top of the rim. Collared pithos Type A has
a neck measuring about 10 cm or greater in height, while Type B has a shorter neck, less than
10 cm high, usually measuring from 5.0 to 7.5 cm. This feature appears to be chronologically
significant and may reflect the typological development of this pithos.



Collared Pithos Type A (fig. 20.1)



Collared Pithos Type A from Beit Shean was recovered from a large public structure in
Area NA (Building B) assigned to Level VII.



11



This building was probably part of a large late
thirteenth–early twelfth century



b.c.



New Kingdom Egyptian fortress overlooking the Jordan
Valley and was destroyed by a conflagration, undoubtedly fueled by the large quantities of
grain that were discovered in a carbonized state during its excavation. The complete collared
pithos was found in the doorway connecting two rooms, together with several Egyptian-style
bowls, an Egyptian-style cooking pot, two imported Egyptian vessels, an imported Cypriot
White Slip II bowl, and several Canaanite-style kraters and storage jars. This stratum has been
dated to the very end of the LB II or early Iron I period (Mazar 1993a, 1993b, 1994a, 1997a,
1997b).



10. This typology is based on Callaway’s confusion with the term “collar,” which he understood to refer to the
folded rim rather than to the ridge at the base of the neck.

11. I would like to thank Professor Amihai Mazar and the Beit Shean Expedition for allowing me to publish this
pithos from Area NA.



long to pull back 2 lines after table

oi.uchicago.edu





381



THE COLLARED PITHOS IN CONTEXT



The collared pithos excavated in this context is characterized by its tall neck, measuring 10
cm in height. The everted rim with a stepped molding on the container from Beit Shean is
similar in concept to rims on LB northern pithoi found in great numbers at Hazor (Yadin et al.
1960: pl. CXLV:3–5; 1961: pl. CCXCVIII:8, 9; note that Rast [1978: 9, fig. 1:1] also comments
that a tall-necked collared pithos rim from Taanach is reminiscent of Late Bronze II pithoi
from Hazor).

The tall-necked collared pithos Type A appears in noteworthy quantities at the fortified
site of Tell el-



º



Umeiri (Herr et al. 1991: figs. 5:5, 6, 7, 10; 6:1, 2, 3, 6, 7; Herr 1997: 237–38;
see, e.g., figs. 4:14; 4:19:5–8; 4:20:1–3, 5–7); however, smaller numbers of short-necked col-
lared pithoi Type B also appear together with Type A. Tall-necked collared pithoi are also
known from late thirteenth century contexts at Tel Nami (Artzy, pers. comm.), Iron I levels at
Hazor (Stratum XII: Yadin et al. 1961: pl. CLXVIII: especially nos. 9–11), and one possible
example at Taanach (Rast 1978: fig. 1:1). Type A appears in much smaller quantities than
Type B, and stratified examples date to the end of the LB and early Iron I period. It has been
recovered mainly in the Jezreel Valley, northern Canaan, and Transjordan. Noteworthy is its
apparent absence at central hill country sites.



Collared Pithos Type B (fig. 20.2)



Giloh, located in the modern neighborhood of that name, southwest of Jerusalem, is excep-
tional because of its barren, waterless terrain and remote location from the main route travers-
ing the central hill country. It is one of a handful of early Iron I (late 13th/12th centuries



b.c.



)
sites located in the vicinity of Jerusalem and has been interpreted as a fortified herdsman vil-
lage (Mazar 1981, 1990, 1994b). I have selected Giloh as a type site because of its distinctive
ceramic repertoire, consisting mainly of collared pithoi (33.7%), storage jars (16.6%), cooking
pots (22.6%), open vessels (9.4%), and small closed vessels (11.8%) (Mazar 1994b: 88), which
is characteristic of other small rural sites in the hill country.

All collared pithoi recovered at Giloh belong to Type B. These pithoi display a great variety
of rim profiles and do not appear to demonstrate any chronological significance (Mazar 1981:
n. 37 for comparative material to the Giloh rim profiles). The overwhelming majority of exca-
vated collared pithoi belong to this general group, especially from sites in the central hill country
such as Mt. Ebal (Zertal 1986–87, see, e.g., figs. 12:1; 13:1; 16:8, 13), Shiloh (Finkelstein, Bu-
nimovitz, and Lederman 1993: figs. 6.48:1, 2, 4; 6.49:3, 4; 6.51:1, 4, 6), Ai (Callaway 1980:
figs. 150:17–28; 154), Bethel (Kelso 1968: pls. 56; 57:1–5), Tell en-Nasbeh (Wampler 1947:
pls. 1:1, 3–11; 2:12–22), Tell el-Ful (Albright 1924: pl. XXVIII:17–24; Sinclair 1960: pl.
20:10–18), Jerusalem (City of David: Steiner 1994: figs. 4–6); Emek Rephªaim: Edelstein and
Milevski (1994: 19–20, fig. 12:1–2), and Beth Zur (Funk 1968: fig. 7). It is also the most popular
type at Dan Strata VI and V (Biran 1989: figs. 4.1:2–5; 4.16:8; 4.18:6), Megiddo Stratum VI
(Schumacher 1908: pl. XLVI: d; Loud 1948: pl. 84:1, Jar 120 and pl. 84:4, Jar 122, which
according to the excavators appear in Strata VIIB–VI; Esse 1992: fig. 3:1), Taªanach periods
IA–B (Rast 1978: figs. 4:1; 9:1; 10:1, 3, 4), ºEn



Ó



agit (Wolff 1998: fig. 3), ºIzbet



Í



ar



†



ah (Finkel-
stein 1986: fig. 9:1, 3, 4), Tell es-Saºidiyeh (Tubb, Dorrell, and Cobbing 1996: fig. 20), and
Deir ºAlla (Franken and Kalsbeek 1969: fig. 47:1–2; 1992: fig. 5–16:26). Collared pithoi Type
B also appear alongside collared pithoi Type A at Tel Nami (M. Artzy, pers. comm.) and Tell
el-ºUmeiri (e.g., Herr et al. 1991: fig. 6:5).

oi.uchicago.edu





382



ANN E. KILLEBREW



FIGURE 20.2. Collared Pithos Type B from Giloh (scale: 1:5; after Mazar 1990: fig. 5).

oi.uchicago.edu

[image: image120.tif]



383



THE COLLARED PITHOS IN CONTEXT



Type B collared pithoi have a long life-span, appearing from the end of the LB, reaching
their peak of popularity during the Iron I, and continuing into the Iron II period. They are
found in large numbers in the central hill country and in several regions of Transjordan (espe-
cially the Madaba Plains), with smaller numbers appearing in the Jezreel and Jordan valleys,
in the north at Tel Dan, and at several sites in the Shephelah and coastal plain.

WARE DESCRIPTION AND PROVENIENCE

Collared pithoi from these two representative sites were petrographically examined under
a polarizing microscope. The petrographic thin sections were formed by affixing a small frag-
ment of pottery to a glass slide and then grinding it with an abrasive powder until it was
reduced to about 0.03 mm in thickness. Most of the minerals in the pottery are transparent at
this stage and can be studied and identified under the petrographic microscope. One of the
main applications of petrographic examination and the identification of minerals is the deter-
mination of the raw material source; this is most effective when minerals of a limited geolog-
ical distribution are identified within the clay body. In these instances, it is a relatively easy
task to pinpoint an area (or areas) from which a particular vessel is most likely to have origi-
nated. In Israel, the usefulness of this technique has proven itself and has become an increas-
ingly essential component of pottery analysis (for recent examples see, e.g., Glass et al. 1993;
Goren 1996; Greenberg and Porat 1996).



Beit Shean



Geologically, Beit Shean is located in a very diverse region. Igneous basalt outcrops
appear in the eastern hills of the Lower Galilee to the north and sedimentary limestone and
chalk deposits belonging to the Eocene period form the Mt. Gilboa range to the south. In the
Beit Shean and Jordan valleys, calcite-rich alluvial sediments rest on Upper Pleistocene trav-
ertine and Lisan deposits. Most of these sediments are exposed within a 1 km perimeter of the
tell (Nir 1989; Shaliv, Mimran, and Hatzor 1991). Most of the Beit Shean ceramics from
Level VII were fashioned out of locally available clays.



Beit Shean Ware Groups



A total of three main ware groups and six additional ware groups have been defined in the
Level VII ceramic assemblage found with the collared pithos in Building B at Beit Shean.
Most of the vessels I examined (Killebrew 1998: 211–15) belong to the three main locally
produced wares designated as Wares BS-A, BS-B, and BS-C. The single collared pithos Type
A uncovered from the latest series of excavations from Beit Shean does not match any other
of the vessel wares and falls into a distinctive ware group termed Ware BS-D, which is differ-
ent from all other sampled wares and is defined by its non-calcareous matrix and large quan-
tities of basalt minerals.



WARE BS-D



(fig. 20.3). The collared pithos was formed from a clean, non-calcareous clay
containing small amounts of silty and sandy quartz (1–3%) and traces of mica. No carbonate
is visible in the matrix. Large quantities of subangular sand-sized basalt and minerals associ-
ated with basalt appear (about 10–20%). Ware BS-D originates in an area with a basaltic for-
mation and could possibly have originated (but not necessarily) in the Beit Shean region.

oi.uchicago.edu





384



ANN E. KILLEBREW



Giloh



Giloh is located to the south of Jerusalem, on a hill opposite and north of Beit Safafa.
Greater Jerusalem is situated in the Judaean hills, part of the central hill country. The area of
Jerusalem sits on bedrock consisting of limestone, dolomite, and chalk layers dating to the
Upper Cretaceous period. Two main groups are visible: the earlier Judea Group and the later
Mount Scopus Group (Gill 1996). Giloh is situated on a hill of the Weradim Formation, over-
lying the exposed Kefar Shaul Formation. These rest on the Aminadav Formation; all are part
of the Judea Group.



12



These formations consist of dolomite (Weradim and Aminadav Forma-
tions), limestone (Weradim, Kefar Shaul, and Aminadav Formations), and chalk (Kefar Shaul
Formation). In the nearby Na



˙



al Rephaim, outcrops of the Mo



˛



a Formation are exposed
(Arkin et al. 1976). It is noteworthy that a dolomite sand quarry, probably where sand-sized
dolomite temper was obtained, was found in Na



˙



al Rephaim (see Edelstein and Milevski 1994
for a description of this quarry which consists of a shaft hewn 5 or 6 m into the hillside; see
also Goren 1996: 51–52).



12. The Judea Group, which forms the Judean hills, comprises mostly hard limestones and dolomites. To the
west, the Judea Group forms the synclinorium of the Shephelah region upon which formations of the Mt.
Scopus/HaShephela Group were laid Buchbinder 1969: 3. See also Gill 1996 for further details regarding
the geology of the Jerusalem area.



short to balance next spread



FIGURE 20.3. Ware Group BS-D (width: 3.5 mm; crossed polarizers): Non-calcareous clay matrix with small
quantities of silt and sand-sized quartz (approx. 1–3%); noteworthy is the subangular sand-sized basalt (4–10%)
grains (e.g., in center and upper left of photograph).

oi.uchicago.edu





385



THE COLLARED PITHOS IN CONTEXT



Giloh Ware Groups



Three main ware groups were defined in the Giloh assemblages (Killebrew 1998: 215–19).
Two main locally available clay types were used in the production of the Giloh collared pithoi:
Ware GI-A has a fine clay matrix consisting of mainly clay-sized particles with considerable
quantities of sandy-sized dolomite temper; Ware GI-B has a dolomitic clay matrix and
includes several types of temper in its paste.



13



WARE GI-A.



This ware group is characterized by its pure clay matrix and varying amounts
of sand-sized dolomite, which is the dominant temper in vessels assigned to this fabric. Ware
GI-A is a well-known and well-documented clay type belonging to the “Mo



˛



a-Aminadav”
(Glass et al. 1993: 272–76) or the “Mo



˛



a clay-dolomitic sand” group (Goren 1996: 51–52). It
is a fine, iron-rich, calcareous clay with coarse dolomitic sand and is documented from sites in
the central hill country, including Tell en-Nasbeh and Raddana and its vicinity from the Chal-
colithic through Iron Ages (Porat 1989a, 1989b; Gilead and Goren 1989; Franken 1990: 79–85;
Goren 1996: 51–52). I have divided this major family into four subgroups based on temper.



GI-A1



(fig. 20.4). The very fine clay matrix is well prepared with densely tempered,
well-sorted, subangular sand-sized dolomite (



>



50%), and occasional mud balls. Quartz is
very rare.



13. The third ware group, GI-C, is a non-calcareous terra rossa assigned to clay Type 3. This clay type was used
mainly in the production of cooking pots (see Killebrew 1999) and occasionally bowls.

FIGURE 20.4. Ware Group GI-A1 (width: 3.5 mm; crossed polarizers): Calcareous clay matrix densely tempered
with well-sorted subangular sand-sized dolomite (



>



50%).

oi.uchicago.edu





386



ANN E. KILLEBREW



GI-A2



(fig. 20.5). This ware group has the same clay matrix as GI-A1 and includes large
quantities of well-sorted dolomite (



>



50%), a few shale fragments, and mud balls.



GI-A3



(fig. 20.6). This storage jar is formed out of a fine clay typical of Ware GI-A. It is
defined by its large quantity of dolomite silt (



>



50%) and small quantity of poorly sorted, sub-
angular, sand-sized dolomite, calcite, and mud balls (1–3%).



GI-A4.



A fourth subgroup includes moderate to large amounts of sand-sized temper and
small quantities of calcite and limestone temper. However, the clay matrix belongs to dolo-
mitic clay, which characterizes Ware GI-B. Ware GI-A4 includes features of both ware groups
GI-A and GI-B.



WARE GI-B



(fig. 20.7). This ware is similar to Group A, except that the matrix of this clay
consists mainly of silty dolomite. The temper includes moderate to large quantities of calcite,
limestone, chalk, and quartz, and small amounts of sand-sized dolomite and quartz temper. The
inclusions are less well sorted than in Ware GI-A. This resembles the Mo



˛



a marl-calcareous
sand clay group described in detail by Goren (1995: 301).

The most noteworthy feature of the Giloh assemblage is that all the pottery sampled,
including the collared pithoi, was formed out of clay available in the Jerusalem area. This is
unusual at most sites and is significant because it may be an indicator of the relative isolation



FIGURE 20.5. Ware Group GI-A2 (width 3.5 mm; crossed polarizers): Calcareous clay matrix with large quan-
tities of well-sorted subangular to subrounded sand-sized dolomite temper (



>



50%); shale fragments, visible in
the lower left corner of photograph, are also common in this ware.

oi.uchicago.edu





387



THE COLLARED PITHOS IN CONTEXT



of the Jerusalem area, especially during the early Iron I period. Pithoi and storage jars vary in
their clay matrix, all local to the Jerusalem area, and are characterized by the appearance of
varying amounts of dolomite temper and by the occasional presence of sand-sized limestone,
quartz, and other minerals in the matrix.



Provenience of Collared Pithoi



Other investigations have dealt with studies of provenience and questions of this vessel’s
origin. The results of limited neutron activation and petrographic analyses at Tel Dan (Yellin
and Gunneweg 1989), Shiloh (Glass et al. 1993), Tell Keisan (Courtois 1980), and Sasa
(Cohen-Weinberger and Goren 1996) indicate that several clay sources were used in the man-
ufacture of pithoi and that these large jars were not all locally produced. It is thus all the more
noteworthy that all the collared pithoi examined from Giloh were produced in the general area
of Jerusalem and none was imported from any distance. The lack of uniformity reflected in the
results of these studies may be due to the type of site and the context of the collared pithoi.
The significance of the diversity in the appearance and provenience of the collared pithoi is
discussed below.



FIGURE 20.6. Ware Group GI-A3 (width 3.5 mm; crossed polarizers): Calcareous clay matrix with a large quan-
tity of dolomite silt (



>



50%) and small quantities of poorly sorted subangular dolomite and calcite grains (approx.
3%); mud balls occasionally appear.

oi.uchicago.edu





388



ANN E. KILLEBREW



MODES OF PRODUCTION

Several models, based mainly on the ethnographic record, have been proposed to describe
the organization of pottery production, its distribution, and demand.



14



These models consider
technological features, such as formation techniques and variability of both raw materials and
products, as well as ecological, economic, and social criteria, such as frequency and seasonality
of production, number of workers, their age, sex, and status, degree of labor division, kind and
extent of investment in space or tools, and proximity of consuming groups (Rice 1987: 183–84).

S. E. van der Leeuw (1976: 394–98, 402–3) has divided pottery production into six differ-
ent states of pottery economy that can be defined by certain technological and economic char-
acteristics. His “states” are based on modes of production and can be divided into two basic
industries: (a)



domestic production,



including (1) household production and (2) household in-
dustries; and (b)



professional production,



including (3) workshop industries, (4) village indus-
tries, (5) large-scale industries, and (6) individual industries (see Killebrew 1998: 253–57 for
a detailed discussion of these modes of production). The latter industry involves pottery



14. Balfet 1965: 162–63; van der Leeuw 1976; 1984; Peacock 1981; 1982; Redman and Myers 1981: 289–90;
Rice 1981; Tosi 1984: 23–24; Santley, Arnold, and Pool 1989; Costin 1991; see Pool 1992: 280–83 for a dis-
cussion of these terms and their relationship to ancient ceramics.

FIGURE 20.7. Ware Group GI-B (width: 3.5 mm; crossed polarizers): Dolomitic clay with a moderate amount of
silty dolomite (4–10%); moderate amounts of large grains (up to 1 mm) of calcite, limestone (subangular grain
on lower left), and chalk (rounded grain on right) appear; mudballs (upper half, center) and quartz sand (approx.
1–3%) are present.

oi.uchicago.edu





389



THE COLLARED PITHOS IN CONTEXT



manufactured by a single specialist: often itinerant potters produce pottery for an entire region
from locally available materials at each village or settlement.

A mixed picture is presented regarding possible modes of production of collared pithoi at
various Iron I sites. Although no workshops or kilns relating to the manufacture of pithoi have
been excavated to date, owing to the quality and level of difficulty involved in the production
of the collared pithos, a professional potters’ craft is indicated (see, e.g., London 1989b). At
Tel Dan (Yellin and Gunneweg 1989) and Shiloh (Glass et al. 1993), several origins are indi-
cated for the collared pithoi jars, while all the pithoi at Giloh were produced in the general
vicinity. The diversity of wares and origins of the collared pithoi indicate that several workshops
and/or numerous potters were involved in their production. Thus, I tentatively propose that the
mode of production used to produce the collared pithos, with its diverse procurement of raw
materials and clay preparation, may indicate the existence of several workshops. At sites such
as Giloh where all the pithoi were produced in the general vicinity but exhibit great variation
in their wares, I suggest production by itinerant potters who traveled from village to village in
the region, perhaps in addition to pithos manufacture in a local workshop or workshops.

This theory was first put forward by London (1989a: 43–45) and was further explored by
Esse (1991: 108–9), who later rejected this mode of production in favor of the hypothesis that
the collared pithoi were produced on the household level within a matrilocal society spread by
exogamy (Esse 1992: 97–100). However, it must be recognized that ethnographic studies of
pithoi production have revealed that this is the most difficult vessel to form and demands the
skills of highly professional potters, and not a domestic household mode of production (e.g.,
London 1989b and references therein). In addition, ethnographic studies of exogamous mar-
riages demonstrate that well over 90% of all marriages take place within a 15 km distance
(Granquist 1931; Government of Palestine 1945; Adams and Kasakoff 1976; Lehmann in prep.),
thus not adequately explaining the dispersion and distribution of the collared pithos in Cis- and
Transjordan.

Ethnographic examples of itinerant potters have been described by Asboe (1946), Hampe
and Winter (1962), Linne (1965), Voyatzoglou (1974), and London (1989b), including
examples of itinerant potters who produce only pithoi. The relatively small number of highly
trained potters producing these pithoi may in part explain their standard size and generally
similar morphological features but also may reflect the diversity in detail, such as rim profile
and choice of materials, within an assemblage produced in the same region. The pithoi from
Giloh are a case in point. Production of collared pithoi by itinerant potters may also explain
the very close similarities between some Iron I assemblages, such as those from Tell el-ºUmeiri
in Transjordan and the sites from the Manasseh region in the hill country. The appearance of
similar incisions on rims or handles from sites in Transjordan and Cisjordan may also be an
indication of specific, perhaps itinerant, potters or workshops (see, e.g., Ibrahim 1978: 120;
1983; Finkelstein 1988: 278–80; Zertal 1986–87: 135, 145–47; Clark 1997: figs. B.13:1,
B.14:1, B.15:1).

FUNCTION, CONTEXT, AND ORIGIN

One of the keys to understanding the phenomenon of the rather sudden appearance of the
collared pithos in the archaeological record is the function, context, and origin of these jars.

oi.uchicago.edu





390



ANN E. KILLEBREW



Much of the interpretation regarding the function of the collared pithos is based on precon-
ceptions that this container was used mainly by subsistence-level villagers associated with the
early Israelites. Zertal (1988) has suggested that their popularity in the region of Manasseh
was due to their use as water containers. Finkelstein (1988: 282–83) rebuts this, based on the
appearance of collared jars in regions with abundant water. He states that these jars were used
as storage containers for other types of liquids in addition to water, including olive oil and
wine (see also Ibrahim 1978: 122 for ethnographic comparisons). In Transjordan, flotation of
the contents of collared pithoi at Tell el-ºUmeiri produced remains of lentils, chickpeas, wheat,
barley, and spices, leading the excavators to suggest that these jars were used as containers for
dry goods (Herr et al. 1991: 159).

However, it is clear from the context of a number of these vessels and from provenience
studies of jars, especially from Shiloh and Dan, that the collared pithos was also used as a
transport jar. The long distance transport of pithoi is well documented in the archaeological and
ethnographic record (Blitzer 1990). Recently, Artzy (1994: 137–38; unpublished manuscript)
presented convincing evidence that these jars could have been transported by camel over long
distances in Canaan at the end of the Late Bronze II period. It is noteworthy that these pithoi
first appeared in Cisjordan at late thirteenth century sites such as Aphek, Tel Nami, and Beit
Shean, which are located along what were Egyptian-controlled trade routes. Preliminary stud-
ies of several of these jars from Tel Nami have indicated a number of different wares for these
jars (Artzy, pers. comm.).

Collared pithoi also served as burial containers at a number of sites, such as Megiddo (Esse
1992: 88), Sahab (Ibrahim 1978: 122–23), Tel Nami (Artzy 1993; 1994: 127–28; 1995), Tel
Zeror (Ohata 1970: pl. 56), and Tell es-Saºidiyeh (Tubb 1995: 142–43; see also Druks 1966).
The multifunctional use of pithoi in the ethnographic record is well documented by London
(1989b) in her study of Cypriot potters.

Thus I conclude that the collared jar served a multitude of purposes. At sites along major
trade routes, where these jars appear in small numbers, or at sites where several clay sources,
including nonlocal, are indicated for the production of these jars, it is very likely that the col-
lared pithos was used as a transport vessel. At sites where large numbers of pithoi were found,
either at larger, sometimes fortified, settlements or at smaller villages, these vessels, often
locally produced, served as long-term storage containers for a variety of goods and liquids. At
several sites, pithoi, probably in secondary use, served as burial containers. A key factor in
understanding the function of these jars at the multitude of sites where they appear is their rel-
ative percentage in the entire ceramic assemblage.

The function, context, and morphological features provide clues regarding the origin of
this vessel type, which appears rather suddenly at the close of the LB. Various stylistic and
typological origins have been suggested for this jar, including a Canaanite origin. Kempinski
(1985: 401–2) and Finkelstein (1988: 283–84) have suggested that Middle Bronze Age stor-
age jars served as the source of inspiration for the collared pithoi. Wengrow (1996: 307) claims
that the collared pithoi was a larger version of the Canaanite storage jar. Kempinski (1992: 6)
proposes that these jars were produced in “Canaanite centres,” especially for use by Israelites
in the hill country. Artzy (1994; unpublished manuscript) points out that the ridge of the “col-

oi.uchicago.edu





391



THE COLLARED PITHOS IN CONTEXT



lared pithos,” as a ceramic technique, is known at other sites in the Aegean and eastern Med-
iterranean at the end of the LB and is not unique to Canaan.

The morphological features of the collared pithoi are very closely related to the LB
Canaanite storage jar, with the addition of the ridge at the base of the neck, which is doubtless
a technological feature of its production lacking any real stylistic significance. The earliest
appearance of the collared pithos in small numbers is at sites in Cisjordan located along LB
trade routes. Large numbers of this container seem to first appear at sites such as Tell el-
ºUmeiri on the Transjordanian plateau, perhaps indicating a Transjordanian origin for this ves-
sel type. It may have later spread to newly established villages in the central hill country of
Cisjordan, perhaps hinting at a close relationship between the peoples of the Transjordanian
plateau and the central hill country of Israel. The possible historical significance of this con-
nection has been recently explored by Herr (1998).

CONCLUSIONS

During the LB, the concept of a pithos container was mainly a northern tradition. The han-
dled collared pithos, similar in its typological features to the Canaanite storage jar but signif-
icantly larger in size, appears for the first time in small numbers in regions south of the Jezreel
Valley at the close of the LB at several sites located along trade routes. The initial appearance
of the collared pithos in Cisjordan at sites such as Tel Nami, Beit Shean, and Aphek was prob-
ably the result of trade relations. The presence of large numbers of collared pithoi, mainly in
Transjordan and in the hill country of Cisjordan, is doubtless a result of the significant socio-
economic changes that Canaan underwent at the close of the LB and Early Iron periods. The
production of the collared pithos is the result of a highly specialized potters’ craft. This jar
might have been manufactured at a number of dispersed workshops, but it is suggested here
that professional itinerant potters may have also produced this vessel. Itinerant potters may
explain the appearance of the collared pithos in very large numbers in villages of the central
hill country, where it is doubtful that in such small settlements there would be sufficiently
trained potters to produce these jars in large quantities. Additional provenience studies, espe-
cially petrographic analysis, may provide the additional information to test this hypothesis.

There are indications that this container, with its typological and stylistic roots in Canaan,
may have developed in Transjordan, though further excavations both in Transjordan and Cis-
jordan are necessary to clarify the chronological and typological development of this jar. It is
clear that the collared pithos served a multitude of purposes—a general storage container for
numerous commodities and substances, a transport jar, and, on occasion, a burial container.
The initial appearance in large quantities of the collared pithos at sites where this jar com-
prises some 20% or more of the assemblage in the highlands of Transjordan and Cisjordan
sometime at the end of the thirteenth or early twelfth century



b.c.



do, in my opinion, reflect
the significant socioeconomic changes that occurred in Canaan at the close of the LB. This is
indicated in the resulting well-defined regionality evident in the material culture of the Iron I
period.

oi.uchicago.edu





392



ANN E. KILLEBREW



REFERENCES



Adams, J. W., and Kasakoff, A. B.
1976 Factors Underlying Endogamous Group Size. Pp. 149–73 in



Regional Analysis



,







Vol. 2



: Social
Systems,



ed. C. A. Smith. New York: Academic Press.

Albright, W. F.
1924



Excavations and Results at Tell el-Ful (Gibeah at Saul).



Annual of the American Schools of Ori-
ental Research 4. New Haven: Yale University Press.

1932



The Excavation at Tell Beit Mirsim,



Vol. 1:



The Pottery of the First Three Campaigns.



Annual
of the American Schools of Oriental Research 12. Cambridge, MA: American Schools of Orien-
tal Research.

1934 The Kyle Memorial Excavation at Beth-el.



Bulletin of the American Schools of Oriental
Research



56: 2–15.
1940 Review of Megiddo II and Notes on the Megiddo Pottery of Strata VI–XX.



American Journal of
Archaeology



44: 546–50.

Arkin, Y.; Braun, M.; Buchbinder, B.; Diamant, E.; Itzhaki, Y.; Lasman, N.; Rot, I.; and Shachnai, E.
1976



Jerusalem and Vicinity, Geological Map 1:50,000.



Jerusalem: Geological Survey of Israel.

Artzy, M.
1993 Nami, Tel. Pp. 1095–98 in



The New Encyclopedia Of Archaeological Excavations in the Holy
Land,



ed. E. Stern. Jerusalem: Israel Exploration Society & Carta and New York: Simon &
Schuster.

1994 Incense, Camels and Collared-Rim Jars: Desert Trade Routes and Maritime Outlets in the Second
Millennium.



Oxford Journal of Archaeology



13: 121–47.
1995 Nami: A Second Millennium International Maritime Trading Center in the Mediterranean. Pp. 17–

40 in



Recent Excavations in Israel: A View to the West,



ed. S. Gitin. Archaeological Institute of
America Colloquia & Conference Papers No. 1. Dubuque, IA: Kendall/Hunt.

Unpublished Collared Rim Jars/Pithoi in the Late Bronze IIB Milieu.
manuscript

Asboe, W.
1946 Pottery in Ladakh, Western Tibet.



Man



46: 9–10.

Bahat, D.
1986 The 1975 Excavations at Sasa. Pp. 85–105 in



The Western Galilee Antiquities,



ed. M. Yedaya.
Haifa: Defense Ministry (Hebrew).

Balfet, H.
1965 Ethnographical Observations in North Africa and Archaeological Interpretation. Pp. 161–77 in



Ceramics and Man,



ed. F. R. Matson. Chicago: Aldine.

Beck, P., and Kochavi, M.
1985 A Dated Assemblage of the Late 13th Century



b.c.e.



from the Egyptian Residency at Aphek.



Tel
Aviv



12: 29–42.

Ben-Tor, A., and Portugali, Y.
1987



Tel Qiri: A Village in the Jezreel Valley: Report of the Archaeological Excavations 1975–1977.



Qedem 24. Jerusalem: Institute of Archaeology, Hebrew University of Jerusalem.

Bienkowski, P.
1992 The Beginning of the Iron Age in Edom: A Reply to Finkelstein.



Levant



24: 167–69.

Biran, A.
1989 The Collared-Rim Jars and Settlement of the Tribe of Dan. Pp. 71–96 in



Recent Excavations in
Israel: Studies in Iron Age Archaeology,



eds. S. Gitin and W. G. Dever. Annual of the American
Schools of Oriental Research 49. Winona Lake, IN: Eisenbrauns.

1994



Biblical Dan



. Jerusalem: Israel Exploration Society.

Blitzer, H.
1990 Storage-Jar Production and Trade in the Traditional Aegean.



Hesperia



59: 675–711.

oi.uchicago.edu





393



THE COLLARED PITHOS IN CONTEXT



Briend, J., and Humbert, J.-B.
1980



Tell Keisan (1971–1976)—une cité phénicienne en Galilée.



Orbis Biblicus Orientalis, Series
Archaeologica 1. Fribourg: Editions universitaires.

Buchbinder, B.
1969



Geological Map of HaShephela



. Jerusalem: Geological Survey of Israel.

Buhl, M.-L., and Holm-Nielsen, S.
1969



Shiloh: The Danish Excavations at Tell Sailun, Palestine in 1926, 1929, 1932, and 1963



. Copen-
hagen: National Museum of Denmark.

Callaway, J. A.
1969 The 1966 Ai (et-Tell) Excavations.



Bulletin of the American Schools of Oriental Research



196:
2–16.

1980



The Early Bronze Age Citadel and Lower City at Ai (Et-Tell): A Report of the Joint Archeologi-
cal Expedition to Ai (Et-Tell),



No. 2



.



Cambridge, MA: American Schools of Oriental Research.

Callaway, J. A., and Cooley, R. E.
1971 A Salvage Excavation at Raddana, in Bireh.



Bulletin of the American Schools of Oriental
Research



201: 9–19.

Clark, D. R.
1997 Field B: The Western Defense System.



Madaba Plains Project 3: The 1989 Season at Tell el-
ºUmeiri and Vicinity and Subsequent Studies,



eds. L. G. Herr, L. T. Geraty, Ø. S. LaBianca, and
R. W. Younker. Berrien Springs, MI: Andrews University Press.

Cohen-Weinberger, A., and Goren, Y.
1996 Petrographic Analysis of Iron Age I Pithoi from Tel Sasa.



ºAtiqot 28: 77–83.

Costin, C. L.
1991 Craft Specialization: Issues in Defining, Documenting, and Explaining the Organization of Pro-

duction. Pp. 1–55 in Archaeological Method and Theory, vol. 3, ed. M. B. Schiffer. Tucson: Uni-
versity of Arizona Press.

Courtois, L.
1980 Examen Petrographique et Caractérisation Materielle de quelques Céramiques de Tel Keisan.

Tell Keisan (1971–1976)—une cité phénicienne en Galilée, eds. J. Briend and J.-B. Humbert.
Orbis Biblicus Orientalis, Series Archaeologica 1. Paris: J. Gabalda.

Dothan, M.
1955 Excavations at ºAfula. ºAtiqot 1 (English series): 19–70.

Druks, A.
1966 A “Hittite” Burial near Kefar Yehoshua. Yediot Bulletin of the Israel Exploration Society 30:

213–20 (Hebrew).

Edelstein, G., and Milevski, I.
1994 The Rural Settlement of Jerusalem Re-evaluated: Surveys and Excavations in the Raphªaim

Valley and Mevassert Yerushalayim. Palestine Exploration Quarterly 126: 2–23.

Esse, D.
1991 The Collared Store Jar—Scholarly Ideology and Ceramic Typology. Scandinavian Journal of

Old Testament Studies 2: 99–116.
1992 The Collared Pithos at Megiddo: Ceramic Distribution and Ethnicity Journal of Near Eastern

Studies 51: 81–103.

Finkelstein, I.
1986 ºIzbet Íar†ah, An Early Iron Age Site Near Rosh Haºayin, Israel. British Archaeological

Reports, International Series 299. Oxford: British Archaeological Reports.
1988 The Archaeology of the Israelite Settlement. Jerusalem: Israel Exploration Society.
1992a Edom in the Iron I. Levant 24: 159–65.
1992b Stratigraphy, Pottery and Parallels: A Reply to Bienkowski. Levant 24: 171–72.

Finkelstein, I.; Bunimovitz, S.; and Lederman, Z.
1993 Shiloh: The Archaeology of a Biblical Site, ed. I. Finkelstein. Monograph Series of the Institute

of Archaeology, Tel Aviv University. Tel Aviv: Institute of Archaeology, Tel Aviv University.

oi.uchicago.edu



394 ANN E. KILLEBREW

Franken, H. J.
1990 The Wares of Manufacture of Pottery. Pp. 77–98 in Excavations in Jerusalem 1961–1967, Vol. 2:

The Iron Age Extramural Quarter on the South-East Hill, eds. H. J. Franken and M. L. Steiner.
Oxford: Oxford University Press.

Franken, H. J.
1992 Excavations at Tell Deir ºAlla. The Late Bronze Age Sanctuary. Louvain: Peeters Press.

Franken, H. J., and Kalsbeek, J.
1969 Excavations at Tell Deir ºAlla, Vol. 1. Leiden: E. J. Brill.

Funk, R. W.
1968 The Bronze Age—Iron I Pottery. Pp. 35–53 in The Excavations at Beth Zur, eds. O. R. Sellers,

R. W. Funk, J. L. McKenzie, and P., and N. Lapp. Annual of the American Schools of Oriental
Research 38. Cambridge, MA: American Schools of Oriental Research.

Gilead, I., and Goren, Y.
1989 Petrographic Analyses of Fourth Millennium b.c. Pottery and Stone Vessels from the Northern

Negev, Israel. Bulletin of the American Schools of Oriental Research 275: 5–14.

Gill, D.
1996 The Geology of the City of David and Its Ancient Subterranean Waterworks. Pp. 1–28 in Exca-

vations at the City of David 1978–1985 Directed by Yigal Shiloh, Vol. 4: Various Reports,
eds. D. T. Ariel and A. De Groot. Qedem 35. Jerusalem: Hebrew University of Jerusalem.

Glass, J.; Goren, Y.; Bunimovitz, S.; and Finkelstein, I.
1993 Petrographic Analyses of Middle Bronze Age III, Late Bronze Age and Iron Age I Pottery Assem-

blages. Pp. 271–86 in Shilo: The Archaeology of a Biblical Site, eds. I. Finkelstein, S. Bunimo-
vitz, and Z. Lederman. Tel Aviv: Institute of Archaeology, Tel Aviv University.

Golani, A., and Yogev, O.
1996 The 1980 Excavations at Tel Sasa. ºAtiqot 28: 41–58.

Goren, Y.
1995 Shrines and Ceramics in Chalcolithic Israel: The View through the Petrographic Microscope.

Archaeometry 37: 287–305.
1996 The Southern Levant in the Early Bronze Age IV: The Petrographic Perspective. Bulletin of the

American Schools of Oriental Research 303: 33–72.

Government of Palestine
1945 Survey of Social and Economic Conditions in Arab Villages, 1944. General Monthly Bulletin of

Current Statistics, July 1945: 426–47.

Granquist, H.
1931 Marriage Conditions in a Palestinian Village. Helsingfors: Societas Scientiarum Fennica.

Grant, E., and Wright, G. E.
1939 Ain Shems Excavations (Palestine): Part V (Text). Biblical and Kindred Studies 8. Haverford,

PA: Haverford College.

Greenberg, R., and Porat, N.
1996 A Third Millennium Levantine Pottery Production Center: Typology, Petrography, and Prove-

nance of the Metallic Ware of Northern Israel and Adjacent Regions. Bulletin of the American
Schools of Oriental Research 301: 5–24.

Hampe, R., and Winter, A.
1962 Bei Töpfern und Töpferinnen in Kreta, Messenien and Zypern. Mainz: Romische-Germanischen

Zentralmuseums.

Herr, L.
1991 Pottery Typology and Chronology. Pp. 232–45 in Madaba Plains Project: The 1987 Season at

Tell el-ºUmeiri and Vicinity and Subsequent Studies, eds. L. G. Herr, L. T. Geraty, Ø. S. LaBianca,
and R. W. Younker. Berrien Springs, MI: Andrews University Press.

1997 The Pottery. Pp. 228–49 in Madaba Plains Project: The 1989 Season at Tell el-ºUmeiri and
Vicinity and Subsequent Studies, eds. L. G. Herr, L. T. Geraty, Ø. S. LaBianca, and R. W. Younger.
Berrien Springs, MI: Andrews University Press.

long

oi.uchicago.edu



395THE COLLARED PITHOS IN CONTEXT

1998 Tell el-ºUmayri and the Madaba Plains Region During the Late Bronze–Iron Age I Transition.
Pp. 251–64 in Mediterranean Peoples in Transition Thirteenth to Early Tenth Centuries b.c.,
eds. S. Gitin, A. Mazar, and E. Stern. Jerusalem: Israel Exploration Society.

Herr, L. G.; Geraty, L. T.; LaBianca, Ø. S.; and Younker, R. W.
1991 Madaba Plains Project: The 1989 Excavations at Tell el-‘Umeiri and Vicinity. Annual of the

Department of Antiquities of Jordan 35: 155–80.

Hunt, M.
1987 The Tell Qiri Pottery. Pp. 139–223 in Tell Qiri: A Village in the Jezreel Valley Report of the

Archaeological Excavations 1975–1977, eds. A. Ben-Tor and Y. Portugali. Qedem 24. Jerusalem:
Institute of Archaeology, Hebrew University of Jerusalem.

Ibrahim, M. M.
1978 The Collared-Rim Jar of the Early Iron Age. Pp. 117–126 in Archaeology in the Levant: Essays

for Kathleen Kenyon, eds. R. Moorey and P. Parr. Warminster: Aris & Phillips.
1983 Siegel und Siegelabdrücke aus Sahab. Zeitschrift des Deutschen Palästina-Vereins 99: 43–53.

James, F. W., and McGovern, P. E.
1993 The Late Bronze Egyptian Garrison at Beit Shean: A Study of Levels VII and VIII. Philadelphia:

University Museum, University of Pennsylvania Press.

Ji, C.-H. C.
1995 Iron Age I in Central and Northern Transjordan: An Interim Summary of the Archaeological

Data. Palestine Exploration Quarterly 127: 122–40.

Kelso, J. L.
1968 The Excavation of Bethel (1934–1960). Annual of the American Schools of Oriental Research

39. Cambridge, MA: American Schools of Oriental Research.

Kempinski, A.
1985 The Overlap of Cultures at the End of the Late Bronze Age and the Beginning of the Iron Age.

Eretz-Israel (Nahman Avigad Volume) 18: 399–407 (Hebrew).
1992 How Profoundly Canaanized were the Early Israelites? Zeitschrift des Deutschen Palästina-

Vereins 108: 1–7.

Killebrew, A. E.
1998 Ceramic Craft and Technology during the Late Bronze and Early Iron Ages: The Relationship

Between Pottery Technology, Style, and Cultural Diversity. Ph.D. dissertation, Hebrew Univer-
sity of Jerusalem.

1999 Late Bronze and Iron I Cooking Pots in Canaan: A Typological, Technological, and Functional
Study. Pp. 83–126 in Archaeology, History and Culture in Palestine and the Near East: Essays
in Memory of Albert E. Glock, ed. T. Kapitan. Atlanta: Scholars Press.

Kjaer, H.
1930 The Excavation of Shiloh 1929. Journal of the Palestine Oriental Society 10: 87–174.

Kochavi, M.
1974 Khirbet Rabud = Debir. Tel Aviv 1: 2–33.

Lehmann, G.
In Reconstructing the Social Landscape of Early Israel Marriage Alliances and Lasting Coalitions
preparation in a Rural Context.

Linne, S.
1965 The Ethnologist and the American Indian Pottery. Pp. 20–42 in Ceramics and Man, ed. F. R.

Matson. Chicago: Aldine.
London, G.

1989a A Comparison of Two Contemporaneous Lifestyles of the Late Second Millennium b.c. Bulletin
of the American Schools of Oriental Research 273: 37–55.

1989b On Fig Leaves, Itinerant Potters, and Pottery Production Locations in Cyprus. Pp. 65–80 in
Cross-craft and Cross-cultural Interactions in Ceramics, eds. P. E. McGovern, M. D. Notis, and
W. D. Kingery. Ceramics and Civilization 4. Westerville, OH: American Ceramic Society.

oi.uchicago.edu



396 ANN E. KILLEBREW

Loud, G.
1948 Megiddo II: Seasons of 1935–39. Oriental Institute Publications 62. Chicago: University of

Chicago Press.
Mazar, A.

1981 Giloh: An Early Israelite Settlement Site near Jerusalem. Israel Exploration Journal 31: 1–36.
1985 Excavations at Tell Qasile Part 2: The Philistine Sanctuary: Various Finds, the Pottery, Conclu-

sions, Appendixes. Qedem 20. Jerusalem: Hebrew University of Jerusalem.
1990 Iron Age I and II Tower at Giloh and the Israelite Settlement. Israel Exploration Journal 40:

77–101.
1993a Beit-Shean, Tel Beit-Shean and the Northern Cemetery. Pp. 214–23 in The New Encyclopedia

of Archaeological Excavations in the Holy Land, vol. 1, ed. E. Stern. Jerusalem: Israel Explo-
ration Society & Carta and New York: Simon & Schuster.

1993b Beit Shean in the Iron Age: Preliminary Report and Conclusions of the 1990–1991 Excavations.
Israel Exploration Journal 43: 201–29.

1993c Giloh. Pp. 519–20 in The New Encyclopedia of Archaeological Excavations in the Holy Land,
vol. 2, ed. E. Stern. Jerusalem: Israel Exploration Society & Carta and New York: Simon &
Schuster.

1994a Four Thousand Years of History at Tel Beit-Shean. Qadmoniot 27 (107–8): 67–83 (Hebrew).
1994b Jerusalem and Its Vinicity in Iron Age I. Pp. 70–91 in From Nomadism to Monarchy: Archaeo-

logical and Historical Aspects of Early Israel, eds. I. Finkelstein and N. Naºaman. Jerusalem:
Israel Exploration Society.

1997a Four Thousand Years of History at Tel Beit Shean: An Account of the Renewed Excavations.
Biblical Archaeologist 60: 62–76.

1997b The Excavations at Tel Beit Shean during the Years 1989–94. Pp. 144–64 in The Archaeology
of Israel Constructing the Past, Interpreting the Present, eds. N. A. Silberman and D. Small.
Journal for the Study of the Old Testament Supplement Series 237. Sheffield: Sheffield Aca-
demic Press.

Nir, D.
1989 Beit-Shean Valley: The Region and Its Challenges on the Fringe of the Desert. Tel Aviv: HaKib-

butz HaMeuchad (Hebrew).

Ohata, K.
1970 Tel Zeror III Report of the Excavation Third Season 1966. Tokyo: Society for Near Eastern Stud-

ies in Japan.

Peacock, D. P. S.
1981 Archaeology, Ethnology and Ceramic Production. Pp. 187–94 in Production and Distribution:

A Ceramic Viewpoint, eds. H. Howard and E. Morris. British Archaeological Reports, Interna-
tional Series 120. Oxford: British Archaeological Reports.

1982 Pottery in the Roman World: An Ethnoarchaeological Approach. London: Longman.

Pool, C. A.
1992 Integrating Ceramic Production and Distribution. Pp. 275–341 in Ceramic Production and Dis-

tribution: An Integrated Approach, eds. G. J. Bey III and C. A. Pool. Oxford: Westview.

Porat, N.
1989a Composition of Pottery—Application to the Study of the Interrrelations Between Canaan and

Egypt during the 3rd Millennium b.c. Ph.D. dissertation, Hebrew University of Jerusalem.
1989b Petrography of Pottery from Southern Israel and Sinai. Pp. 169–88 in L’urbanisation, ed.

P. de Miroschedji. British Archaeological Reports, International Series 527, pt. 1. Oxford: Brit-
ish Archaeological Reports.

Rast, W. E.
1978 Taanach I: Studies in the Iron Age Pottery. Cambridge, MA: American Schools of Oriental

Research.

short

oi.uchicago.edu



397THE COLLARED PITHOS IN CONTEXT

Redman, C. L., and Myers, J. E.
1981 Interpretation, Classification and Ceramic Production: A Medieval North African Case Study.

Pp. 285–307 in Production and Distribution: A Ceramic Viewpoint), eds. H. Howard and
E. Morris. British Archaeological Reports, International Series 120. Oxford: British Archaeo-
logical Reports.

Rice, M.
1981 Evolution of Specialized Pottery Production: A Trial Model. Current Anthropology 22: 219–27.
1987 Pottery Analysis. Chicago: University of Chicago Press.

Santley, R. S.; Arnold, J., III; and Pool, C. A.
1989 The Ceramic Production System at Matacapan, Veracruz, Mexico. Journal of Field Archaeology

16: 107–32.

Schumacher, G.
1908 Tell el-Mutesellim I. Band B. Tafeln. Leipzig: Kommission bei Rudolf Haupt.

Shaliv, G.; Mimram, Y.; and Hatzor, Y.
1991 The Sedimentary and Structural History of the Bet Sheªan Area and Its Regional Implications.

Israel Journal of Earth Sciences 40: 161–79.

Sinclair, L. A.
1960 An Archaeological Study of Gibeah (Tell el-Fûl). Annual of the Amercian Schools of Oriental

Research 34–35: 5–52. New Haven: American Schools of Oriental Research.

Steiner, M. L.
1994 Re-dating the Terraces of Jerusalem. Israel Exploration Journal 44: 13–20.

Stepansky, Y.; Segan, D.; and Carmi, I.
1996 The 1993 Sounding at Tel Sasa: Excavation Report and Radiometric Dating. ºAtiqot 28: 63–76.

Stern, E.
1978 Excavations at Tel Mevorakh (1973–1976), Part 1: From the Iron Age to the Roman Period.

Jerusalem: Institute of Archaeology, Hebrew University of Jerusalem.

Tosi, M.
1984 The Notion of Craft Specialization and Its Representation in the Archaeological Record of Early

States in Turanian Basin. Pp. 22–52 in Marxist Perspectives in Archaeology, ed. M. Spriggs.
Cambridge: Cambridge University Press.

Tubb, J.
1995 An Aegean Presence in Egypto-Canaan. Pp. 136–45 in Egypt, the Aegean and the Levant Inter-

connections in the Second Millennium b.c., eds. W. V. Davies and L. Schofield. London: British
Museum.

Tubb, J. N.; Dorrell, P. G.; and Cobbing, F. J.
1996 Interim Report on the Eighth (1995) Season of Excavations at Tell es-Saºidiyeh. Palestine

Exploration Quarterly 128: 16–36.
van der Leeuw, S. E.

1976 Studies in the Technology of Ancient Pottery. Amsterdam: Amsterdam University Press.
1984 Pottery Manufacture: Some Complications for the Study of Trade. Pp. 55–69 in Pots and Pot-

ters: Current Approaches in Ceramic Analysis, ed. P. M. Rice. University of California Los
Angeles Institute of Archaeology Monograph 24. Los Angeles: University of California, Insti-
tute of Archaeology.

Voyatzoglou, M.
1974 The Jar Makers of Thrapsano in Crete. Expedition 16: 18–24.

Wampler, J. C.
1947 Tell en-Nasbeh II The Pottery. Berkeley: Palestine Institute of Pacific School of Religion and

American Schools of Oriental Research.

oi.uchicago.edu



398 ANN E. KILLEBREW

Weippert, M.
1982 Remarks on the History of Settlement in Southern Jordan during the Early Iron Age. Pp. 153–

62 in Studies in the History and Archaeology of Jordan I, ed. A. Hadidi. Amman: Department
of Antiquities, Amman.

Wengrow, D.
1996 Egyptian Taskmasters and Heavy Burdens: Highland Exploitation and the Collared-Rim Pithos

of the Bronze/Iron Age Levant. Oxford Journal of Archaeology 15: 307–26.

Wolff, S.
1998 An Iron Age I Site at ºEn Óagit (Northern Ramat Menashe). Pp. 449–54 in Mediterranean Peo-

ples in Transition Thirteenth to Early Tenth Centuries b.c., eds. S. Gitin, A. Mazar, and E. Stern.
Jerusalem: Israel Exploration Society.

Yadin, Y.; Aharoni, Y.; Amiran, R.; Dothan, T.; Dunayevsky, I.; and Perrot, J.
1960 Hazor II: An Account of the Second Season of Excavations, 1956. Jerusalem: Magness.

Yadin, Y., Aharoni, Y.; Amiran, R.; Dothan, T.; Dothan, M.; Dunayevsky, I.; and Perrot, J.
1961 Hazor III–IV: An Account of the Third and Fourth Seasons of Excavation, 1957–1958. Plates.

Jerusalem: Magness.

Yadin, Y.; Aharoni, Y.; Amiran, R.; Ben-Tor, A.; Dothan, M.; Dothan, T.; Dunayevsky, I.; Geva, S.; and Stern, E.
1989 Hazor III–IV: An Account of the Third and Fourth Seasons of Excavation, 1957–1958. Text.

Jerusalem: Israel Exploration Society.

Yellin, J., and Gunneweg, J.
1985 Provenience of Pottery from Tell Qasile. Strata VII, X, XI and XII. Pp. 111–17 in Excavation at

Tell Qasile, ed. A. Mazar. Qedem 20. Jerusalem: Institute of Archaeology, Hebrew University of
Jerusalem.

1989 Instrumental Neutron Activation Analysis and the Origin of Iron Age I Collared-Rim Jars and
Pithoi from Tel Dan. Annual of the American Schools of Oriental Research 49: 133–41.

Yogev, O.
1993 Nahariya. Pp.1088–90 in The New Encyclopedia of Archaeological Excavations in the Holy

Land, vol. 3, ed. E. Stern. Jerusalem: Israel Exploration Society & Carta and New York: Simon
& Schuster.

Zertal, A.
1986–87 An Early Iron Age Cultic Site on Mount Ebal: Excavation Seasons 1982–1987 Preliminary

Report. Tel Aviv 13–14: 105–65.
1988 The Water Factor during the Israelite Settlement Process in Canaan. Pp. 341–52 in Society and

Economy in the Eastern Mediterranean (c. 1500–1000 b.c.), eds. M. Heltzer and E. Lipinski.
Leuven: Uitgeverij Peeters.

1991 Israel Enters Canaan—Following the Pottery Trail. Biblical Archaeological Review 17: 28–
49, 75.

oi.uchicago.edu



399



CHIPS AND SHELLS: FLOTATION
AT TEL YAQUSH



Egon H. E. Lass



This paper is written in memory of Douglas L. Esse and our last ecstatic season at Tel
Yaqush. Though he was suffering from his terrible disease, he was always extremely positive
and grateful for everything. Now that he is gone, life can never be the same again.

METHODOLOGY

Flotation is a process of water screening by which both the light and heavy fractions of an
unsifted, unsorted soil sample from an archaeological stratum are recovered. The light fraction
is whatever floats to the surface when the soil sample is poured into water, consisting mostly
of charred botanical remains. The heavy fraction is whatever sinks in water and is caught by
a screen, allowing the silty component to escape.

Many elaborate flotation devices have been developed in areas where water is abundant
(Davis and Wesolowsky 1975; Diamant 1979; French 1971; Limp 1974). The method for arid
countries, used at Tel Yaqush, was developed by Robert Stewart at Tell el-Hesi (Stewart and
Robertson 1973), also practiced at Bâb edh-Dhr⺠(Richardson and McCreery 1978) and Ash-
kelon (Lass 1994). A round barrel is filled with water, and a round tub, having a 1.5 mm win-
dow screen for a bottom, is immersed into the barrel. The unprocessed soil sample is weighed
and then poured into the tub, and the light fraction is skimmed off the top of the water with a
0.5 mm strainer. The silt is then shaken through the bottom of the tub, and the resulting heavy
fraction is laid out to dry in the sun. The light fraction is taken indoors and laid out on news-
papers for a slower drying process. The barrel has to be cleaned out after the processing of
approximately 150 kg of soil (Lass 1994: 24).

At Tel Yaqush, Israel, an Early Bronze Age (EB) site south of the Sea of Galilee (fig. 21.1;
see Esse 1990: 222–23; 1993: 1502–04), a barrel was supplied by Kibbutz Gesher, but the
round tub to be immersed in water was lacking. In its stead an elongated shopping bag made
of wide open plastic webbing and lined with 1.5 mm window screen was used. Because the
soil at Yaqush was very dry and full of oxygen, a constant problem was the formation of foam
which clogged the light fraction and had to be washed away.

Subsequently the dried heavy and light fractions were both weighed and became percent-
ages of the original weight of the unprocessed sample. The heavy fraction was manually sorted
into various categories that were counted and weighed, also becoming percentages or ratios of
the original gross weight. The resulting values were computerized in a dBase IV data file cre-
ated for the flotation project. The entire process was accomplished by one person.

When the material culture content of the EB loci of Tel Yaqush is compared with the vari-
ety found in the later strata of Ashkelon (Lass 1994), it becomes obvious that the cultural
debris in the earlier strata is focused and scanty. In Ashkelon the categories included bone, fish

long



21



oi.uchicago.edu





400



EGON H. E. LASS



scales, shell, pottery, botanical remains, chert, metal, slag, eggshell, plaster, glass, pumice,
bitumen, and red ocher. At Tel Yaqush pottery, bone, shell, chert, and botanical remains were
the main categories. Red ocher was also found, an occasional eggshell, and a rare copper frag-
ment. The botanical remains were too small to be weighed accurately, and the tiny bone frag-
ments showed the heavy attrition of time. Since pottery was collected during excavation even
from the layers that were to be sampled for flotation, a heavy bias was introduced, and pottery
was not used for statistical analysis. The material was manually sorted under low magnifica-
tion to insure completeness of recovery, including tiny shell and flint fragments measuring
about 1 mm in diameter.

Not counting potsherds, which had been biased, botanical remains, which were too insig-
nificant in weight to lend themselves to analysis, or chert fragments larger than 2 cm, which
were excluded for reasons discussed below, from 258 soil samples weighing 1,750.5 kg, a
total of 192,484 items were recovered, including 6,289 bone fragments, 62,822 shell frag-
ments, and 123,373 chert chips.

It is axiomatic that societies leave behind the traces of patterned behavior that can be dis-
covered in archaeological remains (Clarke 1977: 18; Butzer 1980: 419; Wood 1978: 260; Sal-
lade 1989: 410–15; South 1978), and that smaller items are more likely to become the
remaining primary refuse in activity areas (Schiffer 1983: 679; 1972: 161; Baker 1978: 291;
Bradley and Fulford 1980: 85; Rosen 1986: 114; Stevenson 1985: 67). The purpose of this flo-
tation project was to discover any patterns that might have remained in the recovered micro-
artifacts. It did not seek to be a complement to macroartifactual finds, but to present unique
and separate information in its own right (see Dunnell and Stein 1989: 39).



FIGURE 21.1. Map Showing Location of Tel Yaqush.



long

oi.uchicago.edu

[image: image121.tif]



401



CHIPS AND SHELLS: FLOTATION AT TEL YAQUSH



Different researchers give different definitions as to what constitutes a microartifact. For
Fladmark (1982: 205), Hull (1987: 772), and Vance (1987: 58), a microartifact is smaller at its
maximum circumference than 1.0 mm; for Stein and Teltser (1989: 6) and Dunnell and Stein
(1989: 35), it is smaller than 2.0 mm; for Simms and Heath it is smaller than 4.75 mm (1990:
805) and smaller than 6.4 mm (1990: 800); for Rosen it is smaller than 20.0 mm (1993: 141)
and smaller than 30.0 mm (1989: 565). The present study arbitrarily agrees with Rosen (1993:
141) that a microartifact is smaller than 20.0 mm. The lower limit must necessarily be 1.5 mm,
since that was the screen mesh used for flotation, although some smaller artifacts were always
recovered that did not pass through the mesh.

Of the recovered artifactual categories, the largest by far was microdebitage. The tiny
chips were ubiquitous throughout the site, and there was some doubt as to whether they were
cultural debris.

Four control samples from a depth of 35 cm under the surface were taken in locations
beyond the limits of the site, averaging about 7 kg in weight (cf. Hull 1987: 775). From a hill
to the northwest, the sample yielded eight chert chips and eighteen shell fragments; from a hill
to the west, one chert chip and eighteen shell fragments; from a hill to the south (south of the
Wadi Kuraiyim, which drains into the Jordan River and runs parallel to the larger Na



˙



al Tabor;
Esse 1993: 1502), no chert chips and fourteen shell fragments; and from the north bank of the
Wadi Kuraiyim close to the southern border of the site, twenty-three chert chips and 809 shell
fragments. The mean number of chert chips found in a sample from the site was 486, and the
mean number of shell fragments 247. It is seen from these statistics that if there had been a com-
ponent of naturally created chert chips in the layers of the site, their number would have been
negligible in comparison with cultural material. It is of some significance that the sample from
the bank of the Wadi Kuraiyim yielded more than three times the number of shell fragments
than the mean number found in a site sample (see below).

KNAPPING EXPERIMENTS

Yet another effort to prove that the chips were debitage was an experiment in flint knap-
ping. It was first assumed that most of the chips were from the process of retouching. In that
case they had to be distinguished from natural geological specimens by the following morpho-
logical characteristics: prominent bulbs of percussion under faceted platforms; ripple lines and
radial lines that radiate out from the point of impact or pressure (fig. 21.2:B, D–G); and bul-
bar scars located on the bulb of percussion (fig. 21.2:C, G), although these were thought to be
associated only with percussion flaking (Peacock 1991: 350–52). If such features could be
produced empirically, they could be compared with what had been found in the field.

One of the samples contained a stone of flint about 25 by 15 cm in size. Such stones do not
occur naturally at the site of Tel Yaqush and were most probably imported in either from the
Yarmuk River or from outcrops of the “Hordus” Formation to the west of the site which con-
tains, among other things, cobbles of flint (Yaakov Nir, pers. comm. 1994). The stone was
coarse grained and had fibrous and blocky fractures, for which reason it had probably been
rejected as raw material for tool making (Fladmark 1982: 209). Direct percussion was used
with brutal force to obtain a substantial number of edges that could be retouched with a
pointed bone fragment. It was seen that direct percussion for the production of macroflakes
yielded a multitude of microchips. Even the process of retouching with a fingernail resulted

oi.uchicago.edu





402



EGON H. E. LASS



not only in flakes bearing one or more of the required morphological characteristics, but also
a substantial amount of amorphous microdebitage (Stahle and Dunn 1982: 86). During the
retouching process two kinds of microflakes were produced. The initial pressure of bone
against chert produced a microflake that had a bulb on the ventral side and one or more of the
other attendant features, whereas the dorsal side remained flat, being part of the original
opposing surface against which pressure was exerted (fig. 21.2:A, F). The tool maker did not
stop here, but followed through with more insistent pressure, creating a second, larger intru-
sive microflake that also had a bulb and various features on the ventral side, and the negative
of the preceding microflake on the dorsal side, producing a characteristically curved profile
when viewed proximally from above the platform (fig. 21.2:G).

When the experimental flakes had been examined, they were compared with the material
extracted in flotation. Except for a few samples from small whole vessels in which not a single
chip was found, all of the samples that contained chips included microflakes that had a bulb at
the very least, and often one or more of the other required characteristics (fig. 21.2:B–E). These



FIGURE 21.2. Tel Yaqush Microflakes. Ventral side in each figure is on the left, dorsal on the right.



B, C,







D,



and



E



are artifacts from ancient strata;



A, F,



and



G



are the result of modern experimentation.



long

oi.uchicago.edu

[image: image122.tif]



403



CHIPS AND SHELLS: FLOTATION AT TEL YAQUSH



characteristic traits were better and more often represented in the Tel Yaqush finds than in the
experimental material, probably because the stone that was used to produce the microflakes
had been a reject for being too impure and too soft.

ANALYSIS

A grid was imposed on Tel Yaqush in which every 10 m square was named by letter and
number. Within the 10 m square was a fine grid of 1 m squares, so that square, fine grid, and
locus number were enough for object placement within 1 m



2



anywhere on the tell. Flotation
samples were taken from fine gridded floors and from any other locus that showed promise for
yielding microartifacts. Owing to the incomplete state of present research at the site of Tel
Yaqush, this paper cannot address problems of stratigraphy or architecture. The statistical pro-
files extracted from floors are informative regardless of their stratigraphic position or age. It
will be seen that the drastic differences of these profiles cannot be but an indication of differ-
ent kinds of activities.

Figure 21.3 shows what was found across a substantial area located at the top of the tell:
two buildings connected by a broad wall, and the courtyard between them. Dated to EB I, this



FIGURE 21.3. Tel Yaqush Square H5. Walls of two EB I houses connected by a broad wall have been darkened.
In each fine grid, the first column is the ratio of flint chip number to gross weight of sample; the second is the
percentage of microflakes among the debitage; the third is weight percentage of shell; and the fourth is weight
percentage of bone. A broad spectrum of activities is indicated.

oi.uchicago.edu

[image: image123.tif]



404



EGON H. E. LASS



area went through a violent, fiery destruction and was therefore extremely rich in macroarti-
facts (Esse 1993: 1502–3). Comparatively, the number of microartifacts was average. Each
fine grid has its own separate histogram, and though the scale is uniform, the function of the
columns differ. The material was presented in this fashion in order to compare several very
different kinds of remains (Rosen 1993: 142). The first column is a ratio of flint chip number
against gross weight of sample. The second is the percentage of retouching flakes with char-
acteristic morphological elements that were sorted out from the undifferentiated microdeb-
itage. Every one of these flakes positively represents an act of retouching. The third and
fourth columns are the weight percentages of shell and bone respectively. The two columns
representing microdebitage have been blackened to facilitate pattern recognition. It will be
seen that the relationship between the two columns is expressed in a true or reversed L shape.
This is not a direct relationship since one is a ratio and the other is a percentage, but it is useful
nevertheless. In the western building, the majority of L’s are reversed. It means that here there
were relatively few microchips of which a high percentage came from retouching. Chances are
that in this area the focus was more on tool use than tool production. In the yard and building
to the east, the L’s are true, meaning that both areas produced relatively more of the amor-
phous microdebitage. Resharpening remained more or less the same as in the eastern building,
except for a pronounced resharpening station at the north end of the eastern building (fine
grids 31 and 42) where the L’s are reversed. In these areas there might have been a broader
range of activities, including tool production and use (see Wood 1978: 260, who states that an
evenness in the histograms of artifact frequencies denotes an equal weighting of activities).
Twenty-one whole sickle blades were found within the borders of the eastern building alone.

In the areas of Square J6 (fig. 21.4), two superimposed floors are shown in which the L
shapes are either evenly matched or a mix of true and reversed. A broad spectrum of activities



FIGURE 21.4. Tel Yaqush Square J6. Two EB I floors are shown; the left was stratigraphically later and super-
imposed upon the right. A broad spectrum of activities is indicated.



long

oi.uchicago.edu

[image: image124.tif]



405



CHIPS AND SHELLS: FLOTATION AT TEL YAQUSH



is indicated, including tool production and tool use, lasting through a substantial period in the
EB I, which may be evidence for long-term occupation without change of activity.

This is not the case in the areas of Square K9 where the earlier loci (fig. 21.5) suggest a
broad spectrum of activities, whereas the later loci (fig. 21.6), with their relatively high
amounts of amorphous microdebitage, point to a more focused attention to the production of
tools, an indication of change through time in the EB III. This specialized activity during the



FIGURE 21.5. Tel Yaqush Square K9. An EB III floor that is stratigraphically earlier than the one shown in fig-
ure 21.6. A broad spectrum of activities is indicated.

FIGURE 21.6. Tel Yaqush Square K9. An EB III floor that is stratigraphically later and superimposed upon the
floor in figure 21.5. Specialized activity is indicated.

oi.uchicago.edu

[image: image125.tif][image: image126.tif]



406



EGON H. E. LASS



same period is even more pronounced in Square R13 (fig. 21.7) where nondescript micro-
debitage is most voluminous, and evidence for retouching at its lowest (Wood [1978: 260]
states that the greater the disparity among the different artifact frequencies, the smaller the
evenness, the more limited the activities).

The loci of Square Q14, dated to the EB II period (fig. 21.8), are areas where almost no
activities were represented in the microartifactual record. Located at the bottom of the tell slope,
the meager finds of Square Q14 are also proof against the claim that cultural material has moved
downslope through geological agents and become secondary refuse (Butzer 1982: 103; Schiffer
1972: 161).

Figure 21.9 is a summary of trends for all the loci that have been discussed. Except for
Square Q14, one might think that the farther one moves downslope, the greater the amount of
microdebitage as expressed in the histogram for debitage ratio. It is highly unlikely that geo-
logical factors had anything to do with this distribution. Most of the artifacts were surrounded
by walls that would have prevented their migration, and almost all of them were still sharp-
edged, meaning that they had not moved far from their original place of deposit. If there had



FIGURE 21.7. Tel Yaqush Square R13. An EB III floor on which specialized activity is indicated.

FIGURE 21.8. Tel Yaqush Square Q14. An EB II floor on which practically no activities are indicated.

oi.uchicago.edu

[image: image127.tif][image: image128.tif]



407



CHIPS AND SHELLS: FLOTATION AT TEL YAQUSH



been a downslope movement of microartifacts, the bottommost locus, Square Q14, remained
singularly unaffected by it as, indeed, did the slope of the Wadi Kuraiyim.

As can be seen in figure 21.9, there are no sharp differences in the number of retouching
microflakes, although where undifferentiated debitage is highest, microflakes are lowest in
number, indicating a focus on tool production. An enigmatic and fascinating relationship is
the almost perfect fit between debitage ratio and shell weight percentage. Most of the shells
came from snails. If geological factors must be ruled out for flint distribution, the same holds
true for shell fragments. The histograms show that the amount of microdebitage determined
the amount of shell within a certain locus. The reason behind this correlation can only be
guessed at. The idea that something connected to the production and use of tools might have
attracted the snails is not very convincing. Possibly the snails adhered to reeds that were carried
in from the Wadi Kuraiyim. It will be remembered that the sample taken from its bank con-
tained more than three times the number of shell fragments than the mean number found in a
sample from the site. The question passes from one enigma to another: why would the number
of reeds carried in from the Wadi Kuraiyim correlate with the production and use of tools?



FIGURE 21.9. Summation of trends in the artifactual record at Tel Yaqush.

oi.uchicago.edu

[image: image129.tif]



408



EGON H. E. LASS



The distribution of bone weight percentage, being a function of bone size (i.e., mean
weight of bone; fig. 21.9), while not a perfect fit, does accommodate itself more or less to the
shell and microdebitage profiles. When these two categories are abundant, indicating areas of
specialization, the bone percentage also rises. To correlate specialized tool production and
minimum retouch with an accent on meat processing seems contradictory. The relationship
between these various kinds of cultural remains is complex and not fully understood. The rela-
tively large amounts of botanical remains in Square H5 (fig. 21.9) are probably due to the fiery
destruction which took place there, carbonizing and preserving them more than in other areas.

CONCLUSION

The results of this study show that different activities have been detected in contempora-
neous and diachronic strata. The microdebitage indicates areas in which tools (1) were both
made and used, (2) were made exclusively, and (3) were neither made nor used. Diachronic
development is shown to remain relatively stable in one area whereas another area bears tes-
timony to change. Since the material is only a small sample representing a substantial site over
an extended period of time, no major trends may be inferred. However, it may be indicative
that all of the EB I loci show a broader range of activities, whereas all of the specialized activity
areas are confined to the EB III period. More comparative material is needed to show whether
this was a general trend of development. Additionally, the poorly understood relationships
between different material culture categories should be an object of scrutiny for future study.

ACKNOWLEDGMENTS

Grateful acknowlegement is made to SeJin Koh and David Schloen for allowing me access
to the site information that was available to this project; to Yaakov Nir and Giora Parnus for
their helpful suggestions; and to Richard Saley for writing the calculation programs. Any
faulty reasoning or errors are entirely my own.

REFERENCES



Baker, C. M.
1978 The Size Effect: An Explanation of Variability in Surface Artifact Assemblage Content.



American
Antiquity



43: 288–93.

Bradley, R., and Fulford, M.
1980 Sherd Size in the Analysis of Occupation Debris.



Bulletin of the Institute of Archaeology



17:
85–94.

Butzer, K. W.
1980 Context in Archaeology: An Alternative Perspective



. Journal of Field Archaeology



7: 417–22.
1982



Archaeology as Human Ecology: Method and Theory for a Contextual Approach.



Cambridge:
Cambridge University Press.

Clarke, D. L.
1977 Spatial Information in Archaeology. Pp. 1–32 in



Spatial Archaeology



, ed. D. L. Clarke. New
York: Academic Press.

Davis, M. E., and Wesolowsky, A. B.
1975 The Izum: A Simple Water Separation Device.



Journal of Field Archaeology



2: 271–73.

oi.uchicago.edu





409



CHIPS AND SHELLS: FLOTATION AT TEL YAQUSH



Diamant, S.
1979 A Short History of Archaeological Sieving at Franchthi Cave, Greece.



Journal of Field Archae-
ology



6: 203–17.
Dunnell, R. C., and Stein, J. K.

1989 Theoretical Issues in the Interpretation of Microartifacts.



Geoarchaeology



4: 31–42.
Esse, D. L.

1990 Notes and News: Tel Yaqush.



Israel Exploration Journal



40: 222–23.
1993 Tel Yaqush. Pp. 1502–4 in



The New Encyclopedia of Archaeological Excavations in the Holy
Land



, ed. E. Stern. Jerusalem: Israel Exploration Society & Carta and New York: Simon &
Schuster.

Fladmark, K. R.
1982 Microdebitage Analysis: Initial Considerations.



Journal of Archaeological Science



9: 205–20.
French, D. H.

1971 An Experiment in Water-Sieving.



Anatolian Studies



21: 59–64.
Hull, K. L.

1987 Identification of Cultural Site Formation Processes through Microdebitage Analysis.



American
Antiquity



52: 772–83.
Lass, Egon H. E.

1994 Quantitative Studies in Flotation at Ashkelon, 1986 to 1988.



Bulletin of the American Schools of
Oriental Research



294: 23–38.
Limp, F. W.

1974 Water Separation and Flotation Processes.



Journal of Field Archaeology



1: 337–42.
Peacock, E.

1991 Distinguishing between Artifacts and Geofacts: A Text Case from Eastern England.



Journal of
Field Archaeology



18: 345–61.
Richardson, J. B., III, and McCreery, D.

1978 Preliminary Analysis of the Plant Remains from Bâb edh-Dhrâº, 1975. Pp. 55–56 in



Preliminary
Excavation Reports: Bâb edh-Dhr⺠Sardis, Meiron, Tell el-Hesi, Carthage (Punic),



ed. D. N.
Freedman. Annual of the American Schools of Oriental Research 43. Cambridge, MA: Ameri-
can Schools of Oriental Research.

Rosen, A. M.
1986



Cities of Clay: The Geoarchaeology of Tells.



Chicago: University of Chicago Press.
1989 Ancient Town and City Sites: A View from the Microscope.



American Antiquity



54: 564–78.
1993 Microartifacts as a Reflection of Cultural Factors in Site Formation. Pp. 141–48 in



Formation
Processes in Archaeological Context,



eds. P. Goldberg, D. T. Nash, and M. D. Petraglia. Mono-
graphs in World Archaeology 17. Madison, WI: Prehistory Press.

Sallade, J. K.
1989 Ethnoarchaeological Investigations, 1976: Evaluating Assumptions Concerning Spatial Pattern-

ing on the Basis of Data from a Cypriot Village. Pp. 407–24 in



American Expedition to Idalion,
Cyprus 1973–1980



, eds. L. E. Stager and A. M. Walker. Oriental Institute Communications 24.
Chicago: Oriental Institute.

Schiffer, M. B.
1972 Archaeological Context and Systemic Context.



American Antiquity



37: 156–65.
1983 Toward the Identification of Formation Processes.



American Antiquity



48: 675–706.
Simms, S. R., and Heath, K. M.

1990 Site Structure of the Orbit Inn: An Application of Ethnoarchaeology.



American Antiquity



55:
797–813.

South, S.
1978 Pattern Recognition in Historical Archaeology.



American Antiquity



43: 223–30.
Stahle, D. W., and Dunn, J. E.

1982 An Analysis and Application of the Size Distribution of Waste Flakes from the Manufacture of
Bifacial Stone Tools.



World Archaeology



14: 84–97.

oi.uchicago.edu





410



EGON H. E. LASS



Stein, J. K., and Teltser, P. A.
1989 Size Distribution of Artifact Classes: Combining Marco- and Micro-Fractions.



Geoarchaeology



4: 1–30.

Stevenson, M. G.
1985 The Formation of Artifact Assemblages at Workshop/Habitation Sites: Models from Peace Point

in Northern Alberta.



American Antiquity



50: 63–81.

Stewart, R. B., and Robertson, W.
1973 Application of the Flotation Technique in Arid Areas.



Economic Botany



27: 114–16.

Vance, E. D.
1987 Microdebitage and Archaeological Activity Analysis.



Archaeology



40: 58–59.

Wood, J. J.
1978 Optimal Location in Settlement Space: A Model for Describing Location Strategies.



American
Antiquity



43: 258–70.

oi.uchicago.edu



411



THE PROTODYNASTIC/DYNASTY 1 EGYPTIAN
PRESENCE IN SOUTHERN CANAAN:
A PRELIMINARY REPORT ON THE



1994 EXCAVATIONS AT NA



Ó



AL
TILLAH, ISRAEL



Thomas E. Levy, David Alon, Edwin C. M. van den Brink,

Eric C. Kansa, and Yuval Yekutieli



This paper is written in memory of Doug Esse, a dear friend and leading light in the study
of the Early Bronze Age of Israel. Doug tragically died at the young age of forty-two, and he
is sorely missed as friend and outstanding colleague. Doug would have been excited by the
new discoveries at Na



˙



al Tillah, and it saddens us that he is not here to share in the exultation
and thrill of our new research.

THE RESEARCH AREA

Na



˙



al (Hebrew; Arabic = Wadi) Tillah is a small secondary seasonal drainage that
debauches into the larger Na



˙



al Gerar and is located near the interface between the Irano-
Turanian semi-arid and Mediterranean environmental zones of Israel’s northern Negev and
southern Shephelah regions. The Na



˙



al Tillah research project focuses on the excavation and
geophysical survey of cave sites adjacent to Abu Hof, a large Chalcolithic settlement situated
at the beginning of the Na



˙



al Tillah drainage and excavations at the Silo Site, situated on the
nearby eastern slopes of the Tel



Ó



alif Terrace which covers a total area of some thirteen hec-
tares. The



Ó



alif Terrace, first extensively excavated by Seger (Seger et al. 1990) is roughly
three hectares larger than the average size of Negev Chalcolithic settlement centers (Levy
1995). The area is characterized by Eocene chalk hills, dissected by small secondary drain-
ages, with many small valleys in-filled with loessial sediments. Rainfall averages around 400
mm on an annual basis. The



Ó



alif Terrace rises to some 490 meters above sea level and marks
the watershed between the Na



˙



al Gerar in the west and the secondary drainages that flow into
the Biqaªat Yaval (Yaval Valley) in the east. The area dominates an important ancient trans-
portation/trade route east-west from the Mediterranean coast, north-south through the south-
ern Shephelah, and northward through the Judaean mountain system.

PREVIOUS EXCAVATIONS IN THE NA



Ó



AL TILLAH REGION

The Na



˙



al Tillah project grows out of earlier pioneering research in this environmental
contact zone carried out under the direction of J. D. Seger of Mississippi State University and
D. Alon of the Israel Antiquities Authority on the



Ó



alif Terrace. The terrace is located on the



T. E. Levy, D. Alon, E. C. M. van den

Brink, E. Kansa, and Y. Yekutieli



22



oi.uchicago.edu





412



T. E. LEVY, D. ALON, E. C. M. VAN DEN BRINK, E. KANSA, AND Y. YEKUTIELI



eastern side of Tel



Ó



alif (Arabic = Tell Khuweilifeh) near Na



˙



al Tillah on the grounds of Kib-
butz Lahav (Alon 1974; 1977a; 1977b; Seger 1987a; 1987b; 1990; 1991; Seger et al. 1990;
Dessel 1991; see fig. 22.1). Seger’s (Seger et al. 1990) precise excavation work provides an
essential stratigraphic cornerstone that demonstrates the rich evidence for the elusive Chal-
colithic through Early Bronze (EB) I sequence in southern Israel. Naomi Porat (1989) was the
first scholar to make an in-depth study of Egyptian–EB I Canaanite trade relations based on
petrographic studies of pottery from the region. J. P. Dessel (1991) made the first systematic
and in-depth ceramic analysis of material from the



Ó



alif Terrace based on Seger’s excavations
at the site (Sites 101 and 301). Alon and Yekutieli (1995) made a similar study using data from
Alon’s eight probes on the



Ó



alif Terrace carried out in the early 1970s. As a total of ten probes
have been made on the



Ó



alif Terrace, researchers have given separate names—such as Site
101, Site 301 (Seger et al. 1990), and the “Silo Site” (Alon and Yekutieli 1995)—to the differ-
ent excavation operations at the site. Together, the early excavations cover an area of some
1,703 m



2



.



F



IGURE



22.1. Topographic map of the



Ó



alif Terrace.



recto is long by 1 pica.
verso is short for heading.
awful, ain’t it?

oi.uchicago.edu

[image: image130.tif]



413



A PRELIMINARY REPORT ON THE 1994 EXCAVATIONS AT NA



Ó



AL TILLAH



THE 1994 EXCAVATIONS AT NA



Ó



AL TILLAH

Building on the work of these earlier researchers, in the summer (July-August) of 1994,
T. E. Levy and D. Alon initiated a pilot excavation and survey season in the Na



˙



al Tillah
region (fig. 22.1; Levy et al. 1995). This included a large horizontal sounding (about 800 m



2



)
on the



Ó



alif Terrace that was labeled the “Silo Site” as it was located near an earthen fodder
silo established by Kibbutz Lahav. This was also the site of Alon’s original work on the ter-
race. Survey and excavations in 1994 have confirmed a full stratified sequence of Chalco-
lithic, early EB I, and late EB I occupation at the Silo Site. In addition, using geophysical sur-
vey techniques, A. Witten located a significant number of potential EB burial caves in the
research area.

Most important is the widespread evidence for a substantial Protodynastic/Early Dynastic
Egyptian presence on the



Ó



alif Terrace that rivals Tel Erani, long thought to be the center of
Egyptian activity in southern Canaan (cf. Brandl 1989). The 1994 excavations have revealed
large quantities of prestige goods such as Egyptian Late Ware vessels (pl. 22.2; cf. Petrie 1901:
fig. 8), a faience jar, Egyptian storage jars, and administrative artifacts such as a clay



bulla



, or
cylinder seal impression, and a



serekh



with the name of King Narmer (Levy et al. 1995). These
are discussed in more detail below. The discovery of the clay bulla marks the first time one of
these administrative artifacts have been found in situ outside of ºEn Besor in southern Israel
(see Gophna 1995). In the following discussion, we highlight the tomb discovery, the pottery
assemblage, and some of the administrative artifacts found in 1994.

A total of five archaeological strata have been identified at the Silo Site on the



Ó



alif Terrace.
These include

• Stratum I, a disturbed and mixed deposit with a wide range of artifacts spanning the
Middle Bronze through Chalcolithic periods.

• Stratum IIA, a rather ephemeral building phase with late EB I remains. The presence
of an Egyptian clay bulla (see below) most probably links this deposit to the beginning
of Dynasty 1 in Egypt.

• Stratum IIB, which contains the best-preserved architectural remains on the site, dates
to late EB I and has been correlated with the very end of the Proto-Dynastic period (cf.
Levy et al. 1995).

• Stratum III, characterized by ephemeral pit dwellings and installations that date to
early EB I.

• Stratum IV which relates to the very late Chalcolithic period.

The following discussion presents a preliminary study of the ceramic assemblage found in
these strata during the 1994 excavations at the Silo Site on the



Ó



alif Terrace.

PRELIMINARY CERAMIC ANALYSIS OF THE 1994 SILO SITE

ASSEMBLAGE,



Ó



ALIF TERRACE



Methodology of Pottery Analysis



Over 900 kg of pottery (approximately 50,000 sherds) were recovered from the first season
of the renewed excavations at Tel



Ó



alif Terrace Silo Site. Based on prior knowledge of fourth

oi.uchicago.edu





414



T. E. LEVY, D. ALON, E. C. M. VAN DEN BRINK, E. KANSA, AND Y. YEKUTIELI



millennium



b.c.



southern Canaanite pottery in general, and at Tel



Ó



alif Terrace in particular
(Dessel 1991; Yekutieli 1992; Alon and Yekutieli 1995), the sherds were first classified
according to time period (Chalcolithic, EB IA, EB IB, and later periods), and sherd weights
and counts were taken for these categories for each locus. Diagnostic sherds were then sub-
jected to a more rigorous analysis. The variables of vessel form, rim form, handle form, deco-
ration application technique and decorative motif, and temper were recorded with an estimate
of the minimum number represented (again for each locus). The tempers were defined visually,
by the authors and by comparison with examples that had been checked petrographically, prior
to our excavation, by Y. Goren. Each of these variables is extremely important and should be
recorded in order to discern as many trends as possible. The mathematical quantification of all
of these variables for all pottery from each locus will eventually enable more objective and
sophisticated comparisons with other sites. In addition, in this way we can hope to detect
changes and patterns along symbolic, economic, and technological dimensions.



The Na



˙



al Tillah Silo Site Ceramic Assemblage



The pottery of the Na



˙



al Tillah Silo Site appears similar to the general range of Late Chal-
colithic to EB I pottery of southern Israel. The following presents a brief overview of the Silo
Site assemblage.



Chalcolithic



For the Chalcolithic, the ceramic assemblage included some vessel forms that are fairly
particular to the Chalcolithic, such as cornets and churns. The Chalcolithic sample most closely
parallels that of the Patish Valley Chalcolithic tradition, in form, temper, and decoration. Evi-
dence for ceramics of the Chalcolithic tradition is found mostly in the deepest strata, chiefly
IV and III (fig. 22.2). This was a very well-developed tradition with fairly standardized and
finely shaped forms. The primary decorative technique was the application of a red painted
band at the rim of the vessel, and the most common types of clay inclusions were wadi sands
of different sizes (small grains for open vessels, larger for closed forms; see fig. 22.3a).



EB IA



Vessels of the EB IA tradition seem crude compared with those of the Chalcolithic period.
This tradition was the most prevalent in Stratum III (fig. 22.2). They tend to have more vari-
able and less refined forms. They also differ technologically, with “grog and grit” temper
(crushed pottery and sandy inclusions) and a brittle straw temper (found mostly on open
forms) becoming very common (fig. 22.3b). Gone are the special forms of the Chalcolithic
such as the churns and the cornets; instead the very crude straw-tempered hemispherical bowls
become quite common. Ledge handles appear in this tradition, and decoration is more variable.
We see more thumbed rims, plastic decoration, and the occasional application of whitewash,
as well as continued use of red painted bands.

The ceramic repertoire of Stratum III places it within the early subphase of EB IA. It has
parallels at Tel



Ó



alif Terrace sites: Silo Site Stratum III (Alon and Yekutieli 1995); and Site
101 Stratum V, 150 meters uphill from the Silo Site (Dessel 1991: 72–73); at several dwelling
caves at the Northwestern settlement at Lachish (Tufnell et al. 1958); and with some variations
at Site H, lower phase, and Nizzanim Stratum V (Yekutieli 1992).

oi.uchicago.edu





415



A PRELIMINARY REPORT ON THE 1994 EXCAVATIONS AT NA



Ó



AL TILLAH



FIGURE 22.2. Na



˙



al Tillah: Pottery profile by stratum.

oi.uchicago.edu

[image: image131.tif]



416



T. E. LEVY, D. ALON, E. C. M. VAN DEN BRINK, E. KANSA, AND Y. YEKUTIELI



EB IB



In Strata I and II we see two more ceramic traditions that are generally assigned to EB IB
(fig. 22.2). The first of these traditions is the local Canaanite EB IB tradition. This can be
broadly characterized as having more refined forms and shaping than EB IA. It also differs
technologically: grog continues to be used as an inclusion, but straw becomes less common.
Crushed calcite becomes very common for some holemouth jars, probably to absorb thermal
stresses from cooking fires. Dolomitic sand also becomes common, especially for some
medium to small storage jars (fig. 22.3c). Ledge handles continue, and we also see the use of
pillar handles on some storage vessels. Decoration also changes somewhat, with plastic rope
decoration on many holemouths, and red painted band or net patterns on whitewashed storage
jars (and some open forms). Of special interest are the numerous “potter’s marks” on some
sherds, which are well known for this time period. These marks were usually symbols made
with two or three stokes incised on the base or body of the vessel. They have been found pri-



FIGURE 22.3. Clay types of different ceramic traditions: a) Chalcolithic-type Canaanite pottery; b) EB IA-type
Canaanite pottery; c) EB IB-type Canaanite pottery; d) EB IB Egyptian pottery.

oi.uchicago.edu

[image: image132.tif]



417



A PRELIMINARY REPORT ON THE 1994 EXCAVATIONS AT NA



Ó



AL TILLAH



marily on the EB IB vessels, mostly on Canaanite forms, but some were found on Egyptian
forms as well. These intriguing marks invite further investigation.

The Egyptian Naqada III–Early Dynastic wares first appear in Stratum IIB (fig. 22.2). Pet-
rographic analysis done by J. P. Dessel to similar vessels at Site 101, 150 m uphill from our
excavation, indicates that some of these vessels, most commonly storage jars, were produced
in the Nile Valley with hard marl clays (Dessel 1991) (fig. 22.3d). Our own petrographic
analysis (done by Goren [Levy et al. 1995]) on the Narmer



serekh



sherd (hard marl-clay body
sherd of an Egyptian storage jar) has further confirmed this observation. These vessels were
very finely and regularly shaped, but very few of these Egyptian imports were decorated. The
majority of the Egyptian vessels of this fabric belonged to the following types of Petrie’s clas-
sification system (Petrie 1953; also see table 22.1):

1. Large to medium storage jars, class L (late), Petrie’s (1953) types 59–63, and 75.
2. Cylindrical vessels, class W (wavy handled), Petrie’s (1953) type 72.
3. Open vessels, many of which were red burnished and “lotus shaped,” Petrie’s (1953)

types 3 and 4.
4. Globular juglets, Petrie’s (1953) types 87 and 88.

Worthy of note are two tube handles from Egyptian D-ware (Petrie’s classification) vessels
with painted spiral decorations. To date, examples of these have been found only at Taur
Ikbeineh (Oren and Yekutieli 1992). These finds are unfortunately not chronologically signif-
icant to the understanding of our site since they were found in highly disturbed contexts of
Stratum I.

In addition to imported Egyptian wares, we also found much evidence for the local produc-
tion of Egyptian type vessels. Petrographic work done for Dessel’s study (Dessel 1991) has
shown that the Egyptian form vessels with straw tempers were locally produced (fig. 22.3d).
This is especially obvious for the Egyptian bread molds (which would have been highly imprac-
tical to import), made of a very crude mold-shaped straw temper. The majority of the rest of the
locally produced Egyptian vessels were of open forms, particularly large storage basins (table
22.2). As with the imported wares, very few of the locally produced Egyptian vessels were
decorated but were very regularly shaped.

In general the ceramic repertoire of Strata I and II belongs to the Southwestern Canaanite
horizon of the later part of the EB IB period. It resembles that of sites such as Arad Stratum
IV (Amiran et al. 1978), Erani Strata VI–V (Yeivin 1961), Tel Maªa



˙



az Stratum I (Amiran



Table 22.1.



Egyptian Forms Produced with Marl Clays (Imported)



Storage
Cylindrical

Jar

Small
Globular

Jar Juglet Bowl
Lotus
Bowl Holemouth Cyl. Bowl N=



I 52.8% 8.5% 15.5% 0.7% 16.2% 4.9% 0.7% 0.7% 142

IIA 45.8% 0.0% 12.5% 0.0% 25.0% 16.7% 0.0% 0.0% 24

IIA/B 62.6% 6.3% 12.5% 0.0% 12.5% 0.0% 0.0% 6.3% 16

IIB 41.6% 7.8% 11.7% 0.0% 27.3% 5.2% 0.0% 6.5% 77

III — — — — — — — — 0

IV — — — — — — — — 0

oi.uchicago.edu





418



T. E. LEVY, D. ALON, E. C. M. VAN DEN BRINK, E. KANSA, AND Y. YEKUTIELI



and van den of Brink, chap. 3, this volume), and the Azor tombs (Ben-Tor 1975), in both local
and imported Egyptian vessels. The horizon is well dated, by synchronization with Egypt
through the imported vessels, to the end of Dynasty 0 and early Dynasty 1 (Amiran and
Gophna 1992: 358).

Some very interesting patterns can be seen in the examination of the chronological pat-
terns of the assemblage. Figure 22.2 shows the prevalence of the different ceramic traditions
through the occupational history of the site. As expected, the Chalcolithic-type wares are more
common at the earliest occupation phase, Stratum IV. The EB IA types of pottery are most
common in Stratum III, while the EB IB Canaanite and Egyptian wares are most common in
the latest occupation phases of Strata II and I.

An interesting point can be made concerning the nature of the Chalcolithic–EB IA transi-
tion. As seen in figure 22.2, there does not appear to be any sharp break, but a more gradual
transition in the predominance of these two ceramic traditions. The EB IA type ceramics are
not without precedence in Stratum III since they do occur in Stratum IV. Nor do Chalcolithic
type ceramics disappear in Stratum III; they decline in prevalence but are still quite com-
mon. Figure 22.4 further illustrates these trends. Large and small grain wadi sand, which con-
stitute the bulk of the Chalcolithic tradition, are most frequent in Stratum IV and become
somewhat scarcer as the brittle straw temper and grog and grit temper (two clay types found
mostly in the EB IA) become more common in Stratum III. Despite some possible mixing due
to site formation processes, this can be taken as evidence that there was no complete break in
the occupation of the site during the Chalcolithic–EB IA transition, something fairly rare in
the southern Levant. This evidence supports observations of some continuity between the
Chalcolithic and EB IA in southwestern Israel (Wright 1937; Amiran 1985; Joffe 1993;
Gophna 1995). Unfortunately, our excavation sample of Stratum IV was extremely limited,
and any conclusions comparing Strata III and IV must be very tentative at this stage.

It is also interesting to note that there is only one sharp discontinuity in this record,
between Stratum III and Stratum IIB. Between these two occupation phases, we see a large
increase in the amount of the EB IB-type Canaanite pottery, and a corresponding decrease in
the proportions of the EB IA and Chalcolithic types of Canaanite pottery. As illustrated in
figure 22.4, some evidence for EB IB-type Canaanite production exists in the preceding Stra-
tum III with the use of grog, calcite, and dolomitic sand clay types. However, the use of these
clay types was rare in Stratum III and may reflect some experimentation with new tempers, or



Table 22.2.



Egyptian Forms Produced with Hard Straw-Tempered Clays (Assumed locally produced)



Storage
Krater Storage

Small
Globular

Jar Bowl
Lotus
Bowl Bread mold Holemouth N=



I 41.5% 2.4% 8.5% 13.4% 3.7% 25.6% 4.9% 82

IIA 43.5% 0.0% 0.0% 0.0% 26.1% 30.4% 0.0% 23

IIA/B 57.9% 0.0% 0.0% 5.3% 15.8% 21.1% 0.0% 19

IIB 60.7% 0.0% 0.0% 5.4% 5.4% 28.6% 0.0% 56

III — — — — — — — 0

IV — — — — — — — 0

oi.uchicago.edu





419



A PRELIMINARY REPORT ON THE 1994 EXCAVATIONS AT NA



Ó



AL TILLAH



F
IG

U
R

E
2

2.
4.

N
a



˙



al
T

il
la

h
Si

lo
S

it
e

19
94

:
S

er
io

gr
ap

h
of

c
la

y
ty

pe
f

re
qu

en
ci

es
b

y
st

ra
tu

m
.

oi.uchicago.edu

[image: image133.tif]



420



T. E. LEVY, D. ALON, E. C. M. VAN DEN BRINK, E. KANSA, AND Y. YEKUTIELI



some sort sampling error (or both). Clearly, there is marked contrast between Strata IIB and
III in the frequencies of these clay types. Even more dramatic is the introduction of Egyptian
wares in Stratum IIB (fig. 22.2). This is clearly illustrated in figure 22.4, where two new
Egyptian clay types suddenly appear. The two coinciding trends of the great increase in the
EB IB-type Canaanite production and the introduction of Egyptian wares indicate a very dras-
tic change in the ceramic assemblage of the site. This marked contrast supports the observa-
tion that the late phase of the EB IA and the early phase of the EB IB are missing at our
excavated area, and that during this time technological and typological changes occurred.

Further analysis of the ceramics can shed new light on the nature of this interaction.
Figure 22.5 shows the ratio of serving vessels to storage vessels over time. As seen in the
graph, with the introduction of Egyptian vessels at the site in Stratum IIB, we see a reversal
in the ratio of serving to storage vessels. After Stratum IIB, storage vessels become more com-
mon than serving vessels. This is true for both Egyptian-type ceramics and Canaanite-type
ceramics. This ratio is rather unexpected, considering that one usually expects serving vessels
to outnumber storage vessels (both because there is usually a need for more serving vessels,
and that they usually have a much higher breakage rate; Orton, Tyres, and Vince 1993).

This dramatic change may reflect a shift in the demand for different types of vessels, with
storage becoming more important. This evidence for more storage requirements (perhaps far
beyond the household level) may be an indication of increasing economic complexity and an
integration of the Silo Site into large economic systems. The mere presence of large numbers
of imported Egyptian storage vessels is a further indication of this process. This evidence, in
addition to epigraphic evidence of the Narmer



serekh



(fig. 22.14:10), as well as numerous pot-
ter’s identification marks, may point to the emergence of a larger, more sophisticated regional
economy including the Silo Site with the Egyptian periphery.



Some Spatial Interpretations



One interesting question concerns the purpose of the Egyptian ceramics at the Silo Site.
Are they there as a result of trade and used for local consumption by Canaanites? Did some
Canaanites use them to emulate Egyptians and Egyptian prestige symbols for purposes of
social ranking? Or were there actually Egyptians living at the Silo Site, using their own
ceramics and symbolically communicating their separate ethnic identity? The mere presence
of Egyptian pottery at the Silo Site need not mean the presence of Egyptians—pots do not
equal people—so further analysis is needed.

Figures 22.6–22.9 can be interpreted to help answer these questions. For these figures,
mixed contexts such as fills, topsoil, and rubbles were not considered. The contexts that were
considered include pits, surfaces, floors, hearths, and structures. These contexts better pre-
serve information about use and especially disposal behaviors, which could be lost in those
other contexts that are far more affected by other processes. Figures 22.6 and 22.7 show that
the locally produced and imported Egyptian ceramics are found in generally the same places.
This similar use of space may indicate that the locally produced and imported Egyptian
ceramics shared some conceptual and functional similarities.

Such similarity is not at all shared between the Egyptian wares and the Canaanite wares
(figs. 22.8, 22.9). Though our sample is small, we do see a dramatic difference in the distribu-
tion of these two categories, resulting from different disposal and (probably) use behaviors.

oi.uchicago.edu





421



A PRELIMINARY REPORT ON THE 1994 EXCAVATIONS AT NA



Ó



AL TILLAH



The vast majority of the Canaanite pottery concentrated toward the north of the excavated
area, associated with little in the way of recognizable structures. On the other hand, the major-
ity of the Egyptian pottery was found towards the south of the excavated area. Complete ves-
sels were found intentionally buried in the floor of Room 1, while other pottery fragments
were found associated with a large stone feature in the southernmost part of the excavated
area. This spatial segregation of Egyptian and Canaanite pottery probably indicates that these
two traditions served very different conceptual and functional purposes. Most probably the
Egyptian ceramics were not evenly distributed among the population of the site.



FIGURE 22.5. Na



˙



al Tillah, Silo Site 1994: Ratio of serving to storage vessels by stratum.

oi.uchicago.edu

[image: image134.tif]



422



T. E. LEVY, D. ALON, E. C. M. VAN DEN BRINK, E. KANSA, AND Y. YEKUTIELI



FIGURE 22.6. Na



˙



al Tillah, Silo Site 1994: Stratum IIB distribution of locally produced Egyptian
pottery.

FIGURE 22.7. Na



˙



al Tillah, Silo Site 1994: Stratum IIB distribution of imported Egyptian pottery.

oi.uchicago.edu

[image: image135.tif][image: image136.tif]



423



A PRELIMINARY REPORT ON THE 1994 EXCAVATIONS AT NA



Ó



AL TILLAH



FIGURE 22.8. Na



˙



al Tillah, Silo Site 1994: Stratum IIB distribution of Egyptian form pottery
(locally made and imported).

FIGURE 22.9. Na



˙



al Tillah, Silo Site 1994: Stratum IIB distribution of Canaanite pottery.

oi.uchicago.edu

[image: image137.tif][image: image138.tif]



424



T. E. LEVY, D. ALON, E. C. M. VAN DEN BRINK, E. KANSA, AND Y. YEKUTIELI



But who were using these ceramics? As discussed earlier, it seems unlikely that the Egyp-
tian ceramics were used in the same way as the Canaanite wares. Many of the Egyptian vessels
served similar functions as the local wares. Storage, food preparation, and food-serving ves-
sels appear in both the Egyptian and the Canaanite samples. The important difference between
Egyptian pottery and Canaanite pottery, reflected by their spatial segregation, is more likely to
be symbolic. Was this symbolic difference an attempt by local Canaanite elites to differentiate
themselves by displaying Egyptian pottery? This is a possible scenario. The Egyptian pottery
might have had some prestige value that local elites chose to emulate.

However, there is little evidence that the Egyptian pottery had any prestige value. It is dif-
ficult to see how an Egyptian bread mold could convey information of high status: they are too
crude, disposable, and utilitarian. Egyptian wares in general, though well shaped, tended not
to have decoration. This lack of decoration may mean that the bulk of the Egyptian ceramic
repertoire was intended for more utilitarian ends and not primarily to communicate prestige
differences. It seems more likely that their presence in Canaan, found segregated from local
types, communicated ethnic distinctions. (“Ethnic” and “prestige” symbolic communication
are not mutually exclusive; however, one of these dimensions might have been more impor-
tant than the other.) Possibly, ethnically distinct Egyptians, maintaining their own technologi-
cal traditions, were living at the Silo Site. The abundant, locally produced Egyptian pottery,
although functionally similar to local wares, was distinct in form, decoration, and manufactur-
ing technology (dissimilar temper), and thus served to communicate a separate Egyptian
ethnicity from the local Canaanites. Again this interpretation is tentative, and further excava-
tion is needed to further confirm this evidence of spatial segregation between Egyptian and
Canaanite pottery.

THE TOMB DISCOVERY

Excavations at the Silo Site were divided into three fields: Areas A, B, and C respectively.
At the beginning of the excavation season, a large enigmatic pile of stones appeared in Area
C some 30 cm below the site surface. When the stone pile was removed, two well-built stone
walls appeared which were encircled by a hard-pack plastered floor (L. 512 and L. 513) that
lips up around the parallel wall structure. These two parallel walls are situated east-west; the
northern one is labeled Wall 11 and the southern one Wall 12. A standing stone was carefully
set on the outside of Wall 11 in association with a small circular installation of stones, all of
which were set in the plaster floor. By the end of the excavation season, Walls 10 and 11 were
shown to extend for about 9.45 m, with a preserved height of over 2.80 m (fig. 22.10). They
were joined in their western extremity by a curvilinear stone wall which makes this feature
form an elongated U-shape (fig. 22.11; pl. 22.1). After removing a large balk which traversed
these walls in their western extremity, we discovered a finely crafted pair of large stone lintels
2.25 m below the height of the tops of Walls 10 and 11. The lintels (figs. 22.10, 22.12) form
the doorway to a spectacular tomb structure.

This monument marks the discovery of a possible Early Dynastic Egyptian-style tomb in
southern Palestine and points to the great potential of the Silo Site on the



Ó



alif Terrace for pro-
viding important data concerning mortuary practices, human skeletal remains, genetic “finger-
printing,” artifacts that reflect trade and exchange, and so on. As mentioned above, Walls 10

oi.uchicago.edu





425



A PRELIMINARY REPORT ON THE 1994 EXCAVATIONS AT NA



Ó



AL TILLAH



F
IG

U
R

E
2

2.
10

. S
ou

th
s

ec
ti

on
v

ie
w

o
f

W
al

l
11

f
ro

m
T

om
b

1,




Ó



al
if

T
er

ra
ce

. N
ot

e
th

e
la

rg
e

w
el

l-
cr

af
te

d
st

on
e

bl
oc

k
ne

ar
t

he
e

nt
ra

nc
e

to
t

he
t

om
b

in
di

ca
te

d
by

t
he

l
oc

at
io

n
of

t
he

t
w

o
li

nt
el

s
on

t
he

l
ef

t
si

de
o

f
th

e
dr

aw
in

g.

oi.uchicago.edu

[image: image139.tif]



426



T. E. LEVY, D. ALON, E. C. M. VAN DEN BRINK, E. KANSA, AND Y. YEKUTIELI



F
IG

U
R

E
2

2.
11

. P
la

n
vi

ew
a

nd
s

ec
ti

on
s

th
ro

ug
h

To
m

b
1

in
A

re
a

C
, S

il
o

Si
te

,



Ó



al
if

T
er

ra
ce

.

oi.uchicago.edu

[image: image140.tif]



427



A PRELIMINARY REPORT ON THE 1994 EXCAVATIONS AT NA



Ó



AL TILLAH



FIGURE 22.12. Section view of entrance and large lintel leading to Tomb 1,



Ó



alif Terrace.

oi.uchicago.edu

[image: image141.tif]



428



T. E. LEVY, D. ALON, E. C. M. VAN DEN BRINK, E. KANSA, AND Y. YEKUTIELI



and 11 were surrounded by a well-made plaster floor that is clearly dated to the late EB I
period, ca. 3300–3000



b.c.



The fact that Wall 10 cut through an early EB I storage silo pro-
vides more evidence for dating the monument to late EB I. Walls 10 and 11 form a long
passageway to the tomb entrance and contain a sharp incline that appears to be made up of
hard-pack earth and stones forming a stairlike entry down to the tomb entrance framed by the
lintels. The lintels rest on dry masonry walls which are preserved to a height of about 1.20 m
and 1.5 m in length. After passing through the lintel-lined entrance, one enters a tholos or
beehive-shaped dome made of dry masonry. While the tholos is remarkably well preserved,
due to its dry masonry construction, it is unsafe to work under the dome. In the center of the
tholos ceiling is a large crafted stone block held in place by stone arches embedded with small
chinking stones. Accidental removal of these small stone chips could cause a collapse; thus,
excavation was abandoned soon after we discovered this structure.

The floor beneath the tholos







was never reached during the 1994 season. However, a thick
deposit of colluvial varve-like sediment with the consistency of butter was found covering the
entire tomb interior. A section was cut though this fill to a depth of around 80 cm, which
enabled us to stand up in a hunched position beneath the tholos







ceiling. The tholos







was con-
structed on top of the entrance to what seem to be two natural caves, some 3.5 m below the
site surface. The entrance to one of these caves was exposed during our limited excavation in
the tomb interior and found to be sealed by a wall made of dry masonry. This leads us to
believe that here we found the actual burial chamber of the tomb which, due to the dangers
outlined above, were later investigated using a substantial safety strategy (Levy et al. 1997).

Our interpretation of this structure as a tomb is based on its uncanny parallel to a few
stone-built First Dynasty tombs found at Helwan in Egypt by the Egyptian archaeologist Zaki
Saad (1951). Saad excavated several tombs that contain steps and lead below ground to a
burial chamber. The



Ó



alif Terrace example described here is unlike any late EB I tomb known
from Palestine that are either burial caves (cf. Ben-Tor 1975) or shaft tombs like the ones
recorded at Bâb edh-Dhr⺠(Schaub and Rast 1989). The wealth of Egyptian imports found at
the



Ó



alif Terrace (cf. Levy et al. 1995, 1997; Seger et al. 1990) point to the importance of this
tomb and its contextual association with these artifacts for clarifying the nature of the Egyp-
tian presence in this part of southern Israel.

We believe that the passage leading to the tomb which is formed by Walls 10 and 11 was
covered with a corbeled stone roof and visible to the late EB I occupants of the



Ó



alif Terrace.
A huge stone fill was found between these walls that represents the collapsed corbeled roof.
That the builders of the tomb knew how to construct a corbeled roof is seen by the presence
of the impressive tholos. Further evidence can be seen in the cross sections made across the
passage (fig. 22.22.11). These cross sections precisely illustrated how the stone walls (nos. 10
and 11) curve inward at the top, illustrating that they supported a corbeled roof. Why the
inhabitants of the



Ó



alif Terrace chose to construct a monumental tomb within part of their set-
tlement is beyond our interpretive powers. By expanding the excavations around the tomb area
as well as the entire Silo Site on the



Ó



alif Terrace (fig. 22.1), we hope to answer this and many
other questions concerning who is buried in the tomb, whether it is an Egyptian or Canaanite
burial, and what is its sociopolitical significance.

SHORT

oi.uchicago.edu





429



A PRELIMINARY REPORT ON THE 1994 EXCAVATIONS AT NA



Ó



AL TILLAH



ADMINISTRATIVE ARTIFACTS



Chronological Implications



Recent excavations at a number of settlement sites in the Nile Delta, all dating to the late
Pre-Dynastic through Early Dynastic periods (ca. 3400–3000 b.c.), have provided us with a
much better understanding of the final stages of the Lower Egyptian, Chalcolithic Maadi cul-
ture. The period referred to here spans the Naqada IIc–d to Naqada IIIa–c in terms of the relative
chronology of Kaiser (1957, 1990). As is well known, settlement sites (as opposed to mortuary
sites) are extremely rare in the Egyptian archaeological record. The new research provides the
missing link concerning the mechanisms that brought about the transition from a late Chalco-
lithic, fairly homogeneous Lower Egyptian culture labeled by some scholars as the “Maadi-
Buto culture” (von der Way 1992: 1) to the following Proto-Dynastic period, characterized by
a material culture with traits shared by the people of both the Nile Delta and Nile Valley, and
which culminated sometime toward the end of the fourth millennium b.c. with the creation of
the early Egyptian state heralded by the so-called Unification of the Two Lands (i.e., Upper and
Lower Egypt) into a single, very homogeneous Early Dynastic material culture prevalent and
widespread throughout the whole of Egypt.

While it has been customary in the past, in discussing the earliest interaction between
Egypt and its eastern neighbors, to lump them together under the common, nondistinctive,
chronological denominator late Pre-Dynastic/Early Dynastic on the one hand, and Early
Bronze I Age on the other, it is now possible to clarify and distinguish between the various
stages of contact and interrelations between contemporary late Chalcolithic Lower Egypt, late
Naqada IIc–d Upper Egypt, and Early Bronze IA Canaan and between contemporary Proto-
and Early Dynastic Upper and Lower Egypt and Early Bronze IB respectively.

At least four distinct phases in the earliest contacts between Lower and/or Upper Egypt on
the one hand and Early Bronze IA–B Canaan on the other can now be distinguished (cf.
Gophna 1992; 1995; Tutundzic 1993; Hartung 1994). These phases include:

(1) The earliest contacts or exchanges occurred between the Chalcolithic Lower Egyptian
Maadi Culture (Seeher 1990) and those of the Chalcolithic Na˙al Beersheva culture (cf. Perrot
1955: 186; Levy 1992: 350 ff.) as attested by the presence of subterranean buildings at Maadi
in Lower Egypt and at sites such as Bir es-Safadi, Abu Matar, and Shiqmim in southern Israel.

(2) The next phase in this process of ongoing contacts was between the late Chalcolithic
Lower Egyptian Maadi-Buto culture and the Early Bronze IA culture in southern Israel, indi-
cating the beginning of some kind of Egyptian presence in southern Canaan. This is seen by the
significant quantities of Egyptian or Egyptianized pottery alongside of local EB IA ceramics
(e.g., Oren and Yekutieli 1992) as well as certain Egyptian types of flint tools (cf. Rosen 1988)
at sites in southern Israel, and of Canaanite ceramic imports found at Maadi (Seeher 1990) and
in the cemetery at Minshat Abu Omar in Egypt (cf. Kroeper 1989).

(3) The next stage—characterized by a sharp quantitative decline in Egyptian items found
in Israel, but by a sharp increase in imported commodities contained in Canaanite vessels in
Egypt—is more or less fixed between late Naqada IId and IIIa2 (in its later part [the early Proto-
Dynastic period] identifiable with early Dynasty 0. These links have been established at ceme-
tery sites in Egypt such as Umm el-Qaªab, Abydos (especially cemetery U (Dreyer 1992, 1993),

oi.uchicago.edu



430 T. E. LEVY, D. ALON, E. C. M. VAN DEN BRINK, E. KANSA, AND Y. YEKUTIELI

Abusir el Meleq, Amrah, Gerzeh, Minshat Abu Omar, Naqada, and Hierakonpolis (see Hartung
1994: nn. 24–32 for refs.).

(4) The apex of these early contacts is between late Proto-Dynastic Egypt (i.e., Dynasty
0) and the contemporary late EB IB culture in southern Israel. It is in this period that we again
see growing evidence for the presence of an Egyptian population component among the local
Canaanite settlers, apparently sometimes characterized by the establishment of their own Egyp-
tian enclaves separate from the local EB I population. These temporal relations are highlighted
in table 22.3.

Although Canaanite imports in this period are almost absent in Egypt, an Egyptian presence
in southern Israel is clearly indicated by the presence of significant percentages of Egyptianized
materials among the local EB IB ceramic repertoire (as shown in the discussion above of the
Egyptian ceramics from the Silo Site on the Óalif Terrace) and in the application of Egyptian
mudbrick building techniques as seen at Tel Erani, Afridar, and ºEn Besor, Stratum III. Some
local form of Egyptian administration is indicated by the finds of locally stamped, Egyptianized
seal impressions at ºEn Besor Stratum III and possibly at the Silo Site (see discussion below).

The pottery fragment with the Narmer serekh-sign discussed here from the Silo Site, Óalif
Terrace, is a welcome addition to the growing body of data testifying for renewed Egyptian-
Canaanite interconnections during this, the later part of the EB 1B, roughly contemporary with
the Naqada IIIb–c1 period in Proto-Dynastic Egypt (cf. table 22.3) at the end of the fourth
millennium b.c.

The Silo Site, Narmer Serekh and Egypt

A single ceramic sherd (reg. no. IAA 64994, L.14/B.259) found at the Silo Site was shown
by petrographic analysis to have been made of Egyptian marl clay and bears an incised serekh
of (Horus) Narmer. The sherd is described in detail in Levy et al. 1995. Here we describe the
significance of the Silo Site Narmer serekh in relation to ten other known serekh-signs incised
on pottery vessels that can be positively identified with Narmer: seven come from Egypt and
three from Israel. Only two of these are preserved on complete jars (fig. 22.13:1, 5). The
remaining eight are preserved on fragments. Excluded here is the incised serekh on the com-
plete Tarkhan jar T. 1100, ascribed by Kaiser and Dreyer (1982: fig. 14, no. 39) to Narmer which
we believe on typo-chronological grounds an invalid ascription (van den Brink 1996: no. 10).

Four of these Nar(mer) serekhs are surmounted by a falcon (two facing to the left, and two
to the right) (fig. 22.13:1, 4, 5, 10), representing the god Horus, with which the kings of Egypt
identified themselves (Barta 1969, 1990). Although fragmented, enough space above our
serekh has been preserved, however, to exclude the possibility that it was ever surmounted by
a falcon.

Not enough space has been preserved to the left of the serekh to positively exclude the
possible presence of accompanying incised potmarks, as found on at least six other instances
of incised Narmer serekhs (fig. 22.13:1, 3–6, 8).

Many more serekh fragments, all of them found in Israel, have been tentatively ascribed to
Narmer. These include fragments found at Small Tel Malhata, ºEn Besor, Tel Óalif (D. Alon
excavations, unpublished, fig. 22.14:11), Tel Maªa˙az (Stratum I; Amiran and van den Brink,
chap. 3, this volume), and Tel Erani (fig. 22.14:6). Due to its fragmentary state of preservation,
the Small Tel Malhata fragment (reg. no. 14/29/1/1; fig. 22.14:17), one of the few fragments

oi.uchicago.edu



431A PRELIMINARY REPORT ON THE 1994 EXCAVATIONS AT NAÓAL TILLAH

with a sign in the serekh’s upper, or name compartment, which is ascribed by Amiran, Ilan, and
Arnon (1983) to Narmer, could better be attributed to King Nj-Ór (for the serekh of this king,
see Kaiser and Dreyer 1982: fig. 14:7–8). The fragmentary nature and incomplete preservation
of most of the samples mentioned above prevents positive identification with any king name in
particular.

Although the new Narmer serekh from the Silo Site is too small to reveal anything about
the original shape of the vessel it belonged to, except that it was part of a vessel preserved just
below the shoulder, some inferences can be drawn. For example, the preserved sherd indicates
that the class of vessels on which these Proto-Dynastic serekh-signs were incised, usually
before firing, is well known and has been the focus of a recent study concerning its typo-


Table 22.3. General Correlation of Southern Canaan and Egyptian Stratigraphy

Period Lower Egypt Upper Egypt Southern Canaan Period

Egypt Buto Umm el-Gaºab, Abydos: Canaan

Dynasty I Stratum V Tomb Z Horus Djet Tel Arad, Stratum III EB II

(Naqada IIIc2) Tomb O Horus Djer el Maghar

Tomb B10/15/19 Horus Aha ºEn Besor, Stratum II

Late Dynasty 0–
early Dynasty 1

Stratum IV Tomb B17/18 Horus Nar(mer)
Tel Óalif Terrace, Silo
Site, Stratum IIa–b

(Naqada IIIb1–
IIIc1)

Tomb B7/9 Horus Ka Tel Erani, Stratum V EB IB (late)

Tomb B1/2 Iry-Hor Tel Arad, Stratum IV

ºEn Besor, Stratum III

Small Tel Malhata,
Stratum IV

Tel Maªa˙az, Stratum I

Early Dynasty 0 Stratum III Tombs U/f; U/g; U/h Tel Erani, Stratum C EB IB (middle)

(Naqada IIIa1–2) Tomb U/i Azor, cave tombs

Tombs U/k; U/j Scorpion I

Naqada IIc–d2 Stratum II Hierakonpolis: Taur Ikhbeineh EB IB (early)

Tombs 100 (“painted
tomb”), 101

Tel Erani, Stratum D

HK Locality 29A Lachish (NW)

Silo Site, Stratum IIIa

Late
Chalcolithic
(Naqada IIb)

Stratum Ib Matmar: Tomb 3131 Site H EB IA

Taur Ikhbeineh

Tel Óalif Terrace, Silo
Site, Stratum IIIb

Chalcolithic Stratum Ia
Shiqmim Beersheva
sites

Chalcolithic

Tel Óalif Terrace, Silo
Site, Stratum IV

oi.uchicago.edu



432 T. E. LEVY, D. ALON, E. C. M. VAN DEN BRINK, E. KANSA, AND Y. YEKUTIELIProvenience Serekh-Sign References

Lower Egypt 1. Minshat Abu Omar
Tomb 44 complete jar
reg. no. MAO.T 44.3

Wildung 1982, fig. 33; Kroeper
1988, fig. 141. Cf. also van den
Brink 1996, no. 20

2. Tell Ibrahim Awad
Phase 6, two fragments
EAO TIA 199
(field reg. nos. B200/150/192,
B200/150/165)

van den Brink 1992b: 52, fig. 83,
Levy et al. 1995

3. Ezbet el-Tell
Cemetery fragment

Bakr 1988, pl. 1, b (top row,
middle)

4. Zaweyit el-Aryan
Tomb Z486, fragment
reg. no. M.F.A. 112342*
*Museum of Fine Arts, Boston

Dunham 1978: 26, pl. XVIa

5. Tarkan Tomb 414
complete jar
reg. no. UC 16083**
**University College, London

Petrie, Wainwright, and Gardiner.
1913: 9, pl. XXXI, 8; pl. LVI,
76b; van den Brink 1996, no. 23

Upper Egypt 6. Umm el-Gaºab, Abydos
Tomb B1–2
two fragments

Right: Petrie 1900, pl. XLIV, 1;
Left: Kaiser and Dreyer 1982,
fig. 24, 40

7. Hierakonpolis
Surface find, fragment

Garstang 1907: 135, p. III, 1

Israel 8. Tel Erani
Stratum V, fragment
Israel Antiquities Authority
(henceforth IAA) 59-225, field
reg. no. 59/225/303/7

Yeivin 1960: 195, fig. 2 and pl.
24a

9. Tel Óalif Terrace, Silo Site
Stratum IIb fragment
field reg. L. 14/B. 25/9
IAA 64994

Levy et al. 1995

10. Tel Arad
Stratum IV fragments
IAA 85-17

Amiran 1974: 5, fig. 1 and pl. 1

FIGURE 22.13. Incised serekh-signs of (Horus) Nar(mer) from Egypt and Israel.

oi.uchicago.edu

[image: image142.tif][image: image143.tif][image: image144.tif][image: image145.tif][image: image146.tif][image: image147.tif][image: image148.tif][image: image149.tif][image: image150.tif][image: image151.tif]

433A PRELIMINARY REPORT ON THE 1994 EXCAVATIONS AT NAÓAL TILLAH

chronology (van den Brink 1996). The only published exception to this general prefiring rule
is the serekh-sign found at ºEn Besor Stratum III which was applied after firing.

It seems that the serekh-bearing sherd from the Silo Site is part of a tall storage jar with
tapering body, between about 50 cm (the earliest examples from Naqada IIIa2) and the latest
< 100 cm in height (from Naqada IIIc1). The earliest examples have discontinuous, pushed up
(i.e., en haut relief), vestigial wavy handles on the vessel shoulder (fig. 22.15:a–b). Somewhat
later in time, these vestigial handles disappear and are replaced by impressed crescentic deco-
rations around the shoulder (fig. 22.15:c–d). Still somewhat later are the jars where this cres-
centic decoration gives way to a continuous rope decoration around the shoulders, waist, and
base of the jars (fig. 22.15:e–f). They all can be either with or without incised serekh-signs.

Twenty-four such complete jars bearing incised serekh-signs have been found to date, both
within and outside of Egypt. They have a time span, in terms of relative Upper Egyptian chro-
nology, from Naqada IIIa2–IIIc1, a period that spans Dynasty 0 and the early First Dynasty.

On the basis of the serekh-signs and (Horus) names incised on them, they can be attributed
to as many as ten different rulers or kings (some of whose names have not been identified yet),
belonging either to or contemporaneous with Dynasties 0/1. Thus far, the tombs of only four
of these kings have been positively identified. All four are located within the same cemetery
at Umm el-Gaºab, Abydos, Upper Egypt (Kaiser and Dreyer 1982) and relate to Iry-Hor, Ka,
Narmer, and Aha. Both Dreyer (1987) and Kaiser (1987) present recent findings from the
royal necropolis at Abydos with a seal impression mentioning this order: Narmer, Aha, Djer,
Udimu, and Mer-Neit. This confirms Narmer’s chronological position preceding Horus Aha.

Van den Brink’s (1996) study of the serekh-bearing jars is important in that it attempts to
distinguish an earlier (i.e., Type III jars) and a later (Type IV jars) phase in the use of the
incised serekh-bearing jars of King Narmer based on the pottery typology of jars on which
they appear. On the basis of a single fragment, such as the Silo Site serekh discussed here, it
is not possible to attribute it to either one of these vessel types. However, circumstantial con-
siderations indicate a later date in Narmer’s lengthy reign which stretched for a period from
thirty to sixty years (see Emery 1961). These considerations include the following.

(1) One (fig. 22.13:6; Levy et al. 1995) of two parallels to the Silo Site serekh-sign stems
from the immediate vicinity of Narmer’s own tomb in Abydos. Assuming that this particular
Abydos serekh was incised on a vessel made while funerary arrangements for Narmer’s burial
were prepared—that is, around the time of his death—and that the Silo Site serekh (based on the
close similarity to the other one) was perhaps produced around the same time in the same work-
shop, it can be assumed to date from the end of Narmer’s reign.

(2) With regard to the palaeography of the Silo Site serekh, assuming that the earliest
examples of Narmer’s name are the fullest and most explicitly written (i.e., with both signs
N’r and mr) and that only later was it felt sufficient to refer to Narmer only with the first part
of his name, i.e., N’r, this would also point to a time later, not earlier, in his reign.

(3) We may assume, on the basis of the Narmer ceremonial palette (e.g., Fairservis 1991),
that the king had to provide himself first with a strong foothold in the Nile Delta before inten-
sive contacts could be established with more eastward regions such as EB IB Canaan.

oi.uchicago.edu



434 T. E. LEVY, D. ALON, E. C. M. VAN DEN BRINK, E. KANSA, AND Y. YEKUTIELI

Provenience Serekh-Sign References

1. Palma˙im Quarry
Stratum 2 fragment
field reg. no. 1617/2044/5221

Braun et al. (see chapter 4, this volume)

2. Óorvat ºIllin Ta˙tit
Stratum IV, fragment
field reg. no. 1779/220/894/1

Braun et al. in preparation (see chapter 4,
this volume)

3. Óorvat ºIllin Ta˙tit
Stratum IV, fragment
field reg. no. 1179/303/1595/1

Braun et al. (see chapter 4, this volume)

4. Tel Erani
Stratum V, fragment
Israel Antiquities Authority
(henceforth IAA) 59-225, field reg. no.
59/225/303/7

Yeivin 1960: 195, fig. 2, pl. 24a

5. Tel Erani
Stratum V, fragment
IAA 59-347/2 reg. no. D60/296/27

Yeivin 1963: 212, fig. 3, pl. XXXIX:3

6. Tel Erani
Surface find, fragment
field reg. no. D60/403/30 [sic!]

Yeivin 1963: 205, fig. 2, pl. XXXIX:2

7. Tel Erani
Surface find, fragment
IAA 58-142

unpublished; Brandl, 1992: 447

8. Tel Maªa˙az
Surface find, fragment
IAA 81-193

Schulman and Gophna 1981, fig. 1, pl. 28a

FIGURE 22.14. Incised serekh-signs from Israel.

Other Incised Serekh-Signs Found in Israel

To date seventeen incised serekh-signs or fragments thereof have come to light in Israel.
These are illustrated in fig. 22.14. Seven of these await in-depth publication, of which three
could be positively identified as Narmer(s) (fig. 22.14:4, 10, 14). The serekh from the Silo Site

oi.uchicago.edu

[image: image152.tif][image: image153.tif][image: image154.tif][image: image155.tif][image: image156.tif][image: image157.tif][image: image158.tif]

435A PRELIMINARY REPORT ON THE 1994 EXCAVATIONS AT NAÓAL TILLAH

9. Tel Maªa˙az
Stratum I, fragment
IAA 94-3328 field reg. no. 71/1

Amiran, Ilan, and Arnon 1983: 83, note 21;
Amiran and van den Brink (see chapter 3,
this volume)

10. Tel Óalif Terrace, Silo Site
Stratum IIB, fragment
reg. no. 64994 (L. 14/B.259)

Levy et al. 1995

11. Tel Óalif Terrace, fragment
IAA 75-504 reg. no. 75/504/253/1

unpublished (van den Brink in preparation)

12. Tel Óalif Terrace, site 101
Area 100, fragment
IAA 86-740 field reg. no. L. 27/B.
1262

Seger 1990: 5, fig. 4

13. ºEn Besor
Stratum III, fragment
IAA 75-186

Schulman 1976: 25, fig. 2

14. Tel Arad
Stratum IV, fragments
IAA 85-17

Amiran 1974: 5, fig. 1, pl. 1

15. Small Tel Malhata
Stratum 1, fragment
IAA 94-3326 field reg. no. 14/79/691/1

Amiran, Ilan, and Arnon 1983: 80, fig. 7a

16. Small Tel Malhata
Stratum 1, fragment
IAA 94-3327 field reg. no. 14/79/1552/
2

Amiran, Ilan, and Arnon 1983: 80, fig. 7b

17. Small Tel Malhata
Surface find, fragment
IAA 94-3325 field reg. no. 14/79/1/1

Amiran, Ilan, and Arnon 1983: 80, fig. 7c

Excluded from figure 22.14 are Seger 1987a, pl. 24c and Yeivin 1968: 37, fig. 2, pl. 1a–b, because the incised
fragments presented there, in our opinion, do not bear connection to serekh-signs.

Drawings not to scale.

oi.uchicago.edu

[image: image159.tif][image: image160.tif][image: image161.tif][image: image162.tif][image: image163.tif][image: image164.tif][image: image165.tif][image: image166.tif][image: image167.tif]

436 T. E. LEVY, D. ALON, E. C. M. VAN DEN BRINK, E. KANSA, AND Y. YEKUTIELI

is not a surprise per se. However, a combination of factors makes this particular serekh of spe-
cial interest.

First of all, in contrast to most other serekh-signs or fragments found in Israel (either
because they are too fragmentarily preserved, e.g., fig. 22.14:2, 3, 8, 9, 11, 12, 15–17, or
because they are ambiguous serekh-signs, e.g., fig. 22.14:5), the Silo Site serekh can unam-
biguously be attributed to a known king, (Horus) Nar(mer).

Second, contrary to some other serekh-signs that were not found in any stratigraphic con-
text (e.g., surface finds: fig. 22.14:6, 7, 8, 17), the sherd under discussion here was found with
a number of other complete Egyptian ceramic vessels and comes from a solid stratigraphic
context in Stratum IIB. This is a clear local late EB I context. Assuming that our sherd is not
an heirloom, which is highly unlikely, it provides us with a chronological peg at this point
between the beginning of the First Dynasty and the late EB IB.

Thus, because of its controlled stratigraphic context and because of the apparent absence
at the site of later, i.e., Early Bronze II, remains, the end of Narmer’s reign can be safely pos-
tulated as having been before the beginning of the EB II period. The stratigraphic attribution
of at least one of the two other positively identified Narmer serekhs found in Israel from Tel
Erani, Stratum V (fig. 22.14:4) is perhaps less securely dated than previously thought (see
Braun 1996).

Summary of Serekh’s Significance

The Narmer serekh found at the Silo Site, Stratum IIB, on the Óalif Terrace in Israel pro-
vides a welcome, clear-cut chronological peg that links the later part of the reign of Narmer

FIGURE 22.15. Egyptian storage jars with incised serekh-signs: a) Type I, from Abusir el-Meleq, Tomb 1144
(Berlin 19931). Ht.: 49.5 cm (after Scharff 1926: pl. 11, no. 28); b) Type IIB, from Turah, Tomb 16.g.9 (Kunst
Historisches Museum, Wien AS 6808; after photograph in van den Brink 1996: pl. 4a); c) Type III; from Min-
shat Abu Omar, Tomb 160 (MAO 160.1). Ht.: 61 cm (Kroeper 1988: fig. 95); d) Type III; from “Eastern Nile
Delta” (Metropolitan Museum of Art 61:122). Ht.: 68.5 cm (Fischer 1963); e) Type IVa; from Abydos, Tomb B1
(University College 16089). Ht.: 99 cm (Petrie 1900: 4); f) Type IVa; from Tarkhan, Tomb 414 (University Col-
lege 16083). Ht.: 97 cm (Petrie, Wainright, and Gardiner 1913: 9, pl. 31, no. 68; pl. 56, no. 76b). Note that Types
I, IIB, III, and IVa refer to chrono-typology of these jars in van den Brink 1996.

oi.uchicago.edu

[image: image168.tif]

437A PRELIMINARY REPORT ON THE 1994 EXCAVATIONS AT NAÓAL TILLAH

with the end of the late EB IB culture in southern Israel. It also contributes to the fast-growing
body of data compiled for this important period concerning the rise in Egyptian-Canaanite
interaction. This newly discovered serekh from the Silo Site provides an important adminis-
trative artifact for examining the large assemblage of Proto-Dynastic and Early Dynastic
material culture in southern Israel such as the application of Egyptian mudbrick building tech-
niques, ceramics, stone vessels, stone palettes, and lithics. The character of Egyptian admin-
istrative activities is also seen in the use of Egyptian cylinder seals applied to local clays in
order to seal small pottery containers and points to the strength of Egypt in this part of the
eastern Mediterranean at this time. The following section describes an unusual clay cylinder
seal impression that adds to the growing assemblage of administrative artifacts coming to light
in southern Israel.

First Dynasty Clay Bulla

A clay bulla (reg. no. 649/94, L.19, B.266) was found in a stratigraphically secure fill dat-
ing to Stratum IIA in Area A of the Silo Site excavations. The cylinder seal impressions are
preserved on an oval-shaped lump of fired, reddish-brown clay, some 6 cm long and 3.5 cm
wide (fig. 22.16, pl. 22.3). The firing is probably accidental and was not intentionally made
when the vessel was sealed. A petrographic analysis by Y. Goren indicates that the clay is of
local loess clay (i.e., non-Egyptian) and therefore conforms with the findings concerning vari-
ous cylinder seal impressions found at ºEn Besor, Stratum III (Porat 1989: 60, Sub-93 and
appendix 5a). Coincidentally, the Silo Site example is quite similar in shape, dimensions, fab-
ric, and color, but not in glyptics, to the ºEn Besor, Stratum III, cylinder seal impression
no. 4 (Schulman 1976: pl. III, no. 4).

Two short, parallel imprints and a deeper one in the middle, perpendicular to the axis of
the cylinder seal impressions, over the whole width of the otherwise plain backside of the
bulla, indicate the original position of the knotted cord to/over which the bulla was attached
(fig. 22.16c; pl. 22.3c; Kaplony 1963: I: 53–54, type VI.2).

The bulla has three elongated, cylinder seal-impressed faces: a central ridge (fig. 22.16d;
pl. 22.3b) with an adjoining field on each side of it (fig. 22.16a–b; pl. 22.3a–b). The inscription

FIGURE 22.16. (a–c) Drawings of a clay bulla (649/94, L.19, B.266) found in Stratum IIA at the Silo Site, Óalif
Terrace; (d) Reconstructed drawing of the cylinder seal impression.

oi.uchicago.edu

[image: image169.tif]

438 T. E. LEVY, D. ALON, E. C. M. VAN DEN BRINK, E. KANSA, AND Y. YEKUTIELI

preserved consists of a single narrow line made up by three signs. All three signs (a face [˙r]
and two flagpoles [ntr.wy]) occupy the full height of each sealing face. Between the two flag-
poles are two small horizontal strokes (the upper one slightly thicker than the lower one), pos-
sibly space fillers, or perhaps phonetic complements (t and r) to ntr. They are repeated once
more after the second flagpole, although the original cylinder seal was probably slightly dam-
aged at this particular point, resulting in four irregular-shaped “dots” instead of two clear hori-
zontal strokes. At both sides of the face (˙r), in a vertical line, there are three more tiny “dots.”

The face (˙r), one of the two major glyptic components in this cylinder seal impression,
have been carefully executed, showing eyebrows, eyes, and nose. Even the cheeks have been
indicated by two dots. The ears have not been indicated separately. Lacking neck or beard, it
compares stylistically best with sealings deriving from King Djer’s funerary complex at Umm
el-Gaºab (Kaplony 1963/1: no. 39; 1963/3: pl. 20, fig. 43, pl. 37, fig. 125).

Two of the three impressions on this bulla are running parallel, although starting from
different points on the original cylinder seal. The third one (pl. 22.3b) has been impressed
upside-down in comparison with the other two. The full inscription (face, flagpole, flagpole; or
flagpole, flagpole, face) of the cylinder seal is preserved on all three sides, in two cases repeat-
ing its beginning, showing flagpole, flagpole, face, flagpole (on the central ridge [fig. 22.16b;
pl. 22.3b]) and face, flagpole, flagpole, face (fig. 22.16a; pl. 22.3a). Figure 22.16d shows a
reconstruction of the complete glyptic sequence and the first repetitive sign.

As for the reading of the cylinder seal impression, van den Brink suggests Ór-ntr.wy or
possibly Ór.j-ntr.wy. We have asked Prof. P. Kaplony, without seeing the original (only pho-
tographs and drawings) to comment on the Silo Site seal impression. He kindly did so in a let-
ter dated March 30,1995. The relevant passages are translated from the German and quoted
below in extenso:

Your reading ˙r and ntrwj are without doubt right. Otherwise, I do not see any signs;
only slashes/notches used as space fillers. The impression concerns a personal name Ór-
ntrwj, analogous to Ór-Nt (Kaplony 1963: I: 579ff.). A date in the early First Dynasty
would fit, although the inscription in this time cannot occupy the full height of the face
(Siegelband) of the seal. Without having seen the original I presume that the cylinder seal
had been carved with a double line carrying the same inscription.

What do Ór-Nt, Ór-ntrwj mean? The face of the person whose name is mentioned here
is identical with that of the deity, thus putting the face implicitly (as is done in the
Pyramid Texts) under the protection of the deity, “My face is (the face of) Neith” c.q.
(of) the Two Gods. The ntrwj, written with two falcons on standards, do appear as early
as on the City Palette (cf. Kaplony 1968: 65 and n. 126).

The closest parallel for the new inscription is probably Petrie 1900: pl. 10, 9:13.2. I
suspected already for some time that the group ntrw/˙r in that particular inscription
represented a personal name Ór-ntrw: “My face is (the face of the) gods.” This assumption
is now corroborated by your newly discovered seal impression. Ór-ntrw is written in Petrie
1900: pl. 10, 9:13.2 before the Horus name and therefore should be a name of a prince. The
interpretation given by Schott (1951: 27) must be refuted. The sign group to the left of the
Horus name in Petrie 1900; Kaplony 1963/1: 665 is the special title of the prince Ór-
ntrw. Ór-ntrwj of the new inscription does not have a title. This is not surprising
considering the early date of the impression.

short

oi.uchicago.edu



439A PRELIMINARY REPORT ON THE 1994 EXCAVATIONS AT NAÓAL TILLAH

CONCLUSION

We sincerely miss not having Douglas Esse among us to discuss the exciting discoveries
we have made in the new Na˙al Tillah project. Doug’s (Esse 1989) insights concerning the
processes of “secondary state formation” in Israel helped shape the original research design
for the Na˙al Tillah Project. The preliminary results of the 1994 excavations in the Na˙al
Tillah region, in particular on the Óalif Terrace, point to the great potential this area has for
exploring the changing nature of Egyptian-Canaanite interaction at the end of the fourth and
beginning of the third millennium b.c. The preliminary ceramic analysis, the potential contri-
bution of the newly discovered tomb, and the exciting epigraphic data all presented here point
to the importance of the Óalif Terrace (see Seger et al. 1990; Dessel 1991) for monitoring
these changing center-periphery relations.

ACKNOWLEDGMENTS

Financial support for the 1994 excavations at Na˙al Tillah were provided by the C. Paul
Johnson Family Charitable Foundation (Chicago), the Samuel H. Kress Foundation (New
York), and Friends of the University of California, San Diego (UCSD) Judaic Studies Program.
We are grateful to all these individuals and foundations. The fieldwork was greatly aided by the
participation of students from the UCSD Summer Session Archaeological Field School. Addi-
tional support was provided by Prof. A. Biran, Director, Nelson Glueck School of Biblical
Archaeology, Hebrew Union College-Jewish Institute of Religion, Jerusalem and the College-
Institute. Thanks also to the supervisors of the 1994 Silo Site excavations, Yorke Rowan,
Jonathan Golden, and Michael Jasmin. We also thank Prof. A. Holl of the University of Paris,
Nanterre, for his contributions to the field project.

We are grateful for permission to reproduce here for the first time serekhs from the collec-
tions of Ruth Amiran and Eliot Braun. Drawings and photographs of finds were kindly made
by Carmen Hirsch, Alina Pikovsky, and Tsila Sagiv. Laurel Mannen of the UCSD Judaic
Studies program cordially prepared the layout of many of the figures. We are grateful to Prof.
P. Kaplony for sharing his expertise in reading the seal impression published here.

We would like to thank Joe D. Seger of the Cobb Institute of Archaeology, Mississippi
State University, for his collegiality and support of the Na˙al Tillah project. In addition, we are
grateful to Uzi Halamish, Brachah Porat, and the Joe Alon Regional Study Center and the
members of Kibbutz Lahav for their great logistic support during the course of our excavations.

REFERENCES

Alon, D.
1974 Lahav-Tel Óalif. Hadashot Arkheªologiyot 51–52: 28–29 (Hebrew).
1977a Tel Óalif. Hadashot Arkheªologiyot 61–62: 40 (Hebrew).
1977b Tel Óalif. Hadashot Arkheªologiyot 63–64: 57 (Hebrew).

Alon, D., and Yekutieli, Y.
1995 The Tel Óalif Terrace, “Silo Site” and Its Implications for Early Bronze Age I. ºAtiqot 27: 149–89.

Amiran, R.
1974 An Egyptian Jar Fragment with the Name of Narmer from Arad. Israel Exploration Journal 24:

4–12.

oi.uchicago.edu



440 T. E. LEVY, D. ALON, E. C. M. VAN DEN BRINK, E. KANSA, AND Y. YEKUTIELI

1985 The Transition from the Chalcolithic to the Early Bronze Age. Pp. 108–12 in Biblical Archaeol-
ogy Today, ed. J. Amitai. Jerusalem: Israel Exploration Society.

Amiran, R.; Ilan, O.; and Arnon, C.
1983 Excavations at Small Tel Malhata: Three Narmer Serekhs. Interim Report. The Israel Museum

Journal 2: 75–83.

Amiran, R., and Gophna, R.
1992 The Correlation between Lower Egypt and Southern Canaan during the Early Bronze I Period.

Pp. 357–60 in The Nile Delta in Transition: 4th–3rd Millennium B.C., ed. Edwin C. M. van den
Brink. Tel Aviv: Edwin C. M. van den Brink.

Amiran, R.; Paran, U.; Shiloh, Y.; Brown, R.; Tsafrir, Y.; and Ben-Tor, A.
1978 Early Arad: The Chalcolithic Settlement and Early Bronze City I. First–Fifth Seasons of Exca-

vations 1962–1966. Jerusalem: Israel Exploration Society.

Barta, W.
1969 “Falke des Palastes” als ältester Königstitel. Mitteilungen des Deutschen Archäologischen Insti-

tuts, Abteilung Kairo 24: 51–57.
1990 Der Palasthorustitel und seine Vorläufer in der Frühzeit. Göttinger Miszellen 117/118: 55–58.

Baur, M. I.
1988 The New Excavations at Ezbet el-Tell, Kufur Nigm: The First Season, 1984. Pp. 49–62 in The

Archaeology of the Nile Delta: Problems and Priorities, ed. E. C. M. van den Brink. Amster-
dam: The Netherlands Foundation of Archaeological Research in Egypt.

Ben-Tor, A.
1975 Two Burial Caves of the Proto-Urban Period at Azor. Qedem 1: 1–53. Jerusalem: Institute of

Archaeology, Hebrew University.
Brandl, B.

1989 Observations on the Early Bronze Age Strata of Tel Erani. Pp. 357–87 in L’Urbanisation de la
Palestine à l’âge du Bronze Ancien, ed. P. de Miroschedji. British Archaeological Reports, Inter-
national Series 527. Oxford: British Archaeological Reports.

1992 Evidence for Egyptian Colonization of the Southern Coastal Plain and Lowlands of Canaan dur-
ing the Early Bronze I Period. Pp. 441–77 in The Nile Delta in Transition: 4th–3rd Millennium
B.C., ed. Edwin C. M. van den Brink. Tel Aviv: Edwin C. M. van den Brink.

Braun, E.
1996 Cultural Diversity and Change in the Early Bronze I of Israel and Jordan. Ph.D. dissertation, Tel

Aviv University.
Dessel, J. P.

1991 Ceramic Production and Social Complexity in Fourth Millennium Canaan: A Case Study from
the Tell Óalif Terrace. Ph.D. dissertation, University of Arizona.

Dreyer, G.
1987 Ein Siegel der frühzeitlichen Königsnekropole von Abydos. Mitteilungen des Deutschen Archä-

ologischen Instituts, Abteilung Kairo 43: 33–43.
1992 Recent Discoveries in the U-Cemetery at Abydos. Pp. 293–300 in The Nile Delta in Transition:

4th–3rd Millennium B.C., ed. Edwin C. M. van den Brink. Tel Aviv: Edwin C. M. van den
Brink.

1993 Umm el-Qaab. Nachuntersuchungen im frühzeitlichen Königsfriedhof. 5/6 Vorbericht. Mitteil-
ungen des Deutschen Archäologischen Instituts, Abteilung Kairo 49: 23–62.

Dunham, D.
1978 Zawiyet el-Aryan. The Cemeteries Adjacent to the Layer Pyramid. Boston: Museum of Fine Art.

Emery, W. B.
1961 Archaic Egypt. London: Hammondsworth.

Esse, D. L.
1989 Secondary State Formation and Collapse in Early Bronze Age Palestine. Pp. 81–96 in L’Urban-

isation de la Palestine à l’âge du Bronze Ancien, ed. P. de Miroschedji. British Archaeological
Reports, International Series 527. Oxford: British Archaeological Reports.

oi.uchicago.edu



441A PRELIMINARY REPORT ON THE 1994 EXCAVATIONS AT NAÓAL TILLAH

Fairservis, W. A., Jr.
1991 A Revised View of the Na’rmer Palette. Journal of the American Research Center in Egypt 27:

1–20.
Fischer, H. G.

1963 A First Dynasty Wine Jar from the Eastern Nile Delta. Journal of the American Research Center
in Egypt 2: 44–47.

Garstang, J.
1907 Excavations at Hierakonpolis, Esna and in Nubia. Annales du Service des Antiquités de l’Égypte

4: 132–48.
Gophna, R.

1992 The Contacts Between the ºEn Besor Oasis, Southern Canaan, and Egypt during the Late Predy-
nastic and the Threshold of the First Dynasty: A Further Assessment. Pp. 385–94 in The Nile
Delta in Transition: 4th–3rd Millennium B.C., ed. Edwin C. M. van den Brink. Tel Aviv: Edwin
C. M. van den Brink.

1995 Early Bronze Age Canaan: Some Spatial and Demographic Observations. Pp. 269–81 in The
Archaeology of Society in the Holy Land, ed. T. E. Levy. New York: Facts-on-File.

Hartung, U.
1994 Bemerkungen zu Chronologie der Beziehungen ägyptens zu Südkanaan in spätprädynastischer

Zeit. Mitteilungen des Deutschen Archäologischen Instituts, Abteilung Kairo 60: 107–14.

Joffe, A. H.
1993 Settlement and Society in the Early Bronze I and II of the Southern Levant: Complementarity

and Contradiction in a Small-Scale Complex Society. Sheffield: Sheffield Academic Press.

Kaiser, W.
1957 Zur inneren Chronologie der Nagadakultur. Archaeologia Geographica 6: 69–77.
1987 Zum Siegel mit frühen Königsnamen von Umm el-Qaab. Mitteilungen des Deutschen Archäol-

ogischen Instituts, Abteilung Kairo 43: 115–19.
1990 Zur Enstehung des gesamtägyptischen Staates. Mitteilungen des Deutschen Archäologischen

Instituts, Abteilung Kairo 46: 287–99.

Kaiser, W., and Dreyer, G.
1982 Umm el-Qaab. Nachuntersuchungen im frühzeitlichen Königsfriedhof. 2. Vorbericht. Mitteil-

ungen des Deutschen Archäologischen Instituts, Abteilung Kairo 38: 211–69.

Kaplony, P.
1963 Die Inschriften der ägyptischen Frühzeit. Ägyptologische Abhandlungen, Band 8, Vols. I–III.

Wiesbaden.
1968 Steingefässe mit Inschriften der Frühzeit und des Alten Reiches. Brussels: Monographies Reine

Elisabeth.

Kroeper, K.
1988 The Excavations of the Munich East Delta Expedition in Minshat Abu Omar. Pp. 11–46 in The

Archaeology of the Nile Delta: Problems and Priorities, ed. E. C. M. van den Brink. Amsterdam:
The Netherlands Foundation for Archaeological Research in Egypt.

1989 Palestinian Ceramic Imports in Pre- and Protohistoric Egypt. Pp. 402–22 in L’Urbanisation de
la Palestine à l’âge du Bronze Ancien, ed. P. de Miroschedji. British Archaeological Reports,
International Series 527. Oxford: British Archaeological Reports.

Levy, T. E.
1992 Radiocarbon Chronology of the Beersheva Culture and Predynastic Egypt. Pp. 345–56 in The

Nile Delta in Transition: 4th–3rd Millennium B.C., ed. Edwin C. M. van den Brink. Tel Aviv:
Edwin C. M. van den Brink.

1995 Cult, Metallurgy and Rank Societies—Chalcolithic Period (ca. 4500–3500 b.c.e.). Pp. 226–44
in The Archaeology of Society in the Holy Land, ed. T. E. Levy. London: Leicester University
Press.

oi.uchicago.edu



442 T. E. LEVY, D. ALON, E. C. M. VAN DEN BRINK, E. KANSA, AND Y. YEKUTIELI

Levy, T. E.; van den Brink, E. C. M.; Goren, Y.; and Alon, D.
1995 New Light on King Narmer and the Protodynastic Egyptian Presence in Canaan. Biblical

Archaeologist 58: 26–35.

Levy, T. E.; Alon, D.; Smith, P.; Yekutieli, Y.; Rowan, Y.; Goldberg, P.; Porat, N.; van den Brink, E. C. M.; Wit-
ten, A. J.; Golden, J.; Grigson, C.; Kansa, E.; Dawson, L.; Holl, A.; Moreno, J.; and Kersel, M.

1997 Egyptian-Canaanite Interaction at Nahal Tillah, Israel (ca. 4500–3000 B.C.E.): An Interim
Report on the 1994–1995 Excavations. Bulletin of the American Schools of Oriental Research
307: 1–51.

Oren, E. A., and Yekutieli, Y.
1992 Taur Ikhbeineh. Earliest Evidence for Egyptian Interconnections. Pp. 361–84 in The Nile Delta

in Transition: 4th–3rd Millennium B.C., ed. Edwin C. M. van den Brink. Tel Aviv: Edwin C. M.
van den Brink.

Orton, C.; Tyres, P.; and Vince, A.
1993 Pottery in Archaeology. Cambridge: Cambridge University Press.

Perrot, J.
1955 Excavations at Tell Abu Matar, near Beersheba. Israel Exploration Journal 5: 167–89.

Petrie, W. M. F.
1900 The Royal Tombs of the First Dynasty. Part I. London: Kegan Paul, Trench, Trubner.
1901 Diospolis Parva. The Cemeteries of Abadiyeh and Hu, 1898–1899. London: Kegan Paul,

Trench, Trubner.
1953 Corpus of Proto-Dynastic Pottery. British School of Archaeology in Egypt 66. London: Egyp-

tian Exploration Fund.

Petrie, W. M. F.; Wainwright, G. A.; and Gardiner, A. H.
1913 Tarkhan I and Memphis. London: British School of Archaeology in Egypt.

Porat, N.
1989 Composition of Pottery. Applications to the Study of Interrelations between Canaan and Egypt

during the Third Millennium b.c. Ph.D. dissertation, Hebrew University of Jerusalem.
Rosen, S. A.

1988 A Preliminary Note on the Egyptian Component of the Chipped Stone Assemblage from Tel
Erani. Israel Exploration Journal 38: 105–16.

Saad, Z. Y.
1951 Royal Excavations at Helwan (1945–1947). Cahier No. 14 Supplément aux Annales du Service

des Antiquités de l’Égypte. Cairo: Imprimerie de L’institut français d’archéologie orientale.
Scharff, A.

1926 Die archäologischen Ergebnisse des vorgeschichtltchen Gräberfeldes von Abusir el-Meleq.
Leipzig: J. C. Hinrichs.

Schaub, R. T., and Rast, W. E.
1989 Bâb edh-Dhrâº: Excavations in the Cemetery Directed by Paul W. Lapp (1965–67). Winona Lake,

IN: Eisenbrauns.
Schott, S.

1951 Hieroglyphen—Untersuchugen zum Ursprung der Schrift. Abhandlungen der Akademie der
Wissenschaften und der Literatur in Mainz. Jahrg., 1950, no. 24. Wiesbaden: Verlag der Wissen-
chaften un der Literatur.

Schulman, A. R.
1976 The Egyptian Seal Impressions from ºEn Besor. ºAtiqot 11: 11–26.

Schulman, A. R., and Gophna, R.
1981 An Archaic Egyptian Serekh from Tel Maªa˙az. Israel Exploration Journal 31: 165–67.

Seger, J. D.
1987a Tel Óalif—1986. Excavations and Surveys in Israel 5: 45–46.
1987b Tel Óalif—1986. Israel Exploration Journal 37:192–195.
1990 Tel Óalif—1987. Excavations and Surveys in Israel 7–8: 69–71.
1991 Tel Óalif—1989. Excavations and Surveys in Israel 9: 67–68.

Seger, J. D.; Baum, B.; Borowski, O.; Cole, D. P.; Forshey, H.; Futato, E.; Jacobs, P. F.; Laustrup, M.; O’Conner-
Seger, P.; and Zeder, M.

1990 The Bronze Age Settlements at Tell Óalif: Phase II Excavations, 1983–1987. Pp. 1–32 in Bulle-
tin of the American Schools of Oriental Research Supplement 26, ed. W. E. Rast. Baltimore:
Johns Hopkins University.

oi.uchicago.edu



443A PRELIMINARY REPORT ON THE 1994 EXCAVATIONS AT NAÓAL TILLAH

Seeher, J.
1990 Maadi—eine prädynastische Kulturgruppe zwischen Oberägypten und Palästina. Praehis-

torische Zeitschrift, Band 65: 123–56.
Tufnell, O., et al.

1958 Lachish IV: The Bronze Age. London: Oxford University Press.
Tutundzic, S.

1993 A Consideration of Differences between the Pottery showing Palestinian Characteristics in the
Maadian and Gerzean Cultures. Journal of Egyptian Archaeology 79: 33–55.

van den Brink, E. C. M., ed.
1992a The Nile Delta in Transition: 4th–3rd Millennium B.C., ed. Edwin C. M. van den Brink. Tel

Aviv: Edwin C. M. van den Brink.
1992b Preliminary Report on the Excavations at Tell Ibrahim Awad, Seasons 1988-1990. Pp. 43-68 in

The Nile Delta in Transition: 4th–3rd Millennium B.C., ed. Edwin C. M. van den Brink. Tel
Aviv: Edwin C. M. van den Brink.

van den Brink, E. C. M.
1996 The Incised Serekh: Serekh-Signs of Dynasties 0–1. Part 1: Complete Vessles. Pp. 140–58 in

Aspects of Early Egypt, ed. J. Spencer. London: British Museum.
von der Way, T.

1992 Excavations at Tell el-Faraª in/Buto in 1987–1989. Pp. 1–10 in The Nile Delta in Transition: 4th–
3rd Millennium B.C., ed. Edwin C. M. van den Brink. Tel Aviv: Edwin C. M. van den Brink.

Wildung, D.
1982 Ägypten vor den Pyramiden. Münchner Ausgrabungen in Ägypten. Mainz: von Zabern.

Wright, G. E.
1937 The Pottery of Palestine from the Earliest Times to the End of the Early Bronze Age. New

Haven: American Schools of Oriental Research.
Yeivin, S.

1960 Early Contacts between Canaan and Egypt. Israel Exploration Journal 10: 193–203.
1961 First Preliminary Report on the Excavations at Tel “Gat” (Tell Sheykh ºAhmed el-Areyney),

Seasons 1956–1958. Jerusalem: Tel Gat Expedition.
1963 Further Evidence of Narmer at “Gat.” Oriens Antiquus 2: 205–13.
1968 Additional Notes on the Early Relations between Canaan and Egypt. Journal of Near Eastern

Studies 27: 37–50.
Yekutieli, Y.

1992 The Early Bronze 1a of Southwestern Canaan-Settlement, Economy and Society. M.A. thesis,
Tel Aviv University (Hebrew).

oi.uchicago.edu



444 T. E. LEVY, D. ALON, E. C. M. VAN DEN BRINK, E. KANSA, AND Y. YEKUTIELI

PLATE 22.1. Overview of Egyptian-style tomb found in Area C, Silo
Site, Óalif Terrace. These two walls form a passageway with an
incline that leads to the tomb entrance.

oi.uchicago.edu



445A PRELIMINARY REPORT ON THE 1994 EXCAVATIONS AT NAÓAL TILLAH

PLATE 22.2. Example of a Late Ware imported Egyptian
vessel found in Area A at the Silo Site, Óalif Terrace.

PLATE 22.3. Series of photographs illustrating the three impressed sides of the clay bulla found in Stratum IIA
at the Silo Site, Óalif Terrace (649/94, L.19, B.266). Plate 22.3c shows the impression of a cord.

oi.uchicago.edu



446A PRELIMINARY REPORT ON THE 1994 EXCAVATIONS AT NAÓAL TILLAH

oi.uchicago.edu



447



ON THE SIGNIFICANCE OF THE
EARLY BRONZE III GRANARY BUILDING

AT BEIT YERA



Ó



Amihai Mazar



In his Ph.D. dissertation (1982), later published as a book (Esse 1991), Douglas Esse
combined archaeological data with environmental, economic, and social factors to recon-
struct the economic basis and social structure of the Early Bronze Age (EB) communities in
northern Palestine. An important part of this research emphasized the place of Beit Yera



˙



in
the settlement system of the period. The present paper is an attempt to elaborate on this sub-
ject, focusing on the Granary Building at Beit Yera



˙



and its implications. The paper is dedi-
cated to the memory of Doug, a close friend, an excellent archaeologist, and an exceptionally
noble person.*

INTRODUCTION

The Granary Building at Beit Yera



˙



(Khirbet Kerak) was uncovered by M. Stekelis and
M. Avi-Yonah in 1945–46 in the uppermost EB stratum (Maisler, Stekelis, and Avi-Yonah
1952: 223–28, pls. 17–19; for an updated description of the site and its exploration, see Esse
1991: 33–53). Though only a preliminary report on its excavation was published, this building
can be used as a source for the reconstruction of socioeconomic and perhaps religious and
political life in the largest EB Age city that has been excavated in Israel. Concerning this
building Esse wrote:



. . . the very existence of a public granary of such enormous size indicates not only town
planning, but also economic and social planning. Such a structure, unmatched so far at
any other EB city, indicates a high level of organization and undoubtedly reflects a
strong central authority. (Esse 1991: 53)



In this paper I deal with the function of the building in light of its plan and contents and try to
reconstruct its role in the economic life of the city.

As observed by its excavators, this building is unique; no similar building is known from
any period in the entire ancient Near East.

THE BUILDING



Description



The brief description of the building that follows is based on the excavators’ preliminary
report and on a study of the 1:50 plan, prepared by I. Dunayevsky and kept in the Institute of



* A Hebrew version of this paper was published in



Shnaton



,



An Annual for Biblical and Ancient Near Eastern
Studies,



ed. M. Weinfeld, 10 (1990): 123–36, Jerusalem.



23



oi.uchicago.edu





448



AMIHAI MAZAR



FIGURE 23.1. Plan of the Granary Building at Beit Yera



˙



.

oi.uchicago.edu

[image: image170.tif]



449



THE EARLY BRONZE III GRANARY BUILDING AT BEIT YERA



Ó



Archaeology of the Hebrew University. The dimensions in the following description were
taken directly from this plan and differ somewhat from those given in the preliminary report.

The outer dimensions of the building are 31.25 (western outer wall) x 41 m. The eastern and
northern facades were not preserved entirely, but they can be reconstructed as measuring
approximately 35 m and 32 m respectively. The area of the building is about 1200 m



2



(fig. 23.1).
It is well constructed, though its outer corners do not form right angles, and it is trapezoid in
plan. The building was surrounded by stone-paved streets, evidence of a sophisticated level of
urban planning. The entrance to the building was from the east, through a corridor 3.3 m wide
and 14 m long. (This length is in fact the width of the eastern wall.) At the end of the corridor
was a threshold built with a single line of rather large stones (seen in Maisler, Stekelis, and
Avi-Yonah 1952: pl. 19A lower left). The corridor leads into an inner courtyard 11 m wide and
6.9 m long (76 m



2



in area). A part of the courtyard near the entrance, and a 2 m wide strip along
the inner (western) part of the courtyard, were paved with pebbles (seen in the above mentioned
photograph).



1



Along the southern wall of the courtyard were two stone constructions: the east-
ern one (1 x 1 m) was interpreted by the excavators as a flight of five steps leading to the top
of the building’s southern wall, which they called a “platform.” In fact, only two steps can be
seen in the plan, and this construction might have been a small podium. The western installation
consists of two large stones. This one could have been a step leading to a short corridor, which
led to the circle to its south.

Three ovens were found in the courtyard: two of them were located, one near the other, just
inside the entrance, yet in such a way that they would not disturb the entrance. The area of the
two ovens was deliberately separated from the entrance to the courtyard by elongated stones.
The eastern of these two ovens was unusual in form: it is described as consisting of two ovens
one inside the other. The larger oven had a 0.5 m wide opening divided by a “column which
was joined by arches to the two side walls.” In the oven was found “a pottery fragment with
four holes, apparently the key-piece of the vaulted roof.” Inside the oven were found fragments
of a large stand of the Khirbet Kerak family (now exhibited in the Israel Museum), as well as
a bowl of the same family. The third oven was located in the southwestern corner of the court-
yard, in a narrow niche between the eastern stone steps and the front of the enclosed hall.

In the western part of the courtyard, a wide entrance (2.1 m) led to a broad hall with inner
dimensions of 4.4 x 11.2 m. Two unworked stone slabs found along the long axis of the hall
were probably bases for wooden pillars that supported the roof. The preliminary report
describes a pebble floor in this hall, but such a floor is not shown in the detailed plan (though
the pebbles in the courtyard are definitely shown). The two circles at the western corners of
the building could be approached from this hall through 0.8 m wide diagonal corridors,
entered exactly at the southwestern and northwestern corners of the hall.

The outer stone walls of the building are about 10 m wide (except the eastern facade which
is 14 m wide). The tops of the walls are leveled. These wide walls contained sunken circles 7–
9 m in diameter. (Out of six measurable circles, four are 8 m in diameter: one is 9 m, and one
is 7 m.) The excavation revealed the remains of eight circles; a ninth can be reconstructed in
the unexcavated corner of the building. The stone floor of the circles is sunken about 10 cm



1. The picture is wrongly explained in Maisler, Stekelis, and Avi-Yonah 1952: 227 as showing the eastern side
of the courtyard and the entrance to the enclosed hall at its east (pointed out to me by P. de Miroschedji).

oi.uchicago.edu





450



AMIHAI MAZAR



below the top of the walls. Each circle is divided by four narrow partitions that do not reach
the center of the circle. These walls are oriented to the points of the compass. Narrow passages
led from two of the circles into the hall in the western part of the building.

The building is generally well planned and constructed with great care. The circles are exact,
and the partition walls inside each of them are oriented exactly to the points of the compass, as
if deliberately planned.



Reconstruction



Several attempts to reconstruct this building show each circle as a foundation for an indi-
vidual cylindrical or cone-shaped silo. (Mellaart 1966: 75, fig. 28, followed by Kempinski
1992: 77, fig. 10, show free-standing conical silos, while Kempinski 1978: fig. 9, reconstructs
free-standing cylindrical silos. For a general discussion of the “beehive granaries” of Pales-
tine, see Currid 1985.) These reconstructions were probably inspired from the Egyptian depic-
tions of free-standing conical silos in what may be defined as “beehive granaries” (Currid
1985: 105–6, figs. 2–5). Yet in these reconstructions, the role of the wide, solid stone foun-
dation walls in which the circles were embedded (denoted by the excavators as a “platform”)
remains obscure. Currid claims that the “the thick foundation wall certainly served to prevent
rodent intrusion, above-ground water seepage, and the influx of moisture from the ground”
(Currid 1985: 104, n. 37). This certainly explains the very existence of a raised solid stone
foundation. Yet this does not necessarily mean that the foundations did not carry a superstruc-
ture over the entire area. In my opinion, these wide walls with flattened upper levels were the
foundations for a massive, solid mudbrick superstructure which covered the entire space
between the circles, so that from the outside the building was seen as a massive structure.
Each individual silo was probably cylindrical with straight sides and a flat or domed ceiling.
The massive mudbrick superstructure would give the individual silos a solid outer frame. The
four piers in each circle must have supported the ceiling. The upper part of each silo could
have protruded above the main part of the superstructure and might have had a domed top, as
is shown in cylinder seal impressions from Susa, dating to the “Proto-Urbaine” period (late
fourth millennium



b.c.



) (fig. 23.2). There probably was an opening in the ceiling for filling the
granary, and the grain was taken out through openings at the base at the structure, as shown in



FIGURE 23.2. Susa cylinder seal impression showing granaries (adapted from Collon 1987: 146, no. 622).

long

oi.uchicago.edu

[image: image171.tif]



451



THE EARLY BRONZE III GRANARY BUILDING AT BEIT YERA



Ó



Egyptian depictions of silos (Currid 1985: 105, fig. 2). The corridors found leading to two of
the circles probably enabled access to these openings through the solid superstructure that
covered the entire stone foundation around the circular silos. This reconstruction suggests that
the granaries rose to a considerable height, which might have been equal to or slightly below



FIGURE 23.4. Early Helladic model granary from the island of Melos.

FIGURE 23.3. Suggested reconstruction of the Granary Building at Beit Yera



˙



(drawing by Anna Yamim).

oi.uchicago.edu

[image: image172.tif]



452



AMIHAI MAZAR



the diameter of the circles. A 6 to 8 m height for each of the silos is a reasonable estimate.
(Compare the average height of 5 m for the 2–3 m diameter of Egyptian silos; Currid 1985:
104.) The thick, solid walls would have been able to withstand the pressure of the grain. (At
the narrowest point the width of the granary wall was 1 m.) These considerations stand at the
basis of the reconstruction shown on figure 23.3.



The Finds



A destruction layer which covered the floor of the courtyard contained many pottery sherds,
including a large amount belonging to the Khirbet Kerak family. Many “broken and blackened
bones of animals” are reported as having been found “in several loci, especially in the open
court near the ovens, together with carbonized olive kernels, soot and other traces of fire.”

The finds from this building were never published in detail, and I have been unable to
locate any field diaries or other records.

Two unusual cult objects were published in the preliminary report (Maisler, Stekelis, and
Avi-Yonah 1952: 227, pl. 19:B). The first (fig. 23.5, pl. 23.1b) is a fragment of a thick clay ring
that bears two clay bull heads modeled in three dimensions. This object has been reexamined
by Amiran (1989: 31–33).



2



The second object (fig. 23.6, pl. 23.1) is a fragment of a zoomorphic libation vessel.



3



It is
hollow, and the mouth of the animal served as a spout. The vessel was made of coarse ware,
with plenty of large grits. The fragment is 10 cm long and 9.2 cm high; since it depicts about



2. I would like to thank Ruth Amiran for permission to republish the drawing of this object here. It is now
exhibited in the collections of Beit Gordon at Kibbutz Deganya.

3. The object was found in the storage area of the Rockfeller Museum, among other material from the Beit
Yera



˙



excavations. My thanks to J. Zias for locating it. The photograph was supplied by the Israel Antiqui-
ties Authority and the drawing prepared by Mrs. S. Helbreich.

FIGURE 23.5. Pottery ring with two bull figurines found in the Granary Building at Beit Yera



˙



.

oi.uchicago.edu

[image: image173.tif]



453



THE EARLY BRONZE III GRANARY BUILDING AT BEIT YERA



Ó



half of the animal, the complete vessel was probably about 20 cm long. The animal can be
identified as a lion, as shown by the shape of the head, the paws, and the delicate incisions on
the head which probably represent the mane. The ears are broken. One might suggest that the
broken ears were in fact horns, and thus interpret the animal as a bull. A prominent vertical
ridge down the center of the animal’s chest may strengthen this interpretation. Nevertheless,
it seems to me that the fragments at the front of the head were originally ears, and that this ani-
mal is indeed a lion. The back of the vessel has not been preserved. The entire preserved part
was painted with red stripes in a crisscross pattern. The piece is particularly important, since
it is the sole EB II–III zoomorphic vessel known from the Land of Israel. Depictions of lions
are well known in third millennium



b.c.



art of the ancient Near East, such as on Meso-
potamian cylinder seals depicting lions attacking domestic animals or in representations of
lions in wood carvings and seal impressions from Ebla (Matthiae 1981: pls. between pp. 80–
81). In the Land of Israel, lions are a familiar motif on cylinder seals from EB (Ben-Tor



FIGURE 23.6. Lion-shaped libation vessel found in the Granary Building at Beit Yera



˙



.

oi.uchicago.edu

[image: image174.tif]



454



AMIHAI MAZAR



1978: 52, fig. 15; 72–73, fig. 21). The lion motif became more common in Canaanite art in the
second millennium



b.c.



, often appearing in a ritual context.
Another object that probably came from the Granary Building is a figurine of an animal

(sheep?), of a type found at other EB sites (pl. 23.1c, right).



Function



The excavators made several suggestions as to the function of the building: a public gra-
nary, a fort, a palace, or a temple. They settled on the last one. The suggestion that the building
might be a granary was discounted in the original Hebrew publication because an oven had
been found in one of the sunken circles (Stekelis and Avi-Yonah 1947: 60), though in fact no
oven was found in the circles themselves. The unique oven in the building’s courtyard was
interpreted by the excavators as an installation for burning incense. In the English version of
the publication, Avi-Yonah added a note (Maisler, Stekelis, and Avi-Yonah 1952: 228 n. 12)
in which he presented a parallel to the building in the steatite model structure from the island
of Melos (fig. 23.4), which dates to the Early Helladic period (Marinatos 1946: 342, fig. 4;
Buchholz and Karageorghis 1973: 90, no. 1122). On the basis of the Melos model, which was
interpreted as representing a public granary, Avi-Yonah wrote:



This find increases the probability that the building was a granary after all and the ovens
a later addition, once it went out of use. If this hypothesis is accepted, it bears striking
evidence of the agricultural development and social structure of the EB settlements in the
Jordan Valley. It presumes a plentiful grain crop stored in a public building connected
with either a religious or secular authority.



The relevance of the Melos example was reinforced by the data on links between the
Aegean world and the Land of Israel in the EB III period (Hennessy 1967: 82–83). Kilian sug-
gested a resemblance between the Early Helladic II round building at Tiryns which he inter-
preted as a granary and the building at Beit Yera



˙



(Kilian 1986: 68), yet this is a very different
building, much reconstructed, and its function is still elusive.



4



The identification of the struc-
ture at Beit Yera



˙



as a granary was generally accepted in the archaeological literature. R. Ami-
ran noted parallels to third-millennium circular silos in northern Anatolia and Trans-Caucasia,
at the sites of Yanik Tepe, Shengavit, and elsewhere (Amiran 1965: 165–67). However, in
these examples, the round silos appear side by side and do not form part of a single building.
Some of them exhibit internal partitions, formed by four piers which do not meet in the center
of the circle, as in the Beit Yera



˙



building. In light of the origin of Khirbet Kerak ware in the
Kura-Araxis culture, these parallels seem to be relevant, despite the fundamental difference in
plan mentioned above.

Several scholars claimed that the “Granary Building” was a temple (Mellaart 1966: 74–76;
Amiran 1989: 32; Busink 1970: 379–80; Wright 1985: 218). In my opinion, these two expla-
nations do not need to contradict one another: the building can be interpreted as a temple, with
the granaries serving an economic role. The single large hall inside the building is a “broad-
house” structure reminiscent of the acropolis temple at Ai and the “White Building” at Tel
Yarmuth. Both of these structures have been identified as EB III temples, contemporary with



4. My thanks to Sharon Zuckerman for this reference.

oi.uchicago.edu





455



THE EARLY BRONZE III GRANARY BUILDING AT BEIT YERA



Ó



the Beit Yera



˙



structure (Miroschedji 1988: 35–41; Kempinski 1992: 53–59). The inner
dimensions of the broadhouse at Yarmuth (4.75 x 11.50 m) are almost identical to those of the
hall inside the “Granary Building,” though the Yarmuth example has four column bases along
the hall’s longitudinal axis, while only two were found at Beit Yera



˙



. The hall at Ai is larger
(inner dimensions 6.00 x 17.50 m). Remains of a dais were found opposite the entrance to the
hall at Yarmuth, but no traces of this kind were found at Ai or at Beit Yera



˙



. The unusual cult
objects found at the Beit Yera



˙



building, which are rare in the EB, and the unique oven con-
taining a large pottery stand found in the inner courtyard, reinforce the possibility that this
building was indeed a temple. However, two points weaken our argument that the inner hall
was a sanctuary: (a) No focal point of cult was found, such as the evidence for a dais found at
Yarmuth opposite the entrance; and (b) the approach to three of the circles could only be from
inside this inner hall. This can hardly be accommodated with the interpretation of the hall as a
sanctuary. Despite these difficulties, I would not exclude the possibility that the inner hall was
a sanctuary. Moreover, it may be suggested that the model structure from Melos represented a
temple as well since it would otherwise be difficult to account for its facade, which has a ped-
iment covered with rich spiral ornamentation. It thus may be suggested that both Beit Yera



˙



and Melos possessed temples with attached granaries, which were part of the temple’s role as
an economic power.

SOCIOECONOMIC INTERPRETATION

In the following section I attempt to widen the discussion of the Beit Yera



˙



structure’s
function into its significance for understanding the city’s socioeconomic system. With this
goal in mind, the following questions are considered:

1. What was its potential grain storage capacity?
2. What as the relationship between the amount of grain stored in the structure and the

city’s available agricultural land?
3. What was the relationship between the amount of grain stored in the building and the

city’s population?
4. What was the nature of the agricultural, socioeconomic, and organizational systems in

which the “Granary Building” functioned?

Clear answers to these questions remain elusive. Nevertheless, partial answers can be pro-
vided and several possible hypotheses raised.

(1) The basic hypothesis that the stone circles served as foundations of granaries should
be accepted. Their average diameter of 8 m implies that their superstructures were quite high,
between 6 and 8 m (see above). For the purpose of calculating the volume of the granaries, I
assume that they were 6 m high. This is in my view a minimal height for a round structure
with a diameter of 8–9 m. Obviously, any alteration in this estimated height will affect the fol-
lowing calculations considerably. The average floor area of each granary, minus the area taken
up by the inner piers, is 42 m



2



. The volume of an average granary would thus be 250 m



3



. The
entire Granary Building would thus have been able to store about 2,250 m



3



of grain. The
weight of 1 m



3



of grain can be calculated according to various coefficients. The most common

oi.uchicago.edu





456



AMIHAI MAZAR



for wheat is 770 kg per 1 m



3



(see references in Rosen 1986: 172; Schwartz 1994: 26, table 2),
while the coefficients for barley differ from 610 kg per 1 m



3



(Rosen 1986: 172) to 714 kg (ref-
erences in Schwartz 1994: 26, table 2). Thus if the building was fully utilized, its maximal
capacity was somewhat above 1,700 tons of wheat and between 1,370 and 1,600 tons of bar-
ley. This capacity of the building exceeds earlier estimates, such as the estimation of 800 tons
brought by Kempinksy’s (1978: 29, cited by Esse 1991: 100).

(2) What was the relationship between the building’s storage capacity and the city’s agri-
cultural potential? It can be assumed that the Granary Building was intended to store the max-
imum harvest possible, and that in bad or mediocre years the granary was not completely
filled. A hectare (10 dunams or 2.5 acres) of land worked by traditional methods in the Levant
yields between 300 and 1,000 kg of wheat, and the average in Arab peasant agriculture is
about 700 kg (Avitzur 1977: 44–63; Rosen 1986: 172; Rosen uses an average of 650 kg per
hectare of wheat and 800 kg per hectare of barley).



5



In the first year that Kibbutz Deganya was
settled just south of Beit Yera



˙



, its land yielded 830 kg of grain per hectare (Avitzur 1977).
This figure rose to 1,180 kg of wheat and 1,500 kg of barley per hectare in subsequent years.
These numbers are in accord with figures based on Sumerian documents: 1,254 kg of grain per
hectare in pre-Sargonic Girsu and 700 kg in Ur III texts (Adams 1981: 86, 146; Schwartz
1994: 28). In the Baqºah Valley of Lebanon, where conditions are much more similar to the
Jordan Valley, the estimation is just 300 kg per hectare (Marfoe 1979: 5). It may be suggested
that harvests in the EB were on the average not much smaller than those obtained from Arab
peasant agriculture. We can therefore calculate that in order to fill the Granary Building with
1700 tons of wheat, an area of about 2,430 hectares (6,075 acres) must have been under culti-
vation (based on an average yield of 700 kg per hectare). Beit Yera



˙



’s best agricultural lands
lie mainly to the south, in a triangle bounded by the Yarmuk River, the Jordan River, and the
Sea of Galilee up to a distance of about 8 km from the city. This is the farthest that farmers
would have been likely to cover on foot to reach their fields. This triangle, known as the Kin-
rot Valley, covers roughly 2,700 hectares (6,750 acres). This area comprises the northern end
of the Irano-Turanian ecosystem of the Jordan Valley, which enjoys an average rainfall of
300–400 mm. This allows grain to be grown with maximum yields in dry farming. No other
EB sites are known from this area and its periphery, except for Tell el-Hama (Hamat Gader),
which extends over about 1 hectare and whose inhabitants probably worked some of the same
land (see the distribution map in Esse 1991: 149, fig. 29). The closest EB site to the south is
Tel Yaqush (excavated by Esse), which is far beyond the potential lands of Beit Yera



˙



. To the
north, along the Sea of Galilee, there was another EB city at Tel Raqat (Khirbet Quneitra,
north of Tiberias), but its people would never have competed with the population of Beit
Yera



˙



for land in the Jordan Valley south of the Sea of Galilee. It thus may be assumed that
the agricultural land in the triangle defined above was cultivated by the inhabitants of Beit
Yera



˙



itself and not by the people from villages or satellite towns in the area (fig. 23.7). The
people of Beit Yera



˙



could also cultivate land on the Yavneel plateau, just west of Beit Yera



˙



,
though these were less accessible lands.



5. I thank Baruch Rosen for the time and effort he spent in discussing this subject with me. In the following
calculations I utilized Rosen’s methodology in analyzing the ºIzbet Sartah silos (Rosen 1986).



short

oi.uchicago.edu





457



THE EARLY BRONZE III GRANARY BUILDING AT BEIT YERA



Ó



FIGURE 23.7. Map showing the agricultural hinterland of Beit Yera



˙



.

oi.uchicago.edu

[image: image175.tif]



458



AMIHAI MAZAR



(3) In order to answer our other questions, I make use of data and methods employed by
B. Rosen and M. Broshi in their research on similar topics (Broshi 1979, 1986; Rosen 1986).
The working hypothesis is that data collected in anthropological research carried out in a similar
environment at a similar technological level (the Near East, especially late nineteenth-century
Palestine) are valid for earlier periods.

The area of the city of Beit Yera



˙



is estimated at 20 hectares (50 acres). Studies in ancient
and current traditional demography led to various estimates of density coefficients: from 100
persons per built-up hectare to 200–250 people per built-up hectare in the EB sites (as well as
other periods) in the Land of Israel. A figure of 200 or 250 persons was utilized for the Bronze
Age and the Iron Age (Marfoe 1979: 21; Broshi and Gophna 1984). Esse pointed out the dif-
ficulties in using a fixed constant of a number of people per hectare, due to the differences in
types of sites and in their settlement density (Esse 1991: 130–35). However, Beit Yera



˙



was a
densely built city in the EB III, and thus we utilize the 200 people per hectare factor and esti-
mate that Beit Yera



˙



’s population would amount to about 4,000 people (the high factor of 250
people per hectare would give a figure of 5,000). For the sake of simplifying the following cal-
culations, we assume that the Granary Building was used for storing wheat alone; if barley
was also stored there, the calculations would be only slightly affected.



6



The average annual consumption of wheat in wheat-based cultures in general, and in the
traditional economy of the Near East in particular, can be estimated as 150–250 kg per indi-
vidual; the higher figure (200 kg) is more generally accepted. Higher numbers (as much as
around 500 kg per person) are also cited (Rosen 1986:173; Schwartz 1994: 26, table 2, where
ten different estimations are cited, the highest being 558 kg per year and the lowest 143 kg per
year). Obviously, in calculating consumption, it has been taken into account that while wheat
(and perhaps barley) was the basis of the population’s food, they did have access to other
sources of nutrition (legumes, various fruits and vegetables, fish, dairy products, beef, and
mutton). A population of about 4,000 would thus consume about 800 tons of wheat per year.
A loss of about 30% during storage should be taken into account, caused by the depredations
of rodents and the need to provide seeds for the next year’s crop. Higher figures for this loss,
as much as 50%, are also cited (see Rosen 1986: 173; Schwartz 1994: 27, table 2).

Fodder for livestock has not been taken into account in these calculations since it is
assumed that cheaper forms of food and straw were used for this purpose and were not stored
in the central Granary Building. Broshi reckoned 300 kg of wheat per individual per annum,
including loss, seed corn, and animal fodder. In order to provide 800 tons of grain, it would
thus be necessary to store about 1,040 tons. This amount could easily have been stored in the
Granary Building, which according to my calculations, could store up to 1,700 tons of grain.
It is thus obvious that the building could store far more grain than the yearly consumption of
the entire city’s population.

This can be explained in several different ways. It may be that the building was designed
to store bumper harvests from especially abundant years (as Rosen suggested for the silos at
ºIzbet Sartah), while in other years the building’s maximum capacity was not used. Thus in a
bumper year, when the granaries were full, it was possible to store grain in preparation for



6. Barley was the principal crop grown in the third millennium



b.c.



in Mesopotamia and was used for making
bread and brewing beer (Jacobsen 1982: 38–51).

oi.uchicago.edu





459



THE EARLY BRONZE III GRANARY BUILDING AT BEIT YERA



Ó



poor years—since grain can be stored in this way for three to four years (though with an
increasing rate of loss). The building could thus have served to even out the food supply for
the city’s inhabitants over good and bad years. Another possibility is that the surplus grain
was used for trade with other cities or regions in the country, whose economies were based on
different crops or products.

One of the most likely possibilities is that the people of Beit Yera



˙



traded with pastoralists
living on the Golan and Issachar Heights. The EB III sites (so-called enclosures) found in
these areas point to the existence of a permanent population whose economy was based on
animal husbandry (cattle, sheep, and goat) as well as horticulture (Kochavi 1989: 21–25; con-
tra Esse 1991: 156–62, who raised the hypothesis of a dimorphic society in the Land of Israel
in the EB III, with reciprocal relations between the settled populations of the Jordan and Hula
Valleys and pastoralists in the Golan and the Galilee).

It is also probable that the inhabitants of Beit Yera



˙



traded with the people of the central
highlands of western Palestine, where horticulture was the basis of the economy (Stager 1985;
Esse 1991: 100–102). Esse raised the question whether the granary building reflects a redis-
tributive or market exchange economic system (Esse 1991: 100). A lack of written records
prevents, in his view, a definite conclusion on this point. Yet he describes Beit Yera



˙



as a cen-
tral site in northern Palestine: “The diversity of its resource base and the practice of successful
mixed economy” (p. 100) were the basis for its economic power, as well as its



. . . location straddling the border between two major eco-zones: the Mediterranean and
the Irano-Turanian steppe . . . although Beth Yera



˙



surely participated in interregional
and international trade, it is the level of intraregional trade that the city’s existence comes
into focus most sharply . . . the city became a true focal point, the nexus between two
very diverse ecological zones with different potentials. (Esse 1991: 100)



Esse thus defined the Beit Yera



˙



granary as a center of interregional trade, a redistribution
center, which is expected in a border zone between two different ecological regions.

The number of people occupied in growing this amount of grain can be very roughly esti-
mated. In traditional agriculture in the Land of Israel, a peasant (



fellah



) family of about five
people with their animals could cultivate an area of about 8–20 hectares (20–50 acres) of dry
farming grain crops; the average area per family was 12–15 hectares (30–37 acres).



7



This area
supplied a family with its agricultural produce. It is difficult to establish how many people
were needed to cultivate grain with the technological methods and the sociopolitical organiza-
tion of EB, and we do not know the division of labor in the city: Was there a degree of “spe-
cialization” and did only some inhabitants work on grain cultivation; or was it the main
occupation of most of the population?

The data presented above indicate that, in order to cultivate the 2,430 hectares necessary
to provide enough grain to fill the Granary Building, about 200 family units would be
required, or about 1,000 people. This number gives a very general idea of a 1:4 ratio between
the families who were involved in the cultivation of the amount of grain that could be stored
in the Granary Building as against the entire population in the city. Other inhabitants of the



7. “The agricultural unit necessary to provide for a peasant family with the technical ability to farm was cal-
culated at a



fadan



—the ability to work with a pair of oxen. It was usually of an area of 12 to 15 hectares.
The



fadan



provided for the family at a fairly low level of subsistence . . .” (Avitzur 1977: 36–38).

oi.uchicago.edu





460



AMIHAI MAZAR



city (about three-quarters of the families) would be occupied with other agricultural activities,
such as growing olives for the olive oil production, legumes, etcetera. There was also fishing
in the Lake of Galilee, and perhaps also cattle, sheep, and goat breeding (Esse 1991: 98–100).
Others would be professional craftsmen, military men, civil servants, priests, and so on.

If it is assumed that part of the grain needs of the inhabitants was grown on private allot-
ments and not stored in the public Granary Building, then we may conclude that some of the
grain kept in the building was used for trade even in years of average harvests. The territory
calculated as necessary to provide the amount of grain stored in the Granary Building formed
the greater part of the land within a reasonable walking distance of the city (the triangle
bounded by the Jordan and Yarmuk Rivers, as described above). According to our calcula-
tions, about 400 hectares in this area would have been left over for harvests other than grain,
or for producing grain that was not stored in the public granary. This would have been too
small an area to supply the needs of those inhabitants who were not linked to the economic
system represented by the Granary Building. It is therefore possible that some of the Beit
Yera



˙



agricultural lands were in more distant plots, for example in the Yavneel Valley to the
west or south of the confluence of the Yarmuk and the Jordan Rivers.

(4) The above calculations are evidence for a sophisticated system of centralized admin-
istration and careful planning of the agricultural-economic system at Beit Yera



˙



. The latter
entailed the concentration of agricultural produce and its redistribution to all or a large part of
the city’s inhabitants, and most probably its use in intraregional trade (with other cities or
regions in the country). Long-term storage of grain was probably also planned, in order to
overcome the effects of drought years. It may be assumed that the authority which erected and
managed the Granary Building had considerable political and economic power. But we have
no way of knowing how the ownership of land and agricultural labor in the city were orga-
nized. Did the city possess a “specialized economy” with only a quarter of its inhabitants
engaged in grain cultivation, while the rest carried on other forms of cultivation, fishing, herd-
ing, and various craft activities? Or were the different forms of agricultural work shared
equally among all the city’s inhabitants? Did the authority that owned the Granary Building
also possess the land on which the grain stored there was grown, with the agricultural workers
as serfs? Or did the building store grain that had been grown on privately owned allotments
and collected by a powerful government, either by legal means or by force of religious tradi-
tion? If the suggestion that the core of the building was a temple is accepted, it is possible that
most of the city’s land was controlled by this temple. This phenomenon is reminiscent of the
situation in Sumerian cities, where the temple had enormous economic power. However, even
in the case of Sumer, where we possess written sources, scholars disagree over the extent of
temple control of the city lands (Kramer 1963: 73–77; Gadd 1971: 126–31).



8



At Ebla, which
is the chronologically and geographically closest archive to the period of the Granary Building
at Beit Yera



˙



, economic power seems to have been in the hands of a secular authority—the
king of the city (Matthiae 1981: 163–86).



8. According to documents from the temple of Baba at Lagash, tenants and hired laborers worked the temple
lands at Lagash in the Early Dynastic period. The grain was stored in the temple granaries and was also
ground there. The salaries of the temple staff and its various suppliers were paid mostly in grain. For detailed
information on grain cultivation in Mesopotamia at different periods, see Jacobsen 1982. However, these data
concern irrigation agriculture, which is very different from the dry agriculture of the Land of Israel.

oi.uchicago.edu





461



THE EARLY BRONZE III GRANARY BUILDING AT BEIT YERA



Ó



A more detailed reconstruction of the socioeconomic system related to the Granary
Building would be totally speculative. Yet the building provides evidence for a complex eco-
nomic system that existed in the Land of Israel at the zenith of the EB urban culture. Features
such as specialization, advanced architectural planning on a large scale, concentration of food
resources and their redistribution according to need and ability, long-term food storage, and
interregional trade characterized this period and must have necessitated a complex bureau-
cratic system.



9



Whether this system was secular or related to temple administration is impos-
sible to know.

The “Granary Building” at Beit Yera



˙



was found near the surface and definitely belongs
to the latest stage of the EB city’s existence. Thus the city of Beit Yera



˙



did not show any evi-
dence for decline toward the end of the period of its existence.



10



Like at other EB III cities, it
appears that the end of the city occurred at a time when the EB III urban system in Palestine
was at its zenith. The demise came abruptly; however, its causes allude us.



11



Despite the high degree of speculation and uncertainty in the above discussion, it seems to
me that the Granary Building at Beit Yera



˙



can shed light on the socioeconomic structure of
large city-states in the Land of Israel on the eve of the demise of urban culture at the conclu-
sion of the EB III period.



9. The high degree of hierarchy and central authority in the EB III period is now exemplified superbly by the
well-planned palace and other public buildings and fortifications at Tel Yarmuth (Miroschedji 1999: 2–19).

10. The urban architecture of EB III at Beit Yera



˙



can be seen in the complex plans of structures there, like that
published by Eisenberg (Eisenberg 1981: 11–13).

11. Various suggestions were published concerning the reasons for the end of the urban culture at the end of the
EB III period. I would add that a plague should be taken seriously as a possible reason for the demise of this
culture when it was at its floruit. Plagues were one of the major reasons for abrupt changes in human history.
The abrupt end of the EB III culture hints in my view to such a cause.



REFERENCES



Adams, R. M.
1981



Heartland of Cities.



Chicago: University of Chicago Press.

Amiran, R.
1965 Yanik Tepe, Shangavit and the Khirbet Kerak Ware.



Anatolian Studies



15: 165–66.
1989 Re-Examination of a Cult and Art Object from Beth-Yera



˙



. Pp. 31–33 in



Essays in Ancient Civi-
lization Presented to Helene J. Kantor,



eds. A. Leonard and B. B. Williams. Studies in Ancient
Oriental Civilization 47. Chicago: Oriental Institute.

Avitzur, S.
1977



Changes in the Agriculture of the Land of Israel 1875–1975.



Tel Aviv: Milo (Hebrew).

Ben-Tor, A.
1978



Cylinder Seals of Third Millennium Palestine.



Bulletin of the American Schools of Oriental
Research Supplement Series 22. Cambridge, MA: American Schools of Oriental Research.

Broshi, M.
1979 The Population of Western Palestine during the Roman-Byzantine Period.



Bulletin of the Ameri-
can Schools of Oriental Research 236: 1–10.

1986 The Carrying Capacity of the Land of Israel in the Byzantine Period and Its Demographic Sig-
nificance. Pp. 49–56 in Man and Land in the Ancient Land of Israel, eds. A. Oppenheimer,
A. Kasher, and A. Rappaport. Jerusalem: Yad Ben Zvi Institute (Hebrew).

oi.uchicago.edu



462 AMIHAI MAZAR

Broshi, M., and Gophna, R.
1984 The Settlement and Population of Palestine during the Early Bronze Age II–III. Bulletin of the

American Schools of Oriental Research 253: 41–53.

Buchholz, H. G., and Karageorghis, V.
1973 Prehistoric Greece and Cyprus. New York: Phaidon.

Busink, Th.
1970 Der Tempel von Jerusalem: von Salomo bis Herodes, Der Tempel Salomos I. Leiden: Phaidon.

Collon, D.
1987 First Impressions. London: British Museum.

Currid, J.
1985 The Beehive Granaries of Ancient Palestine. Zeitschrift des Deutschen Palästina-Vereins 101:

97–109.

Eisenberg, E.
1981 Tel Bet Yera˙. Hadashot Arkheªologiyot 76: 11–13 (Hebrew).

Esse, D. L.
1991 Subsistence, Trade, and Social Change in Early Bronze Palestine. Studies in Ancient Oriental

Civilization 50. Chicago: Oriental Institute.

Gadd, C. J.
1971 The Cities of Babylonia. Pp. 93–144 in Cambridge Ancient History, Vol. 1, pt. 2, 3rd ed.

Cambridge: Cambridge University Press.

Hennessy, B.
1967 The Foreign Relations of Palestine during the Early Bronze Age. London: Quaritch.

Jacobsen, Th.
1982 Salinity and Irrigation Agriculture in Antiquity. Bibliotheca Mesopotamica 14. Malibu: Undena.

Kempinski, A.
1978 The Rise of an Urban Culture. Israel Ethnographic Society Studies 4. Jerusalem: Israel Ethno-

graphic Society.
1992 Fortifications, Public Buildings and Town Planning in the Early Bronze Age. In The Architec-

ture of Ancient Israel, eds. A. Kempinski and R. Reich. Jerusalem: Israel Exploration Society.

Kilian, K.
1986 The Circular Building at Tiryns. Pp. 65–71 in Early Helladic Architecture and Urbanization,

eds. R. Hagg and D. Konsola. Studies in Mediterranean Archaeology 76. Göteborg: Paul Aström
Verlag.

Kochavi, M.
1989 The Land of Geshur—Regional Archaeology on the Southern Golan in the Biblical Period.

Qadmoniot 22: 21–26 (Hebrew).

Kramer, S. N.
1963 The Sumerians. Chicago: University of Chicago Press.

Maisler, B.; Stekelis, M.; and Avi-Yonah, M.
1952 The Excavations at Beth Yera˙ (Khirbet Kerak) 1944–1946. Israel Exploration Journal 2:

218–29.

Marfoe, L.
1979 The Integrative Transformation: Patterns of Sociopolitical Organization in Southern Syria. Bul-

letin of the American Schools of Oriental Research 234: 1–42.

Marinatos, S.
1946 Greniers de l’Helladique Ancien. Bulletin de Correspondance Hellenique 7: 337–51.

Matthiae, P.
1981 Ebla: An Empire Rediscovered. London: Hodder and Stoughton.

Mellaart, J.
1966 The Chalcolithic and Early Bronze Age in the Levant and Anatolia. Beirut: Khayats.

oi.uchicago.edu



463THE EARLY BRONZE III GRANARY BUILDING AT BEIT YERAÓ

Miroschedji, P. de
1988 Yarmuth 1. Paris.
1999 Yarmouth: The Dawn of City-States in Southern Canaan. Near Eastern Archaeology 62: 2–19.

Rosen, B.
1986 Subsistence Economy of Stratum II. Pp. 156–185 in ºIzbet Sartah, ed. I. Finkelstein. British

Archaeological Reports, International Series 299. Oxford: British Archaeological Reports.

Schwartz, G. M.
1994 Rural Economic Specialization and Early Urbanization in the Khabur Valley, Syria. Pp. 19–36

in Archaeological Views from the Countryside, eds. G. M. Schwartz and S. E. Falconer. Wash-
ington, DC: Smithsonian Institution.

Stager, L. E.
1985 The First Fruits of Civilization. Pp. 172–88 in Palestine in the Bronze and Iron Ages: Papers in

Honour of Olga Tufnell, ed. J. N. Tubb. London: Institute of Archaeology.

Stekelis, M., and Avi-Yonah, M.
1947 The Excavations at Beth Yera˙, Second Preliminary Report. Bulletin of the Jewish Palestine

Exploration Society 13: 53–60 (Hebrew).

Wright, G. R. H.
1985 Ancient Building in Syria and Palestine. Leiden: E. J. Brill.

oi.uchicago.edu



464 AMIHAI MAZAR

PLATE 23.1. Cult objects from the Granary Building at Beit Yera˙.

oi.uchicago.edu



447



ON THE SIGNIFICANCE OF THE
EARLY BRONZE III GRANARY BUILDING

AT BEIT YERA



Ó



Amihai Mazar



In his Ph.D. dissertation (1982), later published as a book (Esse 1991), Douglas Esse
combined archaeological data with environmental, economic, and social factors to recon-
struct the economic basis and social structure of the Early Bronze Age (EB) communities in
northern Palestine. An important part of this research emphasized the place of Beit Yera



˙



in
the settlement system of the period. The present paper is an attempt to elaborate on this sub-
ject, focusing on the Granary Building at Beit Yera



˙



and its implications. The paper is dedi-
cated to the memory of Doug, a close friend, an excellent archaeologist, and an exceptionally
noble person.*

INTRODUCTION

The Granary Building at Beit Yera



˙



(Khirbet Kerak) was uncovered by M. Stekelis and
M. Avi-Yonah in 1945–46 in the uppermost EB stratum (Maisler, Stekelis, and Avi-Yonah
1952: 223–28, pls. 17–19; for an updated description of the site and its exploration, see Esse
1991: 33–53). Though only a preliminary report on its excavation was published, this building
can be used as a source for the reconstruction of socioeconomic and perhaps religious and
political life in the largest EB Age city that has been excavated in Israel. Concerning this
building Esse wrote:



. . . the very existence of a public granary of such enormous size indicates not only town
planning, but also economic and social planning. Such a structure, unmatched so far at
any other EB city, indicates a high level of organization and undoubtedly reflects a
strong central authority. (Esse 1991: 53)



In this paper I deal with the function of the building in light of its plan and contents and try to
reconstruct its role in the economic life of the city.

As observed by its excavators, this building is unique; no similar building is known from
any period in the entire ancient Near East.

THE BUILDING



Description



The brief description of the building that follows is based on the excavators’ preliminary
report and on a study of the 1:50 plan, prepared by I. Dunayevsky and kept in the Institute of



* A Hebrew version of this paper was published in



Shnaton



,



An Annual for Biblical and Ancient Near Eastern
Studies,



ed. M. Weinfeld, 10 (1990): 123–36, Jerusalem.



23



oi.uchicago.edu





448



AMIHAI MAZAR



FIGURE 23.1. Plan of the Granary Building at Beit Yera



˙



.

oi.uchicago.edu





449



THE EARLY BRONZE III GRANARY BUILDING AT BEIT YERA



Ó



Archaeology of the Hebrew University. The dimensions in the following description were
taken directly from this plan and differ somewhat from those given in the preliminary report.

The outer dimensions of the building are 31.25 (western outer wall) x 41 m. The eastern and
northern facades were not preserved entirely, but they can be reconstructed as measuring
approximately 35 m and 32 m respectively. The area of the building is about 1200 m



2



(fig. 23.1).
It is well constructed, though its outer corners do not form right angles, and it is trapezoid in
plan. The building was surrounded by stone-paved streets, evidence of a sophisticated level of
urban planning. The entrance to the building was from the east, through a corridor 3.3 m wide
and 14 m long. (This length is in fact the width of the eastern wall.) At the end of the corridor
was a threshold built with a single line of rather large stones (seen in Maisler, Stekelis, and
Avi-Yonah 1952: pl. 19A lower left). The corridor leads into an inner courtyard 11 m wide and
6.9 m long (76 m



2



in area). A part of the courtyard near the entrance, and a 2 m wide strip along
the inner (western) part of the courtyard, were paved with pebbles (seen in the above mentioned
photograph).



1



Along the southern wall of the courtyard were two stone constructions: the east-
ern one (1 x 1 m) was interpreted by the excavators as a flight of five steps leading to the top
of the building’s southern wall, which they called a “platform.” In fact, only two steps can be
seen in the plan, and this construction might have been a small podium. The western installation
consists of two large stones. This one could have been a step leading to a short corridor, which
led to the circle to its south.

Three ovens were found in the courtyard: two of them were located, one near the other, just
inside the entrance, yet in such a way that they would not disturb the entrance. The area of the
two ovens was deliberately separated from the entrance to the courtyard by elongated stones.
The eastern of these two ovens was unusual in form: it is described as consisting of two ovens
one inside the other. The larger oven had a 0.5 m wide opening divided by a “column which
was joined by arches to the two side walls.” In the oven was found “a pottery fragment with
four holes, apparently the key-piece of the vaulted roof.” Inside the oven were found fragments
of a large stand of the Khirbet Kerak family (now exhibited in the Israel Museum), as well as
a bowl of the same family. The third oven was located in the southwestern corner of the court-
yard, in a narrow niche between the eastern stone steps and the front of the enclosed hall.

In the western part of the courtyard, a wide entrance (2.1 m) led to a broad hall with inner
dimensions of 4.4 x 11.2 m. Two unworked stone slabs found along the long axis of the hall
were probably bases for wooden pillars that supported the roof. The preliminary report
describes a pebble floor in this hall, but such a floor is not shown in the detailed plan (though
the pebbles in the courtyard are definitely shown). The two circles at the western corners of
the building could be approached from this hall through 0.8 m wide diagonal corridors,
entered exactly at the southwestern and northwestern corners of the hall.

The outer stone walls of the building are about 10 m wide (except the eastern facade which
is 14 m wide). The tops of the walls are leveled. These wide walls contained sunken circles 7–
9 m in diameter. (Out of six measurable circles, four are 8 m in diameter: one is 9 m, and one
is 7 m.) The excavation revealed the remains of eight circles; a ninth can be reconstructed in
the unexcavated corner of the building. The stone floor of the circles is sunken about 10 cm



1. The picture is wrongly explained in Maisler, Stekelis, and Avi-Yonah 1952: 227 as showing the eastern side
of the courtyard and the entrance to the enclosed hall at its east (pointed out to me by P. de Miroschedji).

oi.uchicago.edu





450



AMIHAI MAZAR



below the top of the walls. Each circle is divided by four narrow partitions that do not reach
the center of the circle. These walls are oriented to the points of the compass. Narrow passages
led from two of the circles into the hall in the western part of the building.

The building is generally well planned and constructed with great care. The circles are exact,
and the partition walls inside each of them are oriented exactly to the points of the compass, as
if deliberately planned.



Reconstruction



Several attempts to reconstruct this building show each circle as a foundation for an indi-
vidual cylindrical or cone-shaped silo. (Mellaart 1966: 75, fig. 28, followed by Kempinski
1992: 77, fig. 10, show free-standing conical silos, while Kempinski 1978: fig. 9, reconstructs
free-standing cylindrical silos. For a general discussion of the “beehive granaries” of Pales-
tine, see Currid 1985.) These reconstructions were probably inspired from the Egyptian depic-
tions of free-standing conical silos in what may be defined as “beehive granaries” (Currid
1985: 105–6, figs. 2–5). Yet in these reconstructions, the role of the wide, solid stone foun-
dation walls in which the circles were embedded (denoted by the excavators as a “platform”)
remains obscure. Currid claims that the “the thick foundation wall certainly served to prevent
rodent intrusion, above-ground water seepage, and the influx of moisture from the ground”
(Currid 1985: 104, n. 37). This certainly explains the very existence of a raised solid stone
foundation. Yet this does not necessarily mean that the foundations did not carry a superstruc-
ture over the entire area. In my opinion, these wide walls with flattened upper levels were the
foundations for a massive, solid mudbrick superstructure which covered the entire space
between the circles, so that from the outside the building was seen as a massive structure.
Each individual silo was probably cylindrical with straight sides and a flat or domed ceiling.
The massive mudbrick superstructure would give the individual silos a solid outer frame. The
four piers in each circle must have supported the ceiling. The upper part of each silo could
have protruded above the main part of the superstructure and might have had a domed top, as
is shown in cylinder seal impressions from Susa, dating to the “Proto-Urbaine” period (late
fourth millennium



b.c.



) (fig. 23.2). There probably was an opening in the ceiling for filling the
granary, and the grain was taken out through openings at the base at the structure, as shown in



FIGURE 23.2. Susa cylinder seal impression showing granaries (adapted from Collon 1987: 146, no. 622).

long

oi.uchicago.edu





451



THE EARLY BRONZE III GRANARY BUILDING AT BEIT YERA



Ó



Egyptian depictions of silos (Currid 1985: 105, fig. 2). The corridors found leading to two of
the circles probably enabled access to these openings through the solid superstructure that
covered the entire stone foundation around the circular silos. This reconstruction suggests that
the granaries rose to a considerable height, which might have been equal to or slightly below



FIGURE 23.4. Early Helladic model granary from the island of Melos.

FIGURE 23.3. Suggested reconstruction of the Granary Building at Beit Yera



˙



(drawing by Anna Yamim).

oi.uchicago.edu





452



AMIHAI MAZAR



the diameter of the circles. A 6 to 8 m height for each of the silos is a reasonable estimate.
(Compare the average height of 5 m for the 2–3 m diameter of Egyptian silos; Currid 1985:
104.) The thick, solid walls would have been able to withstand the pressure of the grain. (At
the narrowest point the width of the granary wall was 1 m.) These considerations stand at the
basis of the reconstruction shown on figure 23.3.



The Finds



A destruction layer which covered the floor of the courtyard contained many pottery sherds,
including a large amount belonging to the Khirbet Kerak family. Many “broken and blackened
bones of animals” are reported as having been found “in several loci, especially in the open
court near the ovens, together with carbonized olive kernels, soot and other traces of fire.”

The finds from this building were never published in detail, and I have been unable to
locate any field diaries or other records.

Two unusual cult objects were published in the preliminary report (Maisler, Stekelis, and
Avi-Yonah 1952: 227, pl. 19:B). The first (fig. 23.5, pl. 23.1b) is a fragment of a thick clay ring
that bears two clay bull heads modeled in three dimensions. This object has been reexamined
by Amiran (1989: 31–33).



2



The second object (fig. 23.6, pl. 23.1) is a fragment of a zoomorphic libation vessel.



3



It is
hollow, and the mouth of the animal served as a spout. The vessel was made of coarse ware,
with plenty of large grits. The fragment is 10 cm long and 9.2 cm high; since it depicts about



2. I would like to thank Ruth Amiran for permission to republish the drawing of this object here. It is now
exhibited in the collections of Beit Gordon at Kibbutz Deganya.

3. The object was found in the storage area of the Rockfeller Museum, among other material from the Beit
Yera



˙



excavations. My thanks to J. Zias for locating it. The photograph was supplied by the Israel Antiqui-
ties Authority and the drawing prepared by Mrs. S. Helbreich.

FIGURE 23.5. Pottery ring with two bull figurines found in the Granary Building at Beit Yera



˙



.

oi.uchicago.edu





453



THE EARLY BRONZE III GRANARY BUILDING AT BEIT YERA



Ó



half of the animal, the complete vessel was probably about 20 cm long. The animal can be
identified as a lion, as shown by the shape of the head, the paws, and the delicate incisions on
the head which probably represent the mane. The ears are broken. One might suggest that the
broken ears were in fact horns, and thus interpret the animal as a bull. A prominent vertical
ridge down the center of the animal’s chest may strengthen this interpretation. Nevertheless,
it seems to me that the fragments at the front of the head were originally ears, and that this ani-
mal is indeed a lion. The back of the vessel has not been preserved. The entire preserved part
was painted with red stripes in a crisscross pattern. The piece is particularly important, since
it is the sole EB II–III zoomorphic vessel known from the Land of Israel. Depictions of lions
are well known in third millennium



b.c.



art of the ancient Near East, such as on Meso-
potamian cylinder seals depicting lions attacking domestic animals or in representations of
lions in wood carvings and seal impressions from Ebla (Matthiae 1981: pls. between pp. 80–
81). In the Land of Israel, lions are a familiar motif on cylinder seals from EB (Ben-Tor



FIGURE 23.6. Lion-shaped libation vessel found in the Granary Building at Beit Yera



˙



.

oi.uchicago.edu





454



AMIHAI MAZAR



1978: 52, fig. 15; 72–73, fig. 21). The lion motif became more common in Canaanite art in the
second millennium



b.c.



, often appearing in a ritual context.
Another object that probably came from the Granary Building is a figurine of an animal

(sheep?), of a type found at other EB sites (pl. 23.1c, right).



Function



The excavators made several suggestions as to the function of the building: a public gra-
nary, a fort, a palace, or a temple. They settled on the last one. The suggestion that the building
might be a granary was discounted in the original Hebrew publication because an oven had
been found in one of the sunken circles (Stekelis and Avi-Yonah 1947: 60), though in fact no
oven was found in the circles themselves. The unique oven in the building’s courtyard was
interpreted by the excavators as an installation for burning incense. In the English version of
the publication, Avi-Yonah added a note (Maisler, Stekelis, and Avi-Yonah 1952: 228 n. 12)
in which he presented a parallel to the building in the steatite model structure from the island
of Melos (fig. 23.4), which dates to the Early Helladic period (Marinatos 1946: 342, fig. 4;
Buchholz and Karageorghis 1973: 90, no. 1122). On the basis of the Melos model, which was
interpreted as representing a public granary, Avi-Yonah wrote:



This find increases the probability that the building was a granary after all and the ovens
a later addition, once it went out of use. If this hypothesis is accepted, it bears striking
evidence of the agricultural development and social structure of the EB settlements in the
Jordan Valley. It presumes a plentiful grain crop stored in a public building connected
with either a religious or secular authority.



The relevance of the Melos example was reinforced by the data on links between the
Aegean world and the Land of Israel in the EB III period (Hennessy 1967: 82–83). Kilian sug-
gested a resemblance between the Early Helladic II round building at Tiryns which he inter-
preted as a granary and the building at Beit Yera



˙



(Kilian 1986: 68), yet this is a very different
building, much reconstructed, and its function is still elusive.



4



The identification of the struc-
ture at Beit Yera



˙



as a granary was generally accepted in the archaeological literature. R. Ami-
ran noted parallels to third-millennium circular silos in northern Anatolia and Trans-Caucasia,
at the sites of Yanik Tepe, Shengavit, and elsewhere (Amiran 1965: 165–67). However, in
these examples, the round silos appear side by side and do not form part of a single building.
Some of them exhibit internal partitions, formed by four piers which do not meet in the center
of the circle, as in the Beit Yera



˙



building. In light of the origin of Khirbet Kerak ware in the
Kura-Araxis culture, these parallels seem to be relevant, despite the fundamental difference in
plan mentioned above.

Several scholars claimed that the “Granary Building” was a temple (Mellaart 1966: 74–76;
Amiran 1989: 32; Busink 1970: 379–80; Wright 1985: 218). In my opinion, these two expla-
nations do not need to contradict one another: the building can be interpreted as a temple, with
the granaries serving an economic role. The single large hall inside the building is a “broad-
house” structure reminiscent of the acropolis temple at Ai and the “White Building” at Tel
Yarmuth. Both of these structures have been identified as EB III temples, contemporary with



4. My thanks to Sharon Zuckerman for this reference.

oi.uchicago.edu





455



THE EARLY BRONZE III GRANARY BUILDING AT BEIT YERA



Ó



the Beit Yera



˙



structure (Miroschedji 1988: 35–41; Kempinski 1992: 53–59). The inner
dimensions of the broadhouse at Yarmuth (4.75 x 11.50 m) are almost identical to those of the
hall inside the “Granary Building,” though the Yarmuth example has four column bases along
the hall’s longitudinal axis, while only two were found at Beit Yera



˙



. The hall at Ai is larger
(inner dimensions 6.00 x 17.50 m). Remains of a dais were found opposite the entrance to the
hall at Yarmuth, but no traces of this kind were found at Ai or at Beit Yera



˙



. The unusual cult
objects found at the Beit Yera



˙



building, which are rare in the EB, and the unique oven con-
taining a large pottery stand found in the inner courtyard, reinforce the possibility that this
building was indeed a temple. However, two points weaken our argument that the inner hall
was a sanctuary: (a) No focal point of cult was found, such as the evidence for a dais found at
Yarmuth opposite the entrance; and (b) the approach to three of the circles could only be from
inside this inner hall. This can hardly be accommodated with the interpretation of the hall as a
sanctuary. Despite these difficulties, I would not exclude the possibility that the inner hall was
a sanctuary. Moreover, it may be suggested that the model structure from Melos represented a
temple as well since it would otherwise be difficult to account for its facade, which has a ped-
iment covered with rich spiral ornamentation. It thus may be suggested that both Beit Yera



˙



and Melos possessed temples with attached granaries, which were part of the temple’s role as
an economic power.

SOCIOECONOMIC INTERPRETATION

In the following section I attempt to widen the discussion of the Beit Yera



˙



structure’s
function into its significance for understanding the city’s socioeconomic system. With this
goal in mind, the following questions are considered:

1. What was its potential grain storage capacity?
2. What as the relationship between the amount of grain stored in the structure and the

city’s available agricultural land?
3. What was the relationship between the amount of grain stored in the building and the

city’s population?
4. What was the nature of the agricultural, socioeconomic, and organizational systems in

which the “Granary Building” functioned?

Clear answers to these questions remain elusive. Nevertheless, partial answers can be pro-
vided and several possible hypotheses raised.

(1) The basic hypothesis that the stone circles served as foundations of granaries should
be accepted. Their average diameter of 8 m implies that their superstructures were quite high,
between 6 and 8 m (see above). For the purpose of calculating the volume of the granaries, I
assume that they were 6 m high. This is in my view a minimal height for a round structure
with a diameter of 8–9 m. Obviously, any alteration in this estimated height will affect the fol-
lowing calculations considerably. The average floor area of each granary, minus the area taken
up by the inner piers, is 42 m



2



. The volume of an average granary would thus be 250 m



3



. The
entire Granary Building would thus have been able to store about 2,250 m



3



of grain. The
weight of 1 m



3



of grain can be calculated according to various coefficients. The most common

oi.uchicago.edu





456



AMIHAI MAZAR



for wheat is 770 kg per 1 m



3



(see references in Rosen 1986: 172; Schwartz 1994: 26, table 2),
while the coefficients for barley differ from 610 kg per 1 m



3



(Rosen 1986: 172) to 714 kg (ref-
erences in Schwartz 1994: 26, table 2). Thus if the building was fully utilized, its maximal
capacity was somewhat above 1,700 tons of wheat and between 1,370 and 1,600 tons of bar-
ley. This capacity of the building exceeds earlier estimates, such as the estimation of 800 tons
brought by Kempinksy’s (1978: 29, cited by Esse 1991: 100).

(2) What was the relationship between the building’s storage capacity and the city’s agri-
cultural potential? It can be assumed that the Granary Building was intended to store the max-
imum harvest possible, and that in bad or mediocre years the granary was not completely
filled. A hectare (10 dunams or 2.5 acres) of land worked by traditional methods in the Levant
yields between 300 and 1,000 kg of wheat, and the average in Arab peasant agriculture is
about 700 kg (Avitzur 1977: 44–63; Rosen 1986: 172; Rosen uses an average of 650 kg per
hectare of wheat and 800 kg per hectare of barley).



5



In the first year that Kibbutz Deganya was
settled just south of Beit Yera



˙



, its land yielded 830 kg of grain per hectare (Avitzur 1977).
This figure rose to 1,180 kg of wheat and 1,500 kg of barley per hectare in subsequent years.
These numbers are in accord with figures based on Sumerian documents: 1,254 kg of grain per
hectare in pre-Sargonic Girsu and 700 kg in Ur III texts (Adams 1981: 86, 146; Schwartz
1994: 28). In the Baqºah Valley of Lebanon, where conditions are much more similar to the
Jordan Valley, the estimation is just 300 kg per hectare (Marfoe 1979: 5). It may be suggested
that harvests in the EB were on the average not much smaller than those obtained from Arab
peasant agriculture. We can therefore calculate that in order to fill the Granary Building with
1700 tons of wheat, an area of about 2,430 hectares (6,075 acres) must have been under culti-
vation (based on an average yield of 700 kg per hectare). Beit Yera



˙



’s best agricultural lands
lie mainly to the south, in a triangle bounded by the Yarmuk River, the Jordan River, and the
Sea of Galilee up to a distance of about 8 km from the city. This is the farthest that farmers
would have been likely to cover on foot to reach their fields. This triangle, known as the Kin-
rot Valley, covers roughly 2,700 hectares (6,750 acres). This area comprises the northern end
of the Irano-Turanian ecosystem of the Jordan Valley, which enjoys an average rainfall of
300–400 mm. This allows grain to be grown with maximum yields in dry farming. No other
EB sites are known from this area and its periphery, except for Tell el-Hama (Hamat Gader),
which extends over about 1 hectare and whose inhabitants probably worked some of the same
land (see the distribution map in Esse 1991: 149, fig. 29). The closest EB site to the south is
Tel Yaqush (excavated by Esse), which is far beyond the potential lands of Beit Yera



˙



. To the
north, along the Sea of Galilee, there was another EB city at Tel Raqat (Khirbet Quneitra,
north of Tiberias), but its people would never have competed with the population of Beit
Yera



˙



for land in the Jordan Valley south of the Sea of Galilee. It thus may be assumed that
the agricultural land in the triangle defined above was cultivated by the inhabitants of Beit
Yera



˙



itself and not by the people from villages or satellite towns in the area (fig. 23.7). The
people of Beit Yera



˙



could also cultivate land on the Yavneel plateau, just west of Beit Yera



˙



,
though these were less accessible lands.



5. I thank Baruch Rosen for the time and effort he spent in discussing this subject with me. In the following
calculations I utilized Rosen’s methodology in analyzing the ºIzbet Sartah silos (Rosen 1986).



short

oi.uchicago.edu





457



THE EARLY BRONZE III GRANARY BUILDING AT BEIT YERA



Ó



FIGURE 23.7. Map showing the agricultural hinterland of Beit Yera



˙



.

oi.uchicago.edu





458



AMIHAI MAZAR



(3) In order to answer our other questions, I make use of data and methods employed by
B. Rosen and M. Broshi in their research on similar topics (Broshi 1979, 1986; Rosen 1986).
The working hypothesis is that data collected in anthropological research carried out in a similar
environment at a similar technological level (the Near East, especially late nineteenth-century
Palestine) are valid for earlier periods.

The area of the city of Beit Yera



˙



is estimated at 20 hectares (50 acres). Studies in ancient
and current traditional demography led to various estimates of density coefficients: from 100
persons per built-up hectare to 200–250 people per built-up hectare in the EB sites (as well as
other periods) in the Land of Israel. A figure of 200 or 250 persons was utilized for the Bronze
Age and the Iron Age (Marfoe 1979: 21; Broshi and Gophna 1984). Esse pointed out the dif-
ficulties in using a fixed constant of a number of people per hectare, due to the differences in
types of sites and in their settlement density (Esse 1991: 130–35). However, Beit Yera



˙



was a
densely built city in the EB III, and thus we utilize the 200 people per hectare factor and esti-
mate that Beit Yera



˙



’s population would amount to about 4,000 people (the high factor of 250
people per hectare would give a figure of 5,000). For the sake of simplifying the following cal-
culations, we assume that the Granary Building was used for storing wheat alone; if barley
was also stored there, the calculations would be only slightly affected.



6



The average annual consumption of wheat in wheat-based cultures in general, and in the
traditional economy of the Near East in particular, can be estimated as 150–250 kg per indi-
vidual; the higher figure (200 kg) is more generally accepted. Higher numbers (as much as
around 500 kg per person) are also cited (Rosen 1986:173; Schwartz 1994: 26, table 2, where
ten different estimations are cited, the highest being 558 kg per year and the lowest 143 kg per
year). Obviously, in calculating consumption, it has been taken into account that while wheat
(and perhaps barley) was the basis of the population’s food, they did have access to other
sources of nutrition (legumes, various fruits and vegetables, fish, dairy products, beef, and
mutton). A population of about 4,000 would thus consume about 800 tons of wheat per year.
A loss of about 30% during storage should be taken into account, caused by the depredations
of rodents and the need to provide seeds for the next year’s crop. Higher figures for this loss,
as much as 50%, are also cited (see Rosen 1986: 173; Schwartz 1994: 27, table 2).

Fodder for livestock has not been taken into account in these calculations since it is
assumed that cheaper forms of food and straw were used for this purpose and were not stored
in the central Granary Building. Broshi reckoned 300 kg of wheat per individual per annum,
including loss, seed corn, and animal fodder. In order to provide 800 tons of grain, it would
thus be necessary to store about 1,040 tons. This amount could easily have been stored in the
Granary Building, which according to my calculations, could store up to 1,700 tons of grain.
It is thus obvious that the building could store far more grain than the yearly consumption of
the entire city’s population.

This can be explained in several different ways. It may be that the building was designed
to store bumper harvests from especially abundant years (as Rosen suggested for the silos at
ºIzbet Sartah), while in other years the building’s maximum capacity was not used. Thus in a
bumper year, when the granaries were full, it was possible to store grain in preparation for



6. Barley was the principal crop grown in the third millennium



b.c.



in Mesopotamia and was used for making
bread and brewing beer (Jacobsen 1982: 38–51).

oi.uchicago.edu





459



THE EARLY BRONZE III GRANARY BUILDING AT BEIT YERA



Ó



poor years—since grain can be stored in this way for three to four years (though with an
increasing rate of loss). The building could thus have served to even out the food supply for
the city’s inhabitants over good and bad years. Another possibility is that the surplus grain
was used for trade with other cities or regions in the country, whose economies were based on
different crops or products.

One of the most likely possibilities is that the people of Beit Yera



˙



traded with pastoralists
living on the Golan and Issachar Heights. The EB III sites (so-called enclosures) found in
these areas point to the existence of a permanent population whose economy was based on
animal husbandry (cattle, sheep, and goat) as well as horticulture (Kochavi 1989: 21–25; con-
tra Esse 1991: 156–62, who raised the hypothesis of a dimorphic society in the Land of Israel
in the EB III, with reciprocal relations between the settled populations of the Jordan and Hula
Valleys and pastoralists in the Golan and the Galilee).

It is also probable that the inhabitants of Beit Yera



˙



traded with the people of the central
highlands of western Palestine, where horticulture was the basis of the economy (Stager 1985;
Esse 1991: 100–102). Esse raised the question whether the granary building reflects a redis-
tributive or market exchange economic system (Esse 1991: 100). A lack of written records
prevents, in his view, a definite conclusion on this point. Yet he describes Beit Yera



˙



as a cen-
tral site in northern Palestine: “The diversity of its resource base and the practice of successful
mixed economy” (p. 100) were the basis for its economic power, as well as its



. . . location straddling the border between two major eco-zones: the Mediterranean and
the Irano-Turanian steppe . . . although Beth Yera



˙



surely participated in interregional
and international trade, it is the level of intraregional trade that the city’s existence comes
into focus most sharply . . . the city became a true focal point, the nexus between two
very diverse ecological zones with different potentials. (Esse 1991: 100)



Esse thus defined the Beit Yera



˙



granary as a center of interregional trade, a redistribution
center, which is expected in a border zone between two different ecological regions.

The number of people occupied in growing this amount of grain can be very roughly esti-
mated. In traditional agriculture in the Land of Israel, a peasant (



fellah



) family of about five
people with their animals could cultivate an area of about 8–20 hectares (20–50 acres) of dry
farming grain crops; the average area per family was 12–15 hectares (30–37 acres).



7



This area
supplied a family with its agricultural produce. It is difficult to establish how many people
were needed to cultivate grain with the technological methods and the sociopolitical organiza-
tion of EB, and we do not know the division of labor in the city: Was there a degree of “spe-
cialization” and did only some inhabitants work on grain cultivation; or was it the main
occupation of most of the population?

The data presented above indicate that, in order to cultivate the 2,430 hectares necessary
to provide enough grain to fill the Granary Building, about 200 family units would be
required, or about 1,000 people. This number gives a very general idea of a 1:4 ratio between
the families who were involved in the cultivation of the amount of grain that could be stored
in the Granary Building as against the entire population in the city. Other inhabitants of the



7. “The agricultural unit necessary to provide for a peasant family with the technical ability to farm was cal-
culated at a



fadan



—the ability to work with a pair of oxen. It was usually of an area of 12 to 15 hectares.
The



fadan



provided for the family at a fairly low level of subsistence . . .” (Avitzur 1977: 36–38).

oi.uchicago.edu





460



AMIHAI MAZAR



city (about three-quarters of the families) would be occupied with other agricultural activities,
such as growing olives for the olive oil production, legumes, etcetera. There was also fishing
in the Lake of Galilee, and perhaps also cattle, sheep, and goat breeding (Esse 1991: 98–100).
Others would be professional craftsmen, military men, civil servants, priests, and so on.

If it is assumed that part of the grain needs of the inhabitants was grown on private allot-
ments and not stored in the public Granary Building, then we may conclude that some of the
grain kept in the building was used for trade even in years of average harvests. The territory
calculated as necessary to provide the amount of grain stored in the Granary Building formed
the greater part of the land within a reasonable walking distance of the city (the triangle
bounded by the Jordan and Yarmuk Rivers, as described above). According to our calcula-
tions, about 400 hectares in this area would have been left over for harvests other than grain,
or for producing grain that was not stored in the public granary. This would have been too
small an area to supply the needs of those inhabitants who were not linked to the economic
system represented by the Granary Building. It is therefore possible that some of the Beit
Yera



˙



agricultural lands were in more distant plots, for example in the Yavneel Valley to the
west or south of the confluence of the Yarmuk and the Jordan Rivers.

(4) The above calculations are evidence for a sophisticated system of centralized admin-
istration and careful planning of the agricultural-economic system at Beit Yera



˙



. The latter
entailed the concentration of agricultural produce and its redistribution to all or a large part of
the city’s inhabitants, and most probably its use in intraregional trade (with other cities or
regions in the country). Long-term storage of grain was probably also planned, in order to
overcome the effects of drought years. It may be assumed that the authority which erected and
managed the Granary Building had considerable political and economic power. But we have
no way of knowing how the ownership of land and agricultural labor in the city were orga-
nized. Did the city possess a “specialized economy” with only a quarter of its inhabitants
engaged in grain cultivation, while the rest carried on other forms of cultivation, fishing, herd-
ing, and various craft activities? Or were the different forms of agricultural work shared
equally among all the city’s inhabitants? Did the authority that owned the Granary Building
also possess the land on which the grain stored there was grown, with the agricultural workers
as serfs? Or did the building store grain that had been grown on privately owned allotments
and collected by a powerful government, either by legal means or by force of religious tradi-
tion? If the suggestion that the core of the building was a temple is accepted, it is possible that
most of the city’s land was controlled by this temple. This phenomenon is reminiscent of the
situation in Sumerian cities, where the temple had enormous economic power. However, even
in the case of Sumer, where we possess written sources, scholars disagree over the extent of
temple control of the city lands (Kramer 1963: 73–77; Gadd 1971: 126–31).



8



At Ebla, which
is the chronologically and geographically closest archive to the period of the Granary Building
at Beit Yera



˙



, economic power seems to have been in the hands of a secular authority—the
king of the city (Matthiae 1981: 163–86).



8. According to documents from the temple of Baba at Lagash, tenants and hired laborers worked the temple
lands at Lagash in the Early Dynastic period. The grain was stored in the temple granaries and was also
ground there. The salaries of the temple staff and its various suppliers were paid mostly in grain. For detailed
information on grain cultivation in Mesopotamia at different periods, see Jacobsen 1982. However, these data
concern irrigation agriculture, which is very different from the dry agriculture of the Land of Israel.

oi.uchicago.edu





461



THE EARLY BRONZE III GRANARY BUILDING AT BEIT YERA



Ó



A more detailed reconstruction of the socioeconomic system related to the Granary
Building would be totally speculative. Yet the building provides evidence for a complex eco-
nomic system that existed in the Land of Israel at the zenith of the EB urban culture. Features
such as specialization, advanced architectural planning on a large scale, concentration of food
resources and their redistribution according to need and ability, long-term food storage, and
interregional trade characterized this period and must have necessitated a complex bureau-
cratic system.



9



Whether this system was secular or related to temple administration is impos-
sible to know.

The “Granary Building” at Beit Yera



˙



was found near the surface and definitely belongs
to the latest stage of the EB city’s existence. Thus the city of Beit Yera



˙



did not show any evi-
dence for decline toward the end of the period of its existence.



10



Like at other EB III cities, it
appears that the end of the city occurred at a time when the EB III urban system in Palestine
was at its zenith. The demise came abruptly; however, its causes allude us.



11



Despite the high degree of speculation and uncertainty in the above discussion, it seems to
me that the Granary Building at Beit Yera



˙



can shed light on the socioeconomic structure of
large city-states in the Land of Israel on the eve of the demise of urban culture at the conclu-
sion of the EB III period.



9. The high degree of hierarchy and central authority in the EB III period is now exemplified superbly by the
well-planned palace and other public buildings and fortifications at Tel Yarmuth (Miroschedji 1999: 2–19).

10. The urban architecture of EB III at Beit Yera



˙



can be seen in the complex plans of structures there, like that
published by Eisenberg (Eisenberg 1981: 11–13).

11. Various suggestions were published concerning the reasons for the end of the urban culture at the end of the
EB III period. I would add that a plague should be taken seriously as a possible reason for the demise of this
culture when it was at its floruit. Plagues were one of the major reasons for abrupt changes in human history.
The abrupt end of the EB III culture hints in my view to such a cause.



REFERENCES



Adams, R. M.
1981



Heartland of Cities.



Chicago: University of Chicago Press.

Amiran, R.
1965 Yanik Tepe, Shangavit and the Khirbet Kerak Ware.



Anatolian Studies



15: 165–66.
1989 Re-Examination of a Cult and Art Object from Beth-Yera



˙



. Pp. 31–33 in



Essays in Ancient Civi-
lization Presented to Helene J. Kantor,



eds. A. Leonard and B. B. Williams. Studies in Ancient
Oriental Civilization 47. Chicago: Oriental Institute.

Avitzur, S.
1977



Changes in the Agriculture of the Land of Israel 1875–1975.



Tel Aviv: Milo (Hebrew).

Ben-Tor, A.
1978



Cylinder Seals of Third Millennium Palestine.



Bulletin of the American Schools of Oriental
Research Supplement Series 22. Cambridge, MA: American Schools of Oriental Research.

Broshi, M.
1979 The Population of Western Palestine during the Roman-Byzantine Period.



Bulletin of the Ameri-
can Schools of Oriental Research 236: 1–10.

1986 The Carrying Capacity of the Land of Israel in the Byzantine Period and Its Demographic Sig-
nificance. Pp. 49–56 in Man and Land in the Ancient Land of Israel, eds. A. Oppenheimer,
A. Kasher, and A. Rappaport. Jerusalem: Yad Ben Zvi Institute (Hebrew).

oi.uchicago.edu



462 AMIHAI MAZAR

Broshi, M., and Gophna, R.
1984 The Settlement and Population of Palestine during the Early Bronze Age II–III. Bulletin of the

American Schools of Oriental Research 253: 41–53.

Buchholz, H. G., and Karageorghis, V.
1973 Prehistoric Greece and Cyprus. New York: Phaidon.

Busink, Th.
1970 Der Tempel von Jerusalem: von Salomo bis Herodes, Der Tempel Salomos I. Leiden: Phaidon.

Collon, D.
1987 First Impressions. London: British Museum.

Currid, J.
1985 The Beehive Granaries of Ancient Palestine. Zeitschrift des Deutschen Palästina-Vereins 101:

97–109.

Eisenberg, E.
1981 Tel Bet Yera˙. Hadashot Arkheªologiyot 76: 11–13 (Hebrew).

Esse, D. L.
1991 Subsistence, Trade, and Social Change in Early Bronze Palestine. Studies in Ancient Oriental

Civilization 50. Chicago: Oriental Institute.

Gadd, C. J.
1971 The Cities of Babylonia. Pp. 93–144 in Cambridge Ancient History, Vol. 1, pt. 2, 3rd ed.

Cambridge: Cambridge University Press.

Hennessy, B.
1967 The Foreign Relations of Palestine during the Early Bronze Age. London: Quaritch.

Jacobsen, Th.
1982 Salinity and Irrigation Agriculture in Antiquity. Bibliotheca Mesopotamica 14. Malibu: Undena.

Kempinski, A.
1978 The Rise of an Urban Culture. Israel Ethnographic Society Studies 4. Jerusalem: Israel Ethno-

graphic Society.
1992 Fortifications, Public Buildings and Town Planning in the Early Bronze Age. In The Architec-

ture of Ancient Israel, eds. A. Kempinski and R. Reich. Jerusalem: Israel Exploration Society.

Kilian, K.
1986 The Circular Building at Tiryns. Pp. 65–71 in Early Helladic Architecture and Urbanization,

eds. R. Hagg and D. Konsola. Studies in Mediterranean Archaeology 76. Göteborg: Paul Aström
Verlag.

Kochavi, M.
1989 The Land of Geshur—Regional Archaeology on the Southern Golan in the Biblical Period.

Qadmoniot 22: 21–26 (Hebrew).

Kramer, S. N.
1963 The Sumerians. Chicago: University of Chicago Press.

Maisler, B.; Stekelis, M.; and Avi-Yonah, M.
1952 The Excavations at Beth Yera˙ (Khirbet Kerak) 1944–1946. Israel Exploration Journal 2:

218–29.

Marfoe, L.
1979 The Integrative Transformation: Patterns of Sociopolitical Organization in Southern Syria. Bul-

letin of the American Schools of Oriental Research 234: 1–42.

Marinatos, S.
1946 Greniers de l’Helladique Ancien. Bulletin de Correspondance Hellenique 7: 337–51.

Matthiae, P.
1981 Ebla: An Empire Rediscovered. London: Hodder and Stoughton.

Mellaart, J.
1966 The Chalcolithic and Early Bronze Age in the Levant and Anatolia. Beirut: Khayats.

oi.uchicago.edu



463THE EARLY BRONZE III GRANARY BUILDING AT BEIT YERAÓ

Miroschedji, P. de
1988 Yarmuth 1. Paris.
1999 Yarmouth: The Dawn of City-States in Southern Canaan. Near Eastern Archaeology 62: 2–19.

Rosen, B.
1986 Subsistence Economy of Stratum II. Pp. 156–185 in ºIzbet Sartah, ed. I. Finkelstein. British

Archaeological Reports, International Series 299. Oxford: British Archaeological Reports.

Schwartz, G. M.
1994 Rural Economic Specialization and Early Urbanization in the Khabur Valley, Syria. Pp. 19–36

in Archaeological Views from the Countryside, eds. G. M. Schwartz and S. E. Falconer. Wash-
ington, DC: Smithsonian Institution.

Stager, L. E.
1985 The First Fruits of Civilization. Pp. 172–88 in Palestine in the Bronze and Iron Ages: Papers in

Honour of Olga Tufnell, ed. J. N. Tubb. London: Institute of Archaeology.

Stekelis, M., and Avi-Yonah, M.
1947 The Excavations at Beth Yera˙, Second Preliminary Report. Bulletin of the Jewish Palestine

Exploration Society 13: 53–60 (Hebrew).

Wright, G. R. H.
1985 Ancient Building in Syria and Palestine. Leiden: E. J. Brill.

oi.uchicago.edu



464 AMIHAI MAZAR

PLATE 23.1. Cult objects from the Granary Building at Beit Yera˙.

oi.uchicago.edu



465



NOTES ON EARLY BRONZE AGE METROLOGY
AND THE BIRTH OF ARCHITECTURE IN



ANCIENT PALESTINE



*



Pierre de Miroschedji



To the memory of Douglas Esse, a superb archaeologist, a devoted colleague, a fine gen-
tleman, and a dear friend.

INTRODUCTION



Although biblical metrology



1



is well known and has been the subject of several treatments
(Barkay and Kloner 1986; Barrois 1953: 244–47; Ben-David 1978; Kaufman 1984; Paul and
Dever 1973: 173–74; Powell 1992: 898–901; Scott 1958, 1959; Ussishkin 1976; de Vaux 1960:
297–301), pre-Israelite metrology in Palestine has been paid relatively little attention (see
Dever 1974: 43; Milson 1986, 1987, 1988; Toombs and Wright 1961: 33, fig. 12; G. E. Wright
1965: 89; G. R. H. Wright 1985: 118–20). This weak interest is probably the result of a low
confidence in the value of these studies, for in the absence of written documentation, empirical
metrology relies on measurements made on the remains of buildings which, as a rule, were
constructed with dry stone masonry; these measurements are obviously approximate while
metrological studies aim at centimetrical, or even millimetrical precision. Moreover, it is easy
to show that hazardous manipulation of the results of unsystematic measurements can lead to
practically any kind of results. If the data are uncertain and the method for collecting them
questionable, how can we achieve reliable results?

This pessimism is, however, excessive and should not prevent us from attempting to shed
some light on early metrology, provided some clear-cut rules are posited. A large number of
systematic measurements should be made on one or several buildings in order to obtain a sta-
tistical basis: even when dealing with dry stone masonry, it cannot be assumed that all stones
have moved from their original positions. Also, manipulations of the numerical data gained
from these measurements should be as few as possible and based on reasonable and verifiable
assumptions concerning the use of subdivisions of a unit of measure. And more important, the
methodology of the research should be based on a theoretical framework.

This last remark deserves emphasis. A study of early metrology should not be conducted
solely with the legitimate curiosity of discovering what unit of measurement was first used,
whether it was locally invented or borrowed from a neighbor. It should examine the signifi-
cance of the introduction of a metrological system for a better understanding of an ancient



1. In this paper, the word “metrology” is used conventionally to designate a system of linear measures, as only
measures of length are taken in consideration—and not measures of capacity and weight.



* This paper was submitted for publication in January 1996.



24



oi.uchicago.edu





466



PIERRE DE MIROSCHEDJI



society. The appearance of metrology is inseparably linked to that of architectural planning
which accompanies the first manifestations of monumental architecture; the latter, in turn, has
long been recognized as an essential archaeological correlate of complex societies with which
it is invariably associated. Now monumental architecture makes its first appearance in Pales-
tine in the Early Bronze Age (EB). Recent discoveries at Tel Yarmuth offer impressive
examples of palatial architecture and an opportunity to test, by comparison with other contem-
porary sites in Palestine, whether a metrological system had emerged in the third millennium
together with peculiar techniques of architectural planning and a specialized builder to apply
them, the architect.

These are the main questions to be investigated in the following pages. This paper begins
with a sort of



status questionis



of early metrology, both theoretical and empirical, and the ear-
liest techniques of architectural planning; it continues with an examination of the data of Early
Bronze Age sites in Palestine, especially Tel Yarmuth; and it concludes with general com-
ments on the significance of early metrology and the birth of architecture in Palestine.

PRELIMINARY REMARKS ON EARLY METROLOGY

AND ARCHITECTURAL PLANNING



A Theoretical Approach



Until the modern introduction of the metrical system, all measurements were of course
based on portions of the human body: palm, span, and cubit have conveniently been used from
time immemorial as units of lengths, and despite their approximation have satisfied all needs.
As no building can be made without measuring first, building and measuring have developed
together, and this intimate association is attested, for example, in the earliest Mesopotamian
myths.



2



For a long time the builder was no specialist but the owner of a traditional knowledge
inherited from generation to generation: building techniques were simple, planning rudimen-
tary, and several members of the group shared this basic knowledge. Ethnographic studies in
Iran, in Palestine, and elsewhere in the Near East have shown that in most cases, the builder
traced the plan of the building on the ground with chalk or plaster at a 1:1 scale (Aurenche
1981: 95 [with refs.]). As the weight of tradition and technical constraints limited his freedom
of creation, well-tried methods were endlessly repeated.

This method of “freehand planning” was convenient enough as long as simple dwellings,
or small public buildings that were just enlarged versions of the domestic houses, had to be
built. It became inadequate, though, as soon as larger public buildings had to be erected, espe-
cially palatial buildings where the large number of rooms and the need for easy circulation
necessitate the design of elaborate plans. For these more sophisticated buildings, new tech-
niques of architectural planning had to be introduced, involving the ability to trace right
angles and to materialize guidelines for the placement of the walls on a large ground surface.
This could be achieved with the help of a twelve-knotted rope, a surveying instrument in



2. See “Enki and the world order,” where this god, initiator of mankind, is presented, among other things, as
the patron of surveying and building: See Bottéro and Kramer 1989: 176–77 (lines 340–47).

oi.uchicago.edu





467



METROLOGY AND THE BIRTH OF ARCHITECTURE IN ANCIENT PALESTINE



which the distance between each knot corresponds to a unit of length. It has been used from
time immemorial as the most convenient tool to draw right angles by tracing on the ground a
peculiar kind of rectangular triangle—often called the “three-four-five” or “Pythagorean” tri-
angle—whose sides represent respectively three, four, and five units of length.

Recent research has established that these developments took place very early in Mesopo-
tamia: J.-D. Forest has conclusively shown that from the beginning of the Ubaid period, the
plans of large buildings were established on the basis of an orthogonal grid in which the side
of each square corresponds to that of the unit of length used for planning (Forest 1991; see
also Eichmann 1991; Forest 1993). The plan of the building was inscribed within the grid
established with the help of a 12-knotted rope. The grid was not necessarily materialized on
the ground, at least not completely, as it was mainly intended to serve as guidelines for the
placement of the major walls, the partition walls being added in a more empirical way.

The advantages of this system of grid planning are manifold (Forest 1991: 164; 1993: 177–
79). The plan of the building can be conceived and drawn on the ground at actual scale with a
minimum of measurements, regularities are introduced in the proportions and the sizes of the
rooms, courtyards, etc., and symmetry is easy to achieve. There is thus an appreciable differ-
ence between a plan realized in an empirical way and a plan elaborated with the help of a plan-
ning grid. A convenient way for archaeologists to identify the use of such a planning system
is by superimposing on a drawn plan an orthogonal grid with a mesh corresponding to a rec-
ognized or a plausible unit of length, in order to check whether there is an adequate correspon-
dence between the lines of the grid and the position of the main walls of the plan (see Forest
1991; and below, figs. 24.6, 8–9).

A relationship between the introduction of these new planning methods and the adoption of
a metrological system is a plausible hypothesis. It is well known that following the introduction
of molded bricks of standardized size (a rather late phenomenon: see Aurenche 1993), masons
were accustomed to translate measures of length not only into empirical cubits (the distance
from the elbow to the tip of the middle finger, i.e., about 0.45 m), but also into numbers of
bricks (whose sizes were themselves related to that of the empirical cubit); this is still the
method used nowadays by traditional builders. However, the planning grids identified by J.-D.
Forest in Mesopotamia of the Ubaid period seem to have been established on the basis of a fixed
unit of length distinct from the size of the bricks, suggesting that the plan was drawn by using
a conventional and predefined unit of length and not by counting a certain number of bricks
(Forest 1991: 163). Hence it is presumably among architects that the use of a cubit of a con-
ventional and fixed length was first adopted. An indication of this is given by the fact that both
in Palestine and in Egypt, cubits of a fixed value, which were longer than the empirical or
“short” cubits of around 0.45 m—that is, not directly connected anymore to a part of the human
body—were called “builder’s” or “stone cutter’s cubit” (Iversen 1975: 16; Ben-David 1978).

Architectural planning with the help of a grid is a technique especially well adapted to an
illiterate society as it provides a convenient mnemotechnical system to memorize the charac-
teristics of major plans, which could thus be transmitted orally from one specialist to another.



3



3. It should be remembered (contra Arnold 1991: 10) that the introduction of metrology and architectural plan-
ning is independent from the invention of writing, as indicated by the Mesopotamian evidence which shows
that units of measurements and grid planning were used since the beginning of the Ubaid period.

oi.uchicago.edu





468



PIERRE DE MIROSCHEDJI



A good experience with this system could lead ultimately to a familiarity with methods of
tracing simple geometrical figures (squares, rectangles, and triangles of various proportions),
especially those that can be drawn with a twelve-knotted rope. As these methods were also
easy to memorize and transmit, the corresponding geometrical figures are ubiquitous in the
design of early buildings.



4



In literate societies, however, the architect could draw the plan on
a tablet (see Heinrich and Seidl 1967, 1968) or a papyrus (see Arnold 1991: 10). The differ-
ence between grid planning and preplanning is not simply one of literacy but of capacity of
abstract conception: grid planning implies that the plan is conceived and executed on the
ground at actual scale, while preplanning supposes that a plan conceived entirely in abstract is
drawn beforehand and then traced on the ground after a scale translation. This represents a
rather late development, restricted to few areas of the ancient Near East.

Thus metrology appears together with large-scale architectural planning at a rather
advanced stage of development of the art of building, when the need was felt for the construc-
tion of buildings more elaborate than simple dwellings. This occurred at a level of sociopolit-
ical development corresponding with the emergence of a complex society. A more specialized
knowledge was then required, stimulating eventually the appearance of specialists possessing
a specific competence in measuring and planning. Hence the appearance of metrology heralds
both the birth of architecture



stricto sensu,



distinct from the art of building, and the interven-
tion of a new specialist, the architect, different from the traditional builder.



Two Examples from the Ancient Near East



The validity of these general remarks can be sustained in broad terms by archaeological
data derived from several areas of the ancient Near East. Mesopotamia and Egypt are by far
the most informative in this respect as they shed light on the chronology and the modalities of
this process.

Most significant are the data from Mesopotamia. In this area, a significant development in
the art of building can be traced back to the late sixth millennium



b.c.



, with the appearance,
perhaps as early as the Samarra period, of symmetrical buildings with elaborate plans (Forest
1983, 1991). J.-D. Forest has shown that in the Ubaid period (starting in the Ubaid 0 Phase),
the plans of the buildings were established on the basis of a fixed unit of length and laid out
with the help of an orthogonal grid (Forest 1991; see also Kubba 1990; Eichmann 1991: 82–
86). These, however, were dwelling units. Public buildings (“residences of chieftains”) quite
different from contemporary dwellings in size and complexity of plan are well attested in the
late Ubaid period; this elaboration and the standardization of the plan manifest the emergence
of specialized builders (Aurenche 1981: 295; Margueron 1986; 1987a: 15–20). By the Uruk
period, monumental public buildings with a specific plan and peculiar architectural details
(pilaster with niches, columns, etc.) are known from southern Mesopotamia to northern Syria,
over a distance of more than a thousand kilometers, testifying to the widespread distribution



4. The interested reader will find comments on the use of these geometric figures for the design of buildings in
early architecture in Choisy 1983: 51–57, 100, 136–38, 385–99. See also Badawy 1961 for ancient Egypt.
Graeco-Roman examples are dealt with in Vitruvius’



De architectura



(see especially the translation with
comments by Choisy 1971), while European medieval examples will be found in the notebooks of Villard
de Honnecourt (see Bechmann 1993).

oi.uchicago.edu





469



METROLOGY AND THE BIRTH OF ARCHITECTURE IN ANCIENT PALESTINE



of a special architectural tradition and identical planning and building techniques (Margueron
1987a: 20–32); these were evidently carried by itinerant professional architects concerned
with official buildings. By then, metrological preoccupations are a well established fact
(Franck 1975). In the following Jemdet-Nasr period, southern Mesopotamia witnessed the
appearance of palatial architecture at Jemdet-Nasr and at Uruk (Margueron 1982: 23–34, 577–
78; 1987a: 32–36). In the Early Dynastic period, several palaces were built (Margueron 1982:
35–144; Heinrich 1984) whose plan had been designed with the help of a modular grid and
simple geometric figures (Heinrich 1984: 556–57). By then a sophisticated metrological sys-
tem had been elaborated, whose details are known through numerous textual references; based
on the cubit (conventionally assessed at about 0.50 m), it consists of its multiple, the reed (6
cubits), the rod (= 2 reeds or 12 cubits), and the rope (= 10 rods or 20 reeds or 120 cubits)
(Powell l989). By the middle of the third millennium, the status of the architect emerges,
linked to surveying and to mathematical knowledge and placed under the patronage of the
great god Enki (see above, n. 2).

For Egypt, it is generally agreed that a formal metrological system had been already elab-
orated in the Third Dynasty (ca. 2649–2575



b.c



.), when there is both textual and monumental
evidence of the simultaneous use of a



small cubit



, used for the ordinary purposes of crafts and
industries, and a



royal cubit



, reserved for cultic and official buildings and sometimes called
the “the Royal Cubit measuring stones” (Iversen 1975: 16). Longer than the small cubit of
about 0.45 m, the royal cubit prevailed throughout Egyptian history with more or less the
same fixed value, evaluated by Egyptologists between 0.522 m and 0.525 m, the latter value
usually considered the most probable (Iversen 1975: 14–19; Helck 1980; Arnold 1991: 10–23,
251–52; Powell 1992: 898; for other values, see Badawy 1954: 69 [52.4 cm]; Gardiner 1957:
199 [52.3 cm]; and Edwards 1992: 297 [52.2 cm]).

These precise measurements could be checked by scholars because measuring rods resem-
bling wooden rulers one or two cubits long have been found in many tombs, including those
of architects (Arnold 1991: 251). Longer measuring rods existed also: it has been demonstrated
that the base length of the pyramid of Cheops could have been achieved most accurately by
using rods with a length of 4 or 8 cubits (Arnold 1991). For measuring longer distances, sur-
veyors used a rope. The use of knotted ropes—with knots placed at regular intervals corre-
sponding to a conventional number of cubits—is attested in the New Kingdom for measuring
fields (Arnold 1991: 252), but there is little doubt that it was used much earlier for building
purposes. The “stretching of the cord” during the foundation ceremonies for temples has
been frequently represented (see references in Arnold 1991: 23, n. 20): wooden pegs were
hammered into the ground at the place of the knots, and Arnold mentions that “several
examples of the wooden pegs that could have been used for that purpose have been found in
situ” (Arnold 1991: 10–11, n. 21). The knotted rope permitted the tracing of geometric fig-
ures, especially the “three-four-five” (or “Pythagorean”) triangle, which has been systemati-
cally used by architects of the Old Kingdom, in particular for the surface planning and the
elevation of the funerary complexes of Djoser (Third Dynasty), Unas (Fifth Dynasty), and Teti
(Sixth Dynasty) at Saqqara (Meyer-Christian 1987; Labrousse, Lauer, and Leclant 1977: 66–
72; Lauer and Leclant 1972: 51–57).

It is probable that the measures of lengths used in these monuments derived from an older
tradition, for royal mastabas of the First Dynasty already testify to the systematic use of a

oi.uchicago.edu





470



PIERRE DE MIROSCHEDJI



cubit of canonical value (Emery 1938: 3–9; 1939: 10–19; Vandier 1952 [I.2]: 613–89). Fixed
measures of length might have been introduced even earlier, when large-scale monumental
architecture began to appear in the Naqada III period, as suggested by the plan of some
Dynasty 0 tombs in Abydos/Umm el-Qaab Cemetery B (Dreyer et al.1993: 25, Abb. 1), and
by architectural representations or evocations of elaborate palatial buildings with facades
decorated with niches on ivories (Vandier 1952 [I.1]: fig. 373), palettes (Vandier 1952: fig.
392), and on



serekh-



signs incised on pottery (see Kaiser 1982: 260–69). Although much later
than in Mesopotamia and after a much faster evolution, Egypt confronts us again with a situa-
tion where the emergence of monumental architecture is concomitant with that of metrology
and with the appearance of professional builders.



The Case of Palestine



Such spectacular and early developments as in Mesopotamia and in Egypt cannot be traced
in the southern Levant. The Old Testament testifies to the use of a system of measures of length
based on a cubit of approximately 0.525 m, similar to that of Egypt, and smaller units: the span
(half a cubit), the palm (one-third of a span or one-sixth of a cubit), and the finger (one-fourth
of a palm). Based on the testimony of Ezek. 40–48, scholars traditionally differentiate between
a short “cubit of a man” with a length of six palms and a longer “royal cubit” of seven palms
(“a cubit and a handbreadth” in Ezek. 40:5 and 43:13, i.e., six palms plus one); the former is
an empirical cubit, measuring about 0.45 m—the actual distance from the elbow to the tip of
the middle finger—and used for ordinary purposes, while the latter, whose length exceeds that
of the human cubit, was later known as the “builder’s cubit” (Ben-David 1978).

Although the validity of these metrological reconstructions have been put in question and
considered insufficiently reliable to formulate a length estimate expressed in the metric system
(see Powell 1989: 900), they seem to correspond indeed to a cubit of about 0.525 m as demon-
strated by measurements of contemporary monuments, among others Hezekia’s Siloam tunnel
and contemporary royal tombs (Petrie 1892: 28–35; Ussishkin 1976; Barkay and Kloner 1986).

This royal cubit was also in use in Phoenicia (Ben-David 1978: 28). In ancient Israel, it was
the length measure



par excellence



, used to express all distances, even very long (e.g., the
25,000 cubits of the holy district in Ezek. 45:1). But there were also longer units of length
expressing multiples of the cubit, such as the reed (see Ezek. 40:3, 5–8; 42:16–19; but cf. Pow-
ell 1992: 900) and the rope (see 2 Sam. 8:2; 2 Kings 21:13; Jer. 31:39; Amos 7:17; Mic. 2:4;
Zach. 2:5), and a smaller unit, the span or half cubit (see Ezek. 43:17 and 1 Sam. 17:4).

A few indications suggest that in the Levant, the antiquity of the “royal” cubit may go
much beyond the Israelite period, at least as early as the Middle Bronze Age II. In Syria, the
use of a cubit of about 0.525 m has been established at Ebla in the Middle Bronze Age II
Building P3 (Matthiae 1993: 164b), and a study of the other contemporary Syrian palaces will
probably reveal it too (although this was not stated by Margueron 1987b). In contemporary
Palestine, the use of the same cubit was convincingly identified in the layout of the Fortress
Temple at Shechem (cf. Toombs and Wright 1961: 33, fig. 12; G. E. Wright 1965: 89; Dever
1974: 43; contra Milson 1987), and it could probably be demonstrated also through a careful
metrological analysis of Palestinian palaces of the Middle and Late Bronze Ages.

But what can be said about the Early Bronze Age? As recent discoveries have appreciably
increased our knowledge of the architecture of this period, it is appropriate to investigate if

oi.uchicago.edu





471



METROLOGY AND THE BIRTH OF ARCHITECTURE IN ANCIENT PALESTINE



third millennium



b.c.



Palestine had been left entirely out of the metrological and architectural
developments that had taken place in contemporary Mesopotamia and Egypt. The first evi-
dence to analyze is that from Tel Yarmuth.

THE DATA FROM THE EB IIIC PALATIAL COMPLEX AT TEL YARMUTH

Ongoing excavations at Tel Yarmuth have recently uncovered part of a large EB IIIC pala-
tial complex covering about 6,000 m



2



(Miroschedji 1991, 1992, 1993a, 1993b, 1994, 1995a,
1995b. Measuring 84.20 m long and 72.10 m wide, the complex is limited by a thick peripheral
wall presenting, on part of its length, square inner buttresses placed at regular intervals (fig.
24.1). These buttresses indicate the existence of two major courtyards inside the enclosed area,
a large one occupying its southwestern half and a secondary one located to the northeast. The
built-up area covered



grosso modo



the northeastern half of the complex. In the present state of
excavations, it consists of a network of about twenty corridors, halls, and rooms, including
several storerooms filled up with pithoi. In the northeastern part of the complex, which
corresponds to a lower terrace (squares U-W 24–40), most of the walls have been destroyed
by erosion.

A quick glance at the plan of this complex reveals two striking features in addition to
monumentality: a remarkable regularity in the size of the rooms, the walls, the doors, the
buttresses, and their intervals, and an overall plan unusually elaborate for an EB building,
with perfect right angles and with rectilinear walls extending for distances of up to 55 m.
These features immediately suggest the use of a fixed measure of length and large-scale archi-
tectural planning—two possibilities that deserve to be carefully investigated.



Measuring



At an early stage of the excavations of this complex, it was recognized that the walls fall
into different categories according to their thickness. It was also assumed that the thickness of
the partition walls, evaluated at 0.50/55 m, may correspond to a unit of measurement since it
was apparently half of that of the standard walls. The precise value of this unit, conveniently
called a



cubit



, was ultimately calculated at about 0.52 m on the basis of the measurements of
the standard walls, several score of which had been uncovered (Miroschedji 1992: 269, 271;
1993b: 834–36).

Numerous measurements conducted since then have confirmed the validity of this evalua-
tion and have provided a statistical basis so solid as to minimize legitimate doubts. Such mea-
surements have also established the use of a half cubit—a span—and eventually led to the
surprising result that an even more precise evaluation of the length of the cubit—about
0.525m—could be suggested because several measurements, especially those taken on longer
distances, which are better explained by the use of a cubit of this value.

Practically all the measurements of walls, doors, buttresses, and room length and width
made inside the palatial complex conform to multiples of this cubit, or of its half. Although it
is not the aim of this paper to provide a full list of measurements, some synthetic information
should nevertheless be presented here to justify our conclusions.

The walls have standardized thicknesses corresponding to multiples of the cubit. Four
sorts of walls were distinguished (fig. 24.2):

oi.uchicago.edu





472



PIERRE DE MIROSCHEDJI



• partition walls, with a thickness of 1 cubit only (three examples only, measuring
between 0.52 and 0.55 m);

• standard walls (the large majority of the inner walls of the complex), with a thickness
of 2 cubits (average thickness measured: 1.056 m);

• thick walls, with a thickness of 2.5 cubits (about 1.30 m; seven examples only); and
• peripheral walls, which enclose the complex and measure 3.5 cubits in thickness

(average thickness measured: 1.85 m).



5



5. The very thin walls (approx. 0.45 m) which delimit a small rectangular installation in Square R 34 should
be mentioned only for the record, as they are not part of the structure of the palace.

FIGURE 24.1. Yarmuth: General plan of Palace B (Area B, Stratum B-1) and of the adjacent Area G (Stratum
G-2) at the end of the 1993 season of excavations.

oi.uchicago.edu

[image: image176.tif]



473



METROLOGY AND THE BIRTH OF ARCHITECTURE IN ANCIENT PALESTINE



The thickness of the walls seems actually to be determined by the size of the bricks (fig. 24.3).
They measure in average 0.50 x 0.25 x 0.15 m—such as one brick placed crosswise, or two
bricks arranged lengthwise, gave, with the addition of intermediate mortar and lateral plaster,
the thickness of a partition wall.

The width of the more than a dozen doorways cleared or studied up to now corresponds
generally to that of the attached wall, i.e., two cubits (approx.1.05 m) (fig. 24.4). We know of
only one example of a narrower door (0.80 m or about 1.5 cubit) and only two (?) examples
of wider doors (1.30–1.34 m, i.e., 2.5 cubits). Most doors are flanked by two doorjambs made
of sun-dried or baked bricks with a width between 0.25 m and 0.30 m (average: 0.27 m), i.e.,
half a cubit.

The regularity of the size of the buttresses of the peripheral wall and of the intervals sep-
arating them offer the best opportunity for precise measurements (fig. 24.5). Square in plan, the
buttresses measure on each side about 1.82–1.84 m, i.e., 3.5 cubits. They are separated by inter-
vals of about 2.10 m (4 cubits) along the southwest and southeast sides, and about 2.35 m (4.5
cubits) along the northwest side (for a possible explanation of this difference, see below). At
the western and southern corners of the complex, the interval between the inner side of the cor-
ner and the first buttress of the perpendicular peripheral wall is, in both cases, of 4.15–4.20 m,
or 7.5 cubits (3.5 + 4 cubits). The regularity of these measurements is such that the location of
the unexcavated buttresses can be established with quasi-certitude, allowing substantial resto-
ration of the plan (fig. 24.1). Thus, during the 1993 season, narrow soundings (5 m x 2 m) could
be successfully excavated in Squares Q 40 and S 41 at the predicted location of a buttress.

Seven buttresses can be restored along the northwest side of the complex and twenty along
the southwest side. Expressed in cubits, the length of the southwest side is of: (2 peripheral



FIGURE 24.2. Categories of wall attested in Palace B of Yarmuth.

oi.uchicago.edu

[image: image177.tif]



474



PIERRE DE MIROSCHEDJI



walls x 3.5 cu.) + (20 buttresses x 3.5 cu) + (21 intervals x 4 cu.) = 161 cu. that corresponds
to 83.72 m or 84.52 m depending on a value of 0.520 m or 0.525 m for the cubit. As the mea-
sured length of this side (measured with a laser theodolite) turned out to be 84.482 m, a value
of about 0.525 m for the cubit results in the smallest (and insignificant) discrepancy (0.043 m
only) and obviously recommends itself as the best approximation.

The length of all rooms, without exception, conforms to measurements in cubits. Regular-
ities in length and width could be established: for example, the width of several corridors is in
the range of 3.5 cubits (approx. 1.85 m), that of some narrow, elongated rooms is about 4
cubits (approx. 2.1 m), while the lengths of the four rectangular rooms in Q-V 25 seem to be
11.5 cubits each. A possible reason why these particular measurements appear is given below.

Standardized measurements were also observed for other architectural features such as
recesses, short walls flanking doorways, stone slabs used in stairs, location and spacing of pil-
lar bases, and so on.

On the basis of these numerous measurements, it is not too hazardous to conclude that the
layout of the EB IIIC palatial complex of Yarmuth has been achieved with the help of a fixed
unit of length which has been systematically used by its builder. The consistency of the mea-
surements is remarkable and leaves little doubt concerning the value of this “Yarmuthian”
cubit: it is almost certainly in the range of 0.52 m, and possibly close to 0.525 m, the latter
value offering the best working hypothesis to date. Hence it is conventionally retained in the
remainder of this paper.



FIGURE 24.3. Typical structure of a standard wall in Palace B of Yarmuth.

FIGURE 24.4. Plan of a typical doorway in Palace B of Yarmuth.



long

oi.uchicago.edu

[image: image178.tif][image: image179.tif]



475



METROLOGY AND THE BIRTH OF ARCHITECTURE IN ANCIENT PALESTINE



Whether this value of the cubit is “natural” (and liable, therefore, to be found in different
cultural contexts, otherwise unrelated) or “cultural” (and thus necessarily derived from one
source) is a moot point. The fact that Mesopotamian cities used various cubits with local values,
all of which were different from those of the Levant and Egypt (Powell 1989: 462–63), suggests
that the figure of 0.520–0.525 m is culturally significant. It could be further surmised—but
hardly demonstrated—that this value of the cubit used in EB IIIC Yarmuth was inspired by that
of the Egyptian royal cubit, traditionally evaluated at 0.525 m (see above). In any case, it may
well be that we have identified in Yarmuth the earliest evidence known so far of the use of this
cubit in the Levant. Its introduction in third millennium Palestine inaugurates a metrological
tradition that will last for nearly three millennia, at least until the end of the Iron Age.



Planning



Another set of data, admittedly less reliable and more speculative, should be discussed
now: it concerns the planning of the palatial complex, obviously an overall and large-scale
operation. Can we elicit the procedures followed by the builder in planning this complex? The
orthogonality of the layout necessarily implies the use of a twelve-knotted rope: only this sur-
veying instrument could achieve this result. The regular distribution of the main structural
elements of the plan further implies the use of an overall planning method rather than the
application of specific proportions to individual spaces (rooms and courtyards). A logical
assumption is that the builder of the complex used a modular grid system to obtain guidelines
for the placement and the orientation of the walls, a planning method already attested, as we
have seen, in fifth millennium Mesopotamia (see above).

If we admit the possibility that a grid was used to plan the layout of this complex, the
question is to determine the module chosen by the architect to establish such a grid. As the
cubit is a measure of length much too short for convenient planning, especially in the case of
an architectural complex the size of Palace B, we should look for a multiple of the cubit. The



FIGURE 24.5. Size and spacing of buttresses along the outer wall of Palace B of Yarmuth.

oi.uchicago.edu

[image: image180.tif]



476



PIERRE DE MIROSCHEDJI



measurements taken up to now indicate that all lengths attested in the palatial complex are
multiples and/or combinations of the following lengths: 2, 2.5, 3.5, and 4 cubits. This obser-
vation conveniently restricts the range of choice for a module. Actually, the analysis of the
plan shows that only two values recommend themselves: 7.5 cubits (i.e., 3.5 + 4 cubits) or its
double, 15 cubits.

The first value corresponds to the combined length of one buttress (3.5 cu.) plus an interval
between two buttresses (4 cu.). Measurements indicate that while the individual length of a
buttress and that of an interval between two buttresses may vary occasionally, that of their
combined lengths is remarkably stable, implying that 7.5 cubits might be indeed the unit cho-
sen by the builder for planning the complex. It is noteworthy that the length of its southwest
side is precisely twenty-one such units (plus 3.5 cubits for the thickness of the southeastern
wall, which is actually off-plan because of its different orientation). As a matter of fact, many
other measurements can be explained on the basis of this unit. Its use results in a grid illus-
trated in thin lines in figure 24.6, which shows a remarkable correspondence with the plan of
the building. To evaluate this correspondence, it is essential to bear in mind that grid planning
was not designed to provide the builders with the plan of the building, but rather to give them
guidelines for the alignment and placement of the main walls. For that reason, one should not
look for an exact correspondence between the grid and all the walls, nor should the walls be
necessarily positioned on the same side of the lines of the grid; they can be on either side, or
even on the line itself, preferably astride when a wall is comprised of two rows of bricks. In
the case of Palace B, it is clear that the placement of the main walls corresponds closely, and
sometimes even exactly, to that of a grid with a module of 7.5 cubits. For example, the place-
ment of the buttresses along the southwest side was clearly determined with the help of such
a module.



6



Similarly, the layout of the rooms and of the long corridor in Squares Q-X 24–25
was probably achieved in the same way, as the lines of the grid correspond precisely with the
face of the walls. The positioning of the walls in Squares Q-R 24–35 also shows a remarkable
coincidence with that of the grid.

Yet this grid is not entirely satisfactory because on the northwest side, where the intervals
between the buttresses are of 4.5 cubits instead of 4 cubits, the combined length of a buttress
plus an interval is of 8 cubits instead of 7.5 cubits. For this reason it may be appropriate to
look for a module of another value, possibly the double of the preceding one, i.e., 15 cubits
instead of 7.5 cubits. Such a length permitted easier calculation, and it should be remembered
that in contemporary Egypt, architects also favored modules in multiples of five (Lauer and
Leclant 1972: 56–57; Labrousse, Lauer, and Leclant 1977: 71–72).

The resulting grid, illustrated in heavy lines in figure 24.6, shows of course the same
remarkable correspondence with the lines of the main walls of the plan, but it also reveals sev-
eral additional elements that should be mentioned despite their admittedly hypothetical char-
acter. First, the use of a grid with a 15-cubit mesh might offer an explanation of why an
interval of 4.5 cubits instead of 4 cubits was applied along the northwest wall: assuming that
the position of the long NW-SE wall in Q 24–32 was fixed initially, the widening of the inter-
vals between the buttresses was necessary in order to avoid the seventh buttress being built



6. The same module was clearly used for the placement of the buttresses along the southeast wall, but the latter
has a different orientation for reasons still unknown.

oi.uchicago.edu





477



METROLOGY AND THE BIRTH OF ARCHITECTURE IN ANCIENT PALESTINE



right against the wall of either end. Second, it is noteworthy that the straight line of the south-
west wall measures ten units of a 15-cubit length, and that five such units separate the inner
face of the buttresses of the southwest peripheral wall from the 55-meter long wall in R 24–
35, which is almost exactly parallel to the outer southwest wall, thus forming a large rectangle
of ten by five units of 15 cubits each. Third, noting that the west corner of the complex is a
perfect right angle, it could be observed that the distance between the outer wall and the wall
in Q 24–32 is of four such units, suggesting that it may have served as one side of a “3-4-5”
triangle. These three observations might of course be coincidental, but their convergence
strengthens the plausibility of the use of grid with a 15-cubit mesh.



Public versus Domestic Buildings



In addition to size and quality of construction technique, large-scale planning and the use
of a cubit as a measure of length are features that sharply distinguish the palatial complex from



FIGURE 24.6. Yarmuth: schematic plan of Palace B with a superimposed orthogonal grid. The grid is based on
a module of 7.5 cubits (thin lines) and 15 cubits (heavy lines).

oi.uchicago.edu

[image: image181.tif]



478



PIERRE DE MIROSCHEDJI



most other contemporary EB IIIC buildings of Yarmuth. These buildings are well known
through excavations in other areas; especially significant are the domestic buildings of Area G,
Stratum G-2 (fig. 24.1) because they are stratigraphically related to the palatial complex and
have been cleared on an area of about 600 m



2



. With such a large exposure, the database is large
enough to confidently state that none of these dwellings exhibit architectural features compa-
rable to those of the palatial complex. None of them, in particular, betrays the use of a unit of
measurement similar to the cubit used in planning the palatial complex; as a rule, their walls
are only 0.70–0.75 m thick, and their type of masonry is variable. In addition, they were not
planned together with the palatial complex as shown by the fact that they present different lay-
outs and orientations (fig. 24.1).

The difference between public and domestic buildings can be shown at best when compar-
ing the process of construction planning. While the palatial complex is the result of one single
large-scale building operation, the contemporary domestic houses were built without overall
planning. They were added one to another through time, in an agglutinative process which
sometimes necessitated the establishment of awkward junctions between corners of separated
buildings. Two examples from Yarmuth are illustrated on figure 24.7 (see caption for com-
ment). When compared with figure 24.1, they eloquently show the difference between the
building techniques used for domestic or for public architecture. Many other similar examples
can be given in Yarmuth and in other contemporary sites.

How can this difference be expressed, besides the obvious functional dichotomy of the two
building types? Presumably, the EB IIIC domestic buildings and the palatial complex were
planned and built by different people: in one case, we are dealing with dwellings built by their
owners according to elementary building techniques transmitted from father to son; in the
other case, with a public building erected on orders from a professional builder possessing a
specialized knowledge in architecture. An essential aspect of this architectural knowledge—
and actually a sine qua non condition to put it in practice—was the use of a fixed unit of mea-
surement and that of specific planning techniques. In other words, the architectural data from
Yarmuth in the EB IIIC period illustrate an essential dichotomy between the persistence of an
art of building of immemorial antiquity and the emergence of architecture per se.

OTHER EB II–III METROLOGICAL DATA FROM YARMUTH

When did such a dichotomy appear? With several examples of public buildings earlier than
EB IIIC, the Yarmuth excavations offer the possibility of at least a partial answer to this ques-
tion. The use of a cubit for planning is the discriminant feature to look for in order to distin-
guish one building method from another. The pre-EB IIIC public constructions of Yarmuth
belong to two categories, freestanding buildings and fortifications.



Freestanding Buildings



The freestanding public buildings include the remains of Building C (EB IIIB) and those
of the so-called White Building (EB IIIA).

The scattered building remains ascribed to Building C cover possibly as much as 1,000 m



2



and characterize Stratum C-2 in Area C (Miroschedji et al. 1988: 42–43). The walls usually
have a deep foundation trench and appear with two different thicknesses: about 1.05–1.10 m

long

oi.uchicago.edu





479



METROLOGY AND THE BIRTH OF ARCHITECTURE IN ANCIENT PALESTINE



(i.e., 2 cubits), or between 1.30 m (toward the top of the wall) and 1.55–1.60 m (close to its
base at the bottom of the foundation trench), i.e., between 2.5 cubits and 3 cubits. Only two
rooms could be measured for size: Locus 239 is 3.00–3.10 m by 1.80–1.85 m (inner dimen-
sions), i.e., 6 cubits by 3.5 cubits; Locus 240 has the same width and its length is uncertain
(about 11 cubits). Although too few to be conclusive, these measurements raise the possibility
that the metrological system applied in the EB IIIC palatial complex of Area B was already in
use in the preceding phase for the construction of another large public building of palatial
character. Although the plan of this building is too fragmentary to be definite, it should also be
noted that the repartition of its scattered walls suggests that a grid with a module of 7.5 cubits
might also have been used for its layout.

Another freestanding public building dated to EB IIIA is the so-called White Building, prob-
ably a temple (Miroschedji et al. 1988: 35–41). Its walls are a little over 1 m thick (2 cubits),
the main entrance is 1.25–1.30 m wide (2.5 cubits), and the secondary entrance to the south
about 0.8 m (1.5 cubits). These measurements suggest that the EB IIIB–C metrological system
might have been also applied in the planning of this building. The inside measurements of the
main hall are 11.60 m by 4.75 m (22 by 9 cubits), and the outside measurements are thus 26
cubits by 13 cubits. This ratio of 2:1 indicates that the planning of the building was made by



FIGURE 24.7. Two samples of domestic architecture from Yarmuth, Area G, Stratum G-2. The circles call atten-
tion to cases of awkward junctions between initial walls (in hatched lines) and later additions (in gray).

oi.uchicago.edu

[image: image182.tif]



480



PIERRE DE MIROSCHEDJI



using the outside measurements. As this building is not very large, there is no evidence here of
the use of a planning grid.



Fortifications



The other public constructions of Yarmuth uncovered so far belong to the second cate-
gory, fortifications, and date to the EB II and EB III periods (for details and chronology, see
Miroschedji 1990). Unfortunately, fortifications do not lend themselves to easy measurements
because, for reasons inherent to their nature and their method of construction, their thickness
usually varies from place to place with an amplitude of a cubit or half a cubit; hence the met-
rological interpretation of the measurements is hazardous.

Among the striking features of the EB III fortifications are the platforms characterizing
Phase III of the defense system. They remain to be precisely measured, and the few data pres-
ently available indicate that they have different measurements. The length of platform 101 in
Area A could not be ascertained, and its width is 12.8 m (approx. 24 cubits). The width of
platform 302 in Area D is 7.35 m (14 cubits).

The late EB II fortifications are represented by the outer city wall (Wall B) and a city gate.
Due to its peculiar structure as a retaining wall, Wall B is attested with different thicknesses:
2.60 m (Area D), 3.00 m (Area A, square F 21), and 3.50–3.60 m (Acropolis). At the city gate,
however, where it is a freestanding wall built with header and stretcher masonry, its thickness
is 2.40 m (4.5 cubits), on both sides of the passageway, which is itself 2.40 m wide; the simi-
larity of these three measurements may well be significant.

Finally, the thickness of the earliest city wall, Wall A, dated to EB II, is also variable: a
little over 5.60 m (approx. 11 cubits) in Area A (Wall 131, measurement at top); 5.75 m (11
cubits) in Area C north (Walls 693 + 1008, measurement at base); between 3.50 m (at the
base) and 2.85 m (at the top) in Area C northwest (Wall 1018), i.e., between 6.5 and 5.5
cubits; and 2.35 m (4.5 cubits) in Area D (Wall 342). The southern buttress in Area A projects
2.30–2.40 m (4.5 cubits), and the northern buttress in Area D 7.35 m (14 cubits). The length
of the bastion cannot be measured precisely because of its trapezoidal shape; its width is about
13.40–13.50 m (25.5 cubits); it is limited to the east by a wall 3.20 m (6 cubits) thick (Wall
235); to the south by a wall 5.00–5.40 m (= 10 cubits) thick (Wall 188); and to the west by a
wall 3.75 m (7 cubits) thick (Wall 191). The absence of right angles is a remarkable feature of
these structures; and the way the bastion is attached to the rampart is also indicative of rudi-
mentary planning methods.

These measurements are regrettably taken at random and are too variable to be conclu-
sive. Moreover, several of them can be interpreted as multiples of an empirical cubit of about
0.45 m as well. Whether or not a cubit of around 0.52 m was already in use in Yarmuth in the
early EB II is therefore an uncertain possibility that should now be tested against the data
derived from other EB sites.

DATA FROM OTHER EARLY BRONZE AGE SITES IN PALESTINE

This task, however, presents considerable difficulties because other Palestinian sites of this
period lack the quantity of metrological data known from Yarmuth. In most cases, the dearth
of detailed measurements and the small scale of the published plans preclude any generaliza-

oi.uchicago.edu





481



METROLOGY AND THE BIRTH OF ARCHITECTURE IN ANCIENT PALESTINE



tion of the Yarmuth observations. The few examples reviewed below are borrowed from those
sites that have yielded the most impressive remains of EB public architecture, and which, at
the same time, have been published with more details.



Early Bronze Age I



For the EB I, it seems that neither the temples of Megiddo Stratum XIX (Loud 1948: 61;
Dunayevsky and Kempinski 1973: 167–68; Kempinski 1989: 170–75; 1992: 56) and of Hartuv
(Mazar and Miroschedji 1993, 1996) nor the large building of Tel Erani (Kempinski and
Gilead 1988, 1991; Nigro 1995: 7–11) illustrate the use of a cubit of a fixed length. At
Megiddo Stratum XIX, the walls have been measured at 1.20–1.30 m (2.5 cubits ?) to the east
and south, and at 3.00–3.05 m (6 cubits ?) to the west; at Hartuv, the thicker walls are about
1.00–1.10 m, but they show variations in their thickness, and other walls of the same building
are thinner. More significant, the layout of those buildings is rather irregular, with few right
angles, suggesting that no elaborate planning method was used. At Hartuv, though, the size of
Hall 134 has a ratio of 2:1, and Hall 152 might have had the same proportions. On the whole,
it appears that these EB I buildings are comparable to the public buildings of the Chalcolithic
period at ºEn Gedi (Kempinski 1992: 55) and Gilat (Alon and Levy 1989: 166–68, 176–82).



Early Bronze Age II



Observations are also inconclusive for most EB II sites. In the case of the fortifications at
Ai and Arad, the use of a cubit remains uncertain for lack of precise measurements (Callaway
1980: 117, fig. 75; Amiran et al. 1978: 11–13). For those of Tell Taºannak, Lapp has given
some measurements (Lapp 1967: 3–10): the earliest city wall (Wall 58) is 4.20 m thick (8
cubits ?), the latest (Wall 28) is 3.77 m (7 cubits ?), while the later EB III rectangular tower
(43) is 9.85 m thick (19 cubits ?). At Tell el-Fârºah, detailed plans of which were available to
this writer, no evidence of the use of a cubit could be established: although they were built
with mudbricks, the thickness of the city walls and of the city gate’s tower walls appears irreg-
ular (de Vaux 1962: 212–34). The same incertitude applies to the public buildings (temples
and “palaces”) of Tel Arad, which exhibit an irregular layout (Amiran, Alon, and Cohen 1976;
Nigro 1995: 12–16), a priori precluding the use of a large-scale planning operation and sug-
gesting rather that separate units were added one to another.

The situation is different with the so-called palace of Ai, actually a temple of really mon-
umental proportions. Its original layout dates to EB II, and its reconstruction to EB III (Calla-
way 1965; 1969: 1993: 41–42 [with bibliography]). The walls of both phases of this complex
are remarkably well built and regular, with an even thickness (on plan) of about 2.10 m, i.e., 4
cubits; it is interesting to observe here again the coincidence between elaborate building tech-
niques (sophisticated masonry, worked pillar bases with raised top, etc.) and the manifestation
of metrological concerns. The repetition of identical measurements suggests that a modular
system was used for the layout of this temple. But its plan is unfortunately so incomplete that
it is difficult to establish the length of the planning module. With much hesitation, I tentatively
suggest that a 5.5–cubit module was applied to the EB IIIA reconstruction of this temple (as
in Megiddo Strata XVII–XV; see below) because the resulting grid shows a rather good cor-
respondence with the plan of the building.

oi.uchicago.edu





482



PIERRE DE MIROSCHEDJI



Early Bronze Age III



The EB III period offers more spectacular evidence of monumental architecture and more
significant instances of architectural planning.

A famous EB III public building discovered in Palestine is the monumental granary of Beit
Yera



˙



(Maisler, Stekelis, and Avi-Yonah 1952: 223–28; Mazar, chap. 23, this volume). Here
again, only random measurements could be made with uncertain results. The inner circles mea-
sure between 7.6 m and 8.0 m in diameter (about 15 cubits), except for circles II and III, which
are larger, 8.75–8.95 m (about 17 cubits). The dividing walls are approximately 1.00 m (2
cubits). The entrance passage (Locus 7) is about 3.70 m (approx. 7 cubits) wide at the entrance
and about 3.15 m (6 cubits) at its end. The width of Locus 6 is about 6.85 m (13 cubits); that
of Locus 2 is 4.35–4.50 m (about 8.5 cubits). None of these measurements is really precise, and
their evaluation in “Yarmuthian” cubits is therefore dubious, although not impossible. Yet it
remains questionable whether a grid system was used for the layout of this building because
it has few right angles.

Megiddo Strata XVII–XV presents a most interesting example of the EB III public archi-
tecture because of the quality and size of the building remains uncovered and because of their
approximate contemporaneity with those of Yarmuth. They consist of three temples and a
palace.

The EB III temples comprise a single temple (5192) ascribed to Stratum XVII(?)–XVI
and a complex of twin temples (5192 and 5269) related to Stratum XV (Loud 1948: 78;
Dunayevsky and Kempinski 1973; Kempinski 1989: 175–77; Esse 1991: 83–90). These build-
ings can be studied at leisure as they are still standing today. Measurements, however, are at
times imprecise because the walls have been partly reconstructed. Their thickness varies
between 1.90 m and 2.10 m, the majority being in the range of 2.0–2.1 m (4 cubits).



7



The
length of walls can also be converted in cubits, with little approximation: the size of the main
building (external measurements of hall and portico) is between 17.6 and 17.8 m on each side
(33.5 or 34 cubits). Precise measurements were given by the excavators for the two preserved
altars: the altar of Temple 4040 measures 2.60 m x 2.20 m x 1.05 m, corresponding exactly to
5 x 4 x 2 cubits; and the altar of Temple 5192, whose full height was not preserved, measures
5.25 x 3.95 m, i.e., exactly 10 x 7.5 cubits (a ratio of 4:3). Although these data do not offer a
statistical basis, they nevertheless appear coherent enough to sustain the conclusion that a
cubit of 0.525 m was also used for the building of these EB III temples (contra Milson 1988).

Even more interesting is the analysis of their layout. The similarity of their dimensions and
the perfection of their right angles clearly indicate that they are the result of a planning tech-
nique involving a knowledge of elementary geometry. The example of Yarmuth discussed
above suggests that the characteristic elements of their plan can be explained at best by the
assumption that the builders used a modular grid system for their layout. However, application



7. Here is a list of measurements taken in the field:



Temple 4040



: North wall: 1.95–1.96 m; South wall: 2.00–2.05 m; East wall: 2.00 m; West wall: 1.90 m.
Small wall protruding from the western ante: 0.85–0.86 m.



Temple 5192



: North wall: 2.06–2.07 m; South wall: 2.05–2.06 m; East wall: 1.92 m; West wall: 1.85 m to
2.10 m.



Temple 5269



: East wall: 1.82 m to 2.00 m; West wall: 1.92 m to 1.95 m.



a bit long

oi.uchicago.edu





483



METROLOGY AND THE BIRTH OF ARCHITECTURE IN ANCIENT PALESTINE



of the “Yarmuthian” grid with a mesh of 7.5 and 15 cubits does not give a convincing result,
thus indicating that we must search for a module of another value, calculated ad hoc for these
buildings. On the basis of the various measurements taken and their interrelationships, the
value of 11 cubits (i.e., 5.5 cu. x 2) recommends itself as a likely possibility.



8



The resulting grid fits tightly to the position of all the walls, especially in the case of the
twin temples (fig. 24.8). It also reveals the major geometric figures possibly used by the builder
to lay out the plan of each temple: the complete building (hall, portico, and adjacent room) is
inscribed in a rectangle of 4 x 3 units of 11 cubits, the main building (hall and portico) in a
square of 3 x 3 units, while the proportions of the main hall fit a rectangle of 3 x 2 units, in
which two “3-4-5” triangles can be drawn for the construction of the right angles. Since these
geometric figures fit so precisely the main elements of the plan, their conjunction can hardly
be accidental, and the proposed governing grid appears as a likely hypothesis.



9



It is noteworthy that all these figures can readily be constructed with a twelve-knotted
rope and imply a very simple knowledge of geometry, such as can be expected for a builder
of third millennium Palestine.

Other interesting observations can be added. The similarity in plan and proportions of the
twin temples 5192 and 5269 shows that they were constructed together, the grid of Temple 5192
apparently determining the position of the southeastern wall of Temple 5269; however, the grid
of the latter temple, although identical to that of Temple 5192, was slightly offset to the south-
west. On the other hand, Temple 4040 is the result of a distinct building operation, as shown
by the difference in orientation and proportions, and by the fact that the grid fits less precisely
to its structure; this conclusion is of course sustained by the stratigraphic observations of
Dunayevsky and Kempinski which have established that Temple 4040 was built before the twin
temples (Dunayevsky and Kempinski 1973: 165).

Another important EB III public building of Megiddo is Building 3177 of Stratum XVI,
most probably a palace (Loud 1948: 70–76; Kempinski 1989: 30, 155; Esse 1991: 84). Its pub-
lication is so succinct that it is difficult to ascertain on the basis of the thickness of the walls
whether a cubit of the same value as that in Yarmuth and in the Megiddo temples was used.
However, the similarities in architectural details with the palace of Yarmuth are so striking
that this possibility cannot be excluded, and it is retained here as a working hypothesis.

The analysis of the layout of this palace is also difficult because the building was only par-
tially excavated, and its builders could not design it freely as they had to adapt to the adjacent
terrace walls (Walls 4114 and 4045), with the result that walls of different orientations were
built on the western and southern sides of the complex. Application of a grid with a module
of 11 cubits (5.5 cu. x 2), the same as for the temples, shows a rather good correspondence with
most walls of the palace and gives the width of several halls, rooms, and corridors (fig. 24.9).



8. This module was calculated by dividing by 3 the length of the back wall of Temple 4040, when it was rec-
ognized that this wall constitutes apparently one side of 3–4–5 triangle.

9. This conclusion could not be formulated by Milson (1988) because of a basic methodological flaw: the
wrong assumption that the layout of a building is planned on the basis of the inner dimensions of its
rooms—while it is always established with the outer dimensions, using simple geometric figures which can
be constructed with the help of a 12 knotted rope. For that reason, I cannot follow Milson (1986, 1987, 1988)
in his attempts to explain the layout of several Early and Middle Bronze Age buildings of Megiddo and
Shechem through elaborated planning techniques involving the calculation of square roots.

oi.uchicago.edu





484



PIERRE DE MIROSCHEDJI



FIGURE 24.8. Schematic plan of the Megiddo Temples of Strata XVII(?)–XV, duplicated for sake of clarity. The
superimposed orthogonal grids are identical in size and based on a module of 5.5 cubits (thin lines) and 11 cubits
(heavy lines) (plan after Loud 1948: fig. 304).

oi.uchicago.edu

[image: image183.tif]



485



METROLOGY AND THE BIRTH OF ARCHITECTURE IN ANCIENT PALESTINE



To assume that such a grid was used to draw the layout of the palace is thus a plausible
hypothesis. It should be noted that it has the same orientation as that of Temple 4040; this
observation suggests that Temple 4040 and Palace 3177 were built at the same time, as part of
the same large-scale building operation.



10



CONCLUSIONS

This brief review of the evidence of metrology and architectural planning on some EB sites
is admittedly incomplete. It tends, however, to confirm and to specify the major conclusions
reached in the previous sections of this paper, which are summarized here.

The first conclusion concerns the introduction during the EB of a standard unit of length
that heralds the appearance of metrology. Based mainly on the data from Yarmuth and
Megiddo, the evidence establishes firmly the widespread use in the EB III of a cubit of
approximately 0.525 m, longer than the empirical cubit of 0.45 m and presumably borrowed
from Egypt, where its equivalent is attested both earlier and on a much larger scale.

What remains uncertain is the adoption date of this long cubit in Palestine. The evidence of
its use in the EB II is dubious, being based mainly on the hypothetical case of the Acropolis
temple at Ai and the rather problematical measurements of fortifications at several sites. It is
therefore safe to conclude, until proof of the contrary is produced, that the use of this cubit
started sometime during the latter half of the EB II, and that it was well entrenched in the EB III.

The second conclusion of this study is that the appearance of a fixed unit of length was
accompanied by the elaboration of new techniques of architectural design. The basic fact that
modular architectural planning was employed in the Early Bronze Age can hardly be denied:
the EB III data from Yarmuth and Megiddo clearly indicate the use of an orthogonal grid sys-
tem based on a module corresponding to a multiple of the cubit. Earlier evidence of the use of
this technique is based only on the Acropolis temple at Ai, so that the introduction of this tech-
nique can be dated again to the latter half of the EB II at the earliest. There is no reason to
ascribe it to Egypt, although the data briefly mentioned in the first section of this paper indi-
cate that grid planning is well attested in the contemporary Nile Valley, both for monuments
and for works of art.

The concomitant introduction of a fixed measure of length and new techniques of architec-
tural planning was necessary in order to erect the larger public buildings required to satisfy the
ambitions of an emerging political power. A third conclusion is precisely the appearance of
monumental architecture in EB III Palestine. This concept, however, needs to be qualified,
because it is not based only on size. Large public buildings erected for cultic purposes existed
in Palestine as early as the Chalcolithic period and grew increasingly larger in the course of the
Early Bronze Age. But with the possible exception of the Megiddo temples, these buildings
simply expanded the traditional broadroom plan of the earliest Chalcolithic temples, which
themselves imitated the contemporary domestic dwellings (see Miroschedji 1993c: 208–11);
hence, from the point of view of architectural design, these temples were not new and merely



10. This was already implied by Dunayevsky and Kempinski 1973: 169 (sub level XVI) and caption to fig. 7
(Stratum XVII); see also Nigro 1995: 19, fig. 6. Note also that a pillar base with a raised center identical in
shape and technique to those found in Temple 4040 was uncovered in room 5 of Palace 3177.

oi.uchicago.edu





486



PIERRE DE MIROSCHEDJI



continued a well-entrenched building tradition. The same remark applies to non-cultic build-
ings of possibly public character such as the “palace” of Tel Arad (Amiran, Alon, and Cohen
1976: 37; Nigro 1995: 12–16), which consists simply in the grouping of large and well-con-
structed dwellings of ordinary plan.

The appearance of large buildings whose design represents an entirely new formula marks
the EB III period, because nothing like the Beit Yera



˙



granary and the Megiddo and Yarmuth
buildings had ever existed before in Palestine. For these buildings, modular grid planning
offered the necessary tool for both the conception of the layout and its realization on the
ground. The planned building was no longer the simple addition of several components added
one to another to ultimately form a large complex, but rather one single complex conceived
and built as a unit from the very beginning. Moreover, as the layout could be expanded at will
depending on the size of the module, the building was monumental in essence whatever its
actual dimensions.

These developments justify our fourth conclusion: that the EB has seen the birth of archi-
tecture



stricto sensu



, different from the traditional art of building, and the emergence of a new
specialist, the architect, different from the traditional builder. This conclusion is best exempli-
fied by the increasing dichotomy between domestic and public buildings: the differences are
seen not only in the size of the buildings, the quality and nature of their construction techniques
(foundation trenches for walls, particular masonry, floor and wall plastering, etc.), and several
constructional details (doorjambs, thresholds, etc.); but more importantly in the use of a “spe-
cialized” cubit of 0.525 m—longer than the empirical cubit of 0.45 m and later known as the
“builder’s cubit”—and in new designing methods.



FIGURE 24.9. Schematic plan of Megiddo Palace 3177 of Stratum XVII(?)–XVI with a superimposed orthogo-
nal grid identical in size and orientation to that of Temple 4040 on fig. 24.8 (plan after Loud 1948: fig. 304).

oi.uchicago.edu

[image: image184.tif]



487



METROLOGY AND THE BIRTH OF ARCHITECTURE IN ANCIENT PALESTINE



These developments were ultimately the result of expanding urbanization and the growing
complexity of society culminating in the establishment of city-states. The best archaeological
correlate of the emergence of a stronger political power and a deeper social hierarchy is pro-
vided by the existence of palaces, side by side with ordinary dwelling areas, as can be seen
clearly in Yarmuth (see fig. 24.1).

The construction of these prestige buildings necessitated a specialized knowledge. In tra-
ditional societies, technical expertise was shared by a small number of specialists who were
therefore in great demand by local rulers, who used to request or exchange them like gifts (see
Zaccagnini 1983 for analysis of the mobility of craftsmen in the ancient Near East, including
builders). We may suppose that in third-millennium Palestine, specialized builders traveled
also from place to place and transmitted their knowledge orally. Actually, very few itinerant
architects might have been needed per generation to build all the public buildings ever erected
in all the city-states of EB III Palestine. Their mobility presumably contributed to the spread
of the new techniques for measuring lengths and for planning. This hypothesis may explain
why, for example, Palace B of Yarmuth and Building 3177 of Megiddo share so many simi-
larities in constructional details.

In antiquity, the act of measuring and planning was always embodied with a diffuse sym-
bolic meaning. One reason is that all distances measured, because they were based on the cubit,
were harmonically related to the human body, which in turn was perceived as the measure of
the Universe because man had been created in God’s image. Hence the canon and the propor-
tions used in Near Eastern art for the representation of the human body were determined by the
length of a cubit and related to that of an architectural unit of measure (see, among others,
Iversen 1975, 1990; and Badawy 1962 for Egypt; Azarpay 1987, 1995 for Mesopotamia and
Iran). These conceptions received considerable elaboration in classical Greece and provided
the philosophical rationale for the Ionian urban layouts (Martin 1956).

Whether symbolic preoccupations of this nature were ever present among Palestinian
builders of the EB and their sponsors is of course a moot point, impossible to document with
archaeological data. But it cannot be insignificant that the first appearance of metrology and
of modular architectural planning took place, in Palestine as elsewhere, when the first city-
states emerged and their inhabitants had to forge a new relationship with their environment in
both space and time.

REFERENCES



Alon, D., and Levy, T. E.
1989 The Archaeology of Cult and the Chalcolithic Sanctuary at Gilat.



Journal of Mediterranean
Archaeology



2: 163–221.

Amiran, R.; Alon, D.; and Cohen, C.
1976 A Public Building Area in the Ancient Canaanite City of Arad (Report on the Ninth and Tenth

Seasons of Excavations, 1974–75).



Israel Museum News



11: 35–40.

Amiran, R.; Paran, U.; Shiloh, Y.; Brown, R.; Tsafrir, Y.; and Ben-Tor, A.
1978



Early Arad, The Chalcolithic Settlement and the Early Bronze Age City.



I:



First–Fifth Seasons
of Excavations, 1962–1966.



Jerusalem: Israel Exploration Society.

Arnold, D.
1991



Building in Egypt: Pharaonic Stone Masonry.



Oxford: Oxford University Press.

oi.uchicago.edu





488



PIERRE DE MIROSCHEDJI



Aurenche, O.
1981



La maison orientale. L’architecture du Proche-Orient des origines au milieu du quatrième
millénaire.



2 vols. Paris: Geuthner.
1993 L’origine de la brique dans le Proche-Orient ancien. Pp. 71–85 in



Between the Rivers and the
Mountains,



eds. M. Frangipane, H. Hauptmann, M. Liverani, P. Matthiae, and M. Mellink.
Archaeologica Anatolica et Mesopotamica Alba Palmieri Dedicata. Rome: Università di Roma
La Sapienza.

Azarpay, G.
1987 Proportional Guidelines in Ancient Near Eastern Art.



Journal of Near Eastern Studies



46:
183–213.

1995 Proportions in Ancient Near Eastern Art. Pp. 2507–20 in



Civilizations of the Ancient Near East,



ed.-in-chief, J. M. Sasson, volume 4. New York: Scribner’s.

Badawy, A.
1954



A History of Egyptian Architecture



. Vol. 1, ed. A. Badawy. Cairo: Imprimerie Urwand Fils.
1961 The Harmonic System of Architectural Design in Ancient Egypt.



Mitteilungen des Instituts für
Orientsforschung der Deutschen Akademie der Wissenschaft zu Berlin



8: 1–14.
1962 The Modular System of Egyptian Town Plans.



Bibliotheca Orientalis



19: 207–13.

Barkay, G., and Kloner, A.
1986 Jerusalem Tombs from the Days of the First Temple.



Biblical Archaeology Review



12: 23–39.

Barrois, A.
1953



Manuel d’archéologie biblique.



Tome I. Paris: A. Picard.

Bechmann, R.
1993



Villard de Honnecourt: la pensée technique au XIII



e



siècle et sa communication.



Paris: A.
Picard.

Ben-David, A.
1978 The Hebrew-Phoenician Cubit.



Palestine Exploration Quarterly



110: 27–28.

Bottéro, J., and Kramer, S. N.
1989



Lorsque les dieux faisaient l’homme



,



Mythologie mésopotamienne



. Paris: Gallimard.

Callaway, J. A.
1965 The 1964 ºAi (et-Tell) Excavations.



Bulletin of the American Schools of Oriental Research



178:
13–40.

1969 The 1966 ºAi (et-Tell) Excavations.



Bulletin of the American Schools of Oriental Research



207:
41–53.

1980



The Early Bronze Age Citadel and Lower City at Ai (et-Tell), A Report of the Joint Archaeologi-
cal Expedition to Ai (et-Tell): No.2.



Cambridge: American Schools of Oriental Research.
1993 Ai. Pp. 39–45 in



The New Encyclopedia of Archaeological Excavations in the Holy Land



, vol.
1, ed. E. Stern. Jerusalem: Israel Exploration Society & Carta and New York: Simon & Schuster.

Choisy, A.
1971



Vitruve.



Reprint of the 1909 edition. Paris: de Nobele.
1983



Histoire de l’architecture.



Tome I. Reprint of the 1899 edition. Paris-Geneva: Slatkine Reprints.

de Vaux, R.
1960



Les institutions de l’Ancien Testament



. Tome I. 4th edition. Paris: Éditions du Cerf.
1962 Les fouilles de Tell el-Fârºah, rapport préliminaire sur les 7



e



, 8



e



, 9



e



campagnes, 1958–1960
(suite).



Revue Biblique



69: 212–53.

Dever, W. G.
1974 The MB IIC Stratification in the Northwest Gate Area at Shechem. Bulletin of the American

Schools of Oriental Research 216: 31–52.

Dreyer, G.; Hartung, U.; and Pumpenmeier, F.
1993 Umm el-Qaab. Nachuntersuchungen im frühzeitlichen Königsfriedhof, 5/6. Vorbericht. Mittei-

lungen des Deutschen Archäologisches Institut, Abteilung Kairo 49: 23–62.

Dunayevsky, I., and Kempinski, A.
1973 The Megiddo Temples. Zeitschrift des Deutschen Palästina-Vereins 89: 161–87.

a tiny bit short

oi.uchicago.edu



489METROLOGY AND THE BIRTH OF ARCHITECTURE IN ANCIENT PALESTINE

Edwards, I. E. S.
1992 Les pyramides d’Égypte. Paris: Librairie Générale Français.

Eichmann, R.
1991 Aspekte prähistorischer Grundrißgestaltung in Vorderasien. Beiträge zum Verständnis

bestimmter Grundrißmerkmale in ausgewählten neolithischen und chalkolithischen Siedlun-
gen des 9.-4. Jahrtausends v. Chr. (mit Beispielen aus der europäischen Prähistorie). Bagh-
dader Forschungen, Band 12. Deutsches Archäologisches Institut, Abteilung Baghdad. Mainz
am Rhein: von Zabern.

Emery, W.
1938 Excavations at Saqqara: The Tomb of Hemaka. Service des Antiquités de l’Égypte 19. Cairo:

Government Press.
1939 Excavations at Saqqara 1937–1938: Hor-Aha. Service des Antiquités de l’Égypte 20. Cairo:

Government Press.

Esse, D. L.
1991 Subsistence, Trade, and Social Change in Early Bronze Age Palestine. Studies in Ancient Ori-

ental Civilization 50. Chicago: Oriental Institute.

Forest, J. D.
1983 Aux origines de l’architecture obéidienne: les plans de type Samarra. Akkadica 34: 1–47.
1991 Le système de mesure de longueur obéidien, sa mise en oeuvre, sa signification. Paléorient 17:

161–72.
1993 Review of Aspekte prähistorischer Grundrißgestaltung in Vorderasien, by R. Eichmann, Revue

d’assyriologie et d’archéologie orientale 87: 161–81.

Franck, D. R.
1975 Versuch zur Rekonstruktion von Bauregeln und Massordnung einer nordsyrischen Stadt des

vierten Jahrtausends—Untersucht anhand von Grabungsergebnissen der Deutschen Orient-
Gesellschaft in Habuba Kabira. Mitteilungen des Deutschen Orient-Gesellschaft 107: 7–16.

Gardiner, A. H.
1957 Egyptian Grammar. Oxford: Clarendon.

Heinrich, E.
1984 Die Paläste im alten Mesopotamien. Berlin: De Gruyter.

Heinrich, E., and Seidl, U.
1967 Grundrißzeichnungen aus dem Alten Orient. Mitteilungen des Deutschen Orient-Gesellschaft

98: 24–45.
1968 Maß und öbermaß in der Dimensionierung von Bauwerken im alten Zweistromland. Mitteilun-

gen des Deutschen Orient-Gesellschaft 99: 5–54.

Helck, W.
1980 Maße und Gewichte (pharaonische Zeit). Col. 1199–1209 in Lexikon der Ägyptologie, Band III,

eds. W. Helck and W. Westendorf. Wiesbaden: Harrassowitz.

Iversen, E.
1975 Canon and Proportions in Egyptian Art. 2nd ed. (with Y. Shibata). Warminster: Aris & Phillips.
1990 Metrology and Canon. Mitteilungen des Deutschen Archäologischen Instituts Abteilung Kairo

46:113–25.

Kaiser, W.
1982 Umm el-Qaab Nachuntersuchungen in frühzeitlichen Königsfriedhof. 2. Vorbericht. Mitteilun-

gen des Deutschen Archäologischen Instituts, Abteilung Kairo 38: 260–69.

Kaufman, A. S.
1984 Determining the Length of the Medium Cubit. Palestine Exploration Quarterly 116: 120–32.

Kempinski, A.
1989 Megiddo: A City-State and Royal Centre in North Israel. Materialen zur Allgemeinen und Ver-

gleichenden Archäologie, Band 40. Munich: C. H. Beck.
1992 Chalcholithic and Early Bronze Age Temples. Pp. 53–59 in The Architecture of Ancient Israel,

eds. A. Kempinski and R. Reich. Jerusalem: Israel Exploration Society.

oi.uchicago.edu



490 PIERRE DE MIROSCHEDJI

Kempinski, A., and Gilead, I.
1988 Tel Erani, 1987. Israel Exploration Journal 38: 88–90.
1991 New Excavations at Tel Erani: A Preliminary Report of the 1985–1988 Seasons. Tel Aviv 18:

164–91.

Kubba, S. A. A.
1990 The Ubaid Period: Evidence of Architectural Planning and the Use of a Standard Unit of

Measurement—The “Ubaid Cubit” in Mesopotamia. Paléorient 16: 45–55.

Labrousse, A.; Lauer, Ph.; and Leclant, J.
1977 Mission archéologique de Saqqarah II: Le temple haut du complexe funéraire du roi Ounas.

Bibliothèque d’Étude 73. Cairo: Publications de l’Institut Français d’Archéologie Orientale.

Lapp, P. W.
1967 The 1966 Excavations at Tell Taªannek. Bulletin of the American Schools of Oriental Research

185: 2–39.

Lauer, Ph., and Leclant, J.
1972 Mission archéologique de Saqqarah I. Le temple haut du complexe funéraire du roi Téti. Biblio-

thèque d’Étude 61. Cairo: Publications de l’Institut Français d’Archéologie Orientale.

Loud, G.
1948 Megiddo II, Seasons of 1935–39. Oriental Institute Publications 62. Chicago: University of Chi-

cago Press.

Maisler, B.; Stekelis, M.; and Avi-Yonah, M.
1952 The Excavations at Beth Yera˙ (Khirbet Kerak). Israel Exploration Journal 2: 218–29.

Margueron, J.-Cl.
1982 Recherches sur les palais mésopotamiens de l’âge du Bronze. 2 vols. Bibliothèque

archéologique et historique de l’Institut français d’archéologie orientale de Beyrouth 107. Paris:
Geuthner.

1986 Quelques remarques concernant l’architecture monumentale à l’Époque d’Obeid. Pp. 349–77 in
Préhistoire de la Mésopotamie, ed. J.-L. Huot. Colloques internationaux du Centre National de
la Recherche Scientifique. Paris: Éditions du Centre National de la Recherche Scientifique.

1987a L’apparition du palais au Proche-Orient. Pp. 9–38 in Le système palatial en Orient, en Grèce et
à Rome, ed. E. Lévy. Actes du Colloque de Strasbourg, 19–22 juin 1985. Strasbourg: Université
des Sciences Humaines.

1987b Les palais syriens à l’âge du Bronze. Pp. 127–58 in Le système palatial en Orient, en Grèce et
à Rome, ed. E. Lévy. Actes du Colloque de Strasbourg, 19–22 juin 1985. Strasbourg: Université
des Sciences Humaines.

Martin, R.
1956 L’urbanisme dans la Grèce antique. Paris: Picard.

Matthiae, P.
1993 Ebla au IIe millénaire. Pp. 162–65 in Syrie, mémoire et civilisation. Paris: Flammarion.

Mazar, A., and de Miroschedji, P.
1993 Hartuv. Pp. 584–85 in The New Encyclopedia of Archaeological Excavations in the Holy Land,

vol. 2, ed. E. Stern. Jerusalem: Israel Exploration Society & Carta and New York: Simon &
Schuster.

1996 Hartuv, An Aspect of the Early Bronze I Culture of Southern Israel. Bulletin of the American
Schools of Oriental Research 302: 1–40.

Meyer-Christian, W.
1987 Der “Pythagoras” in Ägypten am Beginn des Alten Reiches. Mitteilungen des Deutschen

Archäologisches Institut, Abteilung Kairo 43: 195–203.

Milson, D.
1986 The Design of the Royal Gates at Megiddo, Hazor, and Gezer. Zeitschrift des Deutschen Paläs-

tina-Vereins 102: 87–92.
1987 The Design of the Temples and Gates at Shechem. Palestine Exploration Quarterly 119: 97–

105.
1988 The Design of the Early Bronze Age Temples at Megiddo. Bulletin of the American Schools of

Oriental Research 272: 75–78.

oi.uchicago.edu



491METROLOGY AND THE BIRTH OF ARCHITECTURE IN ANCIENT PALESTINE

Miroschedji, P. de
1990 The Early Bronze Age Fortifications at Tel Yarmuth—An Interim Statement. Eretz-Israel (Ruth

Amiran Volume) 21: 48*–61*.
1991 Notes & News: Tel Yarmut, 1990. Israel Exploration Journal 41(4): 286–93.
1992 Notes & News: Tel Yarmut, 1992. Israel Exploration Journal 42(3–4): 265–72.
1993a Jarmuth, Tel. Pp. 661–65 in The New Encyclopedia of Archaeological Excavations in the Holy

Land, vol. 2, ed. E. Stern. Jerusalem: Israel Exploration Society & Carta and New York: Simon
& Schuster.

1993b Fouilles récentes à Tel Yarmouth, Israël (1989–1993). Compte rendu de l’Académie des Inscrip-
tions et Belles-Lettres 1993: 823–47.

1993c Cult and Religion in the Chalcolithic and Early Bronze Age. Pp. 208–20 in Biblical Archaeology
Today, 1990, Proceedings of the Second International Congress on Biblical Archaeology, eds.
A. Biran and J. Aviram. Jerusalem: Israel Exploration Society.

1994 Notes & News: Tel Yarmut, 1993. Israel Exploration Journal 44: 145–51.
1995a Les premières cités-États cananéennes. Les dossiers d’archéologie 203 (mai 1995): 81–100.
1995b The Excavations at Tel Yarmuth. Qadmoniot 28 (109): 27–38. (Hebrew).

Miroschedji, P. de, et al.
1988 Yarmouth 1: Rapport sur les trois premières campagnes de fouilles à Tel Yarmouth (Israël), 1980–

1982, by Pierre de Miroschedji, with contributions by S. Davis, P. Goldberg, A. Kermorvant,
E. Nodet, S. Rosen, and G. London. Paris: Éditions Recherche sur les Civilisations.

Nigro, L.
1995 L’architettura palatina della Palestina nelle Età del Bronzo e del Ferro. Contesto archeologico

e sviluppo storico. Contributi e Materiali di Archeologia Orientale 5. Rome: Università degli
studi di Roma La Sapienza.

Paul, Sh. M., and Dever, W. G.
1973 Biblical Archaeology. Jerusalem: Keter.

Petrie, W. F.
1892 The Tomb-Cutter’s Cubits at Jerusalem. Palestine Exploration Fund Quarterly Statement 25:

28–35.

Powell, M. A.
1989 Maße und Gewichte. Pp. 457–517 in Reallexikon der Assyriologie und Vorderasiatischen

Archäologie, Band 7, ed. D. O. Edzard. Berlin-New York: de Gruyter.
1992 Weights and Measures. Pp. 897–908 in The Anchor Bible Dictionary, vol. 6, ed. D. N. Freedman.

New York: Doubleday.

Scott, R. B.
1958 The Hebrew Cubit. Journal of Biblical Literature 77: 205–14.
1959 Weights and Measures in the Bible. Biblical Archaeologist 22: 22–40.

Toombs, L.E., and Wright, G. E.
1961 The Third Campaign at Balatah (Shechem). Bulletin of the American Schools of Oriental

Research 161: 11–54.

Ussishkin, D.
1976 The Original Length of the Siloam Tunnel in Jerusalem. Levant 8: 82–95.

Vandier, J.
1952 Manuel d’archéologie égyptienne. Tome I: Les époques de formation. 2 vols. Paris: A. J. Picard.

Wright, G. E.
1965 Shechem: The Biography of a Biblical City. New York-Toronto: McGraw Hill.

Wright, G. R. H.
1985 Ancient Buildings in South Syria and Palestine. Vol. 1. Handbuch der Orientalistik, Abt. 7, Bd.

1, Abschnitt 2, B, Lfg. 3. Leiden: E. J. Brill.

Zaccagnini, C.
1983 Patterns of Mobility among Ancient Near Eastern Craftsmen. Journal of Near Eastern Studies

42: 245–64.

oi.uchicago.edu



492 PIERRE DE MIROSCHEDJI

oi.uchicago.edu



493



STANDARDIZED COLLARED-RIM PITHOI
AND SHORT–LIVED SETTLEMENTS



Avner Raban



This paper is an attempt to reevaluate the actual archaeological data concerning a certain
type of pottery container in its chronological, spatial, and topographic context, in order to esti-
mate its “story.” A score of publications dealing with this type of vessel and all kinds of argu-
mentative articles have been presented since the “ethno-historic” potential of it was suggested
by Albright over sixty years ago (1937: 25).

Biblical archaeologists have either embraced the appearance of the collared-rim pithoi as
one of two major “objective” benchmarks which attest to “Israelite Settlement” (Albright 1949:
118; Aharoni 1970: 264–65; Garsiel and Finkelstein 1978), or refuted it (Ibrahim 1978; Esse
1992). The later “unethnic” trend is corroborated with the more recent “secular” fashion among
some scholars (Finkelstein 1996; contra Dever 1995). This secular trend is rationalized by
socioeconomic paradigms in which this vessel is seen to characterize the subsistence economy
of horticulture-based societies on the one hand (Finkelstein 1988: 285; Zertal 1988b; London
1989; Esse 1992: 95–103) or, on the other hand, is viewed as a product used in long-range,
imperially instigated trade of either spices from across the eastern deserts to the Mediterranean
(Artzy 1994), or the local horticultural products of the central highland—to Egypt, by sea
(Wengrow 1996). Yet, as more data become available from newly published excavations and
surveys, none of these proposals truly fits the ever-growing geographical and environmental
context of the collared-rim pithoi within the Iron Age I. Even the less pretentious assumptions—
that these vessels were made to accommodate the demands for storing potable water in rural set-
tlements at some distance from a stable water source (Zertal 1988b)— would not fit the spatial
distribution of these vessels at urban sites such as Megiddo (Esse 1992), Dan (Biran 1989), and
Beit Shean (A. Mazar 1994); or at sites located near rich water sources, such as Tel Zeror and
Tel Nami, on the coast, and lowland sites, such as Tel Yenoªam and Midrach-ºOz (Raban 1991).

COLLARED-RIM PITHOS AND ITS PHYSICAL PROPERTIES

The type of vessel under discussion is the one known as the “Central Highland” type
(Finkelstein 1988: 275), of the last quarter of the second millennium



b.c.



It differs in size and
manufacturing technique from the less-known contemporaneous “collared-rim jar” (see, e.g.,
Esse 1992: fig. 3:2), which is discussed here, for reasons that are explained below. This central
highland pithos has a body shape resembling contemporary and earlier local Middle Bronze
Age (MB) II and Late Bronze Age (LB) storage vessels, being of basic elongated oval form
that tapers toward its thickened tipped base (fig. 25.1a). It has two simple vertical handles that
are attached to the body either at the shoulder level or just below it, which is the widest part of
the body. The shoulders are rather high and either flattened or slightly curved. The rim is usually



25



oi.uchicago.edu





494



AVNER RABAN



outfolded with a variety of cross sections, of no clear chronological significance (Finkelstein
1988: 277–78). The short, concave neck is attached to the body by a widening ridge that char-
acterizes the vessel as a “collared-rim” one. This ridge has been considered a technical compo-
nent fashioned in order to cover and strengthen the binding line between the handmade body
and the wheel-thrown neck (A. Mazar 1981: 27; Glass et al. 1993: 279). Almost all specimens
of this type that we have studied are made of a prime quality clay, properly levigated with the
addition of carefully selected tempers (see below).

The walls of its oversize body are rather thin for such a bulky container and, in most cases,
of a surprisingly even thickness. These qualities were achieved by a sophisticated, profes-
sional, and highly standardized knowledge employed by the potters who would combine the
techniques of coils, built walls, on a turntable, with final modification on a fast wheel. There
are a series of horizontal grooves on the shoulders, or just below, in some variants (Ibrahim
1978: 117) and the remains of rope marks in other, inferior finished specimens, indicating that
the potters had to fasten the built-up body at some stage, probably to avoid distortion when the
clay was still rather wet and too soft to properly sustain the final shaping on a tournette or a
wheel. All in all, the selection of raw materials, the mixture of several types of clay, the selec-
tion of tempers, and the complicated process of producing the final form were rather demand-
ing and probably more difficult to achieve than any other type of clay vessel of the period
(Cohen-Weinberger and Goren 1996: 80–81; Daviau 1996: 609–12).

The other property of this pithos, which is rather unique and can be distinguished from the
other two contemporary ones—the so-called Cypro-Tyrian (fig. 25.1b) and the Galilean (fig.
25.1c) pithoi—is the high rate of uniformity in shape and external dimensions. Almost all of
the specimens that have been restored are 1.02–1.18 m high; with a maximum diameter of
about half this size—0.51 to 0.60 m; and with the mouth’s inner diameter measuring 0.15–0.19
m. These metrological data are the ones that the potter could use while building the vessel and
were probably the regular means, together with the size of the clay lump, by which a certain



FIGURE 25.1. Typical pithoi.



a:



Central highland country, collared-rim type;



b:



“Cypro-Tyrian” type;



c:



“Gali-
lean” type.



long

oi.uchicago.edu

[image: image185.tif]



495



STANDARDIZED COLLARED-RIM PITHOI



standardized container was then produced (Raban 1981: 2–5; and see appendix below). It is
tempting to suggest that there was a general notion, if not dictated measurements, for the pot-
ters to meet. The required demand was probably for a standard volume that could be achieved
by using the basic units of the “royal cubit”: one for the maximum diameter and two for the
vessel’s height. These units might have been measured inside the vessel (internal measure-
ments) in order to be more accurately relevant for the desired volume. The actual volume of
most restored specimens of this type of vessel have seldom been checked; or if so, they have
not been published. Some scholars have casually mentioned that the capacity of “their” col-
lared-rim pithoi is between 150–200 liters (Ibrahim 1978), or that the “total weight when
loaded is 80–120 kg” (Artzy 1994: 137). Our calculation, for a specimen 1.10 m high and
0.54 m wide, is 110–20 liters.

Yet, with all due respect to the general notion of a highly standardized type of vessel, one
might recall that the specimens come not only from widely dispersed locations, but also from
chronological contexts that span almost three centuries. So, metrological uniformity is relative
and conditional, with a trend toward reduced volume through time, as seems to be the case
for every metrological standard in ancient history (Grace 1965: 12; Will 1977; Raban 1981:
195–227). There are, though, a few exceptional, oversize specimens, such as the single restor-
able pithos from Giloh, which is said to be 1.50 m tall (A. Mazar 1990: 88), although the
calibrated drawing (A. Mazar 1990: fig. 5) does not fit! There are also smaller ones, with a
height of just below 1 meter (see, e.g., table 25.1, nos. 12, 23 in the appendix).

When comparing the metrologic data of the Iron Age I collared-rim pithoi with other types
of allegedly well-standardized containers, such as LB commercial jars, or even Rhodian wine
amphoras, one may find the uniformity rate of this “inferior” type to be clearly higher. This fact
calls into question the notion that it is a typical product of “household women” (Esse 1992),
used for storage of potable water in villages (Zertal 1988b), sedentary household liquids
(Finkelstein 1988: 283), and for use in the horticulture subsistence economy (Finkelstein 1996:
204). Summing up the facts, one might consider the characteristics of the collared-rim pithos
of the early Iron Age period, of the central highland type, as such:

1. It is a pithos by definition—being too heavy to be carried from one place to another
when filled. So it is a storage container rather than a portable one.

2. As a ceamic product, it is of high quality and rather sophisticated, with its raw mate-
rial carefully prepared, the tempers properly selected, and the various stages of creating
the desired form by the potter demonstrating the best professional coordination.

3. All together, shape, external measurements, and volume represent a successful, delib-
erate attempt to meet a specific standard of uniformity, which is quite rare in pithoi
throughout history.

4. Though it is a vessel destined for storage in a fixed location, its shape resembles
portable jars of the period and of the preceding ones in the region.

STYLISTIC AND CONCEPTUAL PROTOTYPES

It is almost unquestionable that the type of pithoi called “Central Highland Collared-Rim”
of the early Iron Age has its closest prototypes among the pithoi of the MB II period (Finkel-
stein 1988: 281; Bonfil 1992). The various subtypes of these earlier pithoi were found in urban

oi.uchicago.edu





496



AVNER RABAN



and rural sites, almost all over Palestine and in Syria, as far north as Ebla and Ugarit. It seems
that there were regional variants, which quite closely resemble the geographic subdivision of
the later early Iron Age’s three different types of pithoi: (a) Bonfil’s Type I, of the MB II
period, matches the territory of the so-called Galilean Iron I pithoi, although stylistically they
more closely resemble the LB II types of Hazor (Bonfil 1992: 26: fig. 9); (b) Bonfil’s Types III
and V were probably in vogue along the coast, in the Jezreel Valley, and in the adjacent hill
country, much the same territory in which the so-called Tyrian, or Cypro-Phoenician, types of
the Iron I period are to be found; (c) Bonfil’s Types IV and VI are characteristic of the central
highland and southern Jordan valley, much like the highland pithoi of the Iron I period. Yet,
one must recall that this geographical division is in fact not definite in both periods. Foreign
types have been found in each geographic unit, both in MB II contexts (Bonfil 1992: fig. 9)
and in the Iron I period—as in the case of Tel Dan, in which all three variants have been
uncovered (Biran 1989). The MB II prototypes are of the same high quality and of the same
conceptual form of oversized jars as the early Iron Age highland collared-rim pithoi. It is true
that in most cases they were of a better technological tradition—being formed on a fast wheel
and furnished with a variety of well-formed rims. But these prototypes lack uniformity in size
and, unlike their successor, they had either no handles or four handles—in most specimens of
the highland Types IV and VI (Bonfil 1992: 28–32). For these differences one would consider
the MB II prototypes as better suited for handling while full, either being lifted by two to four
persons by their four handles, or being slung up in a net, as was the case of the handleless vari-
ants (fig. 25.2).

Of special interest is the similarity in contextual situations of both the MB II and the
Iron I pithoi. These vessels were found mostly in storage rooms next to the fortification sys-
tem of the settlement (Bonfil 1992: 32–3; and for the Iron I pithoi, see below). Others were
found in what might have been temple storerooms, or other religious contexts (Finkelstein
1986: 29, 34), or in burials—either as gifts or as coffins—quite probably in a secondary use
(although pithoi as coffins are known since the Neolithic era, such as in the burial ground of
Karatash-Semayük, and see, e.g., Mellink 1972).

It is true that there are almost no specimens of that conceptual type (i.e., oversized, wheel-
made jars) of pithoi in LB contexts at Levantine sites, but there are a few at sites such as Tell



FIGURE 25.2. Canaanite jars shifted by Egyptian porters to storehouses. From the tomb of Rekhmara, late fif-
teenth century



b.c.



(after Newberry 1900, pl. XIII).



long

oi.uchicago.edu

[image: image186.tif]



497



STANDARDIZED COLLARED-RIM PITHOI



el-Farºah North (de Vaux and Stève 1947: fig. 22: 2) and Aphek (Beck and Kochavi 1985:
fig. 5:2), and probably the collared-rim jar Type 1018 from all three phases of the Fosse Temple
at Lachish (Tufnell et al. 1958: pl. 87). Other collared-rim jars (not pithoi) were found in LB
contexts, as far north as Ugarit (Courtois 1969: fig. 1:c) and as far south as Shallal in Nubia
(Reisner 1910: fig. 304: 1). This type was found also in Palestinian sites such as Megiddo (Guy
and Engberg 1938: pl. 56: 10) and Tel Jerishe (Ory 1940: 57, pl. XII 10). Large jars with four
handles are also known from various sites in Twentieth Dynasty contexts from the late thir-
teenth–early twelfth century



b.c.



, such as Deir-el Balah (T. Dothan 1979: 38–39, figs. 81, 89,
124, 130), Nizzanim (Gophna and Miron 1970: fig. 2: 5), Type 1020 from Lachish (Tufnell et
al. 1958: pl. 87), Tell el-Farºah South burials (see, e.g., Brug 1985: 121, graph 5), and ºIzbet



Í



ar



†



ah (Brug 1985: 133, graph 14). All these specimens clearly represent the typological and
conceptual continuation of the MB II variants IV1 and IV3 (Bonfil 1992: fig. 4), although of
somewhat smaller size. Most of these specimens contain ridges at the base of their short necks
which resemble “collars.” For that reason and for the widespread spatial distribution of both
these MB II and LB prototypes and their collared-rim successors of the twelfth century



b.c.



,
it is quite unreasonable to accept the idea that “new settlers found some complete MB II vessels
at Shiloh and simply copied them” (Finkelstein 1996: 204) as an explanation for the revival of
the pithos as an oversized jar concept.

This is true not only because of the basic conceptual difference between pithoi with four,
or no handles versus those with only two (unsuitable to be shifted when full), but because the
data show quite clearly that the type called central highland collared-rim pithos have been
found in clear “pre-Israelite” thirteenth century



b.c.



contexts elsewhere in Palestine—not in
Shiloh, nor in the highland, but at coastal sites, such as Aphek, in the so-called Egyptian res-
idency (Beck and Kochavi 1985: 34–5, fig. 5:1), Tel Zeror (Ohata 1970: 71, Pl. 56), Tel Dor
(Raban 1995: 323), and Tel Nami (Artzy 1994: 127–28, fig. 9). There are others from the same
late thirteenth century context at valley sites, such as Tel Yenoªam (Liebowitz 1984: 14), Beit
Shean in the Monumental Egyptian Service Building (A. Mazar 1994: 75), and at the temenos
of the temple of Stratum VIIB at Megiddo (Loud 1948: pl. 64:2).

Another set of conceptual prototypes are those earlier pithoi that, like the collared-rim ones,
were destined to be carried empty and then serve as fixed containers, either inside a storage
building or in a hold of a merchantman. These more remote prototypes were usually handmade,
with some kind of a base—either flattened or molded, with either no handles or several pairs,
and almost always with applied molded decoration of horizontal bands, wavy lines, knobs,
meanders, or snakes. The pithoi of that set are of various shapes and sizes and can be traced back
as far as Middle Minoan Crete. Of more comparable moderate size than the huge Minoan pithoi
are the Mycenean specimens from Mycenae and Tyrins, the cargo of the Point Iria wreck (Lolos
1995: fig. 8), the Cypriot ones, of late Cypriot date (fourteenth–thirteenth century



b.c.



), such as
those from Maa-Paleokastro (Karageorghis and Demas 1988: pl. 82), Athienou (T. Dothan and
Ben-Tor 1983: 113–15, fig. 52), Pyla-Kokkinokermos (Karageorghis and Demas 1984: pl. 41),
and many others, including the fourteenth century



b.c.



wreck at Ulu Burun (cf. Bass 1986).
This set of prototypes has typological successors in the LB pithoi of Hazor (Yadin et al.

1958: pls. LXXXVIII:11–12, CIX:4; CXXXIV:8; 1960: pls. CXXII:1–6, CXLV:1–5; 1961:
pl. CXII:9) and good parallels at Dan (Biran 1989: fig. 4:7), Tel Dor (Raban 1995: fig. 9:13;
Stern 1994: fig. 45), Tel Sasa (Stepansky, Segal, and Carmi 1996: 68–70, fig. 7:9), Tell Kei-
san (Puech 1980: pls. 68:2, 69:1), Ashdod (M. Dothan and Porath 1993: figs. 34:3, 41:12),

oi.uchicago.edu





498



AVNER RABAN



Tel Akko (M. Dothan and Raban 1992: 85–86, fig. 4), and the western Jezreel Valley.
Although all these specimens, and the two MB II pithoi from Tel Dan (Ilan 1996: 223, figs.
4.94:l, 4.97:6), are the prototypes of locally made Palestinian pithoi of the Iron I period, they
relate only to the so-called Tyrian and to a lesser degree to the Galilean types and not to the
central highland collared-rim pithoi. The conceptual predecessors of this last type, with its
collared-rim and features of an oversized jar, are clearly Canaanite, with possible affinities
related to the imperial economy of Egypt during the New Kingdom era (Raban 1981: 89, 92).

THE SPATIAL DISTRIBUTION OF THE COLLARED-RIM PITHOI

Where have either complete pithoi or sherds (usually identified when there are pieces of
the rim and the neck) of the “classical” (Esse 1992: 87) collared-rim pithoi been discovered so
far (fig. 25.3)? The answer is always tentative and needs to be updated every few years.
Finkelstein (1988: 281–82) and Ibrahim, ten years earlier (1978: 121–22), defined the geo-
graphical limits of occurrence of that type between the Lower Galilee, in the north, the eastern
edge of the settled area in eastern Jordan, between Ajlun in the north to just south of Amman;
the southern reaches of Judaea’s hill country, just south of Hebron, in the south—excluding
Philistia and the valley of Beersheva; and the central coast, from Tell Qasile to Tell Keisan, in
the west. A few years later it became evident that this type is to be found in many Upper Gali-
lee sites, not only at Tel Dan (Biran 1989), but more characteristically in hilltop strongholds,
such as Sasa (Bahat 1986: 100, 101, 104, etc.; Stepansky, Segal, and Carmi 1996: 66–70, figs.
7:2, 8:4–7), Mount Adir (Davis et al. 1985), Horvat ºAvot (Braun 1993) and Horvat Yavnit
(surveyed by the writer, with A. Berman in 1974). This type was found also in the southern
part of the Golan Heights at Tel Soreg (Kochavi 1989: 7), as far east as the southern Hauran
(Kennedy and Freedman 1995: 49–51), and in other Iron I period sites farther to the north dur-
ing surface surveys (Z. Maªoz pers. comm.).

In Transjordan this type has been found in almost every site of the period, not only in the
Gilead and Ammonite territories (van der Steen 1996: 61), and in a somewhat later context—
as far south as Moab and Edom (Bienkowski 1992: 167). In the south the exclusion of the Beer-
sheeva Valley seems to be somewhat biased in character: the thoroughly excavated sites of Tel
Masos and Tel Esdar, in which no collared-rim pithoi have been recorded, seem to be irrelevant
to the issue of a chronological base. That is to say, they were first settled in a relatively later
period. We still lack a proper study of sites during the transitional LB/Iron I period (late thir-
teenth–early twelfth centuries



b.c.



) in that region. But there are some preliminary indications
for the presence of this type of pithos in that area as well (Govrin 1991: 14*–15*, fig. 2:8).

The more significant new data addresses the spatial distribution of this type of pithos along
the coast and at sites to the west and outside Palestine. In Israel it has been found literally in
every studied site of the thirteenth–eleventh centuries



b.c.



, either at the actual waterfront, such
as at Naharia (Yogev 1993: 1089), Tel Akko (M. Dothan and Raban 1992: 87–88), Tel Nami
(Artzy 1994), Tel Dor (Raban 1995: 322–39, figs. 9.17:10, 20, 21; 9.25:25), or somewhat
inland, on the coastal plain and adjacent to it, such as at Tell Keisan (Puech 1980: 216–17, pl.
68:1), Tel Zeror (Ohata 1970: pl. XV: 12), El-A



˙



wat (Zertal 1996: 44), ºEn



Ó



agit (Wolff 1997:
60), Aphek (Beck and Kochavi 1985: 40), Tell Qasile (A. Mazar 1981: 29), and Tel Mevorach
(Stern 1978: 68, fig. 19:4). But none were found so far in Philistia proper, other than two
collared-rim pithos sherds at Tel Hamid, near Ramla (Wolff, pers. comm.).

long

oi.uchicago.edu





499



STANDARDIZED COLLARED-RIM PITHOI



FIGURE 25.3. Map of Palestine during the Iron I period, including all the sites mentioned in this paper.

oi.uchicago.edu

[image: image187.tif]



500



AVNER RABAN



It is important to note that this type has been found along with the so-called Cypro-Tyrian,
ones at coastal sites of the late thirteenth century



b.c.



in Cyprus, at Maa-Paleokostro (Kara-
georghis and Demas 1988: pl. LXXXII:563), Kalavasos-Ayios Dhimitrios (Schuster 1984: 33–
6, fig. 6:2), and Pyla-Kokkinokremos (Karageorghis and Demas 1984: 52, with refs. for other
sites in n. 7).

Also of importance is the fact that in most cases, the chronological occurrence of those
specimens found at coastal sites, in Cyprus, and in the valleys predate the earliest ones from
the central hill country of Palestine. They came from clear pre-1200



b.c.



contexts at Beit
Shean (Albright 1971: 128), ºAfula (M. Dothan 1955: 43–44, fig. 16:4), Stratum VIIb at
Megiddo (Esse 1992: 87, n. 47), Tel Nami (Artzy 1994: 137), Aphek (Beck and Kochavi 1985:
34–5), and in Cyprus (Karageorghis and Demas 1984). For this reason it is hard to accept
Finkelstein’s notion (1988: 283) that the pithos found in Aphek was brought to the site by the
new settlers who had come to the hilly hinterland, or that it was on its way to Egypt as a con-
tainer for agricultural products that have been stripped off by the imperial governors of the hill
country (Wengrow 1996: 308–09).

Summing up the issue of spatial distribution within its chronological context, one might
say that at our present state of knowledge, the collared-rim pithoi were in use already in the
thirteenth century



b.c.



along the Levantine coast of the Mediterranean and at sites along the
major cross-country trade routes, such as the “King’s Highway” in Transjordan, the “Via Maris”
and the connecting routes between the two: Aphek-Jerusalem-Jericho-Amman; Yoqneºam-
Megiddo-Beit Shean-Northern Gilead; Akko-Lower Galilee-Hazor-Damascus. In this respect
one might point out the relatively early occurrence of these pithoi in building complexes that
have been identified as Egyptian outposts (or “Governors Palaces”), such as at Beit Shean,
Megiddo, and Aphek (and see discussion below).

TYPES OF SITES

Summing up the available data concerning the geographical, topographic, and site-type
context within which collared-rim pithoi have been found, one becomes confused. The geo-
graphical, or spatial, distribution does not correspond to any ethnic framework, nor to a socio-
economic one (Wengrow 1996: 312–19). Collared-rim pithoi have been found at major urban
sites, such as Megiddo (Loud 1948; Esse 1992), Beit Shean (see McGovern 1993: 247), Tell es
Saªidiyeh (Tubb 1993), Akko, Tel Keisan, Dor, Aphek, and Buseirah, on the border between
Moab and Edom (Bienkowski 1992: 167). In many of these urban centers, the pithoi were found
in a context suggestive of Egyptian administrative complexes (Wengrow 1996: 319). Other
sites, such as Shiloh, Dan, and probably also Tell Beit Mirsim, Mount Ebal, Tell en-Nasbeh,
Tell el-Umeiri, and Araq el-Amir, were probably centers during the early Iron I period; eco-
nomic centers—such as Sahab (Ibrahim 1978: 122–23), Deir ºAlla (Franken 1969: fig. 47: 1),
Araq el-Amir (Ji 1995: 122–26), Tel Nami, ºEn



Ó



agit (Wolff 1994), and ºAfula (M. Dothan
1993: 38–39); or political centers—such as Aphek, Beit Shean, and Tell es-Saªidiyeh. Other
less significant sites, of typical rural character, have also yielded collared-rim pithoi, as they
have been found in the regional surveys of the Lower Galilee (Gal 1992: 64–71, 79, fig. 51),
western Galilee (Frankel 1986: 305), the eastern valleys (Zori 1977: 8, 28, 114, 138, 144, 151),
western Jezreel Valley (Raban 1991: 25–26), Transjordan (van der Steen 1996), Samaria region

oi.uchicago.edu





501



STANDARDIZED COLLARED-RIM PITHOI



(Zertal 1988b), and almost every surveyed site of the Iron I period in the central highlands of
western Palestine (Finkelstein 1988: 285) and the central plateau of Transjordan (Herr 1998).

Characteristics of the special type of site where collared-rim pithoi are conspicuous among
the pottery finds include a relatively small size, fortifications, and a noncontinuous occupa-
tional sequence. On the coast this is the case for Tel Nami (Artzy 1994: 123–24), ºEn



Ó



agit
(Wolff 1994, 1997), Nahariya (Yogev 1993: 1089), and the Emek-Hefer region (Porath, Dar,
and Applebaum 1985: 55). In the central highlands, such single-period sites have been exca-
vated at Giloh (Mazar 1981), Mount Ebal (Zertal 1993), Khirbet Raddana (Callaway 1993),
and many others, along the fringes of the hill country, either toward the coastal plain, such as
ºIzbet



Í



ar



†



ah and El A



˙



wat (Zertal 1996); or toward the Jordan Valley (Zertal 1991). A similar
picture seems to emerge from the recently published surveys and excavations in Transjordan
(van der Steen 1996: 62–63; Ji 1997). A series of short-lived, hilltop fortified sites of late thir-
teenth–twelfth centuries



b.c.



date were observed by Glueck in Moab, during his surveys in the
1930s (Glueck 1939: map IIIb). Similar types of sites have been studied in the Upper Galilee
at Sasa, Har Adir, Horvat Yavnit, Tel Harashim, and Horvat ºAvot, sites that are no longer con-
sidered to be Israelite settlements (Golani and Yogev 1996: 56).

These small-scale fortified sites might be subdivided according to four different factors:
(1) Chronological: sites that had been settled already in the LB IIB period and/or contin-

ued to be settled and extended in the Iron II period. According to this factor, it seems as if the
majority of noncontinuous sites along the coastline, the northern valleys, and in the Jordan
Valley belong to the first group, while most of the highland sites, on both sides of the Rift Val-
ley, belong to the second group.

(2) Fortification: hilltop fortified sites, including those with peripheral fortification lines,
with a large open area in the center, of the type generally considered to represent a transitional
settlement between pastoralism and agricultural societies (Finkelstein 1996: 205–06); “forti-
fied outposts,” such as the hilltop sites of the Upper Galilee, Moab, etcetera (Wengrow 1996:
319–20); or sites that include a central “tower,” either round or square (Kempinski 1986;
Mazar 1990: 90–93; Zertal 1995).

(3) Proximity to major trade routes, such as the “King’s Highway” (Amman and its sur-
roundings, Sahab, Umeiri, Buseirah), the “Via Maris” (Aphek, Tel Zeror, El A



˙



wat; ºEn



Ó



agit,
ºAfula, etc.), other crossroads (Beit Shean, Qiri, Midrach-ºOz, Tel Qashish, and many others),
and coastal outlets (Nahariya, ºAkko, Tel Nami, Tel Ifshar, Tell Qasile); or to territorial bound-
aries of political entities (Dan, Har Adir, Sasa, ºIzbet



Í



ar



†



ah, site no. 105 in the southern Hauran,
and the Madaba Plain sites).

(4) Collaboration with ground sites and stone-paved pits of a total volume that is “clearly
out of all proportion to the subsistence needs of a village” (Wengrow 1996: 320). There is
also, of course, the statistical analysis, in which one might demonstrate the relative impor-
tance of the collared-rim pithoi for a single-site context (Finkelstein 1988: 281–82). However,
in most cases the statistical data derive either from partly excavated or only surface surveyed
sites, so that one might question the validity or even the relevancy of such data as a sound
basis for either historical or socioeconomical conclusions (contra Herr 1998: 256). At best, in
cases such as in extensively excavated, relatively small sites such as ºIzbet



Í



ar



†



ah, one may
tentatively suggest that the Stratum III settlement seems to have had more pithoi and fewer
pits than Stratum II.

oi.uchicago.edu





502



AVNER RABAN



The transitional state of research on this subject can be illustrated by the fairly recent
claim that collared-rim pithoi characterize sites in areas of “little Egyptian interest” and are
absent in major Egyptian residencies in Canaan (Esse 1992: 101), while a few years later other
scholars claim that “they were principally valued as export containers by the Egyptian govern-
ment” (Stager 1995: 340–44; Tubb, Dorrell, and Cobbing 1996: 30; Wengrow 1996: 323).

THE ARCHAEOLOGICAL CONTEXT WITHIN THE SITES

To understand the function of collared-rim pithoi, one might survey the contextual loca-
tion of these vessels within the better-excavated sites in which this type of container has been
found. Ibrahim (1978: 122–24) has suggested that these vessels served two initial purposes:
(1) as standard containers for olive oil, some of which would be sent off as trade goods; and
(2) for storage of potable water, much like the present-day z



irs.



There are also somewhat fre-
quent instances of these pithoi being used in a secondary manner as burial containers, both for
adults and infants. Yet, this secondary use of these large storage vessels has only been found
outside the central highland of western Palestine. These burials are known to us from the
coastal area, which was settled at that period by Sea Peoples, at sites such as Azor (M. Do-
than 1961), Tel Zeror (Ohata 1970: pl. 56), and Tel Nami (Artzy 1993). The use of collared-
rim pithoi for burials of adults has been attested also from Jezreel Valley sites, not only at
Megiddo (Esse 1992: 88–93), but also near Kefar Yehoshu’a, in the so-called Hittite burial
(Druks 1966; Raban 1982: 43–45). In the Jordan Valley such burials were found at Tell es-
Saªidiyeh (Tubb 1993: 1299) in a context of either Sea People mercenaries or Egyptian admin-
istrators; and farther east, near the “Kings Highway,” at Sahab. There is an interesting remark
by Ibrahim, who claims that this type of burial was “in use alongside anthropoid coffins made
of clay” (Ibrahim 1978: 123).

As storage vessels, in structures designated as warehouses or storage depots, collared-rim
pithoi were found at Megiddo (Esse 1992: 88) in structure nos. 335 and 306 in Area C; at
Shiloh (Finkelstein 1985: 169); at ºEn



Ó



agit, in Area B (Wolff 1997: 60; 1998: 452–53), and
in Area A at Sahab (Ibrahim 1978: 117). In other instances these pithoi were found in what
might have been residential units, such as in Area B at Sahab, at Sasa (Stepansky, Segal, and
Carmi 1996: 64–68), ºIzbet



Í



ar



†



ah (Finkelstein 1983), and Tell el-Umeiri (Herr 1998: 254–
55). At mountaintop sites such as Giloh and Mount Ebal (A. Mazar 1990; Zertal 1986–87),
these pithoi have been found within a less coherent architectural context. It is important to
recall that such pithoi were found in so-called Egyptian palaces, or administrative complexes
at Deir ºAlla (sanctuary?), Beit Shean, and Aphek.

This brief contextual survey cannot affirm Zertal’s claim that these pithoi were manufac-
tured as drinking water containers (Zertal 1988b). The fact that such vessels are found in well-
watered sites (Finkelstein 1996: 204), one being next to a still-functioning cistern (Ibrahim
1978: 117), makes this most unlikely. It is likely, though, that the inhabitants might have stored
water in these pithoi in certain instances, probably as a secondary usage.

Finkelstein (1988: 284–85; 1986: 204) connected these pithoi with horticulture-based
subsistence economies, with no significant commercial exchange of commodities among neigh-
boring societies. Artzy (1994) and Wengrow (1996) suggest that these vessels are typical
commercial containers, either for long-range camel and seaborne incense trade (Artzy) or

oi.uchicago.edu





503



STANDARDIZED COLLARED-RIM PITHOI



Egypt-bound exportation (or exploitation) of olive oil from the highland settlements of the
Ramesside controlled country (Wengrow). The claim for a subsistence economy does not
correlate with the quantities of collared-rim pithoi from Strata VIIA and VI at Megiddo, nor
with the neutron activation analysis of the pithoi from Tel Dan and Shiloh. Although most of the
analyzed pithoi from Tel Dan were found to be made of either local or a nearby source of clay,
“five additional collared-rim pithoi were not made at Dan, but at four different sites” (Yellin and
Gunneweg 1989: 140). Four of these pithoi were recovered from the same bin (Locus 4349 of
Stratum VI). Three neutron activation analyzed pithoi from Shiloh are also of nonlocal prove-
nance (Yellin and Gunneweg 1989: 139), and petrographic analysis of forty-eight rims of pithoi
from that site suggests that they were made in two different locations, both at a certain geo-
graphic distance from Shiloh (Glass et al. 1993: 279–81; and see also McGovern 1997: 424).

As commercial containers these pithoi are far too heavy and bulky to be carried in quanti-
ties overland even on the backs of camels (Artzy 1994). There is not a single piece of textual
or archaeological evidence to verify that the precious incense was ever transported in such
heavy, fragile, and bulky clay containers, and there is no logic in doing so. Technically it
would be impossible for two stevedores, each one holding a pithos’ handle, to carry it when
full and weighing well over 150 kg, up a gangway, from the quay to the deck of a merchant-
ship at Tel Nami, or any other Levantine port, as has been suggested (Artzy 1994: 138; Wen-
grow 1996: 308–09).

The last type of data concerning the archaeological context in which collared-rim pithoi
have been found is its correlation with Philistine “Bichrome” ware and/or Mycenean IIIC pot-
tery types. As for Philistine pottery, it has been found at almost every site west of the central
highlands in which collared-rim pithoi have been found, and usually in the same stratum, if not
at the same single complex. This is the case at ºIzbet



Í



ar



†



ah (Brug 1985: 128–33); Megiddo
Stratum VI; Beit Shean Stratum 6, and other sites in the Jezreel Valley (Raban 1991); at Akko
and Tell Keisan (Humbert and Briend 1980: pls. 61–80); ºAfula (M. Dothan 1955: 30–35); Tel
Zeror (Ohata 1970: 67–74, pls. XV:1, LX:5); and Dor (Raban 1995: 322–28). The same cor-
relation has been found at Dan (Biran 1989), Sasa (Bahat 1986: 89, 105), Deir ºAlla (Franken
1969: 245; Brug 1985: 102–05), and other sites in the eastern part of the central Jordan Valley
(Dornemann 1983: 79–81). There are sites near and north of Jerusalem where Philistine pottery
has been found in the same occupation phase as the collared-rim pithoi— such as at Tell Beit
Mirsim (Albright 1971: 118), Beth Shemesh (Grant and Wright 1939: 129; T. Dothan 1982:
50–51), Tell en-Nasbeh Stratum 4 (Zorn 1993: 1099), and Bethel (Kelso 1993: 194). Yet, it is
important to note that such vessels have not been found so far at any of the Iron Age I sites
within the original Philistine territory (Stager 1995: 345; Ji 1997: 32).

DISCUSSION

The collared-rim pithos is a special type of ceramic vessel. Unlike other containers, it is
too heavy to be shifted while full. Yet it has the same standard shape, quality, and accuracy of
production as the portable jars of the period. Typologically it is well established within the
local Canaanite repertoire, with conceptual and stylistic forerunners since the early second
millennium



b.c.



It is therefore a conceptual type of local heritage within the geopolitical
sphere of the southwestern part of the Fertile Crescent. Yet, it differs from all its forerunners

oi.uchicago.edu





504



AVNER RABAN



in two significant aspects: (1) It is not a true pithos—being shaped like a portable jar and pos-
sessing only two handles. So it is clear, from a practical point of view, that it had been pro-
duced for being shifted, but only when empty! (2) Unlike most of its predecessors, either jars
or pithoi, it is highly standardized in size, form, and volume. Considering the fact that most
samples of these vessels, among those studied so far, are of high technical quality, finished on
a fast wheel, and produced in regional centers of production, rather than locally made (Yellin
and Gunneweg 1989: 139–40; Glass et al. 1993: 279–81), it is obvious that these vessels rep-
resent a certain form of administrative regulation.

These well-made, properly standardized containers were fashioned to serve as a measured
unit. Being fixed and idle when full, one might call these pithoi “measured, standard-unit
bins.” It is obvious that these containers, whether filled with liquid commodities or dry ones,
had to be installed in place while empty and be filled and emptied with smaller deeper vessels
(cups or juglets). The standard diameter of their mouth would permit extraction of the con-
tents, although dry content (grains) would necessitate reclining the pithos in order to empty it
properly. As mentioned above, the collar at the base of the neck might indicate that these ves-
sels were sealed, or at least covered by some kind of detached cloth or piece of leather. Such
repetitive covering and uncovering of a container’s mouth would be better suited for contents
such as oil rather than grain, for which a simple lid, board, or large sherd would do. For wine
and beer, gradual, repetitive consumption would surely affect the quality of the commodity.
Yet, having only this single type of “standard-unit bin” in many sites, one might argue that its
function was not confined to only one commodity.

Only at sites where other types of pithoi were in use together with the collared-rim ones,
such as Tel Dan, Akko, Dor, Tell Keisan, Midrach-ºOz, Har Adir, Sasa, etcetera, is it tempting
to consider the possibility that each type was used as a container for a specific content. But we
have to recall that these sites are connected with the material culture heritage of the coastal and
even maritime sphere to the west. So, it is quite likely that the so-called Cypro-Tyrian types
were the conceptual and typological offsprings of the Minoan and the Mycenean pithoi, rather
than the MB II Canaanite, or the LB Egyptian ones, and one might argue that the presence
of several types of pithoi at one site in the same occupation level indicates a multiethnic pop-
ulation. But, considering the fact that they have been found together in the same locus at Dor
(Raban 1995: fig. 9.17) and Tel Dan, in Locus 586 (Biran 1989: figs. 4.1, 4.7), this would make
it a rather unlikely explanation. Besides the difference in shape, the so-called Galilean and
Cypro-Tyrian pithoi differ from the collared-rim ones also by their wide range of sizes and by
their inferior quality of production (most are handmade and finished on a turntable). Nonstand-
ardized in volume, these types might be excluded from the putative administratively controlled
system to which the central hill country type belongs.

Being a component of an organized administrative framework, in which commodities were
stored at and consumed from well measured units, it is most improbable that the collared-rim
pithoi had been fashioned and manufactured in order to contain potable water or mere agricul-
tural products of substantial economy. Being too heavy, fragile, and bulky, these containers
would hardly fit cross-country trade and even less so when “caravans plied no longer: men who
had followed the high roads went round by devious paths” (Judg







5:6). But these vessels are
found not only in urban centers (such as Megiddo, Beit Shean, Akko, and Dor) or at roadside
stations (such as ºEn



Ó



agit, Tel Nami, ºAfula, Tell Qiri [Ben-Tor and Portugali 1987: 91], and
Midrach-ºOz), but also at almost every small-scale “proto-village” site of the late thirteenth–

oi.uchicago.edu





505



STANDARDIZED COLLARED-RIM PITHOI



eleventh centuries



b.c.



on both sides of the Jordan. These “proto-villages” and some of the
“roadside stations” have, in most cases, a noncontinuous sequence of occupation, as have most
of the summit and hilltop strongholds in the Galilee and in Transjordan (see section above



Types of Sites



and also Ji 1997: 30–32 with additional up-to-date bibliography there).
There is an interesting coincidence between the exceptional spatial distribution of these

Iron Age I sites and the MB II ones. On many occasions the noncontinuous occupation does
include these two chronologically departed periods. Finkelstein suggests that this repetitive,
noncontinuous spatial distribution is characteristic of a “long term cyclic process of settle-
ment oscillations and rise and fall of territorial entities in the highlands (Finkelstein 1996:
209). Yet, the same “cyclic process” is to be found also at the peninsular site of Tel Nami
(Artzy 1990, 1994), the road sites of ºEn



Ó



agit (Wolff 1994, 1997) and Midrach-ºOz (Raban
1991: 23), the southern Hauran site no. 105 (Kennedy and Freedman 1995: 49–51), and at
summit strongholds facing the Jezreel Valley (Raban 1994: 20), or the northern coastal plain
(sites 36, 142 in Frankel and Getzov 1997). The coincidence is also in the conceptual repeti-
tion of the use of wheel-made pithoi typologically fashioned as oversized jars in both time
periods.

It is therefore not just a socioeconomical model of a long-term oscillation between pas-
toralism and sedentarism; some history might be factored in as well. There were Egyptian
economic and commercial interests. There were the pharaonic enterprises of the Twelfth
Dynasty, the Hyksos, and there were the military units, strongholds, and administration of the
Twentieth Dynasty. There were the newly established Canaanite city-states of the MB II
period, with their dependent rural lands, and with their military elite of



Mariannu



warriors.
There were the



Habiru



(or ºApiru) “Geust Arbeiters” and the mercenaries of the Egyptian
army and the mercenaries of others, such as the “Heroes of David” (B. Mazar 1986: 83–103),
or the owners of the inscribed arrowheads of the Iron Age I period (see recently, Deutsch and
Heltzer 1995: 11–38; Cross 1993). The full-scale historical context of these two periods or
even the much debated later one is well beyond the scope of this paper, but this much is obvi-
ous: once a discussed type of ceramic container has been recognized as illustrating an orga-
nized administrative framework, one must admit that it had its role within a particular chapter
in history, very probably in a similar way that the later



lmlk



jars had (Mommsen, Perlman, and
Yellin 1984).

Many short-lived settlements, such as sites with no water sources or cultivated land
nearby, do not necessarily represent the first stage of a shift from pastoralism to a sedentary
way of life. In fact, many such sites seem to have been selected for their strategic advantage
and therefore should be understood as military outposts. One might wonder if the tower at
Giloh—which has been tentatively identified as Baªal Peratzim and is mentioned in the Bible
(2 Sam







5:20) as a military outpost (A. Mazar 1990: 101)—is not the same type of military
stronghold as the one at Mt. Ebal, as has been suggested by Kempinsky (1986), or at Har Adir
(A. Mazar 1990: 84) and probably also the sites along the western fringe of the Jordan Valley
(see Zertal 1995 for a discussion of the biblical term



migdal



).
Towers of the Early Iron Age period (and somewhat later) are well known from the vicinity

of Amman, probably guarding the “King’s Highway,” and from other roadside and territorial
boundaries. Recently, Zertal extended his study of such “military outposts,” being inspired by
his excavations at El A



˙



wat, which he considered to be a settlement of the



Shardana



(Zertal
1996: 47–49), and has managed to locate a number of others, mainly in the Mt. Carmel region,

oi.uchicago.edu





506



AVNER RABAN



all of which contain sherds of collared-rim pithoi and are short-lived settlements (pers.
comm.). El A



˙



wat seems to have been the short-lived base of a military unit, probably for mer-
cenaries stationed at a convenient distance from the international highway, which led from
Egypt via Aphek, through the dangerous Wadi ºAra pass to Megiddo.

Its setting provides a view of both the western coastal part of the road and the entrance to
the narrow valley to the north. Whether its fortifications were built by



Shardana



mercenaries,
or by others, they have conceptual engineering elements with good parallels in second millen-
nium



b.c.



sites along the European part of the Mediterranean. The location of the site, at the
northeast edge of the Sharon Plain, should be associated with the



Sikils



of Dor (see M. Dothan
1986). Having no accurate date for the initial phase at that early twelfth century



b.c.



site, it is
impossible to tell who the political masters of these mercenaries were. One is tempted to relate
these mercenaries to Ramesses III, who claims in Papyrus Harris I, LXXVI, 8–10: “I settled
them in strongholds, bound in my name. Their military classes were as numerous as hundreds-
thousands, I assigned portions for them all with clothing and provisions from the treasuries and
granaries every year” (Wilson 1969: 262).

“Their assigned portions” should be measured and stored in measurable units, being the
actual salary paid to the mercenaries. Yet, salary paid in portions was due to other functionar-
ies and employees as well: the hired hands in the fields of the city-states and of the “king’s
land” (see, e.g., Albright 1969: 485, n. 7; and the reference to the people of Issachar in Gen
49:15). The priests at the amphictionic







“tent of the presence” in Shiloh may have received
their portions from the measured contributions of the pilgrims, and these might have been
stored in the collared-rim pithoi that were found there. The concept might have been Egyptian,
or Canaanite, but it was easily adapted by the new Israelite settlers and was probably
demanded by mercenaries who served various political entities. This might have been the case
in the garrisons at Har Adir, Sasa, and



Ó



orvat Yavnit which housed employees of the
Sidonians, or at Dan where other Phoenician employees were stationed. One wonders whether
ºIzbet



Í



ar



†



ah was not such a garrison settlement, either guarding the Egyptian stronghold in
Aphek against the bandits in the hill country to the east, and/or employees of the Israelites
against the Danai or Philistine territory of the coastal plain. The correlation with large, numer-
ous stone-paved bins in such sites may indicate that some of these outposts were both a con-
venient and properly protected location for storing the surplus crop from the fields of nearby
villages.

CONCLUSION

Summing up all the archaeological data and the technical facts concerning the collared-
rim pithoi of the southern Levant during the early phases of the Iron Age, the following con-
clusions may be deduced:

(1) This container was a product of a long-lasting tradition of pottery making and was
manufactured by professional potters in order to meet very specific and highly standardized
administrative demands.

(2) Having its conceptual and typological forerunners in the country, since the heydays of
urbanism and the introduction of the fast wheel into every potter’s workshop, it was not a
product of a transitional society of pastoral newcomers.

long

oi.uchicago.edu





507



STANDARDIZED COLLARED-RIM PITHOI



(3) Its spatial distribution does not fit any territorial boundary, either of a single ethnic,
socioeconomic, political, or topographic unit.

(4) Considering what we can learn from the biblical accounts and historical sources of the
period, it seems that during that transitional period, which included the demise of Egyptian
imperial control and of the Canaanite city-states and the emergence of new political units,
there were still many who served as mercenaries or hired workers in the fields (Gen 49:15).

(5) The short-lived settlements, or rather the non–continuously occupied sites in which
this type of vessel had been in use and which characterized that transitional period, might not
be considered an indication for an initial phase of sedentarism, but rather as an attempt to gain
some stability amid the upheaval—to secure trade routes, cultivation of rural land, and the
guarding of existing or newly established political territories. One might compare it to the
attempt made by the languishing Ottoman Empire in the second half of the nineteenth century
to secure the fringes of its territories by settling the Sharkasis at the edge of the Syrian desert
between Amman and Quneitra.

(6) It is suggested that the collared-rim pithoi were made and fashioned as standard mea-
sured containers for portions allocated by the hiring entity to its employees. The concept was
probably Egypto-Canaanite—developed at the Egyptian controlled territories in Asia during
the time of the Nineteenth–Twentieth Dynasties, to be adopted by the emerging Israelites and
Phoenicians of Sidon and Tyre. In the north and along the Levantine coast (including Cyprus),
other types of pithoi were used—the cruder and larger offsprings of the Bronze Age Aegean
tradition, the so-called Cypro-Tyrian type, with the Galilean version as hybrid of the two.
Similar adaptation was probably made by the Ammonites and the Moabites, somewhat later
(Younker 1997).

REFERENCES



Aharoni, Y.
1970 New Aspects of the Israelite Occupation in the North. Pp. 254–65 in



Near Eastern Archaeology
in the Twentieth Century, Essays in Honor of Nelson Glueck,



ed. J. A. Sanders. New York:
Doubleday.

Albright, W. F.
1937 Further Light on the History of Israel from Lachish and Megiddo.



Bulletin of the American
Schools of Oriental Research



68: 22–26.
1949



The Archeology of Palestine.



Harmondsworth: Penguin.
1969 Akkadian Letters, Pp. 482–90 in



Ancient Near Eastern Texts Relating to the Old Testament,



3rd ed., ed. J. B. Pritchard. Princeton: Princeton University Press.
1971



The Archaeology of Palestine.



3rd rev. ed. Gloucester, MA: Peter Smith.

Artzy, M.
1990 Nami Land and Sea Project 1985–1988.



Israel Exploration Journal



40: 73–76.
1993 Nami, Tel. Pp. 1095–98 in



The New Encyclopedia of Archaeological Excavations in the Holy
Land,



vol. 3, ed. E. Stern. Jerusalem: Israel Exploration Society & Carta and New York: Simon
& Schuster.

1994 Incense, Camels and Collared-Rim Jars: Desert Trade Routes and Maritime Outlets in the Second
Millennium.



Oxford Journal of Archaeology



13: 121–47.

Bahat, D.
1986 Saºsa Excavations, 1975. Pp. 85–105 in



The Western Galilee Antiquities,



ed. M. Yedaya. Tel
Aviv: Israel Ministry of Defence Press (Hebrew).

Bass, G. F.
1986 A Shipwreck at Ulu Bürün (Kas): 1984 Campaign.



American Journal of Archaeology



90: 269–96.

oi.uchicago.edu





508



AVNER RABAN



Beck, P., and Kochavi, M.
1985 A Dated Assemblage of the Late 13th Century



b.c.e.



from the Egyptian Residency at Aphek.



Tel
Aviv



12: 29–42.

Ben-Tor, A., and Portugali, Y.
1987



Tell Qiri—A Village in the Jezreel Valley: Report of the Archaeological Excavations 1975–1977.



Qedem 24. Monographs of the Institute of Archaeology. Jerusalem: Hebrew University.

Bienkowski, P.
1992 The Beginning of the Iron Age in Edom: A Reply to Finkelstein.



Levant



24: 167–69.

Biran, A.
1989 The Collared-Rim Jars and the Settlement of the Tribe of Dan. Pp. 71–96 in



Recent Excavations
in Israel: Studies in Iron Age Archaeology,



eds



.



S. Gitin and W. G. Dever. Annual of the Ameri-
can Schools of Oriental Research 49. Winona Lake, IN: Eisenbrauns.

Bonfil, R.
1992 MB II Pithoi in Palestine.



Eretz-Israel



(Biran Volume)







23: 26–37.

Braun, E.
1993 Horvat ºAvot. Pp. 122–23 in



The New Encyclopedia of Archaeological Excavations in the Holy
Land,



vol. 1, ed. E. Stern. Jerusalem: Israel Exploration Society & Carta and New York: Simon
& Schuster.

Briend, J., and Humbert, J.-B.
1980



Tell Keisan (1971–1976): un cité phénicienne en Galilée



. Fribourg: Editions universitaires.

Brug, J. F.
1985



A Literary and Archaeological Study of the Philistines.



British Archaeological Reports, Interna-
tional Series 265.







Oxford: British Archaeological Reports.

Buhl, M.-L., and Holm-Nielsen, S.
1969



Shiloh, The Danish Excavations at Tall Sailun, Palestine in 1926, 1929, 1932 and







1963



. Copen-
hagen: National Museum of Denmark.

Callaway, J. A.
1993 Raddana, Khirbet. Pp. 1253–54 in



The New Encyclopedia of Archaeological Excavations in the
Holy Land,



vol. 4, ed. E. Stern. Jerusalem: Israel Exploration Society & Carta and New York:
Simon & Schuster.

Cohen-Weinberger, A., and Goren, Y.
1996 Petrographic Analysis of Iron Age I Pithoi from Tel Sasa.



ºAtiqot



28: 77–84.

Courtois, L.
1969 Le mobilier funeraire céramique de la Tomb 4253, du bronze récent (ville sud d’Ugarit). Pp. 117–

32 in



Ugaritica, VI,



ed. C. F. A. Schaeffer. Paris: Mission Archéologique de Ras Shamra, Librarie
Orientaliste Paul Geuthner.

Cross, F. M.
1993 Newly Discovered Inscribed Arrowheads of the 11th Century b.c.e. Pp. 533–42 in Biblical

Archaeology Today, 1990, Proceedings of the Second International Congress on Biblical
Archaeology, eds. A. Biran and J. Aviram. Jerusalem: Israel Exploration Society and Israel
Academy of Sciences and Humanities.

Daviau, P. M. M.
1996 Iron Age II Pithoi from Tell Jawa, Jordan: Construction Techniques and Typology. Studies in the

History and Archaeology of Jordan 5: 606–17.

Davis, D.; Maddin, R.; Muhly, J. D.; and Stech, T.
1985 A Steel Pick from Har Adir in Palestine. Journal of Near Eastern Studies 44: 41–51.

de Vaux, R. P. R., and Stève, A. M.
1947 La première campagne de fouilles à Tell el Farºah près Naplouse (suite 1). Revue Biblique 54:

573–89.

Deutsch, R., and Heltzer, M.
1995 New Epigraphic Evidence from the Biblical Period. Tel Aviv-Jaffa: Archaeological Center

Publication.

oi.uchicago.edu



509STANDARDIZED COLLARED-RIM PITHOI

Dever, W. G.
1995 Ceramics, Ethnicity and the Question of Israel’s Origins. Biblical Archaeologist 58: 200–213.

Dornemann, R. H.
1983 The Archaeology of Transjordan in the Bronze and Iron Ages. Milwaukee: Milwaukee Public

Museum.

Dothan, M.
1955 The Excavations at Afula. ºAtiqot 1: 19–70.
1961 Quelques tombes de l’âge du fer ancien à Azor. Bulletin de la société d’anthropologie 12:

79–91.
1986 Sardinia at Akko? Pp. 105–15 in Studies in Sardinian Archaeology, Vol. 2: Sardinia in the Med-

iterranean, ed. M. S. Balmuth. Ann Arbor: University of Michigan Press.
1993 ºAfula. Pp. 37–39 in The New Encyclopedia of Archaeological Excavations in the Holy Land,

vol. 1, ed. E. Stern. Jerusalem: Israel Exploration Society & Carta and New York: Simon &
Schuster.

Dothan, M., and Porath, Y.
1993 Ashdod V— Excavations of Area G. ºAtiqot 23. Jerusalem: Israel Antiquities Authority.

Dothan, M., and Raban, A.
1992 Tel Akko–1989. Excavations and Surveys in Israel 1991 10: 85–88.

Dothan, T.
1979 Excavations at the Cemetery of Deir el-Balah. Qedem 10. Monographs of the Institute of

Archaeology. Jerusalem: Hebrew University.
1982 The Philistines and Their Material Culture. Jerusalem: Israel Exploration Society.

Dothan, T., and Ben-Tor, A.
1983 Excavations at Athienou, Cyprus (1971–1972). Qedem 16. Monographs of the Institute of

Archaeology. Jerusalem: Hebrew University.

Druks, A.
1966 A “Hittite” Burial near Kefar Yehoshua. Yediot, Bulletin of the Israel Exploration Society 30:

213–20 (Hebrew).

Esse, D.
1992 The Collared Pithoi at Megiddo: Ceramic Distribution and Ethnicity. Journal of Near Eastern

Studies 51: 81–103.

Finkelstein, I.
1983 Israelite Settlement in the Foothills, 13th–10th Centuries b.c.e., in the Light of the Excavations

at ºIzbet Íar†ah. Cathedra 27: 33–38 (Hebrew).
1985 Excavations at Shiloh 1981–1984, Preliminary Report. Tel Aviv 12: 123–80.
1986 ºIzbet Íar†ah, An Early Iron Age Site Near Rosh Haªayin, Israel. British Archaeological Reports,

International Series 299. Oxford: British Archaeological Reports.
1988 The Archaeology of the Israelite Settlement. Jerusalem: Israel Exploration Society.
1996 Ethnicity and Origin of the Iron I Settlers in the Highlands of Canaan: Can the Real Israel Stand

Up? Biblical Archaeologist 59: 198–212.

Finkelstein, I.; Bunimovitz, S.; and Lederman, Z.
1993 Shiloh: The Archaeology of a Biblical Site. Monograph Series of the Institute of Archaeology.

Tel Aviv: Institute of Archaeology, Tel Aviv University.

Frankel, R.
1986 Interim Report on the Archaeological Survey of the Western Galilee During 1976–1979. Pp. 304–

17 in The Western Galilee Antiquities, ed. M. Yedaya. Tel Aviv: Israel Ministry of Defence
(Hebrew).

Frankel, R., and Getzov, N.
1997 Archaeological Survey of Israel—Map of Akhziv (1) and Map of Hanita (2). Jerusalem: Israel

Antiquities Authority (Hebrew and English).

Franken, H. J.
1969 Excavations at Tell Deir ºAlla. Vol. 1. Leiden: E. J. Brill.

oi.uchicago.edu



510 AVNER RABAN

Gal, Z.
1992 Lower Galilee during the Iron Age. American Schools of Oriental Research Dissertation Series

8. Winona Lake, IN: Eisenbrauns.

Garsiel, M., and Finkelstein, I.
1978 The Westward Expansion of the House of Joseph in the Light of the ºIzbet Íar†ah Excavations.

Tel Aviv 5: 192–98.

Glass, J.; Goren, Y.; Bunimovitz, S.; and Finkelstein, I.
1993 Petrographic Analysis of MBA II, LB and Iron Age I Pottery Assemblages. Pp. 271– 86 in

Shiloh: The Archaeology of a Biblical Site, eds. I. Finkelstein et al. Monograph Series of the
Institute of Archaeology. Tel Aviv: Institute of Archaeology, Tel Aviv University.

Glueck, N.
1939 Explorations in Eastern Palestine, III. Annual of the American Schools of Oriental Research 15.

New Haven: American Schools of Oriental Research.

Golani, M., and Yogev, O.
1996 The 1980 Excavations at Tel Sasa. ºAtiqot 28: 41–58.

Gophna, R., and Miron, D.
1970 An Iron Age I Tomb Between Ashdod and Ashkelon. ºAtiqot 6 (Hebrew Series): 1–5 (Hebrew

with English summary).

Govrin, Y.
1991 The Nahal Yatir Site—Moladah in the Inheritance of the Tribe of Simeon? ºAtiqot 20: 13*–24*

(Hebrew with English summary on p. 178).

Grace, V.
1965 The Amphoras. Pp. 5–17 in The Antikithera Shipwreck Reconsidered, ed. J. du Plat-Taylor.

Transactions of the American Philosophic Society 55. Philadelphia: The American Philosophic
Society.

Grant, E., and Wright, G. E.
1939 Ain Shems Excavations, Part V. Biblical and Kindred Studies 9. Haverford: Haverford College.

Guy, P. L. O., and Engberg, R. M.
1938 Megiddo Tombs. Oriental Institute Publications 33. Chicago: University of Chicago Press.

Herr, L. G.
1998 Tell el-ºUmayri and the Madaba Plains Region during the Late Bronze-Iron Age I Transition. Pp.

251–64 in Mediterranean Peoples in Transition, eds. S. Gitin, A. Mazar, and E. Stern. Jerusa-
lem: Israel Exploration Society.

Ibrahim, M. M.
1978 The Collared-rim Jar of the Early Iron Age. Pp. 117–24 in Archaeology in the Levant, eds.

R. Moorey and P. Parr. Warminster: Aris and Phillips.

Ilan, D.
1996 The Middle Bronze Age Tombs, Part IV. Pp. 161–330 in Dan I: A Chronicle of the Excavations,

the Pottery Neolithic, the Early Bronze Age and the Middle Bronze Age Tombs, ed. A. Biran.
Jerusalem: Nelson Glueck School of Biblical Archaeology, Hebrew Union College-Jewish
Institute of Religion.

Ji, C. H.
1995 Iron Age I in Central and Northern Transjordan: An Interim Summary of Archaeological Data.

Palestine Exploration Quarterly 127: 122–40.
1997 The East Jordan Valley during Iron Age I. Palestine Exploration Quarterly 129: 19–37.

Karageorghis, V., and Demas, M.
1984 Pyla-Kokkinokermos. Nicosia: Deptartment of Antiquities of Cyprus.
1988 Excavations at Maa-Paleokastro 1975–1986. Nicosia: Deptartment of Antiquities of Cyprus.

Kelso, J. L.
1993 Bethel. Pp. 192–94 in The New Encyclopedia of Archaeological Excavations in the Holy Land,

vol. 1, ed. E. Stern. Jerusalem: Israel Exploration Society & Carta and New York: Simon &
Schuster.

short

oi.uchicago.edu



511STANDARDIZED COLLARED-RIM PITHOI

Kempinski, A.
1986 Joshua’s Altar—An Iron Age I Watchtower. Biblical Archaeology Review 12: 44–49.

Kennedy, D., and Freedman, P.
1995 Southern Hauran Survey 1992. Levant 27: 39–73.

Kochavi, M.
1989 The Land of Geshur Project: Regional Archaeology of the Southern Golan. Israel Exploration

Journal 39:1–17.

Liebowitz, H.
1984 The Late Bronze Age II and Iron Age I Strata at Tel Yenoºam. Qadmoniot 17: 11–15 (Hebrew).

Lolos, Y.
1995 The 1991 Underwater Survey of the Late Bronze Age Wreck at Point Iria, Part II: The Pottery.

ENALIA Annual, 1991. (English edition of the Hellenic Institute of Marine Archaeology 3: 9–16.)

London, G.
1989 A Comparison of Two Contemporaneous Lifestyles of the Late Second Millennium b.c. Bulletin

of the American Schools of Oriental Research 273: 37–55.

Loud, G.
1948 Megiddo II, Seasons of 1935–39. Oriental Institute Publications 62. Chicago: University of

Chicago Press.

Mazar, A.
1981 Giloh: An Early Israelite Settlement Site Near Jerusalem. Israel Exploration Journal 31: 1–36.
1990 Iron Age I and II Towers at Giloh and the Israelite Settlement. Israel Exploration Journal 40:

77–101.
1994 Four Thousand Years of History at Tel Beit Shean. Qadmoniot 27 (107–108): 66–83 (Hebrew).

Mazar, B.
1986 The Early Biblical Period—Historical Studies. Jerusalem: Israel Exploration Society.

McGovern, P. E.
1993 The Late Bronze Age Egyptian Garrison at Beit Shean. Philadelphia: University Museum of

Pennsylvania.
1997 A Ceramic Sequence for Northern Jordan: An Archaeological and Chemical Perspective. Studies

in the History and Archaeology of Jordan 6: 421–25.

Mellink, M.
1972 Excavations at Karatas-Semayük, 1971. American Journal of Archaeology 76: 257–69.

Mommsen, H.; Perlman, I.; and Yellin, J.
1984 The Provenance of the lmlk jars. Israel Exploration Journal 34: 89–113.

Newberry, P. F.
1900 The Life of Rekhmara, Vizier of Upper Egypt under Thothmes III and Amenhotep II (circa B.C.

1471–1448). Westminster: Q. Constable.

Ohata, K.
1970 Tel Zeror III. Tokyo: Society for Near Eastern Studies.

Ory, J.
1940 A Late Bronze Age Tomb at Tell Jerishe. Quarterly of the Department of Antiquities in Palestine

10: 55–57.

Porath, Y.; Dar, S.; and Applebaum, S.
1975 The History and Archaeology of ºEmek Hefer. Tel Aviv: Hakibbutz Hameuchad (Hebrew).

Puech, É.
1980 La Céramique des Niv. 9c–11. Pp. 216–34 in Tell Keisan (1971–1976), un cité phénicienne en

Galilée, eds. J. Briend and J. B. Humbert. Fribourg: Éditions universitaires.

Raban, A.
1981 The Portable Jar in the Ancient Near East: Typology, Distribution and Provenance. Ph.D. disser-

tation, Hebrew University, Jerusalem (Hebrew, with English summary).
1982 Archaeological Survey of Israel—Nahalal Map (28) 16–23. Jerusalem: Archaeological Survey

of Israel (Hebrew and English).

oi.uchicago.edu



512 AVNER RABAN

1991 The Philistines in the Western Jezreel Valley. Bulletin of the American Schools of Oriental
Research 284: 17–27.

1994 Nazareth and ºAfula Maps, Survey. Excavations and Surveys in Israel 1993, 12: 19–21.
1995 Dor-Yam: Maritime and Coastal Installations at Dor in their Geomorphological and Strategic

Context. Pp. 285–354 in Excavations at Dor, Final Report, Vol. 1a: Areas A and C. Introduction
and Stratigraphy, E. Stern (director) et al. Qedem Reports 1. Jerusalem: Hebrew University,
Institute of Archaeology and the Israel Exploration Society.

Reisner, G. A.
1910 The Archaeological Survey of Nubia: Report for 1907–1908, vol. 1. Cairo: National Printing

Department.

Schuster, P.
1984 Kalavasos-Ayios Dhimitrios 1983. Reports of the Department of Antiquities of Cyprus: 31–38.

Stager, L. E.
1995 The Impact of the Sea Peoples in Canaan (1185–1050 b.c.). Pp. 332–48 in The Archeology of

Society in the Holy Land, ed. T. E. Levy. Leicester: Leicester University.

Stepansky, Y.; Segal, D.; and Carmi, I.
1996 The 1993 Sounding at Tel Sasa: Excavation Report and Radiometric Dating. ºAtiqot 28: 63–76.

Stern, E.
1978 Excavations at Tel Mevorach (1973–1976), Part 1: From the Iron Age to the Roman Period.

Qedem 9. Monographs of the Institute of Archaeology. Jerusalem: Hebrew University.
1994 Dor: The Ruler of the Seas. Jerusalem: Israel Exploration Society.

Tubb, J.
1993 Saªidiyeh, Tell es-. Pp. 1295–1300 in The New Encyclopedia of Archaeological Excavations in

the Holy Land, vol. 4, ed. E. Stern. Jerusalem: Israel Exploration Society & Carta and New
York: Simon & Schuster.

Tubb, J. N.; Dorrell, P. G.; and Cobbing, F. J.
1996 Interim Report on the Eighth (1995) Season of Excavations at Tell es Saªidiyeh. Palestine

Exploration Quarterly 128: 16–40.

Tufnell, O.; Baden-Powell, D. F. W.; Bate, D. M. A.; Cerny, J.; Diringer, D.; Giles, M.; Helbaek, H.; Isserlin,
B. S. J.; Murray, M. A.; Parker, B.; Porada, E.; Thompson, F. C.; Todd, E.; and Waecher, J.

1958 Lachish IV. The Bronze Age. London: Oxford University Press.

Van der Steen, E. J.
1996 The Central East Jordan Valley in the Late Bronze and Early Iron Ages. Bulletin of the American

Schools of Oriental Research 302: 51–74.

Wengrow, D.
1996 Egyptian Taskmasters and Heavy Burdens: Highland Exploitation and the Collared-Rim Pithos

of the Bronze, Iron Age Levant. Oxford Journal of Archaeology 15: 307–26.

Will, E. L.
1977 The Ancient Commercial Amphora. Archaeology 30: 264–70.

Wilson, J. A.
1969 Egyptian Historical Texts. Pp. 227–264 in Ancient Near Eastern Texts Relating to the Old Tes-

tament, 3rd ed., ed. J. B. Pritchard. Princeton: Princeton University Press.

Wolff, S. R.
1994 Archaeology in Israel. American Journal of Archaeology 98: 481–519.
1997 ºEn Haggit. Excavations and Surveys in Israel 1995: 59–60.
1998 An Iron Age I Site at ºEn Óagit (Northern Ramat Manashe). Pp. 449–54 in Mediterranean People

in Transition, eds. S. Gitin, A. Mazar, and E. Stern. Jerusalem: Israel Exploration Society.

Yadin, Y.; Aharoni, Y.; Amiran, R.; Dothan, T.; Dunayevsky, I.; and Perrot, J.
1958 Hazor I. An Account of the First Season of Excavations. 1955. Jerusalem: Magnes.
1958 Hazor II. An Account of the First Season of Excavations. 1956. Jerusalem: Magnes.

Yadin, Y.; Aharoni, Y.; Amiran, R.; Dothan, T.; Dothan, M.; Dunayevsky, I.; and Perrot, J.
1961 Hazor III–IV. An Account of the Third and Fourth Seasons of Excavation, 1957–1958. Plates.

Jerusalem: Magnes.

Long

oi.uchicago.edu



513STANDARDIZED COLLARED-RIM PITHOI

Yellin, J., and Gunneweg, J.
1989 Instrumental Neutron Activation Analysis and the Origin of Iron Age I Collared-rim Jars and

Pithoi from Tel Dan. Pp. 133–41 in Recent Excavations in Israel: Studies in Iron Age Archaeol-
ogy, eds. S. Gitin and W. G. Dever. Annual of the American Schools of Oriental Research 49:
Winona Lake, IN: Eisenbrauns.

Yogev, O.
1993 Nahariya. Pp. 1088–90 in The New Encyclopedia of Archaeological Excavations in the Holy

Land, vol. 3, ed. E. Stern. Jerusalem: Israel Exploration Society & Carta and New York: Simon
& Schuster.

Younker, R. W.
1997 Moabite Social Structure. Biblical Archaeology 60: 237–48.

Zertal, A.
1986–87 An Early Iron Age Cultic Site on Mount Ebal: Excavation Seasons 1982–1987. Tel Aviv 13–14:

105–65.
1988a The Israelite Settlement in the Hill Country of Manasseh. Haifa: University of Haifa (Hebrew).
1988b The Water Factor during the Israelite Settlement Process in Canaan. Pp. 341–52 in Society and

Economy in the Eastern Mediterranean (c. 1500–1000 bc), eds. M. Heltzer and E. Lipinski.
Leuven: Uitgeverij Peeters.

1991 Israel Enters Canaan. Biblical Archaeology Review 17: 28–49.
1993 Ebal, Mount. Pp. 375–77 in The New Encyclopedia of Archaeological Excavations in the Holy

Land, vol. 1, ed. E. Stern. Jerusalem: Israel Exploration Society& Carta and New York: Simon
& Schuster.

1995 The Iron Fortresses in the Jordan Valley and the Origin of the Ammonite Circular Towers. Israel
Exploration Journal 45: 253–73.

1996 El A˙wat—A Fortified Site of the Sea Peoples near the ºIron Valley: Preliminary Report of the
First Three Seasons 1993–1995. Haifa: University of Haifa, Institute of Archaeology (Hebrew).

Zori, N.
1977 The Land of Issachar—An Archaeological Survey. Jerusalem: Israel Exploration Society

(Hebrew).

Zorn, J.
1993 Nasbeh, Tell en-. Pp. 1098–102 in The New Encyclopedia of Archaeological Excavations in the

Holy Land, vol. 3, ed. E. Stern. Jerusalem: Israel Exploration Society & Carta and New York:
Simon & Schuster.

APPENDIX

BASIC METROLOGICAL DATA OF COLLARED-RIM PITHOI

The data presented below (table 25.1) augment most of the published information, and
much of still unpublished samples, as measured by the author with the kind permission of the
principal investigators in charge. Unfortunately no final excavation report includes proper
metrological information about these pithoi, and these measurements were taken from draw-
ings to a (sometimes incorrect) scale or from published photographs in which a scale rod is
presented (marked by *). The specimens are presented in chronological order, based on the
dates suggested by the excavators.

The two most significant parameters are the total height of the vessel (H.) and its maximal
external diameter (D.), as these are the ones the potter could most readily control while build-
ing the vessel and are therefore most relevant for establishing standard volume to a certain
typological form (see, e.g., Raban 1981: 195–227; Daviau 1996: 608).

The third measurement is the internal diameter of the mouth (M.). The significance of this
is more tentative, relevant only if one surmises that standard stoppers were in use. All

oi.uchicago.edu



514 AVNER RABAN

measurements are given in cm. All of the specimens visually observed by the author (marked
by ∞) were produced in a similar manner: first the body was formed by hand, of circular coils
of clay mounted one on top of the other; then, upside down, the body was shaped on a tour-
nette and fastened over with ropes in order to prevent it from collapsing under its own weight.
Finally, it was thrown on the wheel, its neck and rim attached, and the final product properly
shaped.

Some Remarks

As one may notice, the vast majority of the specimens are grouped nicely within 6% devi-
ation off the average height of 109 cm.

The average for maximum width of the body is about 55 cm, with a deviation of up to
10%. The deviations in diameter of the mouth are greater—up to 20%. Considering the rela-
tively large chronological and geopolitical span of these specimens, the metrological devia-
tions (table 25.2) are surprisingly meager. This conclusion might be substantiated by the
statistical fact that within the eighteen measured specimens from Shiloh, fifteen of which
come from the same building (Finkelstein, Bunimovitz, and Lederman 1993: 20–31), the met-
ric deviations are not much less than in the entire corpus. Larger deviations have been noticed
by the author after measuring hundreds of so-called commercial jars (Raban 1981: 195–227).

The last group, consisting of five restored pithoi from Tell Jawa, represents what seems to
be the final, decaying phase in the use of this type of vessel. Although these pithoi are of a dif-
ferent style and of much later date and historical context, their metrological data are included
here in order to illustrate the persistence of what seems to have been a rather well-established
code of practice.

oi.uchicago.edu



515STANDARDIZED COLLARED-RIM PITHOI

Table 25.1. Metrological Data of Collared-Rim Pithoi

Site name, reference, and context in the site H. D. M.

A. Late thirteenth century b.c.:

˚1. Aphek, Beck and Kochavi 1985: fig. 5:1 (Locus 1731) C.120(?) 63 ?

˚2. Tel Nami, Artzy (unpublished) cat. no. 13/88 O-112/9 113 69 17.1a

B. Early twelfth century b.c.:

˚3. Dan, Biran 1989: fig. 4-23:1 (Pit 4349, Stratum VI) 105 56 16.5

˚4. Dan, Biran 1989: fig. 4-23:2 (Pit 4349, Stratum VI) 105 59 18.3

˚5. Dan, Biran 1989: fig. 4-23:3 (Pit 4349, Stratum VI) 105 56 17.4

˚6. Dan, Biran 1989: fig. 4-23:4 (Pit 4349, Stratum VI) ? 56 16.1

7. Giloh, A. Mazar 1990: fig. 5 (Locus 80, No. 556) 114 60 16.4

8. Megiddo, Loud 1948: pl. 83:4 (Stratum VIIb–VI?) 116 63 19.2

˚9. Har Ebal, Zertal (unpublished), L. 423, No. 3406) 104 59 15.7

˚10. Har Ebal, Zertal (unpublished), L. 358, No. 2959) 104 55 14.6

˚11. Har Ebal, Zertal 1988a: pl. 11:1 (Stratum B) ? 49 14.3b

C. Second half of the twelfth century b.c.:

12. ºIzbet Íar†ah, Finkelstein 1986: fig. 9:1 (Stratum III, Room 2007) 96 47 15.3

13. ºIzbet Íar†ah, Finkelstein 1986: fig. 9:3 (Stratum III, Room 2007) 110 59 15.8

14. ºIzbet Íar†ah, Finkelstein 1986: fig. 9:4 (Stratum III, L. 753) 108 52 ?

15. Megiddo, Loud 1948: pl. 83:1 (Stratum VIb) 116 58 18.7

16. Megiddo, Esse 1992: fig. 3:1 (Stratum VI) 106 54 17.6

17. Sahab, Ibrahim 1978: fig. 1 (Area A, Sq. 1) 110 56 16.6

˚18. Sasa, Stepansky, Segal, and Carmi 1996: fig. 7:2 (Destruction level, L. 5) 110 60 17.8

D. First half of the eleventh century b.c.:

˚19. Dan, Biran 1989: fig. 1:6 (Stratum V, L. 586) 104 57 17.3

˚20. Dan, Biran 1989: fig. 4-1:5 (Stratum V, L. 586) ? 58 18.9

˚21. Dan, Biran 1989: fig. 4-1:1 (Stratum V, L. 432) ? 53 18.9

22. Sasa, Golani and Yogev 1996: fig. 6:4 (Stratum II, L. 5, 87/1) ? 56 19.2

˚23. Sasa, Bahat 1986: 100, fig. 1 (Stratum I, L. 20) c. 98 (?) 52 17.6

*24. Shiloh, Buhl and Holm-Nielsen 1969: pl. XXII:186 (House A) 103 51 ?

*25. Shiloh, Buhl and Holm-Nielsen 1969: pl. XXII:187 (House A) 102 53 16.8

*26. Shiloh, Buhl and Holm-Nielsen 1969: pl. XXII:188 (House A) 106 55 18.0

*27. Shiloh, Buhl and Holm-Nielsen 1969: pl. XXII:189 (House A) 107 51 17.4

*28. Shiloh, Buhl and Holm-Nielsen 1969: pl. XXII:190 (House A) 108 54 19.3

*29. Shiloh, Buhl and Holm-Nielsen 1969: pl. XXII:191 (House A) 103 56 14.9

*30. Shiloh, Buhl and Holm-Nielsen 1969: pl. XXII:192 (House A) 109 53 15.1

oi.uchicago.edu



516 AVNER RABAN

Table 25.1. (Cont.)

31. Shiloh, Finkelstein, Bunimovitz, and Lederman 1993: fig. 6-48:1 (Area C, Bldg. 335) 110 57 17.2

32. Shiloh, Finkelstein, Bunimovitz, and Lederman 1993: fig. 6-48:2 (Area C, Bldg. 335) 110 55 14.3

33. Shiloh, Finkelstein, Bunimovitz, and Lederman 1993: fig. 6-48:3 (Area C, Bldg. 335) 111 58 18.0

34. Shiloh, Finkelstein, Bunimovitz, and Lederman 1993: fig. 6-49:3 (Area C, Bldg. 335) 106 51 18.2

35. Shiloh, Finkelstein, Bunimovitz, and Lederman 1993: fig. 6-49:4 (Area C, Bldg. 335) 114 54 17.8

36. Shiloh, Finkelstein, Bunimovitz, and Lederman 1993: fig. 6-51:1 (Area C, Hall 306) 114 52 17.1

37. Shiloh, Finkelstein, Bunimovitz, and Lederman 1993: fig. 6-51:4 (Area C, Hall 306) 110 54 16.2

38. Shiloh, Finkelstein, Bunimovitz, and Lederman 1993: fig. 6-51:6 (Area C, Hall 306) 105 52 16.1

39. Shiloh, Finkelstein, Bunimovitz, and Lederman 1993: fig. 6-56:3 (Area E, Instl. 519) ? 55 16.2c

40. Shiloh, Finkelstein, Bunimovitz, and Lederman 1993: fig. 6-56:4 (Area E, Instl. 519) 112 59 17.1c

41. Shiloh, Finkelstein, Bunimovitz, and Lederman 1993: fig. 6-56:5 (Area E, Instl. 519) ? 54 16.8

˚42. Dor, Stern (unpublished), Area G, Phase 8, destruction level (no. 98.23.110) 112 59 16.2d

*43. ºEn Óagit, Wolff 1998: fig. 3 (Area B) 107 55 18.1e

*44. ºEn Óagit, Wolff 1998: fig. 3 (Area B) 105 52 17.8e

*45. ºEn Óagit, Wolff 1998: fig. 3 (Area B) 106 52 17.6e

*46. ºEn Óagit, Wolff 1998: fig. 3 (Area B) 104 52 17.4e

E. Late eleventh century b.c.:

˚47. Tell Qiri, Ben-Tor and Portugali 1987: fig. 31:3 (Stratum 8, L. 1817) 108 (?) 54 17.7

48. Tel Mevorakh, Stern 1978: fig. 19:4 (Stratum VII, L. 233) 109 57 18.4

F. Late ninth century b.c. = Iron Age II

49. Tell Jawa, Daviau 1996: 608, table 1 (A. 13, 53.5) 107 52 14.0f

50. Tell Jawa, Daviau 1996: 608, table 1 (A. 13, 39.2) 111 53 16.0f

51. Tell Jawa, Daviau 1996: 608, table 1 (A. 13, 29.1) 101 50 15.5f

52. Tell Jawa, Daviau 1996: 608, table 1 (A. 13, 29.3) 107 56 15.5f

53. Tell Jawa, Daviau 1996: 608, table 1 (A. 13, 29.2) ? 51 15.0f

a. Beside this single restored pithos from Tel Nami (no. 2, this appendix), there are several more unrestored ves-
sels and half a dozen complete rims with internal diameters ranging from 14.5 to 18.3 cm. I am grateful to Prof.
Michal Artzy for allowing me to take these measurements. One of her research students, Yossi Solomon, is
presently working on his M.A. thesis entitled “The Collared-Rim Pithoi.”

b. Dr. Adam Zertal was so kind as to let me see and measure his specimens from Mount Ebal (nos. 9–11). He has
published an additional ten rims of collared-rim pithoi from that site (Zertal 1988a: pl. 10, 11), which range in
diameter between 14.2 and 21.8 cm. These varied more than the standard deviation of the entire corpus!

c. These two vessels are without handles and are so far the only specimens of handleless collared-rim pithoi
known from publications.

d. This pithos was measured by me with the kind permission of Prof. E. Stern to whom I am most grateful.
e. S. Wolff kindly supplied the author with these measurements.
f. The measurements were taken and published by Michele Daviau (1996: 608; table 1) and drawings of four,

which have been physically reconstructed, are published there as well.

oi.uchicago.edu



517STANDARDIZED COLLARED-RIM PITHOI

T
ab

le
2

5.
2.

D
ev

ia
tio

ns
o

f
M

et
ro

lo
gi

ca
l P

er
im

ite
rs

o
f

A
ll

M
ea

su
re

d
C

ol
la

re
d-

R
im

P
ith

oi
,

C
om

pa
re

d
w

ith
S

om
e

Ty
pi

ca
l “

C
om

m
er

ci
al

J
ar

s”

oi.uchicago.edu

[image: image188.tif]

518 AVNER RABAN

Table 25.2 shows that the rate of metrological deviation within the chronologically and
spatial well-diverted collared-rim pithoi is not greater than that of either one group of pithoi
from a single, contemporaneous context (Shiloh) or from any type of wheel-made standard-
ized commercial containers.

Notes to Table 25.2

a. LB II Canaanite commercial jars of Type III (Raban 1981: 55–58, 187 = table H.8). Total corpus of ten mea-
sured jars from various sites, dated from early fourteenth to late thirteenth centuries b.c.

b. Eight four-handled Judaean jars with lmlk stamps from Lachish – late eighth to mid–eighth centuries b.c.
(Raban 1981: 195).

c. Nine basket-handled jars of Cypro-Tyrian types, dated from early sixth to late fifth centuries b.c. (Raban
1981: 95–98, 202–23 = table H.25).

d. Thirteen Biconical Phoenician jars from the Levant, Cyprus, Carthage, and Spain, dated from early fifth to
mid–fourth centuries b.c. (Raban 1981: 91–94, 199–200 = table H.21).

oi.uchicago.edu



519



EARLY BRONZE AGE STATE FORMATION IN
THE SOUTHEAST DEAD SEA PLAIN, JORDAN



Walter E. Rast



Many archaeological sites with material remains dating to the Early Bronze Age (EB) are
difficult to integrate into the study of the social organization of this period. In the case of large
tells, the remains of the EB lie so far below later cultural overburden that they often can only
be exposed in a limited way in comparison with the remains of subsequent periods. The exca-
vations at Megiddo are an example of the challenges to opening up a sufficient area on which
conclusions regarding settlement type can be based. Here the EB evidence was uncovered only
on the eastern slope (Area BB), in what was apparently the cultic quarter (Loud 1948: 59–84).
Despite the important results of Finkelstein’s and Ussishkin’s recent reinvestigation of the EB
temples of Area BB (Finkelstein, Ussishkin, and Halpern 2000: 25–74), the remains of this
period have yet to be clarified in other parts of the site. The most recent excavations by Hebrew
University and Pennylvania State University may eventually reach some of this earlier mate-
rial, but at this writing the work has been limited to the Iron, Late Bronze (LB), and Middle
Bronze Ages (MB). Similar difficulties as those at Megiddo appeared at the excavations at
nearby Tell Taºannak. Whereas the excavations on the south slope brought to light a substantial
section of the EB wall, the exposure within the city itself was too small to allow for more than
a few remarks about structures (P. W. Lapp 1967: 10–13; 1969: 14–16). Consequently one
would be hard put to say much about social organization at Tell Taºannak during the EB period.
Other examples of this problem abound.

Single-period sites, on the other hand, of which there are various examples dating to the
EB, clearly have advantages for the study of social organization. The extensive excavations at
Arad are well known for the way they have opened up new perspectives on the settlement pat-
terns and social organization of this period (Amiran 1970: 90–95; 1986: 74–75; Amiran and
Ilan 1992: 34–61). Bâb edh-Dhr⺠and Numeira are further examples of sites limited in their
occupation to the late fourth or early third millennia



b.c.



It might also be mentioned that many
other single-period EB sites like those along the southeastern Dead Sea are found in Jordan.
The extensive work sponsored by the University of Tübingen and Yarmouk University at
Khirbet ez-Zeraqun in north Jordan, also a single-period site, has uncovered information of
major importance for the EB settlement in this region adjacent to the Golan Heights and Syria
(Ibrahim and Mittmann 1994: 13). The additions to some of Glueck’s conclusions regarding
the EB in Jordan have also been important for laying the groundwork for a new study of this
period east of the Jordan River (Schaub 1982: 73–74; Mattingly 1983; Sauer 1986: 3–4).

Apart from Schaub’s publication on EB IB (Schaub 1982), however, little has thus far
been attempted in explaining the social implications of the Expedition to the Dead Sea Plain’s
work, although a fair amount of the raw data has been made available either in preliminary



26



oi.uchicago.edu





520



WALTER E. RAST



reports (Rast and Schaub 1978; 1980; 1981; Schaub and Rast 1984; Coogan 1984) or in final
form (Schaub and Rast 1989). At the same time, a fundamental objective of the expedition has
been to retrieve socially relevant data. With the results of a great amount of processing of the
material remains now available and soon to appear in final volumes, some conclusions may be
proposed regarding the nature of Bâb edh-Dhr⺠and Numeira as constituting an early example
of state formation.

In regard to some of the issues discussed here, two recent publications have focused on the
socioeconomic and political developments of the EB as a whole. Douglas Esse’s study (Esse
1991) is a significant contribution in synthesizing new information dealing with the EB
period, while at the same time providing the first full discussion of the Oriental Institute’s ear-
lier work at Khirbet Kerak (Beit Yera



˙



). Esse’s study has special value because of its focus on
the specific region of the upper Jordan Valley and eastern end of Upper Galilee. Since EB data
in general suggest significant regional differentiation, much is to be gained from spotlighting
a particular area as Esse has done.

Alexander Joffe’s research (1993) has produced a broader perspective on the EB since his
investigation has tracked the entire country through the key transitions from the Chalcolithic to
EB I, and from EB I to EB II. To accomplish this, Joffe has taken into account as much of the
EB data on both sides of the Jordan Valley as could be identified. The computer-based evidence
on which his study rests (organized by the Decapolis Data Base at the University of Rome) led
him to divide the country into twenty subregions, each presumably with its unique cultural fea-
tures. Implicit in these regional differences was the issue of interrelations between the regions,
along with the trajectory of development that each followed (cf. Joffe 1993: 73–82).

The problems of the EB social organization can be formulated in several simple questions,
as implied also in the two previous works. Were all settlements of this period founded on a simi-
lar political, social, and economic basis? How were the walled towns of the EB II and EB III
on both sides of the Jordan Valley organized locally? Were these settlements predominantly
self-supporting? That is, were they able to live mostly independently on the basis of their
regional resources? On the other hand, if there was exchange between different regions, what
items were shipped and how, and what is our evidence for interregional exchange? Was there
a centralized power structure during the EB II and EB III periods? Overall, what type of social
organization are we dealing with during the Early Bronze Age?

THEORIES OF SOCIAL ORGANIZATION:

THE VIEWS OF ELMAN SERVICE AND MICHAEL COE

These questions call for theoretical definition (see the discussion of method and theory in
Joffe 1993: 5–21), especially in dealing with terms like “state” or “social organization.” As a
prelude to considering evidence from Bâb edh-Dhr⺠and Numeira, we look at two proposals
dealing with some of the problems of explaining complex forms of social organization (Ser-
vice 1971; Coe 1961). Although the models proposed in the works about to be discussed are
based on areas quite distant from the ancient Near East, they can serve to raise issues and
problems associated with state formation in the southeastern Dead Sea Valley during the EB,
specifically during the EB III period.

oi.uchicago.edu





521



EARLY BRONZE AGE STATE FORMATION IN THE SOUTHEAST DEAD SEA PLAIN



Students of early social organization are indebted to Elman Service’s often-referenced
study (Service 1971). Service’s work, while devoted to the evolution of societies, also help-
fully outlines unique features of different grades of social development, including two with
which we are concerned here, the chiefdom and state. According to Service, while remnants
of the social patterns of bands or tribes may be carried over into a chiefdom, the latter is
marked by new strides into complexity. The higher status attained by the chief and his family
in a chiefdom also means that other members of the society could take on particular roles.
Within a chiefdom society there are personages who serve as craftsmen or cultic functionaries.
A new level of labor management also emerges, along with an array of larger projects such as
irrigation works, terraced slopes, and the construction of monumental buildings (Service 1971:
162). In contrast to the egalitarian nature of tribal or band societies, chiefdom society is in-
egalitarian (Service 1971: 138–40, 164).

A further stage of development (Service argues for the concept of evolution) is repre-
sented in the state, which carries on many of the advancements of chiefdom society such as
management of resources, but in more complex form. Although state society is not exten-
sively dealt with in Service’s discussion, he does have some words to say about it as the next
level beyond the point where his study ends, the chiefdom. The organization of the state, in
comparison with that of a chiefdom, includes a greater amount of cohesion in social organiza-
tion. For our concerns with the ancient Near East, this would include the fact that some sort of
personage like a king would have been an important figure in the society’s structure. Service
also argues that a mechanism of control is a distinguishing characteristic of the state. In his
words, one of the most important attributes of an emerging state is how it sets in place “a con-
sistent threat of force by a body of persons legitimately constituted to use it” (Service 1971:
163, 165). As shown below, this point is an important one in considering the EB III data at
Bâb edh-Dhr⺠and Numeira as advances into state formation.

Michael Coe’s discussion of two very different types of social organization is also valu-
able in considering features of the EB society in the southeastern Dead Sea Valley (Coe 1961).
Coe terms the constrasting types “unilateral” and “organic” societies. A key for Coe is that a
unilateral system of organization does not maintain instruments for the exchange of goods.
The resources of the society are consequently used for its own subsistence, and thus trade is
nonexistent. The absence of trade along with a mercantile class produces a notable lack of dif-
ferentiation in the group. Artisans and other specialists are found in the service of an authori-
tarian control, and in that sense are either diminished in status or may be altogether invisible
to the public. The result is what Coe terms a “mechanical” society, where workers primarily
supply the ruling personage(s) of the group, including a priest-king and religious functionaries
associated with him. Coe’s examples are the classic Khmer of Cambodia and the classic Maya
of Central America, both of which provide examples of societies focused on a ceremonial cen-
ter. As far as the ancient Near East is concerned, Coe sees in Egypt until the Eighteenth
Dynasty an example of such a unilateral society because it was not until the Eighteenth
Dynasty that cities or towns began to appear in Egypt (Coe 1961: 84).

In addition, the unilateral society emerges in an area where little differentiation of environ-
ment exists, and thus regional specializations on which trade is based are minimal if not
absent. This type of society witnesses a lack of incentive for building and maintaining a road
system since interregional relations are negligible. Such societies exhibit a high degree of

oi.uchicago.edu





522



WALTER E. RAST



control, much of which is based on religious sanction. Coe’s point is that these centers are thus
not really urban, although they constitute an alternative type in the development of complex
societies.

The antithesis of the unilateral society is what Coe terms an “organic” society. This type
of society exhibits a feature of special importance when the model is applied to the EB social
organization below, namely, division of labor. Concomitantly the organic society exports and
imports goods necessary for the stabilization of the society, maintaining a system of transport
to do so. The contacts fostered by trade require and support a variety of social roles, including
merchants, traders, and specialists of various types preparing goods for trade. The mecha-
nisms of governance also broaden somewhat since the ruler (a king or someone comparable)
would not retain his authority without the support of the mercantile and artisan classes. In
comparison with Egypt, Mesopotamia produced many examples of the organic type of society
in its various “micro-geographical units” (Coe 1961: 81–84).

According to Coe, basic environmental differences contributed to the formation of these
polar opposites in social organization. Both the Khmer and Maya civilizations appeared in
regionally undifferentiated environments (Coe 1961: 85), and the same thing can be said for
early Egypt. On the other hand, studies of the EB in ancient Palestine, such as those of Esse
and Joffe, demonstrate the great variety in subregions that undoubtedly contributed to the
emergence of a dynamic town life during this period.

BÂB EDH-DHRº AND NUMEIRA AS PRIMITIVE, ORGANIC STATE SOCIETIES

The theoretical discussions of Service and Coe are valuable for reflection on the nature of
the type of society found at the two EB III sites in the southeastern Dead Sea Valley (fig. 26.1).
In reference to Service, the EB III societies at Bâb edh-Dhr⺠and Numeira are better explained
as incipient state societies rather than as chiefdoms (cf. also Renfrew and Bahn 1991: 156–57
on distinctions between the two). At the same time, since the use of the term “city” for the
Dead Sea sites is problematic, it is probably better to avoid the hyphenated expression city-
state (cf. Mazar 1990: 140, who sees the country divided into about twenty “city-states” during
EB). In earlier reports of the work at Bâb edh-Dhr⺠and Numeira, the term “city” was used for
the Dead Sea settlements (Rast and Schaub 1981: 15–23), but further consideration of this
problem has led us to believe that calling them “towns” seems more appropriate (cf. Schaub
1982: 67).

Particularly important for tracing attributes of a primitive state society is the way such a
society manages its resources (see Service 1971: 133–34, who sees this happening already in
chiefdoms). Examining the various forms of managerial activity in relation to resources may
be more promising than focusing on “scale” in attempting to discern the nature of the EB
social organization (cf. Joffe 1993, passim). The question of management highlights human
response and decision in relation to the needs and aims of an emerging state. For the researcher
it also stresses the importance of exploring archaeological data for what they can tell us about
the regulation of natural and human resources in making and implementing decisions in regard
to various construction and preservation projects. To our way of thinking, the emphasis on
scale imposes a nondynamic grid onto what was certainly an actively engaged community, and
it may also obscure the fact that smaller settlements were at least as sophisticated and dynamic

oi.uchicago.edu





523



EARLY BRONZE AGE STATE FORMATION IN THE SOUTHEAST DEAD SEA PLAIN



FIGURE 26.1. Map of the southeast Dead Sea plain with locations of Bâb edh-Dhrâº, Numeira, and Buleida (map
by Jeannine Schonta).

oi.uchicago.edu

[image: image189.tif]



524



WALTER E. RAST



in their organization as those of greater size. For example, Bâb edh-Dhr⺠and Numeira were
smaller than settlements like ºAi or Arad (Bâb edh-Dhr⺠is a little more than 4 ha, Arad about
10 ha), and yet the cultural record shows the Dead Sea sites to have been highly progressive in
their development. In that case it is questionable whether scale really makes a significant dif-
ference in delineating complexity.

Among the rich cultural data of the Dead Sea sites is the great amount of evidence from
the large and continuously used cemetery at Bâb edh-Dhrâº. The treatment of burials dating
to the EB III is a striking complement to the organizational changes occurring during this
later phase of the EB. That a particular societal group enjoyed higher status during the
EB III is indicated by the data from Charnel House Tomb A 22, not only the largest charnel
house thus far discovered but one which also contained several luxury objects, including
gold leaf jewelry, beads, and other luxury items (Rast and Schaub 1980: 34–39). More than
150 individuals were buried in Tomb A 22, and thus it had one of the largest populations of
any of the charnel houses (Rast and Schaub 1980: 38). When the data of Tomb A 22 are
compared with those from the EB III town of Bâb edh-Dhrâº, the picture of a diverse society
emerges. Tomb A 22 indicates enough “assymetry” to meet the minimum requirement of
what Trigger holds to be necessary for the term “state.” At the same time, data at Bâb edh-
Dhr⺠speak for less elitism than Trigger sees necessary for positing the existence of an
ancient state (Trigger 1978: 160).

The various metal weapons found at Bâb edh-Dhr⺠(Wilkinson 1989: 444–50), many of
which are dated to the EB III, are also best viewed in regard to their social importance. Metal
weaponry has been found both in the tombs and at the town site at Bâb edh-Dhrâº. Why not a
single such weapon, by conrast, has yet to appear at Numeira is unclear. Following Service, the
metal weapons at Bâb edh-Dhr⺠suggest a system of control or force that had implications both
externally and internally. Thus these implements need to be interpreted as a mark of the state
at some stage of its development (Service 1971: 163, 165; cf. Adams 1966: 14). In addition,
emblems of the centralized state authority would also be evident in larger public structures such
as temples or palaces (cf. Mazar 1990: 140). Although no remains of a palace have been dis-
covered at Bâb edh-Dhr⺠or Numeira, the sanctuaries of the EB II and EB III periods at Bâb
edh-Dhr⺠belong in this sphere (Rast and Schaub 1980: 30–31; Schaub and Rast 1984: 50–51).

Coe’s use of the term “organic” is also useful for the definition of the type of state orga-
nization found in the emerging EB towns of the Dead Sea. Coe’s description is put in socio-
logical and anthropological terms, but it has similarities to what Jacobsen characterized from
a political vantage point as “primitive democracy” (Jacobsen 1943, 1957). Jacobsen traced
the Mesopotamian examplar of this type of more open and shared organization from its roots
in early villages, and even in nomadic societies, into societies of greater complexity (Jacob-
sen 1957: 99–109). As shown below, this is the trajectory of the emerging EB III society at
Bâb edh-Dhr⺠and Numeira. Although on a smaller scale, the social organization of the Dead
Sea sites is more along the line of third millennium Mesopotamian settlements than the spe-
cialized, “unilinear” society of Egypt in the same period. Occasional items from Egypt con-
tinued to appear at Bâb edh-Dhr⺠during the EB III period (e.g., Wilkinson 1989: 452–56),
but these were luxury pieces that only the well-off could afford, and thus they scarcely indi-
cate any great impact from Egypt on social organization in the southeastern Dead Sea Plain
during the third millennium



b.c.

oi.uchicago.edu





525



EARLY BRONZE AGE STATE FORMATION IN THE SOUTHEAST DEAD SEA PLAIN



THE RELATION BETWEEN BÂB EDH-DHRº AND NUMEIRA DURING EB III

We turn now to consider two specific activities at Bâb edh-Dhr⺠and Numeira: the man-
agement of labor in constructing and maintaining these town sites and the control of the hydrol-
ogy. The latter especially was necessary for successful occupation of the Dead Sea region in
which the towns were established. It is proposed here that these two forms of evidence point to
a high degree of social organization, and thus they are significant for tracing the development
of early state formation in the Dead Sea Valley. At the same time they attest to the close relation
of the two Dead Sea sites to each other.

By this time the chronology of Bâb edh-Dhr⺠and Numeira is well known (see the table
in Schaub 1993: 131). It is clear that these two towns, and the southeastern Dead Sea Valley
as a whole, reached a high point in the EB III period. The destinies of Bâb edh-Dhr⺠and
Numeira were interlocked during EB III, as various data from the recent excavations show.
This interrelation was evident from the fact that the population of the older settlement at Bâb
edh-Dhrâº, having developed from its earliest EB IA phase on through to EB III, found itself
pressed with needs of space and resources by at least the middle of the EB III. To mitigate the
potential hazards created by overpopulation and diminishing resources, colonizers from Bâb
edh-Dhr⺠began to settle the nearby site of Numeira in the last one-and-a-half centuries of
the EB (ca. 2500–2350



b.c.



)
The basis for interpreting Numeira as an extension of Bâb edh-Dhr⺠made by settlers from

the mother site is found above all in results from the petrographic study of pottery found in
several EB III tombs in the cemetery at Bâb edh-Dhrâº. The analysis of wares from the two
Dead Sea sites showed that pottery made from Nubian sandy clays near Numeira was deposited
in tombs at Bâb edh-Dhrâº, presumably accompanying burials of deceased persons from
Numeira in the Bâb edh-Dhr⺠cemetery (Beynon et al. 1986: 33; Schaub 1987: 247–48; 1993:
135). Assuming that burials of outsiders not associated in some way with the occupants at Bâb
edh-Dhr⺠would have been unacceptable, the petrographic studies support the notion of affinity
between the populations of the two Dead Sea sites, and colonization would be a useful way to
understand the process. We may conclude, therefore, that we are dealing with what can be
termed a small, local system involving these two EB III towns situated just eight miles apart,
similarly constructed and similarly positioned in respect to the Dead Sea, both based on a com-
parable economy, and both possessing many likenesses in their cultural record.

SITE ENGINEERING AND SOCIAL ORGANIZATION

Efforts to understand the social organization of the EB III towns in the southeastern Dead
Sea Valley are best begun by focusing on Bâb edh-Dhrâº. Bâb edh-Dhr⺠holds prime place
because it was the larger of the two EB III sites, and its sheer endurance throughout the entire
third millennium shows that it was the dominant settlement. Its local water sources were also
more extensive than those at Numeira since the



ghor



adjacent to Bâb edh-Dhr⺠(Ghor el-
Mazra’a) was watered not only by outflows from Wadi Kerak, but also from Wadi Ibn Hammad,
as well as the springs at Haditha and ºAin es-Sikkin (see the data in Harlan 1981: 156–57).
Thus, if Bâb edh-Dhr⺠can be explained, the related EB III settlement of Numeira and its role
in the southern Ghor social organization falls into place.

oi.uchicago.edu





526



WALTER E. RAST



There is, however, a more important reason why it is important to interpret from this site
outward. This has to do with the unusual natural features of Bâb edh-Dhrâº, which at first
blush are not impressive and which indeed presented great challenges to those who settled
here (pl. 26.1). The native site of Bâb edh-Dhr⺠consists of laminated white marl and gravel
deposits laid down in the Upper Pleistocene period (Donahue 1981: 144–46). The decision to
settle this site demanded expertise and endurance (Rast 1995), and it is in considering how the
occupants dealt with the natural difficulties that we can begin to understand their efforts at
communal organization. Since Numeira replicated Bâb edh-Dhr⺠in so many ways, the data
from there provide additional information on the same problems.

The foremost challenge to settling the marl site of Bâb edh-Dhr⺠was the instability of this
soil type. Consequently it is important to ask how the ancient builders of the EB III town at
Bâb edh-Dhr⺠resolved problems of building on such a location (Donahue 1981: 134). One
response was the development of an extensive brick-making industry, attested in all parts of
the town and cemetery, where sun-dried brick was used for the construction of tombs during
the EB II and EB III periods. Such use of brick is the more remarkable given the fact that a
plentiful supply of stone was available and would have saved time and effort had it been used
as the main material for construction. Large rocks and boulders from the mountainous eastern
slopes rolled through the Wadi Kerak below the site, and thus a storehouse of boulders was
near at hand. Stones of this size were used in constructing the base of the EB III town wall and
in the latest phase of the EB III were also used for the foundation of the Northeast Towers. The
two superimposed sanctuaries in Field XII also made use of stone foundations, but these were
quite small, round, field or wadi stones (Rast and Schaub 1981: 28, fig. 23). Apart from these
examples, however, the construction employed on the interior of the town, consisting of what
were domestic and in some cases small industrial buildings, was almost entirely of brick from
the foundations upward.

The reason that brick was so extensively used was no doubt that it was more effective on a
site so characterized by precarious marl soils. We can consequently posit groups of laborers
involved in brick making––their work being to gather clay and temper, to form the bricks in
molds, and to bake them to rocklike hardness under the hot sun of the southeastern valley. Pot-
ters’ marks consisting of circles or crosses were commonly made, using the thumb, stick, or
stone. While the bricks could have been made independently by individuals, the great number
of common markings on them suggests primitive “companies” engaged in brick manufacturing.
And incidentally, the local conditions promoting the need for brick manufacture would appear
to consign Mesopotamian influence to a secondary level of explanation at best (cf. Yeivin 1934).

Besides being used for buildings, unfired or sun-baked brick was also used to stabilize
areas of soft marl on the natural site. An extensive retaining wall made of brick was built
along the marl slope of Field XIV on the interior of the town. Although over the centuries
parts of this wall had slipped onto the adjacent surface, what survived showed the ingenuity
of the EB III engineers in constructing this system (Rast and Schaub 1981: 23–25, fig. 17).
The large court surrounding the EB III sanctuary in Field XII was also paved with brick to
provide a more solid surface on the natural marl and gravel (Rast and Schaub 1981: 28).

The energies devoted to controlling the natural soils of the site thus indicate that a good
number of people were involved in the construction and maintenance of the EB III town. That
separate groups of laborers were also employed in the erection of the 7 m wide defensive wall

oi.uchicago.edu





527



EARLY BRONZE AGE STATE FORMATION IN THE SOUTHEAST DEAD SEA PLAIN



was evident from the way this wall was laid. Its foundations were laid in sections, so that at
approximately 15 m intervals a segment would be terminated with a transverse face, after
which a new segment was begun. These segments are of interest because the stone foundation
material in most cases was different for each succeeding segment. In some cases smoothed,
rounded, wadi boulders were gathered and used. In other cases, tabular limestone blocks were
collected, either from the plain near the site or from below the escarpment some 5 km to the
east, where numerous limestone slabs had fractured from their beds. The entire upper part of
the defensive wall, estimated to have been at least 8 m high, was made of mudbrick, requiring
a massive expenditure of effort by the brick-making industry as well as those involved in the
construction. Both the lower stone construction and the brick superstructure of the defensive
wall, therefore, are good indicators of the social organization of Bâb edh-Dhr⺠during its
height in EB III period.

A final example of engineering techniques employed at Bâb edh-Dhr⺠is the terraces used
to construct the defensive wall around the edges of the marl hillocks on which the town was
built. Especially on the east and north sides of the site, the marl slopes were particularly pre-
carious. Had the stone foundation simply been planted on the existing surface, the wall would
have given way to erosion and would have easily toppled. To counteract this, terraces were cut
into the slope and the lowest courses were even plastered into place. That sections of these wall
foundations survived until recent excavation testifies to the effectiveness of such a solution.

Since Numeira was a colonized site, it could be expected that techniques of construction
known at Bâb edh-Dhr⺠were also employed at the new site. Indeed, it seems probable that
some of the same engineers who worked at Bâb edh-Dhr⺠were involved in the construction
of this second site. Although the defensive wall here was about half the width of that at Bâb
edh-Dhrâº, the technique of segment construction was employed, only here the segments were
placed at 7 m intervals (Rast and Schaub 1981: 37–39; Rast and Schaub 1980: 42).

Other examples of engineering ingenuity could be pointed to at both sites. At Numeira the
town wall was not added until much of the interior town had been constructed (Rast and Schaub
1981: 40; Coogan 1984: 76). When the decision was made to add the fortifications to this orig-
inally open settlement, they had to be spliced into previously existing structures oriented on a
different axis. At the eastern end of the site, walls of earlier buildings were incorporated into the
defensive wall, giving the appearance of a solid tower without rooms (Coogan 1984: 79–80),
although the main purpose here was apparently to give special strength to the wall on this
vulnerable side of the site.

When examined from the perspective of people’s contribution to the social systems of
these EB sites, therefore, these activities of construction and maintenance open a window to
how the societies were organized. The most important point is that the EB III societies at both
sites were effective in solving their problems and creating conditions that led to a substantial
standard of living.

HYDROLOGY IN THE SOUTHEASTERN DEAD SEA VALLEY

A second feature affecting the social aspects of settlement of the southern Ghor is found
in the way the population dealt with water resources. Although data from survey showed that
during much of the EB III the southern Ghor was still benefiting from a higher rainfall than in

oi.uchicago.edu





528



WALTER E. RAST



the latter third of the third millennium, the area was still subject to water problems and occu-
pants had to take care to protect the natural supplies. The dual sources of runoff and perennial
spring-flow through the Wadis Kerak and en-Numeira provided for the daily needs of occu-
pants while at the same time replenishing the adjacent



ghors



for cultivation (Harlan 1981:
155–59; McCreery 1980: 216–19). These areas for cultivation were indispensable to the suc-
cess of settlement in the region.

Butzer has pointed to the importance of “perception” as a way of understanding how
people dealt cognitively with aspects of the environments in which they chose to locate
(Butzer 1982: 252–57). In the case of Bâb edh-Dhr⺠the question is what inducements led
people to stay permanently at the site, beginning with village life in the EB IB, followed by
the evolution of a more complex society during the EB II and EB III periods. One resource of
undoubted importance was the availability of water. The decision to locate the EB III towns
next to Wadis Kerak and en-Numeira points directly to the importance for social organization
of the management of this resource. Both wadis were good providers in antiquity as they are
today (Harlan 1981: 155–59), and it was the regulation and control of these water sources that
spurred the growth that took place, especially in the EB III period.

Carsten Körber has added a new dimension to the study of water control in the southern
Ghor in his recent explorations of the whole lower area of the Wadi Kerak. Here Körber found
that the large EB site of Buleida, located several kilometers along Wadi Kerak east of the Ghor
el-Mazra’a, is best explained as having served to guard the source of water near the entrance
of the wadi into the southern Ghor. The ceramic evidence shows that Buleida was settled pre-
dominantly during the EB III, so that it dates to the time of the EB III towns of Bâb edh-Dhrâº
and Numeira (Körber 1993: 551, fig. 1, and 553). Since the terrain around Buleida was poorly
suited for agriculture, being located in the foothills of the Jordan plateau and along a very
much incised area of the Wadi Kerak, the site’s function was different from Bâb edh-Dhr⺠and
Numeira where cultivation of the plain was a principal activity. Buleida would thus have func-
tioned as a guard post of considerable size, assuring control of water used in the agricultural
production supporting Bâb edh-Dhrâº.



1



Körber’s researches have turned up two further fortresses east of Bâb edh-Dhr⺠close to the
eastern edge of the plain. One is on the north side, the other on the south side of the Wadi
Waid’a, a tributary of Wadi Kerak. On Körber’s map the sites are designated as



Ó



irbet Wadi
adh-Dhr⺠N and



Ó



irbet Wadi adh-Dhr⺠S (see the map in Körber 1993: 551, fig. 1, and the dis-
cussion on 551–52). Körber has suggested privately that the positioning of these fortresses once
again highlights their function as guard posts overseeing the sources of water, just as in the case



1. I discussed Körber’s findings with him in Amman in March 1995, and at his suggestion visited the EB site of
Buleida (to be distinquished from the Byzantine ruins by the same name north of Wadi Kerak), accompanied
by Henry Cowherd. In my judgment Körber’s explanation is both correct and of great interest. I wish to thank
Carsten Körber for showing me the pottery from Buleida, which he dates to EB III, and for sharing with me
his views on this EB site (most recently in a communication dated April 1, 1996). As best I can calculate,
the site noted as er-Rishi on the map of Worschech (Worschech 1986: 286, fig. 1) is the same as that indicated
by Körber as Buleda (Körber 1993: 551, fig. 1). Both authors described the site about which they were writing
as lying on the south side of Wadi Kerak. Worschech estimated its size as 80 x 60 m and also listed it as his
Site No. 93 (Worschech 1986: 289). In his April 1, 1996, communication Körber confirmed that his site of
Buleda is the same as No. 93 of Worschech. But see also n. 2.

oi.uchicago.edu





529



EARLY BRONZE AGE STATE FORMATION IN THE SOUTHEAST DEAD SEA PLAIN



of EB Buleida.



2



In modern times water has been channeled through a concrete conduit in the
Wadi edh-Dhrâº, beginning opposite the Neolithic site of Dhr⺠and continuing past Bâb edh-
Dhr⺠onto the Lisan peninsula (Bennett 1980: 30; Raikes 1980: 56). A similar channel runs past
Buleida (Körber, pers. comm. on April 1, 1996). Some type of channeling system could have
been directed through the same wadis during antiquity, even one with a plaster lining since sev-
eral types of plaster material were found at Bâb edh-Dhr⺠(cf. the plaster-lined installation in
Field III excavated by Paul Lapp), although no traces that can relate to the EB have yet been
found around these wadis.

The role that water control played in the socioeconomic development of the Dead Sea sites
may also explain the construction of defensive walls around these settlements. Since Numeira
was at first an open settlement, the placement of a defensive wall around the site in a second
phase of construction was likely connected with the protection of the new industry undertaken
here, that of agriculture. The latter involved the production and processing of crops like barley
and grapes, both of which were found in abundance at Numeira. Bâb edh-Dhr⺠likewise was
ringed about by the defensive system referred to above, and it is interesting that Buleida was
apparently also enclosed by a wall during EB III. The success of these sites in water control
and cultivation thus lay them open to raids by outsiders, whoever these latter might have been.

SOCIAL ORGANIZATION AT THE DEAD SEA SITES

On the basis of these two activities of site construction/management and water control, we
can make some inferences regarding social organization at these two settlements during EB III.
In both cases the data point to the involvement of a great number of people. Some of these
workers were skilled and were thus involved in decisions about planning at the sites. Others
had proficiencies in the production of particular items used in construction, such as bricks, and
still others were adept at calculating the best way to transfer water to key areas for cultivation.
That irrigation was used for a good bit of the EB III cultivation in the southern Ghor has been
argued by McCreery, based on his follow-up studies of botanical remains retrieved by flotation
(McCreery 1980: 214–24).

When discussing water management, it is important to stress that there is no evidence in
the southern Ghor of an authoritarian control of the magnitude of Wittfogel’s oriental despo-
tism, nor can it even be suggested that the EB III settlements were hydraulic societies in rela-
tion to Wittfogel’s well-known discussion of this type of social organism (Wittfogel 1957).
From the perspective of human choice and decision as discussed above, there were other per-
ceived resources alongside water that contributed to the mechanisms of settlement in the
southern Ghor. At the same time, the societies of the two Dead Sea sites do not fit at all the
“mechanical society” of Coe, in which activities were performed for a ruling personage or



2. See Körber’s discussion of these sites in Körber 1993: 551–52. Körber generously shared the pertinent infor-
mation from his field notes. For the fortress on the south side of the wadi (



Ó



. Wadi adh-DhrâºS), Körber
recorded a date of EB III. The pottery sherds retrieved consisted mainly of ledge handles and rims of hole-
mouth jars. Körber also noted that the name Jebel er-Risha was later applied to this fortress ruin on the south
side, and that Worschech has mistakenly used the latter name for his site No. 93 which is EB Buleida. See
the map in Körber 1993: 551, fig. 1.

oi.uchicago.edu





530



WALTER E. RAST



family. The segment construction of the town walls at Bâb edh-Dhr⺠and Numeira indicates
a more participatory organization of labor than the kind associated with pyramid building in
Egypt at roughly the same time. The EB III society in the southern Ghor was an organic one,
using Coe’s other type. That is, the various activities of skilled persons or workers contributed
to the common good, more along the lines of the society of the village. In fact, it is reasonable
to suppose that the EB II and III “urban” societies in the Dead Sea Valley emerged from the
more simple, democratic style of the preceding village societies.

This is why it seems best to conclude that in this region of ancient Palestine, at least, we
may have an example of a type of social organization that corresponds more closely to the
model of Jacobsen mentioned above, that of a primitive form of democracy. Other descrip-
tions might be used for it, such as Joffe’s suggestion of communally shared power, which he
finds already in the EB II period (Joffe 1993: 84–86). What could be added to any of the dis-
cussions of terms and models is the notion that the EB III settlements in the southern Ghor
were for the most part self-dependent. Most of what the EB III occupants needed was pro-
duced within their own subsistence parameter. While the Dead Sea sites might have
exchanged with areas somewhat farther away, such as the plateau to the east, the Negev, or the
southern hill country west of the Dead Sea rift, their meat supplies, fruits, vegetables, and
grain could all have been produced in the valley itself.

At the same time, as the population grew it presumably became necessary to import some
items. McCreery (1980: 224) and Harlan (1981: 159) have suggested that the Ghor could have
provided ample supplies of grain for the population of the EB, which during EB III probably
added up to not quite a thousand people. At the same time, it does not seem out of the question
that some grain might have been imported from as far away as Syria since seal impressions on
store jars at both sites bear Syrian motifs, indicating a connection with that area in relation to
trade (cf. the seal on the fully reconstructed jar from Numeira in N. Lapp 1989: 7–9, fig. 7,
which shows affinities to Syrian types).

The question whether there was some overall, centralized authority in Palestine during
EB III is elusive. The many similarities in ceramics and other artifactual assemblages through-
out the country during EB III are striking and certainly show cultural contact. At the same
time, this does not necessarily lead to notions of centralized authority during this period,
which the regional variation of the period would appear to contradict. On such questions the
cross-comparisons of the EB III sites in different regions of the country may be able to offer
some proposals in the future. A small amount of effort has already been made in this direction
by juxtaposing material remains from Khirbet ez-Zeraqun in north Jordan with those from Bâb
edh-Dhr⺠and Numeira in the southern Ghor, and it is planned that this comparative work will
be extended in the future.

In sum, we can see how greatly the picture of Bâb edh-Dhr⺠has changed since the early
days when Albright and others wrote of this site not as a town or city, but as a unique ceremo-
nial center for peoples in the southern Dead Sea Valley (Albright 1926: 61), in their minds, of
course, more accessible to the public than the unilateral ceremonial society envisaged by Coe.
In any case, the conclusion of a ceremonial center has been overturned by nearly two decades
of intensive research devoted to the study of the EB in the southeastern Dead Sea plain. What
we rather find in the case of Bâb edh-Dhr⺠and Numeira are examples of incipient statehood,
in which organizational features deriving from the earlier villages are still present, and in
which localized development is the major dynamic stimulating change. It is not necessary,

long

oi.uchicago.edu





531



EARLY BRONZE AGE STATE FORMATION IN THE SOUTHEAST DEAD SEA PLAIN



therefore, to look to Mesopotamia and Syria to explain these developments, despite the fact
that some diffusion from those areas might have taken place. Perhaps the situation is not
greatly different in other regions in the country during the EB period. As more effort is put
into studying the local data bases, and considering them in relation to social organization, we
should see advances in our understanding of the EB culture, certainly one of the most interest-
ing in the history of this region.

REFERENCES



Adams, R. M.
1966



The Evolution of Urban Society: Early Mesopotamia and Prehispanic Mexico.



Chicago: Aldine.
Albright, W. F.

1926



The Jordan Valley in the Bronze Age.



Annual of the American Schools of Oriental Research
(1924–25). New Haven: American Schools of Oriental Research.

Amiran, R.
1970 The Beginnings of Urbanization in Canaan. Pp. 83–100 in



Near Eastern Archaeology in the
Twentieth Century: Essays in Honor of Nelson Glueck,



ed. James A. Sanders. Garden City, NY:
Doubleday.

1986 The Fall of the Early Bronze Age II City of Arad.



Israel Exploration Journal



36: 74–76.
Amiran, R., and Ilan, O.

1992



Arad, eine 5000 Jahre alte Stadt in der Wüste Negev, Israel.



Neumünster: Karl Wachholtz.
Bennett, C.–M.

1980 Soundings at Dhrâº, Jordan.



Levant



12: 30–39.
Beynon, D. E.; Donahue, J.; Schaub, R. T.; and Johnston, R. A.

1986 Tempering Types and Sources for Early Bronze Ceramics from Bâb edh-Dhr⺠and Numeira,
Jordan.



Journal of Field Archaeology



13: 297–305.
Butzer, K. W.

1982



Archaeology as Human Ecology.



Cambridge: Cambridge University Press.
Coe, M.

1961 Social Typology and the Tropical Forest Civilizations.



Comparative Studies in Society and His-
tory



4: 65–85.
Coogan, M. D.

1984 Numeira 1981.



Bulletin of the American Schools of Oriental Research



255: 75–81.
Donahue, J.

1981 Geologic Investigations at Early Bronze Sites. Pp. 137–54 in



The Southeastern Dead Sea Plain
Expedition: An Interim Report of the 1977 Season,



eds. W. E. and R. T. Schaub. Annual of the
American Schools of Oriental Research 46. Cambridge, MA: American Schools of Oriental
Research.

Esse, D. L.
1991



Subsistence, Trade, and Social Change in Early Bronze Age Palestine



. Studies in Ancient Ori-
ental Civilization 50. Chicago: Oriental Institute Press.

Finkelstein, I.; Ussishkin, D.; and Halpern, B., eds.
2000



Megiddo III: The 1992–1996 Seasons, Vols. I–II



. Tel Aviv: Emery and Claire Yass Publications
in Archaeology, Institute of Archaeology, Tel Aviv University.

Harlan, J. R.
1981 Natural Resources of the Southern Ghor. Pp. 155–64 in



The Southeastern Dead Sea Plain Expe-
dition: An Interim Report of the 1977 Season,



eds. W. E. Rast and R. T. Schaub, Annual of the
American Schools of Oriental Research 46. Cambridge, MA: American Schools of Oriental
Research.

Ibrahim, M., and Mittmann, S.
1994 Excavations at Khirbet ez-Zeraqon, 1993.



Newsletter of the Institute of Archaeology and
Anthropology Yarmouk University



16: 11–15.

oi.uchicago.edu





532



WALTER E. RAST



Jacobsen, T.
1943 Primitive Democracy in Ancient Mesopotamia.



Journal of Near Eastern Studies



2: 159–72.
1957 Early Political Development in Mesopotamia.



Zeitschrift für Assyriologie,



Neue Folge 18: 91–
140. Berlin: de Gruyter.

Joffe, A. H.
1993



Settlement and Society in the Early Bronze Age I and II Southern Levant: Complementarity and
Contradiction in a Small-Scale Complex Society.



Monographs in Mediterranean Archaeology 4.
Sheffield: Sheffield Academic Press.

Körber, C.
1993 Edh-Dhr⺠Survey 1992.



Annual of the Department of Antiquities of Jordan



37: 550–53. Kenneth
Wayne Russell Memorial Volume.

Lapp, N. L.
1989 Cylinder Seals and Impressions of the Third Millennium



b.c.



from the Dead Sea Plain.



Bulletin
of the American Schools of Oriental Research



273: 1–15.

Lapp, P. W.
1967 The 1966 Excavations at Tell Taºannak.



Bulletin of the American Schools of Oriental Research



185: 2–39.
1969 The 1968 Excavations at Tell Taºannak.



Bulletin of the American Schools of Oriental Research



195: 2–49.

Loud, G.
1948



Megiddo II: Seasons of 1935–39. Text.



Oriental Institute Publications 62. Chicago: University of
Chicago Press.

Mattingly, G. L.
1983 Nelson Glueck and Early Bronze Age Moab.



Annual of the Department of Antiquities of Jordan



27: 461–89.

Mazar, A.
1990 Archaeology of the Land of the Bible: 10,000–586



b.c.e.



New York: Doubleday.

McCreery, D. W.
1980 The Nature and Cultural Implications of Early Bronze Age Agriculture in the Southern Ghor of

Jordan: An Archaeological Reconstruction. Ph.D. dissertation, University of Pittsburgh.

Raikes, T. D.
1980 Notes on Some Neolithic and Later Sites in Wadi Araba and the Dead Sea Valley.



Levant



12:
40–60.

Rast, W. E.
1995 Building on Marl: The Case of Bâb edh-Dhrâº



. Studies in the History and Archaeology of
Jordan



5: 123–28.

Rast, W. E., and Schaub, R. T.
1978 A Preliminary Report of Excavations at Bâb edh-Dhrâº, 1975. Annual of the American Schools

of Oriental Research 43: 1–49.
1980 Preliminary Report of the 1979 Expedition to the Dead Sea Plain, Jordan.



Bulletin of the Ameri-
can Schools of Oriental Research



240: 21–61.
1981



The Southeastern Dead Sea Plain Expedition: An Interim Report of the 1977 Season



. Annual of
the American Schools of Oriental Research 46. Cambridge, MA: American Schools of Oriental
Research.

Renfrew, C., and Bahn, P.
1991



Archaeology: Theories, Methods, and Practice.



New York: Thames and Hudson.

Sauer, J. A.
1986 Transjordan in the Bronze and Iron Ages: A Critique of Glueck’s Synthesis.



Bulletin of the
American Schools of Oriental Research



263: 1–26.

Schaub, R. T.
1982 The Origins of the Early Bronze Age Walled Town Culture of Jordan. Pp. 67–75 in



Studies in
the History and Archaeology of Jordan



I, ed. A. Hadidi. Amman: Department of Antiquities.

oi.uchicago.edu





533



EARLY BRONZE AGE STATE FORMATION IN THE SOUTHEAST DEAD SEA PLAIN



1987 Ceramic Vessels as Evidence for Trade Communication during the Early Bronze Age. Pp. 247–
50 in



Studies in the History and Archaeology of Jordan



III, ed. A. Hadidi. Amman: Department
of Antiquities.

1993 Bâb edh-Dhrâº. Pp. 130–36 in



The New Encyclopedia of Archaeological Excavations in the Holy
Land,



vol. 1, ed. E. Stern. Jerusalem: Israel Exploration Society & Carta and New York: Simon
& Schuster.

Schaub, R. T., and Rast, W. E.
1984 Preliminary Report of the 1981 Expedition to the Dead Sea Plain, Jordan.



Bulletin of the Ameri-
can Schools of Oriental Research



254: 35–60.
1989 Bâb edh-Dhrâº



: Excavations in the Cemetery Directed by Paul W. Lapp (1965–67), Reports of
the Expedition to the Dead Sea Plain, Jordan,



vol. 1. Winona Lake, IN: Eisenbrauns.

Service, E. R.
1971



Primitive Social Organization: An Evolutionary Perspective.



2nd ed. New York: Random House.

Trigger, B. G.
1978



Time and Traditions: Essays in Archaeological Interpretation.



New York: Columbia University
Press.

Wilkinson, A.
1989 Objects from the Early Bronze II and III Tombs. Pp. 444–70 in



Bâb edh-Dhrâº: Excavations in
the Cemetery Directed by Paul W. Lapp (1965–67), Reports of the Expedition to the Dead Sea
Plain, Jordan,



vol. 1, ed. R. T. Schaub and W. E. Rast. Winona Lake, IN: Eisenbrauns.

Wittfogel, K. A.
1957



Oriental Despotism: A Comparative Study of Total Power.



New Haven: Yale University Press.

Worschech, U. F. Ch.
1986 The Fourth Survey Season in the North-West Ard el-Kerak, and Soundings at Baluº 1986.



Annual of the Department of Antiquities of Jordan



30: 285–90.

Yeivin, S.
1934 Masonry of the Early Bronze People.



Palestine Exploration Fund Quarterly Statement



(1934):
189–91.

oi.uchicago.edu





534



WALTER E. RAST



PLATE 26.1. The town site of Bâb edh-Dhr⺠looking northwest, seen here (center left) between the road and the

Wadi Kerak. Note the two major erosional incisions that have removed large parts of the inner town. West of Bâb

edh-Dhr⺠the Wadi Kerak turns northward to empty its water first into the Ghor el-Mazraªa and then into the

Dead Sea at the northern end of the Lisan peninsula. (Photograph by Robert Johnston.)

oi.uchicago.edu



535



DETERMINIST OR NOT DETERMINIST?:
CLIMATE, ENVIRONMENT, AND

ARCHAEOLOGICAL EXPLANATION
IN THE LEVANT



Arlene M. Rosen and Steven A. Rosen



INTRODUCTION: A PRELIMINARY DEBATE

At the beginning of the twentieth century, Ellsworth Huntington (1911), one of the world’s
then foremost geographers, concluded that climatic deterioration, in the form of aridification
and desertification, was the fundamental cause of demographic and perceived cultural decline
in Palestine. The contrast between (1) the rural demographic decline of the late Ottoman
period and the harsh modern climate, and (2) the social and economic prosperity and abundant
evidence for agricultural success (taken as a proxy for climatic amelioration) in classical
times, seemed to clinch the case. Palestine was in demographic and cultural decline as a result
of an environment fundamentally incapable of supporting higher populations, and by implica-
tion, higher culture.

The adoption of Huntington’s thesis as official British Mandatory policy (Troen 1989) pro-
vided a rationale for limiting Jewish immigration to Palestine in the period between the world
wars. This in turn stimulated a major and concerted research effort to refute it by the Jewish
Agency (Ben-Gurion and Ben-Zvi 1979 [1918]), resulting in the antithesis that the decline of
Palestine was the consequence of faulty socioeconomic organization, having little or nothing
to do with climatic deterioration. Environmental deterioration, evidentially incontrovertible,
was seen as a result of poor human management (e.g., Reifenberg 1955).

This early debate on the role of climate and environment in the historical development of
the Levant is of interest for many reasons. Troen (1989) points out that it is an early and classic
example of science in the service of politics. It also encapsulates the primary pitfalls in much
of the recent discussion of how climatic and environmental variables fit into our conceptions
of historical dynamics, foreshadowing some of the misconceptions that still plague so many
discussions of historical human ecology. We classify these pitfalls into two primary categories:
misapprehensions of the environmental and climatic data, and misapprehensions of the nature
of cultural adaptations to environment and climate. This paper addresses these two general
issues, with special reference to a case study of EB collapse, attempting to place it into a
broader framework, hopefully making a case for an historical role for environment that is
important, but not deterministic.

THE MISAPPREHENSION OF ENVIRONMENT

The first issue in the comprehension of environment is that of scale, both chronological
and geographical (cf. Butzer 1982). Climate and environment can be reconstructed on a variety

long



27



oi.uchicago.edu





536



ARLENE M. ROSEN AND STEVEN A. ROSEN



of scales, each of which may have significantly different meaning for human-environment
interactions (cf. Butzer 1978). That is, climatic fluctuations can vary in intensity, as in greater
or lesser amounts of rainfall or temperature; they may vary in length of time, from the
extremely short term, in essence weather, to the long term constituting fundamental climatic
belt shifts. They may vary in area affected, from the microenvironmental to the global. It is
possible to view these scalar effects graphically, as in figure 27.1. Although the examples given
in the figure suggest perhaps discrete categories or levels of environmental fluctuation, in fact,
the axes represent graded continua of phenomena.

In addition to these axes of environmental variability, the randomness, or periodicity of
phenomena must also be considered. Environmental fluctuations occurring at regular inter-
vals, such as seasonality, are considerably different in human perception from those that occur
at apparently random intervals. Phenomena that recur over long periodicities may be treated as
random occurrences by societies which do not recognize the periodicity, but as regular events
by those that do. Finally, random events may occur in statistically regular frameworks, allow-
ing them to be treated as regular events, even if the periodicity is not regular. Thus, one may
plan for the one-in-one-hundred-year flood, or the one-in-twenty-year drought, even if one
cannot precisely predict the timing of these events. The interaction between these different
scales is notable as well. Thus, low-intensity change over the short term, for example, a 10%
drop in precipitation for a year, may require little social or economic adjustment, but a similar
drop over the long term would clearly require adaptive responses.

Understanding these differences in scale is crucial to understanding the human response to
environmental changes. All too often Levantine archaeologists conceptualize climate as simply
wet/dry or hot/cold which tells us little about the impact on cultural systems. Cultural and tech-
nological adaptations to climate and environment must be keyed to these environmental scales,
and it is crucial that we distinguish between phenomena operating on different axes, at different
levels.

Specifically, if we categorize, perhaps trivially, the glacial advances and retreats of the
Pleistocene as intense, long-term, global events, these nevertheless should not serve as a base-



FIGURE 27.1. Graphical representation of the effects of scale on environmental conditions, over time (chrono-
logical changes) and area (geographical differences).

oi.uchicago.edu

[image: image190.tif]



537



CLIMATE, ENVIRONMENT, AND ARCHAEOLOGICAL EXPLANATION



line for evaluating all later environmental change. The absence of climatic fluctuations on the
scale of the Pleistocene-Holocene transition is not equivalent to a general absence of significant
climatic or environmental change. Regional and local events of relatively higher intensity over
the medium or long term may have major effects on societies, and even low intensity events,
such as minor shifts in rainfall patterns, if extending over the longer term, may impact signifi-
cantly on human settlement. Thus, Liphschitz’s (e.g., 1986; Liphschitz, Gophna, and Lev-Yadun
1988) repeated claims that Israeli climate/environment was unchanged throughout the Holocene
because it was consistently a Mediterranean environment should in no way indicate an absence
of climatic variation capable of affecting human settlement. There is ample variation within the
general Mediterranean system to allow for major environmental inputs to historical dynamics,
especially in terms of intense regional shorter term episodes.

A second issue is that of the different processes and factors that constitute environment
and climate, and the different means we have of reconstructing them. Although many archae-
ologists tend to view climate, or environment, as a kind of single variable concept, as in
“climatic change,” or “environmental change,” these are actually comprised of a wide range of
interacting phenomena. These include topography, hydrology, rainfall amount and seasonal
distribution, soil types, temperature, and others. Each of these factors constitutes a dynamic
system unto itself, and each exerts its own influence on agricultural production and society.
Thus, topographical effects may exercise as significant an effect on water regime as rainfall,
and soils may be as important as moisture amount for determining agricultural potentials.
Even variations in temperature can influence effective rainfall by modifying the underlying
evapo-transpiration regime, without any change in precipitation. In essence, human settlement
is not affected by “environmental change,” but rather by specific components of environmental
change, and different subsystems within societies react to different components of the general
environmental system.

The issue of the different components of environment and climate is also important in that
they are reflected by different lines of evidence in the paleoecological record. Paleoenviron-
mental and climatic reconstruction is dependent on proxy data for periods and regions where
historical records do not provide direct measures of such variables as mean temperature, pre-
cipitation, vegetation cover, etcetera. The basic problem faced by the historical ecologist is
similar in this respect to that of the archaeologist, who must reconstruct ancient culture from
relict information, modified by later events.

These proxy data require processual understanding of environmental phenomena in order to
be integrated into a general environmental reconstruction. Different proxy data often represent
disparate components of environmental systems that need not be operating in concert. Con-
versely, they may be the results of interacting components which may be quite difficult to sort
out. Furthermore, they may reflect climatic and environmental events at significantly different
scales. Thus, data sets such as terrace sequences, pollen diagrams, isotope series, and lake levels
cannot simply be equated with greater rainfall or cooler temperatures. Each reflects a specific
subset of interacting environmental processes. For example, river terrace sequences are affected
by such variables as rainfall, vegetation cover, river gradient, geological substrate, and channel
shape. Changes in pollen frequencies are affected directly by changes in vegetation patterns
(although not necessarily in linear fashion), but vegetation patterns are determined by a wide
range of factors, including soil types, drainage, rainfall, temperature, hydrological regime, plant

oi.uchicago.edu





538



ARLENE M. ROSEN AND STEVEN A. ROSEN



succession patterns, shifting patterns of seasonality, and, not by any means least, anthropogenic
factors.

Even apparently clear paleoclimatic indicators, such as isotopic measurement of snail
shells reflecting the plant communities consumed by the snails and hence the general climatic
regime under which the snails lived, require the reconstruction of a chain of environmental
processes. Variation in the C3–C4 plant communities as represented in the snail shell isotope
data reflect shifts of the desert margins that may have been effected by a number of factors,
including fluctuations in seasonal distribution of rainfall, changes in evapo-transpiration rates
caused by changes in temperature, and microenvironmental changes caused by any number of
factors. Furthermore, even those shifts in the desert boundaries caused by climatic fluctuations
might have been less significant in terms of human adaptations in areas farther north. That is,
generalization from the snail isotope data may be as difficult as that from other types of data,
depending on the specific issues of environmental reconstruction.

Belaboring the status of each type of proxy data is beyond the scope of this paper. The key
point is that these data have rarely been integrated into any general climatic framework.
Instead models of climatic and environmental change, especially those trotted out for use by
archaeologists, have remained on the simplistic level of wet/dry cycles, with little true envi-
ronmental reconstruction.

The final issue is the technical one of precision in dating and defining environmental
events. Conceptually, the problem of dating environmental episodes is fairly straightfor-
ward—dating of these events is fundamentally difficult. Aside from simple problems, like
standard deviations in



14



C dates, the proxy data often have different reaction times to the
underlying environmental causes, reaction times which may be much slower than human reac-
tion times. Goodfriend (1988, 1990) has convincingly documented shifting plant communities
on the desert margin, based on snail shell isotopes, but in fact we have little data on how long
it takes for one plant community to die out and another to colonize a region. Furthermore,
such colonization time is surely dependent on distance from source communities, rendering
simplistic single number answers inappropriate. Pollen analyses suffer from similar problems.
Lag times between changing vegetation patterns and the original environmental stimuli have
rarely been addressed in paleoecological reconstruction, yet may be crucial for understanding
human adaptations which are presumably more quickly accomplished, at least on some levels.
Even the establishment of the length of an environmental episode must be considered, espe-
cially since time and intensity may often be confused in the proxy data. It is notable that in
some cases, climatic changes have been assumed to correlate with social changes, and the
social changes thus used to date the climatic events.



1



The circular reasoning here is obviously
problematic.

Beyond the issue of dating, the definition of climatic events in terms of the scales and fac-
tors discussed above has rarely even be addressed, let alone resolved. Most paleoecological
data are qualitative or relative, suggesting directional changes in environment, rather than spe-
cific environmental reconstructions. Finally, our handy charts showing sequences of hot/cold,



1. The graph of Dead Sea fluctuations in Neev and Emery 1967: fig. 17 is a good example. The radiocarbon
date is certainly not sufficient to fix all the climatic events, but they are nevertheless conveniently placed
adjacent to major cultural breaks.

oi.uchicago.edu





539



CLIMATE, ENVIRONMENT, AND ARCHAEOLOGICAL EXPLANATION



dry/wet, alluviation/erosion, or trees/grasses, usually modeled on archaeological periodization
schemes, are fundamentally simplistic, and in fact tell us almost nothing about the real events
that influenced human behavior. As above, the point is not a critique of each individual chro-
nology of climatic events, rather that these chronologies require more critical analysis than is
usually undertaken.

THE MISAPPREHENSION OF CULTURE

In much the same manner that archaeologists have tended to view climate and environ-
ment in overly simplistic ways, so natural scientists have tended to reduce history to a single
variable, called perhaps “culture,” or “society.” As with environment, this reduction is so
overly simplified as to be fundamentally wrong. Culture is so complex a concept as to defy
consensus definition by anthropology. However, for the purposes of the examination of the
role of environment in history, several aspects of culture require explication.

As with the scheme presented above for examining environment, so cultures can be
viewed as consisting of component parts, operating on a range of scales (fig. 27.2). The scale
at which one examines cultural phenomena must be explicit before any attempt can be made
to link culture and environment. Again following the scheme suggested for environment, three
axes can be defined: area, time, and population density. Area varies from the smallest unit of
cultural-archaeological analysis, for the purposes of this discussion, usually the site, through
regional analysis, and even onto global evolutionary events. For archaeological purposes, time
varies from the shortest periods definable archaeologically or historically—for example, events
like battles or destructions—through site occupation spans, cultural-archaeological horizons or
periods, and longer civilization spans. Population densities vary from the sparse occupations of
hunter-gathers in peripheral environments, through densely populated urban settings.

This issue of scale is especially important for analysis of culture-environment interac-
tion. Single site abandonments are clearly not equivalent to regional or areal abandonments,



FIGURE 27.2. Graphical representation of the effects of scale on cultural phenomena, over area, time, and popu-
lation density.

oi.uchicago.edu

[image: image191.tif]



540



ARLENE M. ROSEN AND STEVEN A. ROSEN



short-term political fluctuations differ from major political transformations, and small tech-
nological innovations should be distinguished from principal stages of technological develop-
ment. This may seem trivial to scholars accustomed to dealing with culture-historical dynamics,
but environmental stimuli do not operate in equivalent ways at these different scales, an assump-
tion which seems all too prevalent in interpretations of the relationships between culture and
environment.

Periodicities and fluctuations occur in the cultural realm, as they do in the natural realm.
Braudel (1972) has linked some of these to environmental variables and others to economic
and political factors. Another obvious source of historical periodicity is simple generational
replacement, especially among elites. Classic Hegelian and Marxist dialectics (e.g., McGuire
1992) can also be seen as patterned, if not cyclic. Of course, the underlying causes of these
apparent cycles is grist for much historical debate.

Besides operating on a range of scales, over different axes, it is also useful to examine cul-
tural systems as comprised of different component parts, interconnected, but nevertheless dis-
tinct parts of the whole. This systems approach is important since different types of stimuli on
a culture need not act universally on all components equally and at the same time. Thus, while
it is clear that such realms as technology, ideology, politics, and economy are all clearly inter-
linked in any cultural system, they nevertheless respond in different ways and at different
levels to similar stimuli. Thus, political events need have little effect on technological or sub-
sistence systems, although they may effect profound changes on social systems. Short-term
intense environmental events may affect political systems more than religious symbols or
technology. For example, bureaucrats, technocrats, politicians, theologians, and farmers will
all have different and sometimes conflicting responses to drought. Even within the realm of
one system, that of agricultural economy, there are variations in the impact of environmental
change. An extended drought will result in very different responses from subsistence farmers
than from cash-crop farmers. Finally, causalities are often complex. Le Roy Ladurie (1974)
has suggested that the demographic catastrophe of the Black Plague stimulated major techno-
logical innovation, but only after population collapse.

The different cultural or social subsystems are also represented differentially by different
types of archaeological data. It is a truism that such phenomena as religion and belief systems
are less accessible to archaeological reconstruction than more material components of culture,
such as technology or trade systems. However, it is a mistake to equate lesser visibility with
lesser significance. Ideologies almost always act as the intermediaries, or filters, between
external “physical” stimuli and cultural response, and different cultures will respond to stimuli
in different ways, depending on their ideologies, especially over the short term. Corollary to this,
similar archaeological phenomena may result from either different causes, or similar causes of
differing intensities. For example, it is easy to conceive of a variety of causes of site or regional
abandonment (e.g., Cameron and Tomka 1993; also Cordell 1984: 303–26), but it is also
worth considering that cultures with differing levels of technology or social organization may
react to “cultural” or “social” stress in significantly different ways. A cause for abandonment
in one society may well stimulate intensification and apparent cultural florescence in another.

Finally, culture and history are not to be confused with the shorthand used by scholars
which we call “periodization.” The dating of cultural horizons is not the equivalent of the dat-
ing of cultures, but in fact relies on archaeological constructs whose actual links to significant

oi.uchicago.edu





541



CLIMATE, ENVIRONMENT, AND ARCHAEOLOGICAL EXPLANATION



cultural change may be tenuous. Periodization is built on time-sensitive indicators which, in
the Levant, are not consistently employed over the range of archaeological time. Thus, peri-
odization from the Neolithic though recent times has been constructed from such criteria as
technological innovations, social changes, political changes, and ceramic styles. These defin-
ing criteria have not been applied equally in all times and places, although there is usually
general consensus on the actual cutoff points for different periods. For example, the Iron Age
(1200–586



b.c.



) is defined in Israel not by the introduction of iron technology, but by an
assumed historical event, the arrival of the Israelites, ca. 1200



b.c.



, and by the final demise
of the Judaean kingdom in 586



b.c.



Obviously, neither of these historical events has much
relevance beyond the confines of the southern Levant.

Historical process rarely conforms precisely to archaeological periodization, which by its
nature requires material reflection of that process before it can be recognized. Thus, while final
abandonments or destructions of settlements and regions can often be precisely dated, the pro-
cesses effecting those abandonments or destructions may precede them by centuries. Linkages
between date of abandonment and external causes are thereby obviously suspect.

ENVIRONMENT AND ARCHAEOLOGY: TOWARD APPREHENSION

If the above discussions seem overly theoretical, there is nevertheless one point that
should be clear: neither environment nor culture is a simple phenomenon. Neither can be
appropriately characterized by univariate graphs, and both demand critical review for integra-
tion one with the other. Furthermore, they are inextricably interrelated, and integration is of
the essence.

Although often simply ignored in archaeological explanation in the Levant, the role of
environment and environmental change, including climatic change, is crucial to understanding
ancient Levantine cultures. From the synchronic perspective, the Mediterranean zone which
constitutes the heart of the southern Levant, and which is an assumed background for its
archaeology, is a small and rather unique environmental patch in a much larger Near East.
Subsistence practices appropriate to Egypt or Mesopotamia—for example, those utilizing large
scale river irrigation—are not applicable to the Levant; nor are the terrace systems of the
Levant applicable to the great river valleys. Climatically, the Levant is milder than surrounding
areas as well, more appropriate to the mixed Mediterranean economy of grains, fruits (includ-
ing olives and their products), animal husbandry (e.g., Stager 1985) than surrounding regions.
In a very material sense, Levantine cultures are embedded in a specific ecological niche that
provides a baseline for the economic, social, and political history of the region. Levantine cul-
tures adapt, and have adapted, to fluctuations of virtually all orders in that baseline.

There are two issues here: the reality of fluctuations, and the nature of the adaptations. The
first is in the realm of natural science, and the second, in the realm of the historical sciences.
Herein lies the problem of integration.

There can be no question today as to the reality of environmental change within the
Holocene, and within historic times, even in recent history. Events such as the Little Ice Age,
or the Atlantic warm period, show up in the paleoenvironmental record worldwide. Regional
events, sometimes of perhaps lesser spans, but often of greater intensity, are accepted fact
throughout the scientific community.

oi.uchicago.edu





542



ARLENE M. ROSEN AND STEVEN A. ROSEN



Within the Near East the reality of environmental and climatic change cannot be doubted.
Besides the global events, which by definition would have affected the Levant, numerous
scholars have demonstrated the occurrence of episodes of environmental variation, reflected
over the entire span of historical times. This variation is not uniform but is reflected in a range
of proxy data, representing different processes and intensities of fluctuations.

It has been suggested that this variation in the Levant was of an order too low to significantly
affect social process and pattern (cf. Liphschitz 1986; Liphschitz, Gophna, and Lev-Yadun
1988). However, in the sense that the Levantine Mediterranean zone is a kind of large-scale
transition from the arid subtropical belt of Arabia and North Africa to the continental zone of
Anatolia, minor movements of climatic belts might have even greater effects on the Levant
(e.g., S. A. Rosen 1987; Goldberg and Bar-Yosef 1982). Small shifts in the steep north-south
precipitation gradients characteristic of Israel might effect major reorientation of vegetation
communities, especially at the steppe and desert margins. Thus, such shifts of the 200–300 mm
rainfall isohyet, the threshold for dry farming of cereals, could significantly affect the agricul-
tural potentials of the large tracts of land between Beersheva and Qiryat Gat.

The second issue is that of social and cultural response. The first point is that abandonment
and collapse do not constitute the sum total of possible responses to environmental deteriora-
tion, nor is florescence necessarily the standard response to amelioration. The concepts of col-
lapse and florescence are, of themselves, problematic (Yoffee and Cowgill 1988). The political
collapse of Rome was accompanied by economic florescence in some parts of the empire (e.g.,
Bowersock 1988), and technological florescence of the early Iron Age seems to have been a
result of the socioeconomic collapse of Late Bronze Age (LB) systems (e.g., Muhly 1980).

However, beyond the semantic issue of defining collapse, our perceptions of collapses as
sudden or as the result of a predetermined and unalterable sequence of events is a problematic
view of historical causation. Given environmental or climatic input, there is a range of possible
responses that a society or individual may take, any of which may result in significantly different
outcomes. The decisions upon which these outcomes rest are guided by a complex set of factors,
including cultural perception of the environment, and technological and social potentials for dif-
ferent kinds of response. From our Western perspective, a society that responds technologically
to environmental pressures is more likely to survive than one that responds with temple build-
ing. Yet in the early stages of environmental stress, a society responding with temple building
might well appear archaeologically to be undergoing florescence. Furthermore, although we
might assume that ultimately such a temple-building response would result in famine, alterna-
tively, it is possible to hypothesize that the centralization of power established by such acts
might allow more effective utilization of resources and manpower, perhaps enabling a society
to survive environmental deterioration through social adjustments instead of technological ones.
Other responses, well within the potentials of a society, are also possible, such as diversification
of production, intensification, or, of course, technological innovation.

The point, of course, is that historical explanation demands more than assumed causality
based on mere correlation or coincidence. Social and cultural processes mediate at all levels
of human-environment relations. It is this mediation which determines historical outcomes,
and which, in combination with the study of environment, should be the focus of historical
explanation.

short

oi.uchicago.edu





543



CLIMATE, ENVIRONMENT, AND ARCHAEOLOGICAL EXPLANATION



THE CLOSE OF THE MILLENNIUM:

EXAMINATION OF THE END OF THE THIRD MILLENNIUM



B.C.



One period that has traditionally been the subject of much of the debate concerning cul-
ture-environment interactions in the Levant is the third millennium



b.c.



, the Early Bronze Age
(EB) and its successor, the Middle Bronze (MB) I. The reason for the focus on this period is
related in no small measure to the great and abrupt social changes evident in the transition,
and, importantly, the characterization of these changes as “collapse.” Climatic change can be
attributed to period as well. However, the coincidence between sociopolitical collapse and cli-
matic change should not be confused with perfect correlation nor with direct causality.

The reality of climatic change during the fourth–third millennia



b.c.



is indisputable. A
wide range of proxy data strongly indicates an ameliorated climate at the beginning of the EB
and a deteriorated one at the end.

Amelioration consists of the following lines of evidence: high Dead Sea levels (Frumkin
et al. 1994), and increased runoff (Neev and Emery 1967; Neev and Hall 1977), wadi terrace
alluviation in both the Mediterranean zone (A. M. Rosen 1986) and the arid steppe zone
(Goldberg and Rosen 1987), southward extension of C3 plant communities, as indicated by
snail shell isotopic data (Goodfriend 1988, 1990). Deterioration at the end of the period con-
sists of low Dead Sea levels, wadi incision resulting in abandoned floodplains (A. M. Rosen
1991, 1997), and the northward movement of C3 plant communities.

These different lines of evidence amount to more than a simple outline of wetter/cooler or
drier/warmer climate superimposed on a landscape similar to that of today. In terms of ame-
lioration, the fine alluvial deposits of terraces associated with the EB ceramics in the south and
central Israel suggest an aggrading alluvial regime, with more regular river flow and increased
spring flows. Such systems would have been ideal for simple floodwater farming, and indeed
evidence from plant phytoliths points to cereal cultivation in moist alluvial soils during this
time period (A. M. Rosen 1995). These conditions would have been especially notable for the
steppes and desert fringes, allowing more intensive agricultural exploitation of these regions.
For the higher rainfall areas in the north, farming in the periodically flooded valley bottoms
would have provided stability and predictability of yields even in drier rainfall years. Increased
vegetation cover in the Negev highlands suggests enhanced grazing territories. Areas farther
north would achieve generally greater stability in agricultural regimes.

In terms of deterioration, decreased rainfall probably affected both the southern and northern
regions, although to different degrees. Erosion and downcutting of wadis would have rendered
simple floodwater farming impossible, and greater instability in the rainfall regime probably
rendered agriculture a more risky enterprise even in areas where absolute average amounts were
still adequate for dry farming. In general, the desert boundary moved north, rendering the lower
rainfall areas unusable for farming. The higher rainfall areas suffered lower yields which
lowered the overall carrying capacity of the region.

Dating these episodes is problematic. Although there is rough agreement about the general
climatic periods, dating the specific processes, and the thresholds associated with them, is less
easily accomplished. For example, although we know that erosion replaced alluviation as the
dominant hydrological process some time in the latter half of the third millennium, we cannot

oi.uchicago.edu





544



ARLENE M. ROSEN AND STEVEN A. ROSEN



precisely date this shift in different areas. And indeed, it probably differed from north to south.
Nor can we date precisely when streams ceased to flow. Similar difficulties obtain in dating the
northward expansion of the desert zone and associated plant communities.

The correlation of the social collapse with the evidence for environmental change is also
problematic. Although Issar (1995) ties the abandonment of Tel Arad to climatic decline at
2600



b.c.



, this is at least 400 years prior to the collapse of the rest of the EB system, and he
ignores the rise of such southern EB towns as Tell el-Hesi and Tel



Ó



alif (Lahav). Although he
cites the decline in number of settlements in the Beersheva basin during the MB I period (cit-
ing Govrin’s 1991 data) as the result of desiccation, he ignores the MB I settlement floruit in
even more arid areas farther south (e.g., Cohen 1999; Haiman 1986, 1993).

The sociocultural picture is no less complex. The EB city-states consisted of a hierarchical
society ruled by an elite class whose wealth and power most likely derived from control of
trade, in part the exchange of luxury goods such as olive oil and wine (Stager 1985; Joffe
1993: 82–86). The lower classes of society who provided the necessary labor for the produc-
tion of these products were probably at least partially integrated into the system by loyalties
to a temple cult and its perceived influence on the fertility of crops and herds (e.g., Ben-Tor
1992; Amiran 1972; Joffe 1993: 83) as well as a system of grain redistribution in times of
periodic drought. This relationship of labor in exchange for security, plus the buffer effect of
floodwater farming, functioned as a successful adaptation and equilibrium in a semiarid envi-
ronment with periodic droughts. The climate deterioration and resulting hydrological changes
removed the buffer provided by floodwater farming, thus undermining the ability of the soci-
ety to provide for its population in times of drought. This would have dramatically decreased
the carrying capacity of the region without the introduction of a technological change in water
management. The EB city-states were incapable or unwilling to utilize other systems of agri-
cultural intensification. The reasons for this are unique to this particular cultural milieu and
should be investigated as such.

Recent surveys and excavations in the center of the country suggest that although the
urban system of the EB collapsed, in fact a village system, based on subsistence farming and
animal husbandry, replaced it. Change in settlement patterns seems to reflect a shift in agricul-
tural organization and economic realignment, and not the general collapse of farming in favor
of pastoral nomadism.

The dynamics of Canaanite economic systems must also be considered, especially in the
context of a larger Near East. The development of Egyptian sea trade, especially via Byblos,
probably acted as an economic stimulus to northern Canaan, whereas the Egyptian withdrawal
from the Shephelah during the EB I and the virtual cessation of trade at the end of the EB II
may well have precipitated economic crisis. If copper was the fuel of the Aradian economy,
then Egyptian access to other sources, via the sea route, may well have contributed to the Ara-
dian decline.

On the other hand, the growth of Mesopotamian and Syrian states in the later third millen-
nium had its effect on the Levant as well. The general Near Eastern collapse at the end of the
third millennium was an economic catastrophe, even for those areas that might not have been
directly affected by such primary factors as environmental decline.

It is possible to construct a model of the EB collapse that ignores the role of environment,
just as it is possible to construct one that ignores social and economic processes. The point

oi.uchicago.edu





545



CLIMATE, ENVIRONMENT, AND ARCHAEOLOGICAL EXPLANATION



here is that environmental deterioration did occur, and the EB society must have reacted in
some way. Even lack of reaction is a reaction, albeit an unlikely one.

The effects of the environmental changes reflected in the proxy data seem to be that rain-
fed agriculture in both the more temperate and marginal zones would have become progres-
sively less predictable with an increase in drought years and a resulting decline in carrying
capacity. Floodwater farming, which previously could have acted as a buffer for drought years
also would have suffered as a consequence of decreasing water flow and downcutting stream-
beds. Although the timing of these changes is not precise, there is a gradient of environmental
effect—not all regions would have been affected equally. Agricultural collapse was not all-
encompassing, but selective. This suggests a kind of cascade effect on the economic system
triggered by increased pressures in the marginal zones, declining yields in the temperate
zones, and decreases in the economically important tree and vine crop industries. Such an out-
come is not sudden, even if the environmental effects on some regions may have been intense
and relatively rapid. Furthermore, the impact on nomadic populations in the deserts and desert
fringes may have been just as great. Ethnographically, Bedouin seasonal rounds in the Negev
may extend hundreds of kilometers farther north in drought years than in normal years, well
into the Mediterranean heartland. Long-term decline in rainfall in the Negev seems to corre-
spond to more nucleated settlements in the MB I than were present in the EB (florescence, in
a sense!). In the Negev highlands, where the largest of these settlements have been found, they
are almost always located adjacent to springs. The abandonment of these settlements by the
beginning of the second millennium



b.c.



—that is, at the beginning of the MB II—can be seen
as a continuation of a process of general settlement realignment which may in fact have begun
much earlier, in the EB period.

The infiltration and probable eventual settlement of nomads in the better-watered regions
cannot yet be addressed archaeologically, but texts are clear in indicating the movements of
what appear to be nomadic peoples—namely, the Amorites—during the end of the third and
beginning of the second millennium



b.c.



At least in some instances, hostilities ensued. If tying
these migrations evidentially directly to climatic stimuli is not possible simply because of lack
of concrete evidence, surely the coincidence of events cannot be dismissed as totally unrelated
one to the other.

The case of the EB collapse provides us with an effectual laboratory for examining the
seeming dichotomy between environmental determinism and historical process. The facts that
are readily acknowledged are the existence of a degree of environmental deterioration roughly
coincident with the fall of the EB Canaanite civilization and the abandonment of the vast
majority of towns and city sites. The question worthy of examination here lies in the nature of
this causality. Some climatologists and archaeologists point to the strong evidence for envi-
ronmental deterioration which is taken as an unquestioned given for the fall of a civilization.
However, if this were the case, then why is it that later complex civilizations, such as the MB
cities, were able to survive in the same dry environment that caused the demise of the EB? On
the other hand, some archaeologists have completely rejected the environmental explanation
based on facts such as the location of the EB sites within areas that receive adequate annual
rainfall averages for dry farming, even under the modern dry regime; and also that some city-
states (for example Iktanu [Prag 1974] and Khirbet Iskander [Richard 1986] in Transjordan)
did continue to exist in full urban glory even after cities in western Palestine were abandoned.

oi.uchicago.edu





546



ARLENE M. ROSEN AND STEVEN A. ROSEN



The basic misconceptions here lie in the lack of understanding of the dynamics and resil-
ience of cultural systems in the first case, and a misapprehension of the interaction of climate,
temporal distribution of rainfall, landscape changes, and agricultural economies in the sec-
ond. In order to evaluate the effects of environmental change on a given society such as that
of the EB III, it is necessary to open the cultural “black box” and attempt to examine the eco-
nomic and social effects of a climatic change in order to postulate a probable adaptive
response given the unique social, technological, economic, and political parameters of the
system under study.

So the final answer, in one sense, to the question of the southern Levantine EB cultural
response to climatic deterioration is, in fact, collapse. There does appear to have been environ-
mental stress which seems to have overwhelmed some (although not all) of the civilizations of
the third millennium



b.c.



in the Near East, and in the Levant specifically. But such an expla-
nation is so fundamentally incomplete as to verge on being wrong. The EB society was not
bowled over overnight, regardless of recent volcanic scenarios. The agricultural technologies
available to the reurbanizing MB II peoples, such as better water management, wells, inten-
sive terracing, and possible canal irrigation (cf. Miller 1980), which allowed them to adapt
and flourish in the drier environmental circumstances of the second millennium



b.c.



, were no
less available to the people of the EB. These were not inventions, but adoptions. Thus, the
issue of the timing of these adoptions is of the essence, and this is a question of cultural adap-
tations which are structured by social perceptions and receptiveness, political organization,
and economics, and not merely the occurrence of climatic catastrophe (A. M. Rosen 1995).

CONCLUSIONS

The simplistic environment-culture dichotomy has deluded scholars far too long. There is
no dichotomy. Ancient civilizations did not rise and fall in abrupt episodes initiated by cli-
matic catastrophes. Rather, they were fine-tuned to environment, constantly adjusting various
components of society in response to both major and minor environmental fluctuations. The
view of societies as static, stable, and inflexible is based on archaeological constructs such as
rigid periodization schemes which may either mask or overemphasize change through their
terminologies, or may simply lack the precision to define it well.

This is not to say that social systems did not collapse as a result of environmental pres-
sures, but rather that these collapses were the result of social failures: wrong decisions, poor
investments or planning, and inappropriate responses to natural events, not merely the occur-
rence of the natural events. We simply do not notice the successful adaptations, or take them
as natural and obvious. Yet, it is obvious that cultural response to environment cannot always
be inappropriate because no society could survive even minor fluctuations were that the case.

It might be argued that cultures are able to respond only up to certain intensities of climatic
or environmental deterioration, and that the most intense events leave no option but collapse.
Aside from historical examples that could be trotted out in counterargument, even marked envi-
ronmental events that triggered collapses have often been followed by readaptation. The issue
is really response time, not the intensity of the environmental deterioration. There is no set rule
for what environmental stimuli will initiate collapse. It varies with every culture and historical
circumstance. The key issue must be the reconstruction of the interactions between environment
and culture, and not the mere claim of causality.

long

oi.uchicago.edu





547



CLIMATE, ENVIRONMENT, AND ARCHAEOLOGICAL EXPLANATION



ACKNOWLEDGMENTS

We are grateful to Sam Wolff for inviting us to participate in this volume. This paper was
completed in 1995 while we were visiting scholars in the Department of Archaeology at the
University of Calgary, where Steven held a Canada-Israel Academic Exchange Fellowship.
We profited greatly from the intellectual stimulation provided by both faculty and students
there.

We are both also indebted to Doug Esse. His patience in explaining the Levantine Early
Bronze Age to two anthropologists will be remembered, and his friendship always cherished.

REFERENCES



Amiran, R.
1972 A Cult Stele from Arad.



Israel Exploration Journal



22: 86–88.

Ben-Gurion, D., and Ben-Zvi, Y.
1979 [1918]



Eretz Israel in Past and Present



. Jerusalem: Yad Ben Zvi. (Translated from the Yiddish to
Hebrew by D. Niv).

Ben-Tor, A.
1992 The Early Bronze Age. Pp. 81–125 in



The Archaeology of Ancient Israel



, ed. A. Ben-Tor. New
Haven: Yale University Press and the Open University of Israel.

Bowersock, G. W.
1988 The Dissolution of the Roman Empire. Pp. 165–75 in



The Collapse of Ancient States and Civi-
lizations,



eds. N. Yoffee and G. L. Cowgill.







Tucson: University of Arizona Press.

Braudel, F.
1972



The Mediterranean and the Mediterranean World in the Age of Philip II



. New York: Harper
and Row.

Butzer, K. W.
1978 Cultural Perspectives on Geographical Space. Pp. 1–14 in



Dimensions of Human Geography



,
ed. K. W. Butzer. Research Paper 186. Chicago: Dept. of Geography, University of Chicago
Press.

1982



Archaeology as Human Ecology: Method and Theory for a Contextual Approach



. Cambridge:
Cambridge University Press.

Cameron, C. M., and Tomka, S. A.
1993



Abandonment of Settlements and Regions: Ethnoarchaeological and Archaeological Ap-
proaches



. Cambridge: Cambridge University Press.

Cohen, R.
1999 A



ncient Settlement of the Central Negev. Vol. 1: The Chalcolithic Period, The Early Bronze Age,
and The Middle Bronze Age I



. Israel Antiquities Authority Reports No. 6. Jerusalem: Israel
Antiquities Authority.

Cordell, L.
1984



The Prehistory of the Southwest



. New York: Academic Press.

Frumkin, A.; Carmi, I.; Zak, I.; and Magaritz, M.
1994 Middle Holocene Environmental Change Determined from the Salt Caves of Mount Sedom,

Israel. Pp. 315–32 in



Late Quaternary Chronology and Paleoclimates of the Eastern Mediterra-
nean,



eds. O. Bar-Yosef and R. S. Kra.







Tucson: Radiocarbon and the University of Arizona Press.

Goldberg, P., and Bar-Yosef, O.
1982 Environmental and Archaeological Evidence for Climatic Change in the Southern Levant.

Pp. 399–414 in



Paleoclimates, Paleoenvironments, and Human Communities in the Eastern
Mediterranean in Later Prehistory



, eds. J. Bintliff and W. Van Zeist. British Archaeological
Reports, International Series 133. Oxford: British Archaeological Reports.

Goldberg, P., and Rosen, A. M.
1987 Early Holocene Paleoenvironments of Israel. Pp. 23–33 in



Shiqmim I



, ed. T. E. Levy. British
Archaeological Reports, International Series 356. Oxford: British Archaeological Reports.

oi.uchicago.edu





548



ARLENE M. ROSEN AND STEVEN A. ROSEN



Goodfriend, G. A.
1988 Mid-Holocene Rainfall in the Negev Desert from



13



C of Land Snail Shell Organic Matter.



Nature



333: 757–60.
1990 Rainfall in the Negev Desert during the Middle Holocene, Based on



13



C of Organic Matter in
Land Snail Shells.



Quaternary Research



34: 186–97.

Govrin, Y.
1991



Archaeological Survey of Israel Map of Nahal Yattir



(139)



.



Jerusalem: Israel Antiquities
Authority.

Haiman, M.
1986



Archaeological Survey of Israel Map of Har Hamran Southwest



(198). Jerusalem: Israel Depart-
ment of Antiquities.

1993



Archaeological Survey of Israel Map of Har Hamran Southeast



(199). Jerusalem: Israel Antiq-
uities Authority.

Huntington, E.
1911



Palestine and Its Transformation.



Boston: Houghton and Mifflin.

Issar, A. S.
1995 Climatic Change in History of the Middle East.



American Scientist



83: 350–55.

Joffe, A. H.
1993



Settlement and Society in the Early Bronze I and II in the Southern Levant.



Sheffield: Sheffield
Academic Press.

Le Roy Ladurie, I.
1974



The Peasants of Languedoc



. Urbana: University of Illinois Press.

Liphschitz, N.
1986 The Vegetational Landscape and the Macroclimate of Israel during Prehistoric and Protohistoric

Periods.



Mitkufat Haeven



19: 80–90.

Liphschitz, N.; Gophna, R.; and Lev-Yadun, S.
1988 Man’s Impact on the Vegetational Landscape of Israel in the Early Bronze Age II–III. Pp. 263–

68 in



L’urbanisation



, ed. P. de Miroschedji. British Archaeological Reports, International Series
No. 527. Oxford: British Archaeological Reports.

McGuire, R.
1992



A Marxist Archaeology



. San Diego: Academic Press.

Miller, R.
1980 Water Use in Syria and Palestine from the Neolithic to the Bronze Age.



World Archaeology



11:
331–41.

Muhly, J. D.
1980 The Bronze Setting. Pp. 25–68 in



The Coming of the Age of Iron



, eds. T. A. Wertime and J. D.
Muhly. New Haven: Yale University Press.

Neev, D., and Emery, K. O.
1967



The Dead Sea



. Israel Geological Survey Bulletin 41. Jerusalem.

Neev, D., and Hall, J. K.
1977 Climatic Fluctuations during the Holocene as Reflected by Dead Sea Levels. Unpublished Pro-

ceedings of the International Conference on Terminal Lakes, Webster State College, Utah, 1977.

Prag, K.
1974 The Intermediate Early Bronze–Middle Bronze Age: An Interpretation of the Evidence from

Transjordan, Syria and Lebanon.



Levant



6: 69–116.

Reifenberg, A.
1955



The Struggle Between the Desert and the Sown



. Jerusalem: Jewish Agency.

Richard, S.
1986 Excavations at Khirbet Iskander, Jordan: A Glimpse of Settled Life during the “Dark Age” in

Palestinian Archaeology.



Expedition



28: 3–12.

Rosen, A. M.
1986



Quaternary Stratigraphy and Paleoenvironments of the Shephela, Israel



. Geological Survey of
Israel, Reports 25/86. Jerusalem: Geological Survey of Israel.

oi.uchicago.edu





549



CLIMATE, ENVIRONMENT, AND ARCHAEOLOGICAL EXPLANATION



1991 Early Bronze Age Tel Erani: An Environmental Perspective.



Tel Aviv



18: 192–204.
1995 The Social Response to Environmental Change in Early Bronze Age Canaan.



Journal of Anthro-
pological Archaeology



14: 26–44.
1997 Environmental Change and the Human Response in The Early Bronze Age Southern Levant. In



North Atlantic Treaty Organization Advanced Research Workshop on Climatic Change in the
Third Millennium



bc



and the Social Response



, eds. N. Dalfes, G. Kukla, and H. Weiss. Kemer,
Turkey: North Atlantic Treaty Organization.

Rosen, S. A.
1987 Demographic Trends in the Negev Highlands: Preliminary Results from the Emergency Survey.



Bulletin of the American Schools of Oriental Research



266: 45–58.

Stager, L. E.
1985 The First Fruits of Civilization. Pp. 172–88 in



Palestine in the Bronze and Iron Ages: Papers in
Honor of Olga Tufnell



, ed. J. Tubb. London: Institute of Archaeology.

Troen, I.
1989 Calculating the “Economic Absorptive Capacity” of Palestine: A Study of the Political Uses of

Scientific Research.



Contemporary Jewry



10: 19–38.

Yoffee, N., and Cowgill, G. L., eds.
1988



The Collapse of Ancient States and Civilizations.



Tucson: University of Arizona Press.

oi.uchicago.edu





550



ARLENE M. ROSEN AND STEVEN A. ROSEN

oi.uchicago.edu



551



A SERIES OF RADIOCARBON DATES FROM
THE LATE EARLY BRONZE AGE I SITE AT



Ó



ORVAT ºILLIN TA



Ó



TIT



Dror Segal and Israel Carmi



INTRODUCTION

In December 1991 a number of samples for radiocarbon dating were sent to our laboratory
from the excavation of Strata III and IV of



Ó



orvat ºIllin Ta



˙



tit (see Braun et al., chap. 4, this
volume) near Beth Shemesh, Israel. The samples, collected in the summer of 1991, were pro-
vided by the excavator from selected loci, excavated under tight stratigraphic control. A num-
ber of samples were discarded for lack of organic material and the presence of inorganic
contaminants. However, ten remaining samples were found to be worthy of testing. They
include two derived from short-lived plants, especially good for obtaining dates.

THE SAMPLES

A description of the samples is provided in table 28.1



Table 28.1.



Radiocarbon Samples from the Late EB I site at



Ó



orvat ºIllin Ta



˙



tit



Lab. No. Description



1 RT 1567 Charred olive stones (



olea europaea



) derived from a hearth in Locus 273,
Basket 1391

2 RT 1572 Charred wood on the floor of a house with ceramic vessels, in situ, Locus
117, Basket 1065

3 RT 1573 Charcoal from Locus 174, Basket 1342

4 RT 1576 Charcoal from Locus 258, Basket 786

5 RT 1602 Charcoal from Locus 228, Basket 618

6 RT 1603 Charcoal from Locus 283, Basket 1049

7 RT 1604 Charred emmer wheat* found in a cermaic vessel, in situ, on a floor of a
building; Locus 258, Baskets 813 and 976

8 RT 1660 Charcoal (



olea europaea



) from Locus 258, Basket 976

9 RT 1661 Charcoal from Locus 279, Basket 944

10 RT 1662 Charcoal (



olea europaea



) from Locus 258, Basket 965



* Thanks are due to U. Baruch (Israel Antiquities Authority) and N. Lipschitz (Tel Aviv University) for
the identification of these grains.



28



oi.uchicago.edu





552



DROR SEGAL AND ISRAEL CARMI



The measurement procedures included cleaning of the samples with hydrochloric acid,
combustion in oxygen conversion to lithium carbide, acetylen (addition of water), and finally
to methane (addition of hydrogen), which is the counting gas in the proportional counters of
the laboratory (Carmi 1987).

The cleaning and preparation of the samples in the laboratory was done by standard meth-
ods (Mook and Waterbolk 1985: 34–42; Gupta and Polach 1985: 8–12). No special problems
were encountered in this process.

The results of the measurements are shown in table 28.2.

DISCUSSION

Sample RT-1567 was shown to be modern olive pips, undoubtedly related to agricultural
activity on the site during this century. It should be noted that remains of an olive grove still
cover most of the hillside.



a. Conventional radiocarbon age (1950).
b. Standard dendrochronological calibration (Stuiver and Reimer 1993).
c. The results are given at 1 sigma uncertainty, and the probability shows the confidence in percent-

age. When the results fall on a curve, there is likely to be more than one calendaric age indicated.
In that case the results are indicated by their relative reliability by percentage. An alternate possi-
bility is to combine the calendaric ages and get a probability of 100%; e.g., RT-1604 would be
indicated as 3331–3048



b.c.



Table 28.2.



Radiocarbon Determinations from the Late EB I site at



Ó



orvat ºIllin Ta



˙



tit



Sample D







14



C (



o



/00)



13



C (



o



/00) YBP



a



Calendric Age



b



Probability



c



RT-1567 76.4 ±



5.6 -



25 Modern Modern

RT-1572 -



418.2 ±



2.6 -



21.97 4350 ±



35 3029–2973



b.c.



2931–2905



b.c.



65%
35%

RT-1573 -



443.3 ±



4.0 -



21.21 4705 ±



55 3616–3590



b.c.



3526–3374



b.c.



16%
84%

RT-1576 -



419.3 ±



3.8 -



21 4365 ±



50 3031–2914



b.c.



100%

RT-1602 -



446.7 ±



3.7 -



20.4 4755 ±



55 3631–3506



b.c.



3408–3385



b.c.



85%
15%

RT-1603 -



443.7 ±



5.7 (-



22) 4710 ±



80 3621–3578



b.c.



3534–3373



b.c.



22%
77%

RT-1604 -



428.3 ±



3.1 -



25 4490 ±



45 3331–3092



b.c.



3053–3048



b.c.



97%
3%

RT-1660 -



449.7±



3.7 -



21.3 4800 ±



55 3648–3517



b.c.



100%

RT-1661 -



391.4 ±



6.6 -



23.4 3990 ±



90 2615–2337



b.c.



100%

RT-1662 -



411.3 ±



3.8 -



17.1 4255 ±



50 2916–2868



b.c.



2805–2703



b.c.



55%
45%

oi.uchicago.edu





553



A SERIES OF RADIOCARBON DATES FROM



Ó



ORVAT ºILLIN TA



Ó



TIT



During the preparation of sample RT-1661 a number of minor, technical problems were
encountered. However, they should not have affected the results. The age of this sample is less
expected from an EB I occupation to which it is attributed, and it may be ascribed to either
these technical problems or to the intrusiveness of the sampled material.

The remainder of the



14



C test results, based on nine samples, indicates two clusters of
dates. The later cluster, composed of samples RT-1572, RT-1576, RT-1604, and RT-1662,
indicates dates that are, more or less, in agreement with conventionally accepted dates of the
EB I horizon (Mazar 1990: 92; Ben-Tor 1992: 82). When the results of these samples are aver-
aged out, they indicate a date of 4365



±



22 YBP which, when calibrated, yields the range of
3094–2888



b.c.



(1



s



).
The second cluster of dates is surprisingly earlier than expected but does not seem to be an

anomaly. Included are four samples—RT-1573, RT-1602, RT-1603, and RT-1660—which
give calibrated dates ranging from 3375



b.c.



to 3649



b.c.



These dates are several centuries too
early for the generally accepted time span of the material culture of this site (a late, southern
facies of EB I which is associated with the late Predynastic Period-Dynasty 0 of Egypt).



1



This discrepancy in dates can be explained by a number of possibilities. The cluster of later
dates may have included, besides the obviously seasonal grains of emmer wheat (not likely to
have been stored for more than a year or two), brushwood for fires which would accurately
reflect the age of the settlement from which it derived. The second cluster, which does not
include any obviously seasonal, organic materials, could be composed of samples of charred
wood from trees of no little age, especially if derived from inner tree rings. Another possibility
is of the “recycling” of timber from earlier settlements in the vicinity (e.g., Lipschitz and Biger
1992: 19–22).

Thus, preference for samples for



14



C dating should be given to organic materials of a sea-
sonal nature. Greater accuracy can also be insured by the multiplication of samples. A series
of dates from the same context will invariably reflect a more accurate measure of the date of
the context than will a single sample which, conceivably, could be of doubtful stratigraphic
attribution.

Despite the possible shortcomings of the results of radiocarbon dating for archaeological
sites, the method remains an important and independent source of dating as may be deduced
from the



Ó



orvat ºIllin Ta



˙



tit series.



1. For a discussion of the dating of this site, see Braun and van den Brink 1998 and Braun et al., chap. 4, this
volume.



ACKNOWLEDGMENTS

The authors wish to express their thanks to Amir Drori, Director of the Israel Antiquities
Authority, for his permission to publish this information here. Thanks are also due to Eliot
Braun and Yuval Goren for their help in the preparation and editing of this manuscript.

REFERENCES



Ben-Tor, A.
1992 The Early Bronze Age. Pp. 81–125 in



The Archaeology of Ancient Israel,



ed. A. Ben-Tor. New
Haven: Yale University Press and Open University of Israel.

oi.uchicago.edu





554



DROR SEGAL AND ISRAEL CARMI



Braun, E., and van den Brink, E. C. M.
1998 Some Comments on the Relative Dating of Tomb Uj at Umm el Gaºab and Graves 330 and 787

from Minshat Abu Omar with Imported Ware: Views from Egypt and Canaan.



Egypt and the
Levant



7: 71–94.

Carmi, I.
1987 Rehovot Radiocarbon Measurements III.



Radiocarbon



29: 100–114.

Carmi, I., and Segal, D.
1992 Rehovot Radiocarbon Measurements IV.



Radiocarbon



34: 115–32.

Gupta, S. K., and Polach, H. A.
1985



Radiocarbon Dating Practices at Australian National University



. Radiocarbon Laboratory,
Research School of Pacific Studies. Canberra: Australian National University.

Lipschitz, N., and Biger, G.
1992 Secondary and Tertiary Use of



Cedrus







libani



(Cedar of Lebanon) Timber in Construction.



Qad-
moniot



97–98: 19–21 (Hebrew).

Mazar, A.
1990



Archaeology of the Land of the Bible.



New York: Doubleday.

Mook, W. G., and Waterbolk, H. J.
1985



Handbooks for Archaeologists



3:



Radiocarbon Dating.



Groningen: European Science
Foundation.

Stuiver, M., and Reimer, P. J.
1993 Extended C



14



Database and Revised Calib 3.0 Age Calibration Program.



Radiocarbon



25:
215–30.

oi.uchicago.edu



555



PHILISTINE BICHROME PAINTED POTTERY:
SCHOLARLY IDEOLOGY

AND CERAMIC TYPOLOGY



Ilan Sharon



INTRODUCTION

Douglas Esse’s last published works (1991, 1992) were a foray into a hotly contested
question in archaeology: Can ethnic affiliation be identified on the basis of a typological cor-
pus? In biblical archaeology this question has been most extensively discussed in the frame-
work of the emergence of statehood in the Iron I period, when, it has been argued, at least
three material cultures may be clearly delineated in the archaeological record. These may be
equated with the Israelite, Canaanite (later to be Phoenician), and Philistine ethnic groups,
which were competing for ascendancy at the time (see e.g., Mazar 1990: 295–357 for a gen-
eral introduction to the period). These identifications are brought under increasing criticism in
recent years.

Esse concentrated his attack on the use of the “collared-rim jar” as the



fossil directeur



for identifying the Israelites. This article addresses another facet of the same quandary—the
identification of “Philistine pottery” with Philistine ethnicity.

Esse’s work on Israelite ethnicity may be grouped within a growing body of



processual-
functional



critiques of the previously established identity between the spatial and temporal
distribution of given pot-types and the extent of Israelite seizure of the land of Canaan (Edel-
man 1991; Knapp 1989, 1993: 77–88; Finkelstein 1994, 1995a; Bunimovitz 1994; Portugali
1994; Sharon 1994). The present essay emulates the



structure



of Esse’s argument, but invokes
a different theoretical framework. I attempt to build a



symbolic



model to explain the birth of a
“Philistine” ethnos and the appearance of the hand painted bichrome pottery style.

Several points explicitly raised or implicitly implied in Esse’s reasoning are characteristic
of his world-view, and will underlie the present work too:

(1) The formulation of a theory cannot be explained in terms of data alone, nor of ideol-
ogy alone. Esse chose to present the growth of the database supporting (and countering) the
joint distribution of typological form and ethnic identity in the context of the questions that
engaged contemporary research and the ideologies which motivated the researchers at each
stage of the history of the discipline of biblical archaeology. Looking through the bifocal
lenses of the state of knowledge at a given time and the intellectual milieu in which it was
perceived, he analyzed how the connection between the Israelites and the collared-rim jar was
made, and why the model subsequently broke down.

(2) Esse was non-sectarian in his intellectual affiliation. He combined novel approaches
with a healthy respect for old-fashioned scholarship. Accordingly, the model that he offered
was not a total rejection of the views which had prevailed in seventy-five odd years of study,



29



oi.uchicago.edu





556



ILAN SHARON



but rather one of qualified agreement. While accepting most of the data of previous research-
ers and some of the interpretations, he subtly altered the theoretical basis upon which these
interpretations were based. He does perceive a correlation between the distribution of a
ceramic style and ethnic identity, but the relation is not one of isomorphism.

(3) Esse’s model, as mentioned above, uses functional / processual theory to explain
the emergence of new ethnic identities in late second-millennium Canaan. He sees, in the
thirteenth–tenth centuries



b.c.



, the development of a “highland” economy, of which the col-
lared-rim jar is a characteristic vessel. This economy is typical, though not restricted to, the
group which would come to be called “Israelites.” This economic system has extensive in-
teraction with the entrenched “lowland,” predominantly Canaanite, economic structure but
remains detached (at least to begin with) from a “coastal” (Philistine) economy which is
contemporaneously emerging.

Like Esse’s work on the Israelites, this paper does not present any new data. It does
not even, in the ordinary sense of the word, suggest a new explanation for the origins of the
Philistines, but rather provides theoretical justification for an existing one. In the following
pages, I trace the development of the conception equating the biblical Philistines with the
“Sea Peoples” of Egyptian literature and both of these entities with a particular style of pot-
tery decoration. The Dothans’ recent work (Dothan and Dothan 1992) does this admirably,
and I use it as the primary source of the narration. Next, I shall examine the criticism leveled
at this view, mostly in the last fifteen years. I argue that though these critiques have under-
mined the factual basis of the older paradigm, the attempts to reach a new consensus are hin-
dered by lack of theoretical backing. A semiotic view of culture and of ethnicity might
provide such a background.

In the depiction of the ideological background against which the views of Philistine cul-
tures developed over a century-and-a-half, we shall need to consider changes in the percep-
tion two concepts: The first is the significance of archaeological cultures—from a hereditary
view to a normative one, and hence to functional, systemic, and semiotic definitions of
“material culture.” The meaning of the spatial and temporal distribution of given artifact
types or attributes, in terms of historical, social or behavioral import, is a central question
(perhaps



the



central question) in the discipline of archaeology, and one for which every gen-
eration of practitioners had differing answers. Describing these in detail here would be su-
perfluous, as several textbooks (e.g., Trigger 1989, Hodder 1986) have been devoted to the
subject. Some of the relevant properties of such views that have had a bearing on the
definition of the Philistine material culture are mentioned



en passant



. It should perhaps be
pointed out here, that despite its centralistic archaeological reasoning, this polemic is argu-
ably an academic exercise limited to the archaeological community. Nevertheless the changes
in the acceptable definitions reflect general trends in the climate in which this community
operates, and the changing roles that the past was expected to play in contemporary society.

The second theme to be considered is changes in the definition of



ethnos



or nationality.
What is the exact basis for declaring oneself (or others) to be “American,” “Jewish” or “Pal-
estinian?” (to mention several ethnic designations which are more than a tad problematic . . .).
This question, of course, has a much wider and more immediate significance to society at
large, but it is influenced by the very same trends which affect the former. The emic view of
an ethnos, and of the privileges and obligations of the individual towards it, is usually con-

oi.uchicago.edu





557



PHILISTINE BICHROME PAINTED POTTERY



ceived as an extension (actual or symbolic) of the natural notion of a family. Widely divergent
etic definitions have been used for a variety of legal, moral, and political purposes, but the
following four notions play a role in most: common ancestry; shared traditions or ceremon-
ies; shared circumstances or common destiny; and self-determination. “Classic” nationalism
of the mid-nineteenth to mid-twentieth century gave preference to the first two definitions,
while in the post-war era one or the other of the second two are usually emphasized. This shift
has also colored the connection archaeologists see (or do not see) between material culture
and ethnic identity.

BEFORE 1850: BIBLICAL EXEGESIS AND CONJECTURE

Speculation about the Aegean origins of the Philistines is as old as biblical commentary
itself. It is based on two passages, in Jeremiah (47:4) and Amos (9:7), that declare the origin
of the Philistines to be in



rtpk



(



Caphtor



). The term



µyrtpk



(



Caphtorim



) is also mentioned
in Genesis 10:14, as a group related to, but distinct from, the Philistines. Deuteronomy 4:23
notes the settlement of



Caphtorim



around Gaza, overlapping the southern fringes of the
Philistine littoral. There also seems to be some connection between the Philistines and the



µytrk



—



Crethim



(Ezekiel 25:16; Zephania 2:5). The earliest translation of the Bible to
Greek, the Septuagint (probably second century



B



.



C



.



), renders



rtpk



as Cappadocia (in Cen-
tral Asia-Minor—rather strange, in view of the mention in Jeremiah, which specifically notes
that



Caphtor



is an island) and



Crethim



as Crete. Stephanus of Byzantium (sixth century



A



.



D



.) relates a tradition connecting Gaza, the major city of the Philistine pentapolis, with
king Minos (Dothan and Dothan 1992: 8).

In 1747 Etienne Fourmant identified the Philistines of the Bible with Homer’s



Pelasgians



who were said to have inhabited Greece, Asia Minor, and Crete before the coming of the
Greeks. He also was the first to claim that the biblical term for a Philistine ruler—



ˆrs



(



Seren



)
is a Hebraecised version of the Greek



turannoÍ



(



tyrannos



—tyrant). He thus initiated a long
tradition of etymological gymnastics—trying to assign a place of origin to the Philistines on
the basis of matching possible slivers of a putative Philistine language (a few names and
perhaps a couple of words which have infiltrated into the Hebrew language) with similar-
sounding words or names in Greek or other languages (cf. Singer 1988 for a recent review).
In the same vein Ferdinand Hitzig (1845) claimed that the Philistines (= Pelasgians) were
Aryans, drawing on some similarities between supposed Philistine words (



Goliath



,



Seren



,



Ashdod



,



Ashkelon



) and words in Sanskrit, while K. B. Stark (1852) claimed that the Philis-
tines were none but Phoenicians and found Semitic etymologies to most of the terms that
others had pronounced an Indo-European ancestry for (Dothan and Dothan 1992: 9–10).

IDEOLOGY

Some words are called for to explain the obsession about Philistine



origins



, which seems
to dominate the study of Philistine culture since its inception. This fixation seems peculiar to
modern European commentary, and does not necessarily emanate from the biblical source
itself. True, the genealogical lists in Genesis and Chronicles do speculate about the position
of the Philistines in the family of man, and some of the later prophets allude to a Philistine

oi.uchicago.edu





558



ILAN SHARON



homeland in foreign parts. Such assignation of place of origin and ancestries to known
peoples is part and parcel of any mythology. There does not seem to have been any mystery
attached to the origins of the Philistines in particular. The above-mentioned reference in
Amos, for instance, states in the same breath that the Arameans came from



Qir



; and yet



Qir



has not excited European exegetes, and one has yet to hear of any expeditions being orga-
nized to locate it.

Post-enlightenment Europe’s perception of the past was warped by two great magnetic
poles—the Bible and the classics (Bernal 1987: 4). An educated (or uneducated, for that mat-
ter) European’s initiation to biblical literature was at baptism and continued thereafter every
Sunday of his or her life. Formal schooling was largely a matter of learning Greek and Latin.
Whenever one could read the classics in the original, ones’ education was deemed complete.
Also, prior to the archaeological revolution of the late nineteenth century, Homer, Herodotus,
and the Old Testament



were



the only windows through which one could peek at a past more
than two millennia old. Thus the Greeks and the Hebrews loomed larger-than-life in the Euro-
pean’s intellectual conception of the past. The possibility that a connection could be found



between



these two poles ignited the romantic imagination of eighteenth and ninteenth century
public.

There was also a more “practical” side to this concern. As the shadowy civilizations of the
East took on form and history in the nineteenth century, they came to threaten Eurocentric
world views at just the time that Europeans most needed self-justification for applying differ-
ent moral standards abroad than they did at home. One way out of this dilemma was in theo-
ries that posited that all advances in civilization occurred as a result of the imposition of an
inherent maverick creativity of the roving Indo-Europeans on the stability of servile peasant
cultures in the Middle East and around the Mediterranean (Trigger 1989: 158–69; Bernal
1987: 31–33). Thus the discovery that the Hittites spoke an Indo-European language created
great excitement in the beginning of this century, as did speculations as to the Indo-European
origin of the Sumerians, and the supposed Indo-European (if not actually Aryan) aristocracy
ruling the Canaanites. D. Hogarth, secretary of the Palestine Exploration Fund, upon commis-
sioning an expedition to study the origins of the Philistines as late as 1920, set the terms of
this commission to find out “whether it was in (



sic



) virtue of a distinctly higher apparatus of
civilization that the Philistines so long terrorized the Hebrews” (



Illustrated London News



,
quoted in Dothan and Dothan 1992: 43). Upon conclusion of that commission, W. J. Phythian-
Adams hastened to assure his senders that “the finer arts of the Late Bronze Age (LB) were
imposed on the Canaanites from outside” (Dothan and Dothan 1992: 46).

1830–1900: EGYPTIAN ART AND PALEOGRAPHY

The beginning of modern research of Egypt is usually attributed to the scientific expe-
dition which accompanied Napoleon on his invasion of Egypt. Typically, although the evident
relics of high civilization in Egypt far eclipsed anything that was known of the Bronze Age or
Early Iron Age in Greece or Israel, it was possible connections of Egyptian remains to either
the Bible or Greek mythology that created the greatest excitement.

The first description of the great land and sea battles between the Egyptians and for-
eigners wearing feathered or horned head-dresses, depicted on the walls of the great temple

oi.uchicago.edu





559



PHILISTINE BICHROME PAINTED POTTERY



at Medinet Habu dates to this campaign. It was not, however, until Champollion’s expedition
to Egypt (1829), following his decipherment of the Hieroglyphic script, that the structure
was correctly identified as the burial-temple of Ramesses III, the last of the great rulers of the
New Kingdom.

Before his death Champollion (1836: 180) managed to read one of the names of
Ramesses’ foes and identified them as Philistines (literally:



prstw



). It was not until 1856 that
the entire inscription accompanying the Medinet Habu reliefs was transcribed and deciphered
by Emanuel de Rougé (see Rougé 1867 and Dothan and Dothan 1992: 22–23). It reads:



Year 8 under the majesty of [Ramesses III] . . . The foreign countries made a conspiracy
in their islands . . . their confederation was



prstw



,



t



·



krw



,



skrsw



,



dnynw



and



w



·



ssw



lands
united. They lay their hands upon the lands as far as the circuit of the earth, their hearts
were confident and trusting our plan will succeed.” (trans. after Wilson 1950: 262)



Once the names of these groups of Sea Peoples (as they were later dubbed by Maspero) had
been read, it was realized that some of them appear in Egyptian documents as early as the
Amarna period, and Merneptah, Ramesses III’s predecessor, had already fought off one
attempted invasion by them (Phythian-Adams 1923: 21–22). Additional foreigners “of the
sea” are mentioned in these earlier sources (



rkw



,



drdnw



, etc.)
Additional historical evidence may be found in the Harris Papyrus (Breasted 1906, IV,

§403), which adds that after their defeat the Sea Peoples were settled by Ramesses III in
“strongholds” and allotted rations (i.e., were used as mercenaries). Evidence for Sea Peoples
in Egyptian documents postdating Ramesses III is scant. Some



t



·



krw



are mentioned as resid-
ing at Dor, in the northern coastal plain—in the so called “Story of Wen Amon” (Papyrus
Moscow 120)—a literary work of arguable historical accuracy. Even more problematic is the
“Onomasticon of Amenope” (the Golenischeff Papyrus), a cryptic list of geographical and
other terms, that has been taken by some to read that the



t



·



krw



dwelled somewhere north
of the



prstw



, and the



srdnw



still further north. Both of these sources are conventionally dated
ca



.



1100



b.c.



Rougé, like most scholars before him, immediately tried to match the names he read
with names in the



Iliad



, identifying the



drdnw



as Homer’s Dardanians, the



rkw



as Lycians,
and so forth. A curious byline is that despite the great fondness of biblical writers for col-
lecting ethnographic lists, none of the names of the Sea Peoples, except the Philistines, are
found in the Bible (unless one chooses to accept Yadin’s [1968] scintillating suggestion that
the



dnyn



are actually the tribe of Dan).
Rougé could not fail to note that Ramesses’ great battle against the Sea Peoples (1191



B



.



C



.,
according to Rowton’s [1976] chronology, 1175



B



.



C



. according to Wente and van Scilen
[1976]) took place just a few years after the traditional dating for the fall of Troy (1184



B



.



C



.,
according to Thucidides). It is also just a few decades later than the conventional dating of the
entry of the Israelites into Canaan. Rougé proposed in 1867 that the fall of Troy triggered a
whole set of secondary ripples, where one people displaced another, causing them to push
over a third group and so on, and that the end result was the Philistine’s conquest of south-
western Canaan and their abortive attempts to invade Egypt (Dothan and Dothan 1992: 22–
26). This is probably the earliest formulation of the “wave theory,” which will follow the study
of the Sea Peoples henceforward.

oi.uchicago.edu





560



ILAN SHARON



This theory established a seemingly satisfying triple synchronism between the Bible,
Egyptian literature, and Greek mythology. “Seemingly” should be stressed here, because out of
these three dates one, at least, is completely unfounded, and another doubtful at best. It failed
to account for the mentions of Sea Peoples in Egypt before Ramesses III, as well as references
to Philistines as early as the book of Genesis in the biblical narrative. This did not prevent it
from being widely accepted and quoted. Thus, as of the mid-eighteenth century, the equiva-
lence of the Philistines of the Bible, the Sea Peoples of Egyptian literature, and Aegeans of
the Homeric age was deemed secure. The only issue which remained in debate was which
Aegeans exactly took part in the great eastward migration.

EARLY DATA COLLECTION 1899–1930: ESTABLISHING A CERAMIC TYPOLOGY



The Identification of Philistine Pottery



As the nineteenth century drew to a close and scientific archaeology got under way, more
was being learned about material culture both in Israel and the Aegean. The pottery chronol-
ogy of the Aegean Bronze Age, from expeditions such as Schlieman’s in Mycenae and in
Troy, and Evans’s in Knossos was summarized by Furtwängler and Loeschcke (1886), while
a preliminary chronology based on stratigraphic excavation in Palestine was being worked
out by Petrie and his students. Tangible evidence of the much sought-after struggle between
David and Goliath (as an allegory to the early contact between the Bible and the classics) was
not yet forthcoming, however. Nevertheless, possibility of locating such was sufficiently
exciting to stimulate the Palestine Exploration Fund (PEF) to commission several expedi-
tions to excavate the



tells



of the southern Shephelah in the beginning of the century, with the
specific purpose of locating such contacts (Dothan and Dothan 1992: 29–32). To aid in this
enterprise the PEF contracted the foremost authorities in Aegean archaeology, such as Dun-
can Mackenzie, Evans’s chief assistant on the Knossos excavations.

Early attempts by these expeditions to isolate Philistine material culture attributes were
unsuccessful. Several tombs at Gezer, claimed by Macalister (1912: 297) to be Philistine
burials, were misdated, as we know today, by some 800 years (Stern 1982: 73). Bliss and
Macalister had consulted F. B. Welch, one of the foremost authorities of the day in Aegean
pottery, about several Mycenaean potsherds found in their excavations. Welch verified the
identification of these sherds but agreed with the excavators that their number was not suffi-
cient to warrant the assumption of local manufacture. As it turned out, both Welch and the
excavators overlooked some evidence from this very same expedition, which we commonly
identify today as “Philistine” pottery.

In 1908 Hermann Thiersch proposed that a family of decorated pottery, first found at Tell
es-Safi in 1899 by Bliss and Macalister (1902: 89–97) was, in fact, the pottery manufactured
by the Philistines. Bliss and Macalister had actually noted these distinctive red-and-black
painted potsherds, but they dated them to their “Late Pre-Israelite” period, i.e., earlier than the
supposed Philistine migration. Welch had even noted their similarity to Aegean forms and
decoration, ascribing them to a debased Mycenaean tradition (1900: 347–48). Thiersch argued
that this painted pottery was similar to Furtwängler and Loeschcke’s “fourth style”—the final
phase of Mycenaean pottery (as then known)—while the Mycenaean sherds discussed by

oi.uchicago.edu





561



PHILISTINE BICHROME PAINTED POTTERY



Bliss, Macalister, and Welch were of the earlier “third style.” Therefore, it is this type of
pottery, rather than earlier ones, which heralded the arrival of the Sea Peoples in the Middle
East, on the wake of the Trojan war, the Dorian invasions, or some such catastrophe (Dothan
and Dothan 1992: 31–32). This attribution was quickly and almost universally accepted. The
common name for this ware is “Philistine Pottery” (for a detailed definition see below), or,
more recently, Bichrome Philistine pottery (in order to distinguish it from its monochrome
predecessor—Mycenean (Myc) IIIC/“Wheel Made White Painted”—Furtwängler and Loe-
schcke’s “fourth style”). In order to avoid culturally laden terminology I designate this type of
pottery “Early Iron Age wheel-made Bichrome Painted by hand” pottery (or BPh for short).
This designation may set it apart both from the afore-mentioned monochrome style, and from
the other contemporary bichrome ware, the so-called “Phoenician” bichrome, in which the
decoration was instrumentally applied. By the beginning of the century, then, the identifica-
tion of this pottery with the biblical Philistines, and of the latter with the Sea Peoples of Egyp-
tian literature seemed secure.



Philistines and Achaeans



The next attempt to investigate the Philistines was Duncan Mackenzie’s expedition to
Beth-Shemesh and Ashkelon in 1911–1912 (Mackenzie 1912). He noted that in both of these
sites the appearance of BPh pottery postdates the Mycenaean pottery. Mackenzie compared
the division of the decorative field on some of the BPh pots into metopes with “panel style”
pottery which he had seen in Phylakopi and which he believed to be post-Mycenaean. He
thought the “panel style” was brought into Greece by “Achaean” invaders and identified the
Philistines with these same “Achaeans.” This attribution typically ignores the fact that division
into metopes, while it does appear as a somewhat esoteric element in the Mycenaean decora-
tive tradition, was rather commonplace in the local Canaanite repertoire. Phythian-Adams
(1923) argued that the origins of the “panel style” and hence of the “Achaeans” should be
sought in the Balkans, since it was supposed that they had come into Greece from the north
and into Asia Minor and Palestine from the west, a theory whose main support (apart from
some more name etymologies) was the almost total ignorance of LB cultures in the Balkans at
the time.

The debate about the origins of the Philistines was thus early on connected to another
major controversy of the times—the debate about the origins of the Greeks themselves, or
more specifically, the argument about the Greekness of the Mycenaeans. Blegen and Haley’s
(1928) suggestion that the Mycenaeans were Greek was generally ignored. Arthur Evans saw
the Mycenaean culture as a provincial offshoot of the Minoan civilization, which itself was
held to have oriental origins. Thus the prevailing view was that the first Greeks (whether as
Achaeans or Dorians or whatever) were northern invaders from the Balkans who had brutally
put an end to Mycenaean civilization. Ventris’s decipherment of Linear B, in the early 1950s,
as a Greek dialect, came as a rude shock to many.

The “Achaean origins” theory for the Philistines was severely challenged when Heurtley
and Wace (Heurtley 1936) found out in their excavations at Mycenae that the “panel style”
was not post-Mycenaean, nor was it necessarily northern in origin. As a matter of fact, the
very latest type of pottery still ascribed by Heurtley to the Mycenaeans (the so-called granary
style) took on an aspect not reflected in the BPh pottery. Ergo, the Philistines had left the

oi.uchicago.edu





562



ILAN SHARON



Greek mainland



before



the final demise of the Mycenaeans, and the Philistines were



part



of
the Mycenaean littoral, rather then its exterminators. Moreover, Heurtley pointed out that the
BPh pottery includes attributes from all of the known regional substyles of Mycenaean pottery
and yet bears a close similarity to none. Hence “it is unlikely that it is the pottery of the
(unknown) homeland from which the Philistines came . . . The most we can say is that it may
have been made to satisfy a demand by the newcomers for something which had a Mycenaean
look, and infer from its composite character that they were familiar with the whole Aegean”
(Heurtley, 1936: 109).



Other Excavations



Three more excavations in the beginning of the century, which were to establish parame-
ters for future discussion, are now discussed:

One is W. M. F. Petrie’s excavations at Tell Jemmeh and at Tell el-Farºah (S) (Petrie 1928;
Petrie and Tufnell 1930; Macdonald, Starkey, and Harding 1932). Petrie accepted the attribu-
tion of the BPh pottery to the Philistines, but not Mackenzie’s (1912) and Phythian-Adams’s
(1923) dating of it to the Iron Age, which did not concur with the biblical references to Philis-
tines prior to the Israelite conquest of Canaan, or with references to Sea Peoples in the Eigh-
teenth-Dynasty Egyptian texts. He conveniently found Egyptian scarabs at the same layers to
date the BPh pottery he found at these sites as contemporary with the Eighteenth Dynasty, that
is, to the height of the Late Bronze Age. This accorded well with his view that Tell Jemmeh
was Abraham’s Gerar, where dwelt “Abimelech, the king of the Philistines” (Genesis 26:1).
Petrie saw the invasion of 1186



B



.



C



. as a



second wave



of Aegean migration (Dothan and
Dothan 1992: 64).

The most significant find at Tell el-Farºah were five large rock-cut tombs, four of which
had Philistine pottery in them, as well as anthropoid coffins of distinctive style. Petrie (Petrie
and Tufnell 1930: 7) whimsically named these caves “the tombs of the five lords of the Phil-
istines.” Similar coffins had already been uncovered at Beit Shean (which has its own Phil-
istine connotations: the Philistines hung Saul’s body from its walls, after their victory at the
battle of Gilboa). No Philistine pottery was found with the latter coffins, however (and
hardly anywhere else at Beit Shean, for that matter), though one Myc IIIC cup was discov-
ered (Hankey 1966, and see below for the significance of that find).

Petrie attempted a relative seriation of the Tell Farºah tombs, mainly by the types of scar-
abs found in them. He placed Tomb 542, the richest in finds, as the earliest (dating it to 1320



B



.



C



.) and Tomb 532, in which there is a scarab bearing the name of Ramesses XI (1113–1085



B



.



C



.) late in the series. These dates, while perfectly in synch with Petrie’s own beliefs of the
antiquity of the Philistines, presented a problem to those who accepted the BPh as an Iron
Age phenomenon. Thus, there was a series of attempts at redating these tombs (Starkey in
Macdonald, Starkey, and Harding 1932: 31; Albright 1932; Furumark 1941b: 121; T. Dothan
1982: 29–33; McClellan 1979). Albright (1932: 299–301) accepted the sequence proposed
by Petrie but modified his absolute dates. He pointed out that nowhere did Petrie find Philis-
tine pottery associated with Mycenaean or Cypriot imports, which rules out a Late Bronze
Age date. Scarabs of the great New Kingdom rulers, which Petrie set such store by, often
continued to be manufactured decades (and sometimes centuries) after their namesake’s
death. Although Albright had no absolute dates of his own to offer, he approximated the date

oi.uchicago.edu





563



PHILISTINE BICHROME PAINTED POTTERY



of tomb 542 (the earliest, according to both Petrie and himself) to the beginning of the
twelfth century



B



.



C



., thus reinforcing the seeming connection between the appearance of the
BPh pottery and the campaigns of Ramesses III.

The second significant discovery was the finding of BPh pottery at the Chicago Oriental
Institute’s excavations at Megiddo, arguably the most influential single excavation ever held
in this country. Much of the absolute dating of the Late Bronze Age–Early Iron Age transition
rests on evidence from Megiddo. The published BPh sherds from Megiddo are few in number
but they will crop up in any discussion of the BPh pottery henceforward. The excavators
reported BPh sherds from Strata VIA, VIB, VIIA, VIIB and even one example from Stratum
VIII (T. Dothan 1982: 70–80). Stratum VIII at Megiddo is most definitely LB II. Stratum VII
is problematic: architecturally it is a continuation of the LB city of Stratum VIII; typologi-
cally, it contains LB artifacts (including a hoard of LB ivories in Stratum VIIA, as well as
Mycenaean and Cypriot imports), but also some types characteristic of the Iron Age I, namely
the above mentioned BPh sherds. Stratum VIIA is securely dated by one of the ivories men-
tioned above, which bears a cartouche of Ramesses III (Loud 1939: 9–10). Stratum VIB is the
earliest town that is definitely Iron Age in material culture. The clear destruction at the end of
Stratum VIA is usually attributed to David, although there is no direct evidence (either archae-
ological or historical) for that (for a recent polemic



against



this date see Finkelstein 1996).
Thus, the Megiddo finds as published contradicted the emerging consensus, as BPh

apparently preceded Ramesses III, and appeared in strata that also had Mycenaean and Cyp-
riot, as well as local, Bronze Age artifacts. The Megiddo report, however, is notorious for
misassigning artifacts (and sometimes complete structures), and the stratigraphy of Megiddo
has been reworked many times since its initial publication. Thus T. Dothan (1982: 70–80)
attempts to bring the Megiddo finds into line with the accepted view in a detailed critique, in
which she reassigns all of the BPh sherds found in Strata VIII and VIIB into VIIA, and some
of the obvious LB types from Strata VIIA into VIIB, thereby arguing that the transition
between LB and Iron I at Megiddo is between Strata VIIB and VIIA. If this rearrangement is
accepted, it may be argued that at Megiddo, too, there is a correlation between Ramesses III
and the appearance of BPh pottery.

The third excavation to discuss is William Foxwell Albright’s (1932–43) project at
Tell Beit Mirsim during 1926–1932. Only a few BPh sherds were found in this inland site
(T. Dothan 1982: 43–44). Its importance is due to the fact that it was the first project which
attempted to implement the new standard chronological scheme for Palestine, agreed on in
1922, as well as Albright’s innovations in excavation methods and pottery typology. It was
used thereafter as a type site for many issues relating to Bronze and Iron Age chronology.

Stratum B at Tell Beit Mirsim covers the first half of the Iron Age. BPh pottery appears
only in Phase B2. Phase B1 is devoid of any Mycenaean or Cypriot imports (this is the reason
Albright judged it to be post–LB) and does not have any BPh pottery either. BPh pottery was
the



fossil directeur



for Phase B2. Phase B3 was dated to the tenth century and was character-
ized by hand-burnished red-slipped ware.

This fitted well with Albright’s historic interpretation: Albright thought that the Late
Bronze Age Canaanite civilization succumbed under two waves of invasion. The first was the
Israelite conquest, which he saw as a concerted attack by militant desert nomads in the second
half of the thirteenth century, capturing the hill country and the inland valleys. The Israelites

oi.uchicago.edu





564



ILAN SHARON



were already well entrenched, in his opinion, by the time of Merneptah’s 1232 (1207) cam-
paign. This is the culture evident in Tell Beit Mirsim Stratum B1 according to Albright. The
second wave is the devastation of the coastal plain by the Sea Peoples, just prior to their
attempted invasion of Egypt in 1191 (1175). While the main thrust of this invasion was the
replacement of Canaanites (and their Egyptian overlords) by Philistines, they did extend their
dominion over the lower foothills of Judaea, taking over sites such as Tell Beit Mirsim from
the Israelites. This cultural change is evident, according to Albright, in Stratum B2. Stratum
B3 exemplifies the final Judean takeover, when David squashed the Philistine hegemony.

IDEOLOGY

Archaeological thinking at the end of the nineteenth and beginning of the twentieth cen-
tury was dominated by the



Kulturgeschichte



(culture history) approach, several versions of
which were promulgated in Scandinavia by Oscar Montelius, in Germany by Friedrich Ratzel
and Gustaf Kossina, in England by E. B. Tylor, G. Elliot Smith, and the early works of V. Gor-
don Childe, and in America by Franz Boas (Trigger 1989: 148–206).

The principal notions common to all of these strands were the identity of physiological
and behavioral properties, and hence of biological and social genetics; an analogy of whole
societies to individual organisms; and the notion that cultures and peoples diffuse or flow from
point of “high culture” to “low culture.” Cultural or cognitive traits, just like physiognomy,
were seen as being transmitted by heredity. Specific ranges of vessel types, just like gene
pools, were perceived as the property of



racial lineage groups



, the hierarchical components of
which were “family,” “ethnos,” and “race.” Moreover, the entire ethnos was regarded as a sin-
gle unitary organism.

Now, a rose is still a rose whether its limits be defined by shape, color, texture or aroma.
If the analogy between a rose and a culture holds, it follows that a single distinctive trait is
enough to define the whole. This is the gist of the



fossil directeur



approach to archaeology.
The idea of cultural unitarianism could (and was) carried even further. It was held that a

culture, just like an organism, grows, flowers, and dies. Cultural change was conceived in
terms of “childhood,” “adolescence,” “maturity,” and “decadence,” as in the conventional divi-
sion of artistic styles to “Archaic,” “Classic,” and “Baroque.” By the same reasoning as above,
such a process would be visible with every single facet of that culture. Thus, if a decorative
style is debased (in the eye of the beholder), it would mean that the culture of which it is a
manifestation was decaying, and vice-versa. H. R. Hall (1928), who saw the Myc IIIC pottery
(Furtwängler and Loeschcke’s “fourth style”) as the product of a declining culture, character-
ized its style as “degenerate, vulgar and tasteless.” As the BPh pottery was supposedly deriva-
tive of this style, it could not be anything but decadent.

If culture is inherited like the color of one’s eyes, then the only possible explanation for
similar cultural attributes appearing in two different places is physical transhumance of a lin-
eage group from one place to another. Cultural change was, if radical enough, perforce the
result of total genocide of the incumbent peoples by migrating invaders. If gradual, it was seen
as the result of intermarriage between newcomers and the indigenous lineage groups. The
main question to be answered, once cultural change was demonstrated in the material record,
was where these newcomers came from. The methodology by which this question was ad-

oi.uchicago.edu





565



PHILISTINE BICHROME PAINTED POTTERY



dressed was to look for some place where similar traits appear. If some sort of a connection,
however tenuous, between these two cultures could be established, the conquest of one by
another was as good as proven. If not, (as, say, in the case of pyramids appearing in Egypt and
in Mexico), it might just mean one had not looked hard enough.

Another logical corollary of the unity of culture and of its immutability is the concept of



culture centers



and cultural periphery (Trigger 1989: 160; Harris 1968: 374). While the typi-
cal explanation of cultural change was an influx of a new population from elsewhere, one had
to concede that innovation had to exist



somewhere



, or else all cultures would have been iden-
tical. At least one “authority” (von Däniken; see Epstein 1987) followed hyper-diffusionist
theory to its logical end and posited that all significant innovation in human history came
from outer space. Others have had to accept that evolutionary change



could



occur, but only
in special places and under extraordinary circumstances. Thus the “children of the sun” theory
(Perry 1923; Smith 1923, 1933) proposed that all cultural traits were diffused from Egypt,
while Graebner and Schmidt argued that a hypothetical core in central Asia was periodically
spewing out new cultural stimuli (Harris 1968: 382–92). These formulations were arguably
extreme even for their time, but most scholars shared the conceptual view of culture spreading
from cultural centers like ripples in a pond. Viewed from the periphery, cultural stimuli would
appear in



waves



.
Gustaf Kossina (1911), the most influential figure in central European archaeology before

World War II, combined the beliefs that cultures were a reflection of ethnicity with the racial
division of peoples into



Kulturvolker



, who alone were capable of creativity, and Naturvolker,
or culturally passive people. He promoted a view in which successive waves of Aryan inva-
ders from the north built the various civilizations of the ancient world. Interbreeding with the
indigenous populations diluted their racial purity, however, thus dooming each of these civ-
ilizations in turn to decay. Quite naturally, Kossina’s teachings became (posthumously) the
official version of history in Nazi Germany (Bahn 1996: 136–38).

Kossina’s emphasis on (blond) racial purity as the engine for cultural change could not
find much sympathy in France or England. John Myers (1911) and Arthur Evans (who was
indirectly involved with the “Philistine question”) as well as the early writings of V. Gordon
Childe (1925, 1926, 1929) took the opposite stand: it was the interbreeding of different racial
stocks (such as would have occurred following an encounter between Indo-European inva-
ders and Semitic populations) which produced the rare episodes of cultural innovation. Note,
however, that while reaching opposite conclusions, both the German and the British schools
accepted the same basic premises outlined above.

It would be easy (and wrong) to identify the excesses of Kulturgeschichte doctrine as a
shameful bunion sprouting in some concealed niche of “normal” science. Curiously enough,
this scenario offered something to everyone: It was embraced by white supremacists and Jews,
classicists and semiticists, scientists and clerics. Even the broadest minds of the time could be
caught in an unguarded remark, such as:

It was fortunate for the future of monotheism that the Israelites of the Conquest were a
wild folk, endowed with a ruthless will to exist, since the resulting decimation of the
Canaanites prevented the complete fusion of the two kindred folk which would almost
inevitably have depressed Yahwistic standards to a point where recovery was impossible.
(Albright 1940: 214)

oi.uchicago.edu



566 ILAN SHARON

The insistence on the conservative nature of culture needs some further explanation. In the
first place, it answered some deep-seated doubts many Europeans had when they confronted
the “Orient”: if Western European culture was as self-evidently superior to any other as they
wished to believe, how is it that it was not immediately and universally accepted? For Kossina
the answer was simple: “Naturvolk” did not change their culture because they were biologi-
cally incapable of doing so. Less crudely racist theories had to seek more complex answers.

The assumption that culture is essentially hereditary explains well the homogeneities
encountered in human behavior. It runs into trouble trying to explain variation and change.
Given an (even limited) capacity of humans to change their own behavior, and the examples
of manifestly more successful modes—would not all cultures evolve to the point where they
are indistinguishable? Once this point had been reached, would further development be at all
possible?

On the other side of the organistic analogy (in biological genetics), this dilemma was
called the paradox of regression to the norm, and it nearly proved the downfall of Darwinian
theory in the nineteenth century, before being brilliantly solved by Mendelian genetics. One of
the assumptions of the Darwinian model is that natural selection operates against a back-
ground of virtually unlimited variation in nature (i.e. that every variation that can happen, will
happen, sooner or later). Yet common experience dictated that in uncontrolled breeding the
offsprings of two different purebreds was a mongrel—that is, some sort of an average between
the two parents. This meant that if, say, by chance a long-necked mutant is born in a herd of
short-necked giraffes, her offspring (by short-necked partners) ought to be semi-long necked.
The next generation would be at best semi-long necked and for the most part quarter-lengthed,
and so on. It was not seen how, except in mutations that offer immediate life-or-death advan-
tage or disadvantage, natural selection could compete against this strong normative power.

In “social evolution,” and in archaeology too, many thought that this “regression to the
norm” was a power to be reckoned with. I already mentioned that Kossina (and others) argued
that dilution of bloods invariably led to regression to mediocrity and hence to eventual degen-
eration and to the downfall of cultures. The future of human civilization depends, according to
Kossina, upon the existence of a pool of untainted Nordic genes.

Despite the specter of regression, observation tells us that in society (and biology) diver-
sity seems to be the rule. An assumption that extreme conservatism is in the nature of any cul-
ture may partly explain why two cultures, upon contact, do not fuse into a “mongrel” culture
within a short period of time.

Another way of solving the paradox of “regression to the norm” is the assumption of
Volksgeist. A tenet that each culture or race has an essential spirit, a core of primal attributes
and mental qualities which remains unadulterated as long as that culture survives, no matter
how the outer manifestations of that culture may change with the passage of times and for-
tunes. Such an assumption puts a lower bound on miscegenation and ensures that diversity of
cultures will never decrease (however each of these cultures may change) except in cases of
actual genocide.

One archaeological manifestation of this theory is the fossil directeur, a single type of
artifact whose appearance (and it alone) is considered diagnostic of the essence of that cul-
ture, however divergent the rest of the assemblage in which the fossil appears to be. Locating

short

oi.uchicago.edu



567PHILISTINE BICHROME PAINTED POTTERY

the temporal and geographical limits of this single artifact type would define the homeland
and history of “a people” and hence the extent of a specific language, mythology, cuisine, and
any other cultural trait. Thus, in the case of the Philistines, Phythian-Adams expressed a typi-
cal sentiment in the statement that:

It must be stated at the outset that the data . . . are concerned almost exclusively with
a single type of vase, the bowl fitted with a horizontal loop handle . . . [which is] the
Philistine vase par excellence . . . We must hold, then, that as the Achaeans and their
allies brought this bowl with them [to Greece] before the close of the Bronze Age, so the
Philistines and their companions brought it [to Palestine] in the transitional period that
followed. In other words, we must assume that the vase in question was one designed
. . . by the inhabitants of the regions from which the sea rovers and their kinsmen of the
Troad originally came. But we have seen in the case of the latter that the evidence points
very strongly to the Balkans as being the homeland in question, and since the whole
group seems to have been bound together by ties of a common religion, it is at least
probable that the former will also be found there.” (Phythian-Adams 1923: 23–24, italics
mine)

This, then, was the background against which the view that the BPh pottery reflected
waves of Indo-European invaders was developed. Before going back to data collection, I
should like to stress again the decisive role that theory had in formulating this view. Contem-
porary archaeologists are sometimes vain enough to suppose that archaeological theory is an
invention of the 1960s. We are perhaps helped in this misconception by the fact that practi-
tioners of former generations tended to disguise their theoretical convictions. This was done
because under inductive empiricism, the prevailing philosophy of the day, one was supposed
to “face the facts” indiscriminately, and admitting bias, theoretical or otherwise, was as un-
acceptable as “lack of explicit theory” is today. That archaeology at the turn of the century
consisted of data collection for its own sake and of “low level generalizations” based on
these facts is as much a myth as the notion that the way we proceed today is by objective test-
ing of hypotheses deducible from putative “laws.” The Philistines are a case in point.

The equation of BPh pottery with the Philistine people, and their association with the
Sea Peoples of Ramesses III, was accepted not because it fitted the empirical data available
at the time, but in face of well-known evidence to the contrary, as gathered in what were
generally accepted to be the best field studies of the day. At the time, the only indications
connecting the appearance of the BPh pottery with the activities of Ramesses III (rather than
with earlier events) were Phythian-Adams’s unpublished investigations at Ashkelon and Mac-
kenzie’s observations at Beth Shemesh (an excavation not highly regarded for its method
even by the standards of the day). Petrie’s excavations, and most certainly the Chicago Ori-
ental Institute’s excavations at Megiddo, were generally given higher marks for reliability,
and yet they uncovered information which, on the face of it, contradicted the Philistines =
BPh = prst equation. Nevertheless, this hypothesis was hardly challenged, and elaborate
schemes were devised to explain away apparent inconsistencies between it and the data. I
contend that this was because it fitted deep-rooted theoretical convictions of most scholars
of the day. Even the one authority who promoted a different view, W. M. F. Petrie, let his
own preconceptions guide (or sway, as the case may be) his empirical findings. (For a rather
similar assessment of methodological “progress” in the “hard” sciences, see Feyerabend 1978).

oi.uchicago.edu



568 ILAN SHARON

DATA COLLECTION AND ANALYSIS 1930s–1970s:

CONSOLIDATION OF A PARADIGM

The next period of research is characterized by several new excavations which revealed
new and (for the first time . . .) reliable data about the BPh pottery, as well as several extensive
syntheses, most of which served to further and flesh out the equation presented above. These
were the “golden years” for large scale excavations in Israel. Of the many excavations held,
quite a few produced BPh pottery, or shed light on other facets of the Philistine problem (Tel
Mor, Lachish, Beth Zur, Gezer, Azor, Aphek, Tel Zeror, Afula—to name but a few). The most
pertinent to our discussion were B. and A. Mazar’s excavations at Tell Qasile, M. Dothan’s
excavation at Ashdod, and T. Dothan’s excavations at Deir el-Balah.

Excavation: Tell Qasile

Benjamin Mazar excavated at Tell Qasile, a small mound on the north bank of the Yarkon
River, from 1949 to 1951, and again in 1956, 1962 and 1963. This excavation is unpublished,
except for a preliminary report of the first two seasons (B. Mazar 1951), and a short sum-
mary in the Encyclopedia of Archaeological Excavations (T. Dothan and Dunayevsky 1978).
B. Mazar distinguished twelve occupations at the site, from the early Iron Age to the Arab
period. The first three (Strata X–XII) were characterized by BPh pottery. Strata XII to X rep-
resent a continued period of growth of the community, from transient settlement on bedrock
(Stratum XIIB) to a full-fledged town (Stratum XI). This continuity ended abruptly in Stra-
tum X, which was destroyed by heavy fire. The following settlement (Strata IX–VIII) no
longer had BPh pottery in primary contexts. B. Mazar interpreted the destruction of Stratum X
as the sacking of the Philistine town by David, and the following strata as an Israelite occu-
pation (B. Mazar 1951: 67–68; A. Mazar 1980: 9–12).

Following a salvage excavation in 1971, Amihai Mazar continued his uncle’s project and
excavated what proved to be a sanctuary from 1971 to 1974. Three successive sacred struc-
tures were found, corresponding to B. Mazar’s Strata XII, XI, and X. The sanctuary was
rebuilt after a fashion in Stratum IX (which casts some doubt on B. Mazar’s hypothesis of
ethnic and religious upset following the destruction of Stratum X—A. Mazar 1980: 11).
A. Mazar promptly published this excavation (first as a Ph.D. dissertation and then as a final
report: 1980, 1985a). To date it remains the only properly excavated quantitative sample of a
BPh assemblage of sufficient size in the literature. BPh pottery typically comprises 10–25%
of the total ceramic assemblage in the relevant strata at this site (A. Mazar 1985b: 105).

Mazar went to considerable trouble to search for parallels to the nature of the cult at Tell
Qasile. Architecturally, he characterizes the Tell Qasile sanctuaries as belonging to a group of
“irregular” sanctuaries. Such structures occur in LB in the Levant (though they present a
deviation from the regular symmetrical monumental temple tradition of MB–LB). Parallels
may be found in the Aegean in LB, and (in my opinion, less convincingly) in Cyprus in the
thirteenth–eleventh century B.C. (A. Mazar 1980: 62–68). The cultic vessels mostly represent
local types (although some display unique stylistic elements). The single vessel which may
point to a foreign type of cult practices is a ceramic lion’s head rhyton, which may have an-
tecedents in metal rhytons in the Aegean (A. Mazar 1980: 119–21). Several types of BPh
Mycenaean-style figurines which occur elsewhere are notably missing at Tell Qasile.

oi.uchicago.edu



569PHILISTINE BICHROME PAINTED POTTERY

Tell Qasile is material to the discussion of the Philistines in several respects. First, it put
an end to the unilateral view of the Philistines as destroyers of cultures, because Tell Qasile
is a Philistine foundation. This is the inception of quite an opposite view of Philistines as pre-
servers of urban culture in troubled times. Secondly, as there was no Bronze Age occupation
at Tell Qasile at all, it displayed the clear association of BPh pottery with early Iron Age con-
texts. Also, this was the first time that several successive BPh-bearing strata were recognized
(other than at Megiddo, where at least some of the BPh finds were obviously out of context).
This enabled a chrono-typological subdivision of the Philistine culture, other than its very
earliest phases, which supposedly are missing, since the site was taken to be the product of an
expansion from the Philistine heartland northwards. Finally, of course, it afforded a glimpse
into Philistine cult.

Excavation: Ashdod

Ashdod was the first of the cities of the Philistine pentapolis to be extensively excavated
(by Moshe Dothan), other than Garstang and Phythian-Adams’s deep trench at Ashkelon.
Unfortunately the publication (M. Dothan and Freedman 1967; M. Dothan 1971; M. Dothan
and Porat 1982) is much sketchier than that of Tell Qasile. The findings have been summa-
rized as follows: The last Bronze Age town is Stratum XIV, characterized by Late Cypriote
and Myc IIIB imports. It was destroyed by fire, which the excavator attributes to an early
wave of Sea Peoples during the reign of Merneptah (M. Dothan 1971: 20; Dothan and
Dothan 1992: 165). The next stratum, XIIIB, contains no Cypriote import or BPh pottery, but
does have significant amounts of Myc IIIC, while Stratum XIIIA has both Myc IIIC and BPh.
Strata XII and XI contain “typical Philistine” (i.e., BPh) pottery in them and are thus roughly
equivalent to Tell Qasile Strata XII–X. Stratum X at Ashdod marks the introduction of new
“Judean” elements and the disintegration of the indigenous BPh repertoire and is thus dated
by the excavator to the final days of Philistine hegemony. Its destruction is dated by the exca-
vators either to David’s conquests, or (more likely) to Siamun’s campaign against Philistia,
ca. 960 B.C. (M. Dothan 1971: 21). The latter possibility would place the destruction of Ash-
dod Stratum X somewhat later than that of Qasile Stratum X, which seems to better fit the
evidence.

The prime importance of Ashdod is in the finding of large amounts of Myc IIIC for the
first time (though some pieces were previously recognized at Beit Shean, as mentioned
above) and demonstrating its connection with BPh. The finds at Ashdod supplement Tell
Qasile in adding the early part of the sequence of the development of the BPh pottery and
reinforce the latter part of the sequence as seen there. The stratigraphy is also rather close to
the one observed by Albright at Tell Beit Mirsim, although Ashdod, being a coastal site, had
Myc IIIC pottery in the “gap” between the end of the Late Bronze Age and the appearance of
the BPh, whereas inland Tell Beit Mirsim had “Israelite” pottery in it. This is perhaps why
Moshe Dothan dated his strata similarly. If Stratum XIV ended ca. 1230, then the Philistine
invasion culminating in the raids on Egypt and subsequent repulsion and resettlement after
Ramesses III’s 8th year must have occurred sometime during Stratum XIII. It therefore
seemed natural to equate these events with the appearance of BPh pottery in Stratum XIIIA.
This leaves the Myc IIIC of Stratum XIIIB as evidence of the earlier wave of Sea Peoples
(the ones Merneptah fought off). It has been remarked (e.g. by Sandars 1978: 171–72) that

oi.uchicago.edu



570 ILAN SHARON

there is no compelling reason to attribute the destruction of Ashdod Stratum XIV to Merne-
ptah. Attributing it to Ramesses III, for instance, would put a rather different light on things.

Synthesis: Furumark and the End of Mycenaean Civilization

Great strides were being made, too, in refining archaeological knowledge at the putative
homelands of the Sea Peoples. In particular, the typology and chronology of Mycenaean pot-
tery was much refined, as summarized by Arne Furumark (1941a, 1941b). Furumark intro-
duced the high standard of art historical analysis to the study of Aegean pottery, breaking
down the “style” to a typological facet (vessel form), an execution facet (“rude,” “simple,”
“elaborate,” etc.), a semantic facet (motifs: pictorial and geometric), and a syntactic facet
(composition); and then tracing its development both on the temporal axis (chronology) and
spatial ones (regional styles).

Furumark (1941b: 17) argued that Furtwängler and Loeschcke’s (1886) “four style”
system had no chronological value, except for the “fourth style” (about which see below).
He also opposed Evans’s suggestion to simply extend the Minoan designations to the main-
land. Instead, he used Blegen’s system of dividing the LB pottery styles to “Late Helladic” I,
II, and III; together with Benton and Hutchingson’s division of LH III to LH IIIA, B, and C
(Furumark 1941b: 17–19). It is the last of these which concerns us here.

The Mycenaean Late Helladic IIIC (which I henceforward refer to as “Myc IIIC”) com-
prises Furtwängler and Loeschcke’s “fourth style” (Myc IIIC:1), as well as the so-called sub-
Mycenaean” pottery (Myc IIIC:2). The “fourth style” had already been previously divided (by
Wace) to “close style” and “granary style.” I already mentioned that Heurtley had connected
the BPh style particularly with the “close style,” an identification accepted by Furumark (and
all subsequent researchers).

Chronologically, the Myc IIIC:1 is again subdivided into three periods: IIIC:1a is a tran-
sition period, with late variants of Myc IIIB “open style” still appearing besides some IIIC
“close style”; IIIC:1b is characterized by mature “close style” and the beginning of the “gra-
nary” class “simple style”; while in IIIC:1c the “close style” all but disappears and the “sim-
ple style” predominates, developing into the IIIC:2 “sub-Mycenaean.”

Furumark characterizes the development of the Myc IIIC style by three contrasting pro-
cesses—simplification and geometrization of the motifs on the one hand, and elaboration of
the composition on the other, coupled with a tendency to linearize the execution. A concur-
rent process is the fragmentation of the relatively homogeneous IIIB style to several regional
variants, each of which followed its own line of development.

As part of this process, the BPh (and the Cypro-Geometric!) are seen by Furumark (1941b:
118–20) as “Oriental derivative wares.” The inspiration for most BPh motifs is from early
(though not the earliest) Myc IIIC. He further subdivides the BPh into two chronological
groups, reflecting successive deterioration of the original Mycenaean motifs, to which he
added a third, degenerate group, that he called “sub-Philistine.”

Synthesis: Desborough and the End of Mycenaean Civilization

The Myc IIIC pottery, and to a lesser extent other attributes of material culture, were the
subject of a detailed study by Vincent R. d’A Desborough (1964). His main archaeological
results seem to indicate an even greater diversity and overlap than suggested by Furumark.
“Myc IIIC” is not a single style, but rather a group of different styles with complex interrela-

Long

oi.uchicago.edu



571PHILISTINE BICHROME PAINTED POTTERY

tions. The “close style” and the “granary style,” both of which are seen as emerging from
Mycenae, are rather dissimilar in appearance, representing two divergent lines of develop-
ment from a common Myc IIIB stock. An “octopus style” may have originated in Crete,
where it is used together with a “fringed style,” but is also found in Attica and in Rhodes
(where, however, it appears together with a “sub IIIB open style” survival, no “close” or
“fringe” style at all, and very few atypical “granary” vases). The “simple style” of Cyprus
resembles (but is in no way identical) with the late “granary style” of Mycenae. On the tem-
poral axis, there seem to be connections, if not actually overlap, between “close style” and
“granary style”; “granary style” and “sub-Mycenaean”; and “sub-Mycenaean” and “proto-
Geometric” (Desborough 1964: 9–28).

The evidence of site distribution is as confusing. Many sites were destroyed towards the
end of the LH IIIB stage (Mycenae was even damaged twice). Some were abandoned, but
some (including Mycenae) were reoccupied. Others (e.g., Athens, Miletus, Iolkos) were ap-
parently untouched while yet others (e.g., Perati, the region of Achaea) were actually first
occupied during the LH IIIC. Final abandonment of those LH IIIC sites that did not survive
into the historical period also differs, with Mycenae itself being somewhere toward the mid-
dle of the sequence. Coupled with this evidence should be the dearth of what might be con-
strued as foreign influences, although exceptions can be found in the first iron implements to
make their appearance during this period, the introduction of cist burial and hand-made pots
in the Macedonian style, which were spread (to regions where they were not already present
in the LH IIIB), late rather than early in the sequence.

And yet, when considering all this evidence, Desborough (1964: 219–20) rejects out of
hand the hypothesis that it essentially reflects a process of local development (or degenera-
tion, as the case may be). Instead, he posits at least two waves of external invasions, each
with several undercurrents and secondary ripples. The assumed direction of both is overland
from the northwest, although the possibility of a northeast seaborne origin, at least for the first
one, is considered (not least because this is, according to Desborough’s belief, where Egyp-
tian and Hittite sources place the origins of the Sea Peoples—1964: 222–23).

The first invasion put an end to the thriving “Mycenaean empire,” pushing refugees into
undamaged areas (Achaea and Kephallenia on the west, and as far as Cyprus on the east). It
also created the diversity of Myc IIIC styles, inasmuch as it broke the (preconceived) political
hegemony of Mycenae over the rest of the Helladic “kingdoms,” which was the reason for the
uniformity of the IIIB “koiné” style to begin with (Desborough 1964: 218–20; 225). In
explaining the lack of evidence for any sort of external contacts at this time he considers two
explanations: First, that the “archaeologically rather elusive invaders” did not leave “a single
object or custom . . . in any of the areas through which they passed” because “the culture of
the invaders was probably primitive, and anyway far inferior to that of the Mycenaeans; their
artifacts may for the most part have been of perishable materials, such as wood and leather,
and thus no trace would be left of them” (Desborough 1964: 224). To his credit, he rejects this
possibility, and concludes that the invaders simply did not stay to enjoy the spoils of their
conquest (in which case why the need for the extensive relocation of refugees? [I. S.])

The second wave started, according to Desborough (1964: 230–32) about a century later,
and was perhaps less violent, though more prolonged of duration and even more influential
in its aftermath. It consisted of migrations of peoples from the Balkans into central main-
land Greece, and later into the Peloponnese, causing a fresh surge of displaced persons to set

oi.uchicago.edu



572 ILAN SHARON

sail across the Mediterranean, this time bearing late “granary style” pottery with them. The
fusion of the newcomers with the few remaining Mycenaean descendants would, in the full-
ness of time, produce Greek civilization.

After pondering over the confused and conflicting mythological traditions about “Dorian,”
“Thessalian,” “Ionian,” “Dryopian,” and “Eolian” migrations, Desborough (1964: 246–48,
255) concludes that “at many stages the traditional account is not supported by the archaeo-
logical evidence. The dates may be questioned, the interrelation of the stories is open to doubt,
even the facts may not stand up to examination.” With these limitations in mind, he proposes
that these stories reflect the second, and not the first “wave” of invasions, because the first
wave was not followed, in his opinion, by colonization and because of the traditional dating
of the Dorian invasion (60–80 years after the fall of Troy, which in Desborough’s opinion can
be dated “at no other time” than 1250–1230 B.C., on the basis of Herodotus and a Hittite ref-
erence to a “king of Ahiyawwa” present in person on the Asian mainland at this time).

What does all of this have to do with our Philistines? Desborough begins his discussion of
them by stating:

It is a fact that the Peleset, an important group of the raiders who were defeated by
Ramesses III on the borders of Egypt in the eighth year of his reign, are the Philistines
of the Bible, and it is also true that they settled in the southern part of Palestine after
their defeat. (Desborough 1964: 209)

To this he adds the “virtual certainty” that it was they who destroyed Ugarit just a short
time before (1964: 207). Note how tenuous identifications coalesced, after being repeated
for sixty years (without much addition of data), into unshakable truths. Desborough points
out that the Philistines could not have come into Egypt with their Philistine pottery already
at hand, for that pottery is derived from Myc IIIC:1b, and no Myc IIIC or BPh was found at
Ugarit. It is curious that a lack of any finds attributable to the invaders did not stop Desbo-
rough from positing a Balkan invasion of Argolis. Perhaps this is because it was unthinkable
that in the case of a Mycenaean “invasion” of Canaan, too, the “the culture of the invaders
was probably primitive, and anyway far inferior” to that of the Canaanites, so that “no trace
would be left of them.”

Be that as it may, Desborough assumes that the Philistines who sacked Ugarit and raided
Egypt were, or at least

. . . had been joined by a powerful group of Mycenaeans . . . who fled after the cat-
astrophic invasion of the Mainland at the end of LH IIIB . . . the best organized force
[of the refugees] took to their ships and sailed eastwards, making common cause with
other disturbed groups . . . [They] will have been responsible, together with non-
Mycenaean ethnic groups, for the first of the two destructions at Cyprus. At this stage,
either the whole Mycenaean group or part of it decided to establish itself at Cyprus,
perhaps to be followed shortly after by other Mycenaean groups. . . . The rest of the
raiders than moved on to Syria (where it was quite natural for LH IIIB pottery still to
be current) and joined forces with other groups coming overland from Asia, thus
creating the formidable body which went southwards, by land and sea, towards Egypt.”
(1964: 238–39)

Thus, Desborough concludes that the Philistines adopted the BPh style in pottery only
some years after they were already settled in Canaan. Support for this statement is sought

Long

oi.uchicago.edu



573PHILISTINE BICHROME PAINTED POTTERY

from the fact that anthropoid coffins at Beit Shean (which he assumes must belong to Phil-
istine mercenaries) do not have BPh pottery associated with them. At this point he invokes
a suggestion previously made by Benson (1961), that the entire BPh repertoire, except the
“decadent” sub-Philistine ware, was produced by one man, or a single family of potters. He
proposed that this potter arrived from Cyprus at about 1165 B.C., some fifteen years after the
initial settlement of the Philistines.

Obviously, many of the assumptions underlying this line of reasoning are open to objec-
tion. Two points from it, however, will crop up again. One is the chronological discrepancy
engendered by the assumption that BPh pottery was already being produced at the beginning
of the twelfth century. Also, this is the first of several suggestions that seek to resolve this
discrepancy by dissociating the BPh pottery, in one way or another, from the Philistines.

Synthesis: Dothan and the Origins of the BPh Style

Trude Dothan’s synthesis (1967, and a revised English edition in 1982) undoubtedly
marks the high point of the “Sea Peoples = Philistines = BPh pottery” paradigm. It is by far
the most detailed analysis of BPh pottery extant and, whatever the changes in interpretation
of the evidence, it forms the basic text for all subsequent critiques, present one included. It
is curious that, although people had been talking of “Philistine pottery” for some sixty years
before this study, the term was never before formally defined. Following Dothan, this is per-
haps an appropriate point to formally define the BPh ceramic family:

Typology
The BPh is a family of decorated tableware. It includes bowls, small kraters and small

closed vessels (stirrup jars, pyxides, jugs, and bottles). Dothan identifies seventeen forms, of
which eight, in her opinion, have Mycenaean prototypes, two are borrowed from the Cypriote
repertoire, one is an Egyptian form, four follow local Canaanite traditions, and two are late
forms, one of which has no known antecedents and one which appears in the contemporary
Israelite culture (1982: 96).

Decoration
The BPh pottery is decorated by painting, usually in two colors (red/brown and black) on

thin white wash. The painting is always linear and done in freehand (unlike the “Phoenician”
bichrome, which is probably instrumentally applied and favors thick bands). Bichrome
painted decoration has a long history in Canaan, from the end of the Middle Bronze Age and
throughout the Late Bronze Age. The BPh, as well as other decorated groups of the Iron I
(such as the aforementioned “Phoenician” bichrome) actually form the tail end of this tradi-
tion. Bichrome decoration is totally foreign to the Mycenaean tradition, though it does appear
in Cyprus.

Technology
The BPh vessels are manufactured in the local Palestinian tradition (which suffered

successive loss of quality throughout LB). The white wash is possibly a crude attempt to emu-
late the very light, well-levigated ware typical of the Myc IIIC.

oi.uchicago.edu



574 ILAN SHARON

Motifs
The motifs painted on the BPh pottery are either geometric or pictorial (of which the

bird motif is by far the most common). Dothan recognizes fourteen motifs, of which twelve
appear in the Mycenaean repertoire, one (with four variants) is typically Egyptian, and one
has purely local Canaanite antecedents (T. Dothan 1982: 198).

Function
The BPh class does not form a complete assemblage, as it contains only fine table wares

(and possibly some cultic vessels). For other functions (storage, cooking, trade) the Phil-
istines seemed to have used local-tradition coarse wares. (One exception which does not
materially change this picture are recently-identified cooking vessels in the Mycenaean tra-
dition, which were used at Tel Miqne/Ekron together with the Myc IIIC and perhaps briefly
thereafter—Dothan and Dothan 1992: 241; Killebrew 1998: 397).

Geographic Distribution
The distribution of BPh pottery, as noted by Dothan (the * marks additions to the cor-

pus since Dothan’s 1982 survey, which are included here for the sake of completeness) is
as follows: Excavated sites in the core area where BPh pottery appears in high frequencies
(although quantitative data is unavailable for any site other than Tell Qasile) are Tell Farºah
(S), Deir el-Balah*, Tell Jemmeh, Tel Seraª, Ashkelon, Ashdod, Tel Mor, Tell es-Safi/Gat,
Tel Miqne/Ekron, Azor, Tel Jerishe, and Tell Qasile. These sites are all located within “clas-
sical” Philistia (the southern coastal plain, between Wadi Ghazza [Na˙al Besor] and the
Yarkon River) and include all of the “five cities of the lords of the Philistines” (except for
Gaza, which was not excavated). A secondary zone where significant amounts of BPh pot-
tery occur is the foothills between Judea and Philistia, at sites like Tel Batash (Timna)*,
Beth Shemesh, Tel ºEitun*, Gezer, and Aphek. Groups of BPh vessels were also found at
other places, namely Beth Zur, Bethel, Tell Beit Mirsim, Lachish, and Tell en-Nasbeh in the
highlands of Judea; and Acco, Megiddo, ºAfula, and Dan in the north. Quantitatively these
sites contrast sharply with the two former groups inasmuch as the BPh pottery from them
form a minute percentage of the total assemblage. Handfuls of BPh sherds were found in
many other Iron I sites and are not mentioned here (cf. Brug 1985: 66–96, fig. 18 for a fairly
up-to-date list).

Dothan followed Furumark’s methodology, as well as, to a great extent, his stylistic
conclusions about the BPh pottery, which she attempted to bolster with stratigraphic evi-
dence. At the time of the publication of the Hebrew edition of her book this evidence rested
on the (then unpublished) stratigraphy of Tell Qasile, Petrie’s seriation of the Tell el-Farºah
(S) tombs (as redated by Albright) and the Megiddo evidence, which had to be extensively
reworked to confirm to the supposed sequence, as we have noted above. Dothan divides the
BPh into three chronological substyles, which denote progressive deterioration of the deco-
rative elements and increasing variation from the Mycenaean prototypes. These parallel
Furumark’s division to two substyles plus the “Sub-Philistine” phase.

Thus (at least as of 1982, but see second thoughts in Dothan and Dothan 1992: 258, and
what amounts to a retraction in Dothan 1998: 159–60) Dothan saw the Philistine pottery as
appearing suddenly, in its fully developed form, and then progressively regressing to the

oi.uchicago.edu



575PHILISTINE BICHROME PAINTED POTTERY

Levantine average. While she correctly stresses the affinities between the BPh pottery and the
Mycenaean (especially the IIIC “close style”) her careful survey illuminates also the many
affinities of this group to non-Mycenaean styles. Perhaps the most important implication of
this study is that the BPh pottery can no longer be passed off as another local Myc IIIC variant
but is an entirely different phenomenon. Dothan saw in the appearance of BPh pottery clear
evidence of migration and the many non-Mycenaean elements in it as reflecting the route
which these newcomers took:

It appears that Philistine pottery was one of the local ramifications that developed after
the collapse of the Myc IIIB pottery koine style of the Late Bronze Age. It can be stated
with confidence that this pottery was not the product of a people coming directly from
their country of origin with a homogeneous tradition, but rather reflects the cultural
influences picked up along the way in the long, slow, meandering migration from their
Aegean homeland. (T. Dothan 1982: 217)

Synthesis: Sandars and the Sea Peoples

One more attempt at a broad popular synthesis of the Sea Peoples and the crisis of the
beginning of the first millennium, by Nancy K. Sandars (1978), should be mentioned last, for
in the theoretical facet it invokes both the hitherto favorite culture historical/diffusionistic ex-
planation framework, and the functionalistic mode of explanation to which the next section is
devoted.

Sandars’s analysis is based primarily on historical references (primarily, again, name
etymologies) to Sea Peoples in Egyptian literature, Ugaritic, and Hittite sources, and Greek
mythology; and the dress code and physical characteristics of Sea Peoples or possible Sea
Peoples in the artwork of the Near East and the Aegean. Archaeological considerations (mainly
bronzes; Sandars 1963)—and only then pottery—are secondary.

Sandars paints a very volatile picture of “nations on the move” at the end of the Bronze
Age, hurling peoples across the Mediterranean at the slightest provocation. Thus the srdnw
are identified (mainly by their horned headdress) as originally coming from Syria (as are the
dnynw), more specifically from the region of Ugarit. They emigrated first to Cyprus (on the
basis of ashlar construction at Early Iron Age Enkomi and the horned “ingot god” from
the same site—Sandars 1978: 151–55) and elsewhere at the end of the Late Cypriot period (=
our early Iron Age I) and later they emigrated to Sardinia, to which they gave their name
(based mainly on the eighth century bronze figurines of “horned warriors”). The prstw, in her
opinion, were not primarily a maritime nation either, but uprooted Anatolian farmers (ac-
cording to the depiction of their ox-drawn carts in the Egyptian monuments and admittedly
slender linguistic etymologies which point, in her opinion, to Luwian rather than Greek ante-
cedents—Sandars 1978: 166). They settled mainly harbor-less sites on the southern coastal
plain of Israel. The t·krw who later inhabited the ports of northern Israel (based on the Wen-
Amon story) have more right to the title “People of the Sea.” She favors their identification
with the Teucri of Homer’s Troy, rather than the Sikels who inhabited southeastern Sicily in
the eighth century. The latter, she proposes, are the descendants of the skrsw, who also came
there from Anatolia. Coupled with all of these is a southward movement of Danubian peo-
ples, into the Balkans, northern Italy, and the Troad.

oi.uchicago.edu



576 ILAN SHARON

As for BPh pottery, Sandars believes it should be renamed “Sea People” pottery, because,
in her opinion, it was not specific to the Philistines. On the contrary, its Mycenaean anteced-
ents point away from the (in her opinion) landlubber prstw and towards other groups (1978:
166–67). Sandars emphasizes the non-Mycenaean aspects of the BPh pottery, and its eclectic
nature:

What we find is evidence for an intermingling of peoples from the north among whom an
Anatolian element was very strong (?—I. S.). . . . This whole phenomenon is of mixed
Aegean, Anatolian and native Canaanite elements, and the product of a settled people
who had, at least for a time, found a homeland. Further than this I do not think it is safe to
go.” (Sandars 1978: 169)

This, then, is possibly the “wave theory” at its most elaborate, with cross currents wash-
ing both west to east and east to west. The methodology followed in tracing these hypo-
thetical movements, too, is reminiscent of diffusionism at its heyday. However, Sandars was
also much impressed by Braudel (1972, 1980) and the Annales school of social history,
which sees, behind the intricate tapestry of historical minutiae, a shadowy but constant ma-
terialistic determinism.

Based upon the infrastructure of traditional mixed Mediterranean subsistence economy,
the LB elites in Egypt, Syria, Anatolia, and the Aegean, and even further north up to the
Danube, built a fragile superstructure dependent on long-range trade in metals and luxury
items. To facilitate this trade a class of merchants and sea captains was formed, and to protect
their trade routes and their luxurious lifestyle, the princes increasingly depended on a warrior
class of fighting nobility and/or mercenaries. Once a link in this precarious structure broke, the
luxury goods stopped flowing, and these privileged classes would revert to piracy and freeboot
warfare to obtain their share of the suddenly scarce luxury goods. Thus local stresses on each
one of the Mediterranean powers threatened, and eventually brought down, the entire edifice
(Sandars 1978: 197–98). This theory is a harbinger of processual-functional explanations,
which is considered next.

IDEOLOGY RECONSIDERED

After World War II, no scholar of repute would endorse Kossina-type theories. Many set-
tled for a less virulent version of the culture-history paradigm, commonly known as norma-
tive theory, and a related world view—historical particularism. Perhaps the best definition of
normativism was given by Childe (1964: 15–26; 1965: 25–27): A culture is seen as a set of
norms or rules governing the “proper” way for a member of the cultural group to behave in
any given domain. Thus the difference in pot type between two different cultures reflects the
difference between what “we” think is the proper utensil to eat from, cook in, or whatever;
and the way “they” do. An individual is born into a culture and internalizes its value system
as part of the process of socialization. Once assimilated, these norms form part of one’s sub-
conscious and are rarely, if ever, subject to introspection, much less revision. The “culture
shock” attendant upon the changing of norms is traumatic to the individual (Childe 1965: 22),
and therefore culture is passive and conservative in its nature (Childe 1964: 30)—it will stay
static unless change is forced upon it.

oi.uchicago.edu



577PHILISTINE BICHROME PAINTED POTTERY

The basic model of a culture is still genetic, that is, a “culture” is the attribute of “a
people” and “a people” are conceived of as an extension of the family or a subdivision of the
biological species. The culture is still regarded as an organic unit: Once the “cultures” are
defined, the object of research is to explain the similarities or differences between them, rather
than the components of the individual culture (Watson, LeBlanc, and Redman 1971: 62). The
main difference between the normative paradigm and racist views of culture is that the bonds
between individuals and their culture are seen as social ties. The organic or genetic model is
seen as an analogy rather than a literal description of physical reality.

Almost all the tenets of Kulturgeschichte live on, in expurgated versions. Thus in describ-
ing his methodology, Furumark strongly rejects the theory of Montelius’s “Swedish typol-
ogy” school, which equates cultural development with organic growth, and yet claims that “in
order to ascertain the relation between morphological elements [of pottery types and] . . .
human activity, we must apply a principle of arrangement that corresponds to the general psy-
chological factor constituting that relation. This factor is the inertia of the human mind, man-
ifesting itself in conservatism” (1941a: 3, italics mine).

Desborough, while rejecting the racial interpretation to the demise of the Mycenaean cul-
ture (1964: 242) equates stylistic homogeneity with political adhesion (and hence, by default,
the spatial distribution of a style with national borders):

The effect of [political] disaster following on disaster and of conditions of major unrest
will probably be reflected in the degeneration of the LH IIIB style in the Argolid; and in
so far as this is happening, and as the political dominance is removed, the style is likely
to disappear. . . . It is under these conditions that a gradual change from one style to
another is probable, and also that homogeneity will be seriously affected. Something
new will arise, presuming there is not simply a progressive degeneration . . . . Any new
style, unless there are racial changes, is likely to be founded on the preceding one, but
[with the authority of the Argolid now fundamentally shaken] several new and distinct
styles could emerge.” (Desborough 1964: 5)

Furumark considers Hall’s (1928) judgment that Myc IIIC style is “degenerate, vulgar, and
tasteless” to be too harsh, but he pronounces the Myc IIIC:1 to be “a ‘baroque,’ which is a
natural successor of the ‘Classicism’ of Myc IIIB style” (Furumark 1941a: 571) while in the
succeeding “sub Mycenean” IIIC:2 style “standard motives (sic) had attained their simplest
possible form. Some of them are mere survivals . . . others are highly conventionalized, de-
generate, and muddled” (Furumark 1941a: 576). In a similar vein Desborough declares that
“In the normal way, the continuation of a style will depend on its vitality, and after some time
there will come a period of stagnation” (Desborough 1964: 4).

A peculiarity of the BPh pottery style, which is derived from this theoretical outlook, is
that it is all progressively “degenerate.” Since the BPh is supposedly a derivative of the Myc
IIIC style, which is itself already a “baroque” to the “classic” Myc IIIB, all its “development”
is of necessity downhill. Thus in the chronological subdivision of BPh pottery the first phase
is designated as possessing the highest artistic quality, and succeeding phases are character-
ized by progressive degeneration (Furumark 1941b: 120; T. Dothan 1982: 96). When elabo-
rate examples are found in late contexts (as is the case with the “Orpheus vase” in Megiddo
Stratum VIA (Dothan 1982: 78) or several lavishly decorated pieces in Qasile Stratum X
(A. Mazar 1985: 104) they are regarded as anomalies which need to be explained away.

oi.uchicago.edu



578 ILAN SHARON

I have already noted that for Pythian-Adams the bell-shaped krater (which he calls
“bowl fitted with a horizontal loop handle”) was a fossil directeur to the Volksgeist of the
Sea Peoples—but so is Sandars’s (1978: 88–100) Naue Type IIa sword.

One way of addressing the paradox of regression to the norm, without assuming either
that [some] people were biologically incapable of improvement or the existence of an irre-
ducible Volksgeist to every ethnos, is an assumption of (random) partial reception of cultural
traits. This is exactly how Mendel solved the regression paradox in genetics. This is also the
assumption made by Franz Boas and other promoters of Historical Particularism. If it is not
“culture” as a whole which is diffused from the “cultural center” to the “periphery,” but
rather individual cultural traits. If such traits can only be “exchanged” at a point of contact
between two cultures, and at each such “exchange point” only a small random selection of
traits passes from the donor to the recipient, then given a wide enough list of traits, an almost
unlimited mosaic of cultures can be created. Each culture will have its unique combination of
traits, most of which it has inherited from its forefathers and some of which it has acquired
from its neighbors (Harris 1968: 376–79).

If one accepts the assumptions above, then by cataloging the trait lists of various cultures
and studying the diffusion maps of individual traits, one can recreate the genealogy of each
individual culture in terms of what other cultures it had been in contact with (direct, or n-
times removed). The question why trait x was adopted at a given point of contact, and not trait
y is not, however, deemed answerable. This is the type of reasoning behind the attempts to
trace the route which the Philistines traveled from their unknown homeland to Philistia by
enumerating the “source” of individual decorative motifs on BPh pottery, as is very apparent
in the works of the Dothans, Desborough, and Sandars.

While historical particularism could provide a liberal alternative to the excesses of Kul-
turgeschichte and to unilinear evolution, it had the effect of reducing the nomothetic nature
of archaeological explanation. Whatever the moral odium of the racial approach to anthro-
pology, it was “scientific” in the sense that it held that there are immutable laws which shape
the course of human history, and that one could predict the direction in which a culture would
develop at any given point of time, given its past record and the state of the world at that
point. According to historical particularists one might be able to show how a given culture
arrived at a certain point (i.e., came to possess a given trait list) but not why. Detailed study
might map out the course of development of a single culture, but that knowledge will not aid
us in promulgating any general rules which would apply to other cultures at other points in
time, nor will it enable us to predict what would befall this culture in the future.

Another drawback of historical particularism can be illustrated by the fortunes of the Phil-
istine “wave theory.” When each change in the material culture at a given spot is explained
away by a contact with another culture, explanations tend to dissolve into incredibly intricate
pseudo-histories.

In the case in point, we started out with a seemingly elegant account, linking events
across the Mediterranean—a two-pronged “Achaean” invasion from the Balkans into Greece
on the one side and Anatolia and the Levant on the other. It soon becomes clear, however,
that the new arrivals in the Levant are not post-Mycenaean “Achaeans” and so a domino-
effect gambit is used (“Achaeans” displacing “Myceneans” and pushing them overseas). To
account for biblical and egyptological descriptions we need, however, not one, but at least

oi.uchicago.edu



579PHILISTINE BICHROME PAINTED POTTERY

two invasions, and so an additional wave is now posited (one bringing Myc IIIC in its wake
and the other the BPh). The first wave needs to be divided into at least two ripples if we wish
to account for the development of a local “simple style” of after the initial arrival of Myc
IIIC-bearing peoples. Soon, even these are not enough and Desborough needs to invent a rov-
ing band of potters settling in Philistia after both these “waves” to account for the appearance
of BPh pottery. Meanwhile, in Greece, it is clear that one invasion is not enough to account
for both the change from Myc IIIB to IIIC and from IIIC to proto-Geometric. We need two
(or even three) separate invasions, each of which presumably pushed fresh “waves” of dis-
placed persons in an outward-bound course. These cannot be equated with the two “waves”
which presumably washed out on the shores of Palestine, because no “granary style” or
proto-Geometric ever arrived here. Accepting Finkelstein’s “ultra-low” chronology means
positing another wave of pre-Myc IIIC-manufacturing marauders (Falkenstein 1998: 143,
see discussion below); not only are changes in material culture symptomatic of a wave of
peoples but there are invisible waves, too. And we have not even begun to account for at
least two waves in Cyprus (a wave of destructions by Sea Peoples at the end of the Late Cyp-
riot II period and colonization by “Achaeans” during the LC III [e.g., Hood 1973, Desbor-
ough 1964: 238–39]); we’ll need an additional “wave” to account for Cypriot “granary style”
(Kling, 1989: fig. 26) and perhaps another one for the beginning of Cypro-Geometric (Kara-
georghis 1992, 1994); and what about Sandars’s westward-bound cross currents?

A point which becomes clear when we review the above is that if we allow the positing of
a new “wave” in response to any anomaly in our data, we soon reach what is called a “vacuous
explanation” in logic or a “saturated model” in statistics. The wave theory is vacuous not
because it is necessarily untrue, but on the contrary, because it can never be disproved. Any
new evidence at all (and even lack thereof) might be accommodated within it.

DATA COLLECTION AND ANALYSIS 1970s–1990s: BREAKDOWN OF CONSENSUS

The publication of Trude Dothan’s synthesis in 1967 might be seen as the high point of
the “normal science” period (to use the well-known “Kuhnian” terminology) in the study of
the Philistines. Up to this point, almost all the evidence supported (or at least was taken to
support) the basic equation introduced in 1908:

Prstw (of Egyptian literature) = Philistines (of the Bible) = BPh pottery

There were, however, some cracks in this monolithic view (which were already hinted at
before, see Desborough’s chronological reservations). From the late 1970s onward a trickle
of critiques started to appear, which by the mid-1980s had become a torrent. Some of these
were dependent on the results of new excavations (some of which recovered new evidence
of BPh pottery and some of which by rights should have but did not). Other critiques were
based not on new finds, but on reexamination of old data, that is, looking at the same things
through new ideological glasses.

Excavation: Deir el-Bala˙

BPh pottery was found in the final phase of the Egyptian border post at Deir el-Bala˙,
excavated by Trude Dothan in 1972–1982 (T. Dothan 1979; Dothan and Dothan 1992:

oi.uchicago.edu



580 ILAN SHARON

205–8). The significance of this excavation, however, is not in locating further attributes
defining Philistine culture, but rather those which do not. The excavation started as a result
of finding a cemetery containing anthropoid coffins of the type previously attributed to the
Philistines. The evidence from both the cemetery and the settlement is that, at least at Deir
el-Bala˙, the anthropoid coffins are not associated with the early Iron Age settlement, but
with the previous LB Egyptian fort. This effectively lays to rest long-standing speculation
about the association of this Egyptian type of burial rite with Philistines. It apparently was
reserved to Egyptian garrisons at the end of LB and (perhaps) the beginning of the Iron Age
(which is not to say that occasionally a Philistine mercenary might not be buried in this man-
ner, nor that an Egyptian official’s tomb might not occasionally contain some BPh pottery).
Thus anthropoid burial can be removed from the “Philistine equation” (Bunimovitz 1990:
217; Stager 1995: 341–42).

Excavation: Ekron

By far the most significant excavation for the “Philistine question” in recent years is
Trude Dothan’s and Seymour Gitin’s excavation at Tel Miqne/Ekron, the third city of the
Philistine pentapolis to undergo systematic excavation. The data here can be briefly summa-
rized as follows.

The most complete sequence of the relevant phases was obtained in a section on the
northeastern corner of the mound. A modest Canaanite settlement (Stratum VIII) is replaced
by a huge fortified enclosure in Stratum VII. This new town is characterized by the massive
appearance of Myc IIIC “close style” pottery, making up as much as 50% of the contents of
typical assemblages. The Myc IIIC of the next phase (Stratum VI) is not as elaborate, and can
perhaps be best characterized as a “simple style” (though whether it is the same as the Cyp-
riot or Greek mainland “simple” styles is an open question). BPh pottery makes its first
appearance in Stratum VI but does not completely replace the monochrome Myc IIIC until
Stratum V (T. Dothan 1992: 94–95). Stratum IV is coeval with the final, degenerate, phase of
BPh pottery (i.e., roughly equivalent to Qasile Stratum X). The city lost its prominence and
was severely limited in its area at the beginning of the Iron II period (Ekron Strata III/II).
It was not reoccupied, in its full extent, till the end of the Iron Age (Stratum I).

An additional area, in the middle of the tell, confirms the sequence as described above as
well as the fact that the entire enclosure was occupied in the Iron I period. Also, a public
building was found, with several unusual architectural elements that are possibly Aegean in
origin (Dothan and Dothan 1992: 236–38, Dothan 1998: 155–57, cf. also Karageorghis 1998).

The picture at Ekron thus serves to confirm the one previously obtained by Moshe Dothan
at Ashdod and to fill in more details. The implications of these finds are deferred until the end
of the next subsection.

Compositional Analysis: Myc IIIC Pottery at Ashdod and Ekron

Two neutron activation analysis studies of Myc IIIC sherds from Ashdod (Asaro, Perlman,
and Dothan 1971) and Ekron (Gunneweg et al. 1986) have conclusively shown that the Myc
IIIC pottery at these sites was, like the later BPh pottery, locally made. On the other hand,
petrographic analysis of several Myc IIIC sherds from Mazar’s excavation at Beit Shean (see
below) indicated that these sherds are not local to Beit Shean, nor are they similar to the Ash-

long

oi.uchicago.edu



581PHILISTINE BICHROME PAINTED POTTERY

dod-Ekron group (Mazar 1997: 159). Neutron activation analysis of a Myc IIIC vessel from
Tell Keisan, a Phoenician site in the Akko valley, shows that it was made in Cyprus (Humbert
1993: 864).

Compositional analysis of pottery in Cyprus indicates that the Myc IIIB wares of the LC
II period (our LB), including the so-called “Levanto-Helladic” styles, were manufactured in
Greece. On the other hand, the Myc IIIC of the LC III (our early Iron Age I) are made of
different clays, and are probably by-and-large locally made (Kling 1989: 91–93). This has not
laid to rest the long-standing debate about Mycenaean potters at Crete in LB, for some Myc
IIIC-type vessels, with matt-colored finish unlike the glossy Myc IIIB, already do appear in
late LC II contexts.

At this juncture the question naturally arises whether we should keep calling what we
now know to be a local ware by the name “Mycenaean.” In Cyprus, the tendency in recent
years has been to group the local “Myc IIIC-like” wares together with other wares subsumed
under the technological definition of “Wheel Made White Painted” III wares. Obviously, the
very terms we use to describe empirical observations are theoretically laden. Other propos-
als, like “Monochrome Sea Peoples Ware” (Stager 1995) or “Monochrome Philistine” are
just as loaded, so I will keep using “Myc IIIC” in this work.

The difference between Myc IIIC and BPh is not, then, the result of different prove-
nance, or even of lack of knowledge (on the part of “Philistine” potters) of the technique of
making “true” Mycenaean IIIC ware. It is a deliberate change in technology, decorative tech-
nique, and range of motifs; the Myc IIIC (imported as well as local) being still a pure Ae-
gean style, while the succeeding (and partially contemporary) BPh style incorporates non-
Mycenaean elements.

The implication from this find, and from the typological sequence as revealed in Ashdod
and Ekron is that the Sea Peoples did not come with an already developed “Philistine” pottery
from abroad; but its entire development took place locally, with its stages being: a) imported
Myc IIIC pottery brought by the first settlers; b) locally made Myc IIIC; c) Myc IIIC and
“incipient” BPh; d) BPh. Such a reconstruction means that the non-Mycenaean elements in the
BPh pottery (except for Canaanite ones) could not have been “picked up along the way” by
the wandering Sea Peoples. It also introduces some serious problems of chronology but these
are described below.

The evidence indicates that Myc IIIC pottery from Beit Shean and Phoenicia is genuinely
imported and that the sporadic appearance of this ware in the north is part of an entirely dif-
ferent phenomenon, one which arguably has nothing to do with the settlement of Sea Peoples.

Other Excavations with BPh Pottery

One other project that is potentially crucial for the understanding of the Philistines is the
renewed excavation at Ashkelon, by Lawrence Stager, in which Doug Esse served as associ-
ate director and as director of the laboratory in Jerusalem at the beginning of these excava-
tions during 1986. Reaching the Philistine strata at that site has proven a formidable task.
While some Iron I levels and BPh pottery were found during this excavation, exposure of the
relevant strata thus far has been limited. The two material amendments to the rudimentary
sequence established by Garstang and Phythian-Adams in the twenties are that at Ashkelon,
too, there is a Myc IIIC Monochrome phase, which succeeds the LB destruction and precedes

oi.uchicago.edu



582 ILAN SHARON

the introduction of BPh pottery. Also, inasmuch as the limited exposure is indicative of the
whole site, then at Ashkelon, too, the transition from the Late Bronze Age to the Iron Age is
accompanied by a burst of urban activities, exemplified by the erection of public buildings
(Stager 1995: 345–46). Another city of the Philistine Pentapolis, Gat, identified by most
scholars as Tell es-Safi, is currently being excavated by Aren Maeir and Carl Ehrlich. The
last season has revealed, under a vast destruction layer dated to the Iron II period, an Iron I
phase with considerable amounts of BPh. The preceding phase seems to be of Late Bronze
Age already, and no Myc IIIC was found thus far (Maeir and Ehrlich in press).

At the southeastern edge of the Philistine littoral, Eliezer Oren has been excavating the
two nearby sites of Tel Seraª and Tel Haror. In the late Bronze Age Tel Seraª was an Egyptian
administrative center (Oren 1984). The last Egyptian stronghold, Stratum IX, has no Cypriot
or Mycenaean imports of any kind but does have several Egyptian hieratic ostraca. One bears
a fragmentary date (in regnal years) of 20+. The only probable candidate who reigned that
long is Ramesses III. In the following stratum, VIII, there are both BPh and “Ashdodian”
black-on-red pottery. It thus represents a somewhat late Philistine phase, comparable to Qasile
Stratum X and/or XI. The following occupation, Stratum VII, contains typical tenth century
Judean pottery (Oren 1993b: 1331). At Tel Haror, on the other hand, there are several pits,
allocated to phases B4–B2, which contain both Myc IIIC and BPh. These, then, are probably
contemporaneous with Miqne Stratum VI and Ashdod Stratum XIIIA (Oren 1993a: 582–83).
The big question is whether Haror Stratum B4–B2 are contemporaneous with Seraª Stratum
IX, or are later—that is, whether they fit into the hiatus between Seraª Strata IX and VIII.
Oren (1984) is inclined towards the latter answer. The problem of the absolute chronology of
the beginning of local Myc IIIC and of the BPh styles are discussed further below.

“Sea People” Sites Without BPh Pottery?

Philistine Bichrome pottery appeared in several other recent excavations, which are not
discussed here as they do not change the picture as presented thus far. Perhaps more revealing
to the structure of this picture are a number of excavations where the lack, or scarcity, of BPh
pottery is surprising:

Dor, identified with Tell el-Burj on the northern Sharon coast, is associated with the t·krw
according to the “Tale of Wen Amon” (see Scheepers 1991 for a recent review and full bibli-
ography). Limited exposure, due to the great overburden of later strata, has plagued Ephraim
Stern’s excavation at this site. After great efforts during the last few seasons, however, fairly
wide extents of each of the several Iron I strata at that site were cleared in several areas, and
abundant samples of the local Iron I assemblage was collected (Stern 1994: 85–104). It
includes a single complete jug which might be a rather atypical BPh or some other Myc IIIC
derivative, and only about a score of BPh sherds (mostly tiny fragments from unstratified con-
texts), a miniscule proportion within the Iron I collection. On the other hand, at Dor, as in
Ekron, Ashdod, and possibly Ashkelon, the Iron I is a period of urban flourit, rather than
urban decline, as in most other sites in Israel. (However, the Phoenician littoral north of Dor
also flourished during the Iron I period.) A cautious assessment of the results thus far would
indicate that if any significant t·krw population was present on these strata, it did not manu-
facture BPh pottery or use it to a higher degree than would be expected by normal trade.

oi.uchicago.edu



583PHILISTINE BICHROME PAINTED POTTERY

The new excavations at Megiddo (Finkelstein, Ussishkin, and Halpern, eds. 2000) have
uncovered new exposures of the Phase VIA destruction. For the first time, quantitative data is
available for that site, confirming that the proportion of BPh at this site is small.

Renewed excavations at Beit Shean, first by Yadin and Geva (1986), and then by Amihai
Mazar (1993), have revealed more of the Early Iron Age strata, which are crucial for the dat-
ing of the Bronze Age–Iron Age transition because of the abundance of dated Egyptian finds
from that site. No BPh pottery was found by either of these expeditions, although several
sherds of Myc IIIC (in addition to the pot found by the Pennsylvania excavators in the 1920s)
were found by Mazar in phases probably parallel to “early Stratum 6” and “late Stratum 7” in
the original excavators’ terminology (Wolff 1994: 493; Mazar 1997: 159). It is hard to assess
the chronological significance of these finds until well-stratified, absolutely datable finds are
published by the new expedition, as the spectacular finds of the 1920s’ excavation were noto-
riously misascribed.

Another excavation where BPh pottery is significantly lacking is David Ussishkin’s exca-
vation at Lachish (1985). Lachish was first excavated in the late 1930s by Starkey and subse-
quently published by Olga Tufnell (Tufnell, Inge, and Harding 1940 and Tufnell 1953). The
final LB stratum at that site, Stratum VI, ended with a violent destruction, after which there
was a gap in occupation. The excavators dated this destruction to the reign of Merneptah,
although Tufnell argued that it could actually be as late as Ramesses III. The following phase,
Stratum V, did not have any BPh pottery in it, though some was found in a cave on the slope
of the tell (Tufnell et al. 1958: 292–93). T. Dothan had argued, based on this pottery and two
anthropoid coffins found in Cemetery 500, that there was some occupation of the town during
Iron Age I, but it had not been located by the excavators (1982: 86–87, 276–79).

The renewed excavations reaffirmed, in the main, the results of the previous one, specifi-
cally that Stratum VI dates to the very end of LB, and that Stratum V is not any earlier than
the tenth century. The notable new find was a bronze trapping of the Stratum VI city gate,
which was found in the destruction debris of the gatehouse and that had a cartouche of
Ramesses III on it. Moreover, other written objects (unfortunately with no royal name on
them) from the same destruction layer make it unlikely that this destruction occurred early
within Ramesses III’s reign.

On the basis of the new Lachish information Ussishkin (1985) concluded that the date
of the end of LB is not 1230 b.c. (Merneptah) or 1200 b.c. (Tewoserat), but well into the
reign of Ramesses III, if not even later—that is, 1150 or even later. This is especially signif-
icant in view of the fact that Lachish is located just a few miles southeast of Tel Miqne/
Ekron. The entire sequence of development of the BPh from Myc IIIC must have occurred
after the destruction of Lachish Stratum VI, during the gap of occupation at Lachish, for no
trace of it is visible there. This means that the first appearance of BPh pottery must well
postdate the eighth year of Ramesses III.

Reworking Old Data: The Tell Farºah Cemeteries and the “Low Chronology” for BPh Pottery

A very similar conclusion had already been reached by Thomas McClellan some years
before (1979). As part of his Ph.D. dissertation McClellan ran computer seriation on Petrie’s
data from Tell Farºah (S), which is discussed above. Contrary to the order accepted until
then—which was Tomb 542 (earliest), Tombs 552, 532, 562 (latest); see, e.g., Dothan 1982:

oi.uchicago.edu



584 ILAN SHARON

29–32—McClellan put Tomb 552 as earliest, and then Tombs 562, 532, and 542 as latest in
the series.1

This new order has several implications. First, it undermines the “successive degenera-
tion” model of the relative chronology within the BPh sequence, as it is generally held that
Tomb 542 has the most elaborate BPh examples in it (Dothan 1982: 30–32). Moreover,
McClellan (1979: 66–67) claims that Cemetery 900, as a whole, precedes the Cemetery 500
series of tombs. This would mean that the appearance of BPh pottery at Tell el-Farºah is
later than the reign of Ramesses VIII, as scarabs ascribed to Ramesses III, IV, and VIII were
found in Tombs 934 and 984.

McClellan (1979: 73) argued, based on these findings, that BPh pottery was not really
introduced until the second half of the twelfth century. Therefore, he maintained, it really had
nothing to do with the Philistines, who fought Ramesses III and were settled by him in the
southern coastal strip. Rather, the BPh pottery was brought by yet another, hitherto unknown
“wave” of immigrants, who came from Cyprus 30–40 years after these events, and were not
Philistines at all.

Reworking Old Data: The “Ultra Low Chronology”

As we have seen above, the basic reasoning behind the “low chronology” (Mazar 1985a,
1985b, 1988; Stager 1985, 1995; Oren 1984, 2000) is that the pottery manufactured by the
original Sea Peoples at the time of Ramesses III was the Myc IIIC and that it took at least a
generation for the BPh to develop from that. A series of articles in the 1990s (Ussishkin
1995, 1998; Finkelstein 1995b, 1998) argues for an even lower date for the entire sequence.
The gist of the dispute is whether contemporanity is possible between late “Late Bronze
Age” Egyptian strongholds and “Iron Age” strata with Myc IIIC (Singer 1994: 290–94).
T. Dothan (1989) argues for a “coexistence” between Miqne Stratum VII and Lachish Stra-
tum VI. Finkelstein (1998: 141) maintains that the fact that not one Myc IIIC sherd has
traveled from Miqne to Lachish (a distance of only 25 km) or from Tel Haror to Tel Seraª
(less than 10 km), precludes such coexistence. Indeed, he notes that contrary to pap. Harris
(which states that Ramesses III settled the vanquished Sea Peoples in Egyptian strongholds)
Myc IIIC pottery is never found in conjunction with evidence of Egyptian occupation.
Therefore, Lachish Stratum VI and Seraª Stratum IX, both of which postdate the last Myc
IIIB imports, are earlier than Miqne Stratum VII and Haror Phases B2–4. Now, according to
the cartouche of Ramesses III from Lachish Stratum VI and a hieratic inscription “year 22+”

1. Some years ago, this author undertook to recheck McClellan’s results. I ran his numbers through several
seriation techniques different from the ones he had used, under several different distance functions, and
consistently got the same order McClellan obtained. It has to be pointed out, however, that McClellan used
Petrie’s (or, rather, Duncan’s) century old typology. It is certainly possible that reclassifying all of the ves-
sels (now in the British Museum) according to current typological criteria will produce substantially dif-
ferent results. Another possibility, that had not been considered by either Petrie or McClellan, is that at
least some of these tombs had a considerable chronological range. Serious consideration should be given
to this possibility in view of the current trend to date all anthropomorphic coffins to LB. If these tombs
were used from LB and into Iron Age I then quantitative seriation might mean nothing more than an indi-
cation of the amount of LB pottery (vs. Iron Age pottery) in each tomb (see also a very similar critique in
Brug 1985: 138–39).

oi.uchicago.edu



585PHILISTINE BICHROME PAINTED POTTERY

[presumably of Ramesses III, since no other monarch in the Twentieth Dynasty exceeded
that] in Seraª Stratum IX (Goldwasser 1984) the Egyptian administration in Palestine was
still active in 1160 (using the low chronology of Wente and van Siclen 1976: 217–61). The
first Myc IIIC-using immigrants could not have settled prior to that date, and the BPh could
not have started until at least a generation later.

Evaluation of negative evidence is always problematic. There are any number of late Late
Bronze Age and early Iron Age sites where Myc IIIC was not found. Were they all aban-
doned at the time? The alternative would be to argue that Myc IIIC pottery was produced
strictly for self consumption. Indeed, Bunimovitz (1998: 107) sees this as the result of a
deliberate “containment” policy on the part of the Egyptians—the first ever use of boycott as
an economic weapon to reach political goals(?)—since he sees both the Egyptians and the
Sea Peoples as ruling elites, one must wonder whether the Egyptians forbade “their” Canaan-
ites any interaction with the “Philistine” Canaanites, or just banned the visible display of
seditious decorative elements.

T. Dothan (1998: 151) counters with two arguments: a) Miqne Stratum VIII has Cypriot
as well as Anatolian “gray Minian” imports (and is therefore earlier than Lachish Stratum VI
and Seraª Stratum IX) and there is no occupational gap between Miqne Strata VIII and VII
(and hence Stratum VII must be contemporary with Lachish Stratum VI). This assertion is
belied by her own statement that the constructional fill beneath the Stratum VII floors con-
tains Myc IIIC and “transitional forms” which are lacking in the Stratum VIII destruction
debris (Dothan 1998: 151). This means, on the face of it, that some time must have passed
between the destruction of Stratum VIII and the construction of Stratum VII. There seems,
moreover, to be some disagreement within the excavation team on the factual merits of this
position (see Killibrew 1998: 383).

Be that as it may, the phenomenon of extensive manufacture and use of Myc IIIC at Tel
Miqne and Ashdod (and probably Ashkelon, too) is, by and large, highly localized and crisply
delineated, isolated examples of Myc IIIC in sites outside the core area of the Philistine
pentapolis notwithstanding (see Finkelstein 1998: 142 for definite and putative instances).

Finkelstein runs into an even bigger problem in reckoning this theory with documentary
evidence: if Myc IIIC was not produced at the southern coast until ca. 1130, who were the
Sea Peoples subjugated by Ramesses III in 1175? He is forced at this point, like many before
him, to posit yet another, “invisible” wave of marauders, the only indication for which is a
wake of vandalism (Finkelstein 1998: 143). He further surmises that the subsequent settle-
ment of these “invisible sea peoples” was not in Philistia but in some hitherto unexplored
parts elsewhere.

All of these limitations and more apply to Ussishkin’s (1995, 1998) attempt to lower the
arrival date of the Philistines according to the finds of Stratum VII at Megiddo. The conten-
tion is that Megiddo was an Egyptian stronghold until ca. 1130 b.c., and, as such, would have
prevented the southward advance of the Sea Peoples had it occurred prior to that date. The
scenario offered by Ussishkin is highly precarious. All that may be said is that if the statue-
base of Ramesses VI is indeed to be associated with Stratum VII (the excavator, for one, says
it was found beneath a Stratum VII wall, Loud 1948: 135, note 1); and if the incursion of the
Philistines took the form of an overland military invasion from the north, and if all by-passes
to the ºIron valley were impassable at that precise point in time, and if the Egyptian garrison

oi.uchicago.edu



586 ILAN SHARON

at Megiddo took this invasion to be hostile to Egyptian interests in Canaan, than yes, this
script is not outside the bounds of possibility.

Reworking Old Data: The Tell Qasile Temple and the Nature of BPh Pottery

Two additional critiques which appeared in the mid-1980s propose different views of the
Philistines and of their connection to the BPh pottery.

Shlomo Bunimovitz (1990, 1998) proceeds from an assumption that the Philistines were
a ruling elite, governing a population which was predominately Canaanite. In order to iden-
tify a material attribute as “Philistine,” one must demonstrate that it was brought by the Sea
Peoples from abroad and was used by the Philistine sub-culture, rather than by other cul-
tures which shared the same chronological and geographical space with it. The local Myc
IIIC is, according to Bunimovitz, a distinctively Philistine product. The BPh, on the other
hand, is a copy of the Myc IIIC by local potters. It was used indiscriminately by all of the
population and so is not exclusively Philistine (though it is, of course, probable that it was
used by Philistines too). It was distributed along the coast and occasionally inland by the
usual mechanisms of trade and imitation.

A case in point is the sanctuary at Tell Qasile. There is nothing in its cultic furniture, and
precious little in the architecture, to suggest that the cult practiced in it is any other than
Canaanite. On the contrary, the few indications that we posses of what might be termed true
Philistine religion (e.g., “Ashdoda” type figurines) are conspicuously missing at Tell Qasile.
Here is a clear case, claims Bunimovitz, of a Canaanite revival, possibly under Philistine
political hegemony. The BPh vessels found in this temple (together with quite a number of
“Phoenician” bichrome vessels and even “Israelite” collared-rim jars) simply reflect the
choice of “exotic” wares available to the population at that time.

Reworking Old Data: Amihai Mazar on Megiddo, Tell Qasile,
and the Nature of Philistine Culture

As we have seen above, the chronological framework of the BPh pottery is heavily depen-
dent on the Tell Farºah cemeteries, Megiddo, and Tell Qasile. McClellan undermined one of
the foundations of this framework, and Ussishkin produced independent evidence in support.
Amihai Mazar (1985a, 1985b, 1988) proceeds to demolish the other foundations of this
framework. In his final report on the pottery from the temple at Tell Qasile, Mazar (1985a:
104–5) contradicts some of the previous claims about the subdivision of the BPh repertoire
into chronological phases. While conceding that, in general, the earlier phases of Philistine
culture show greater affinities with Aegean prototypes, and a higher percentage of lavishly
decorated pieces, he does not agree that Stratum X is a “degraded” or “sub-Philistine” phase.
There are several examples of elaborately decorated vessels that belong to this stratum. Con-
versely, some types previously claimed to start only in the “sub-Philistine” phase, such as the
“Ashdodian” style of painting black bands on red background (usually interpreted as some
sort of amalgamation of Philistine technique with the red slip tradition of Judea), actually
already appear in Stratum XI.

Restudying yet again the stratigraphy of Megiddo, Mazar (1985b: 96–97) claims that all
of the BPh sherds recovered from Strata VII and VIII are intrusive (and not just the ones

short

oi.uchicago.edu



587PHILISTINE BICHROME PAINTED POTTERY

from Strata VIII and VIIB, as previously claimed by T. Dothan). This leaves Stratum VII,
which in his opinion is securely dated to Ramesses III, as earlier than the appearance of BPh
pottery, in line with the findings at Lachish and elsewhere.

Reworking Old Data: Brug’s Literary and Archaeological Study of the Philistines

The year 1985 seems to have been a turning point in the assessment of Philistine culture.
In addition to Ussishkin’s (1985) chronological conclusions and Mazar’s (1985a, 1985b)
reassessment of Tell Qasile and Megiddo, that same year saw the appearance of John Brug’s
(1985) study of the Philistines based on his doctoral dissertation which appeared a year ear-
lier. Brug essentially reviews the same material that Trude Dothan did only three years
before, but his conclusions are rather different.

Trude Dothan ascribes a Mycenaean source to a BPh pot type or decorative motif when-
ever its original provenance could be traced to a Mycenaean prototype, whether or not that
type had already previously existed in the Canaanite repertoire. Brug, on the other hand, limits
“Mycenaean” types only to those which could not possibly have been picked up by the Philis-
tines in Palestine itself. Thus, whereas Dothan sees eight BPh forms (out of eighteen) as
descendant from Mycenaean types, Brug classifies only four as having definite Mycenaean
roots, and four as “doubtful” (Brug 1985: 113). The origin of even these four forms is not nec-
essarily Mycenaean, because by the time they arrive in Israel they are also “naturalized” in
Cyprus (vs. Dothan’s ascription of only three forms to Cypriot prototypes). Also, having
defined the BPh group on the basis of form, technique, colors, and motifs, Dothan tends to
take a maximizing or inclusive view, classifying as “Philistine” vessels that have only two or
three, and sometimes even only one, of these traits. Brug takes a rather restrictive view, that
is, any vessel which is not one of the eighteen “classic” forms, and/or does not have one of the
distinctively Philistine motifs drawn in red and black on a white background, is not deemed
Philistine. Thus the percentages of “true” BPh within total assemblages according to Brug are
much depressed (1985: 55–57). The same methodology is followed elsewhere. Brug (in agree-
ment with Bunimovitz) sees little of the Mycenaean, or none at all, in Philistine cultic prac-
tices, architecture, and the like—not because there are no shared traits with the Aegean in
these facets, but because parallels can also be found in the local, or Cypriot, littoral.

It is therefore hardly surprising that Brug concludes that the Philistine culture is compos-
ite (in which he and Dothan are in complete agreement), but that the Canaanite element is
predominant. As for the sources of the Aegean elements in this complex, one need go no fur-
ther than Cyprus. These foreign elements, though naturally the most distinguishable, are also
the most transient. It is the Canaanite elements of Philistine culture that endure into the Iron
II Age. Brug stresses that “migrations” can (rarely) take the form of mass population move-
ments, like that of Slav peoples into the Balkans, but more commonly are small scale settle-
ment of individuals or of families, establishing independent colonies or carving out vassal
fiefs. Brug points out an apt analogy for the latter kind in the settlement of Norsemen in
England (and other coastal parts of Europe) in the Middle Ages; bringing a short period of
radical change in the material culture of these regions, but quickly assimilating into the local
culture, blending into it a few names, decorative motifs, and myths (Brug 1985: 201–5).

oi.uchicago.edu



588 ILAN SHARON

An Assessment of the Critiques

Absolute chronology is quite central to the debate about the character of Philistine cul-
ture, as several additional critiques which support the late dating of the beginning of BPh
(Oren 1984: 55–56; Stager 1985: 62) also point out. A brief discussion is therefore merited.
Although it is generally agreed that Myc IIIB was still being produced during the reign of
Merneptah (ca. 1236–1223 / 1212–1202 b.c.), many tend to lower the terminus post quem for
the end of Myc IIIB to ca. 1200 / 1185 b.c., due to the recent finding of a cartouche of queen
Tewosret (1209–1200 / 1186–1175 b.c.) at Deir Allaª in a LB context with Myc IIIB pottery.
Assuming that BPh pottery was being manufactured during the eighth year of Ramesses III,
or even shortly thereafter, would leave barely a decade for Myc IIIC:1a to appear in Myce-
nae, evolve into the mature “close style,” cross the Mediterranean and get established in local
workshops in the Levant, and after a period of congruence develop into BPh. A period of at
least half a century seems more appropriate for this sequence.

One way out of this quandary is to assume a period of chronological overlap wherein
Egyptian-dominated LB culture at some sites or regions co-existed with Philistine (and
Israelite) sites displaying Iron I cultural attributes (T. Dothan 1993: 97; Stager 1995: 344). In
order to corroborate such a hypothesis some evidence which unequivocally demonstrates
such contemporaneity (e.g., some Myc IIIC from an “LB” site) will have to be forthcoming.
Stager (1995) points out that another long-held belief, that of the settlement of Philistines as
Egyptian mercenaries, is also incompatible with this view. Most recently excavated assem-
blages with distinct Egyptian presence (Lachish Stratum VI, Tel Seraª Stratum IX) do not
have either Myc IIIC or BPh in them, and vice-versa.

Lest it be thought that all this fiddling with a decade here or two decades there is quib-
bling, it has to be pointed out that if the BPh pottery did not appear until thirty or forty years
after the documented appearance of the Philistines, than one is justified in questioning its
“Philistinicity.” This is indeed what McClellan and Bunimovitz, and, to a lesser extent, Mazar
and Desborough, have done.

A. Mazar (1985a: 119–20; 1988: 257) also proposes to slightly change the definition of
“Philistine culture,” but not exactly in the same sense as Bunimovitz. He sees the BPh pottery
as a hybrid of Aegean and Canaanite influences, which came into being on Palestinian soil
only a generation or two after the initial settlement of the Philistines in the days of Rames-
ses III. Mazar, Brug, and Bunimovitz actually present two views which are quite close. The
essence of the difference between them is that Bunimovitz sees Philistine culture as a com-
posite entity, in which (at least) two sub-cultures can be delineated, while Mazar (1985a:
120) sees it as a single atomic “Philisto-Canaanite” culture.

Taken together, all of these critiques point to dissonance in the factual basis of the previ-
ously accepted paradigm. In one way or another all of them also express doubt as to the basic
tenability of the equation of pottery types with an ethnic identity. In this, they reflect a gen-
eral trend in the ideology of archaeologists in the last decades, as I attempt to demonstrate in
the next section. They do have two glaring drawbacks—they do not explain why, if it did not
appear immediately, did a unique pottery style evolve a generation after the Philistine settle-
ment in Canaan; and while they seem to reject the normative explanation of culture, they do
not offer a coherent alternative. In the last section of this work, I try to build such a theoret-

short

oi.uchicago.edu



589PHILISTINE BICHROME PAINTED POTTERY

ical defense to the model suggested by Mazar. First, though, a survey of the main ideological
stances in archaeology in the last decades is in order.

PROCESSUAL EXPLANATIONS

Functional Explanations of Culture

Several different (and sometimes contradictory) developments promoted the “new archae-
ology” revolution, several strands of which developed from the late 1960s and onwards. The
“price” for adopting a normative, particularistic view of material culture was the loss of gen-
erality and of a nomothetic framework for human development. As prestige (and economic
superiority) became a matter of “high tech,” and academic disciplines were either sciences or
nothing at all, these came to be seen as intolerable defects. Such pressures were felt worst in
North American archaeology, which was traditionally located in departments of social studies
rather than in the humanities. It was patently clear that historical particularism would never be
able to “predict the past.” An “explicitly scientific” alternative, it was hoped, might be able to
do that. Incompatible definitions of just what is “explicitly scientific” were being offered at
the time by Hempel (1965), Popper (1959, 1969), and Kuhn (1962). So as not to seem inhos-
pitable, archaeologists tended to agree with them all (Wylie 1985: 483).

Another concurrent development was the rise of environmentalism to a position of dom-
inant paradigm. The threat of atomic holocaust, dwindling resources, and rising pollution
eroded the faith in “progress” as a panacea for the human condition. The lessons that the past
was supposed to teach us had shifted from how “progressive” civilizations eradicate “primi-
tive” ones (to the general glory of mankind), into lessons of how to live in harmony with
nature and that upsetting the equilibrium leads to catastrophe. Archaeologists increasingly
find themselves enlisted into the “green militias,” by inclination and by guild affiliation. As
salvage and conservation come to dominate the archaeological work market, operating
within park services or financed by environmental impact projects, it is only natural that
archaeologists adopt the prevailing idiom in their theoretical discourse.

Yet another source of distaste with the “old archaeology” was its myopic view of the
world through ethnic or racial glasses. The civil rights movement in America in the 1960s, the
economic rise of the far east, increasing multi-culturalism at both edges of society, the aca-
demic and corporate elites on the one side and migrant workers on the other, as well as more
recent developments, such as the rise of the communication age and the European union
movement, had all served to blur the definition of what ethnic or racial identity is, and under-
mine the notion that it is a useful (much less singular) classificatory criterion. Ethnic strife
has become a deplorable third-world phenomenon, hardly Byron or Shelly’s romantic ideal.
Racial bias is reserved for rednecks and skinheads. In lecture halls or boardrooms you mess
with these subjects at your peril (except, of course, to condemn others of hypocrisy).

A completely new definition of the essence of “material culture” seemed called for. In
particular, devastating critiques were being leveled at the view of culture as a set of norms, at
the view of “cultural traits” diffusing from privileged “cultural centers” to passive “peripher-
ies,” and at the view of “cultures” as atomic units.

oi.uchicago.edu



590 ILAN SHARON

The definition of “culture” which most “new” archaeologists adopted was that of the func-
tionalistic school in social anthropology (Harris 1968: 514–26, 635–50). It is really a revival
of the evolutional schemes which were popular in seventeenth and eighteenth century world-
views, in a new pseudo-Darwinian guise. The most often quoted definition is Leslie White’s
(1959: 9) “man’s extra-somatic means of adaptation.” “Culture,” and material culture in par-
ticular, is conceived of as a toolkit, developed by man to help him survive and exploit his envi-
ronment in an optimal fashion. This simple enough definition has far reaching implications:

(1) In the first place, the task of the anthropologist is not deemed to document the con-
tacts between “cultures,” but to explain how each culture is uniquely fitted to its particular
environment, and why new cultural attributes (whether acquired by diffusion or by internal
development) are accepted or rejected, on the basis of their suitability to that environment.

(2) Culture is basically seen as adaptive. Society will tend to change its culture to suit its
physical and social environment. Societies which ignore the environmental directive and are
behaviorally or organizationally inefficient will not be able to survive in the long run.

(3) The boundaries of an observed material “culture” define an ecological niche (White
1959: 284), or a subsistence system rather than an “ethnos” or “a people.” Although it is pos-
sible (indeed probable) that in the long run, a shared way of life will also define the boundaries
of the “us,” and hence provide the base for the definition of social groupings.

(4) Cultural change is usually taken to be indicative of environmental (ecological or eco-
nomical) transformation, or of evolutional processes affecting that society and altering its
relationship with its environment, hence the name “processual archaeology” which was given
to this school.

(5) The model or analogy at the base of the functional paradigm is taken from neo-
Darwinism and from biological ecology (Watson, LeBlanc, and Redman 1971: 88–107). “Cul-
ture” is seen as a direct continuation of biological evolution.

Under such a paradigm, it is understandable that cultural diffusion in general, and
migrations in particular, fell into disrepute as explanations for cultural change. “Blaming” the
appearance of each material attribute on an arbitrary trajectory bringing one “culture” into
contact with another and the stochastic “borrowing” of cultural attributes in these contacts was
seen as begging the question of causative explanation of material culture phenomena.

The essentially synchronic (or, rather, a-chronic) outlook of the functional school in
anthropology led to the dethroning of the apt historical literary reference as the sine quo
non of valid archaeological inference. Pride-of-place was accorded, instead, to ethno-
graphic analogies. The tacit reasoning being that particularistic historical trajectories are
not as relevant as the environmental imperative which dictates similar survival strategies in
similar circumstances.

The processual paradigm became very popular in the 1970s and 1980s, in some places
(northwestern Europe and North America, for instance) to the exclusion of other paradigms.
In the archaeology of the Levant, due to its traditionally close ties to the humanities, it came
rather late, and apart from programmatic statements (e.g., Dever 1981), did not make much of
an impact on the actual agenda of archaeology until the mid-1980s. Since then, however,
there has been a growing body of processual explanations. These include several radically
new approaches to the enigma of the “Sea Peoples.”

short

oi.uchicago.edu



591PHILISTINE BICHROME PAINTED POTTERY

The “Nomads of the Sea”

The model proposed by Michal Artzy (1997, 1998) is essentially an elaboration upon the
one introduced by Sandars (1978) some twenty years earlier. Artzy argues that the Late
Bronze Age lifestyle depended on overseas trade in luxury goods. This promoted the estab-
lishment of a caste of deep sea-going sailors, freebooting traders, mercenaries, and itinerant
artisans who had only transient allegiance to the powers-that-be. They would be the first to
be injured at any recession. Being but marginally attached to the nuclear commonwealth
they would be most inclined to switch from safeguarding its lifelines to preying upon them,
thus initiating a negative feedback cycle that would sever more webs of commerce, further
weaken the established order, and add deflationary pressures to the economy. This is the
process which she sees at the root of the “thirteenth century crisis.” Esse (1991, 1992) sug-
gested that a very similar dis-equilibrium was concurrently evolving between the urban
lowland civilization and highland pastoralists and farmers, a model further developed by
myself several years ago (Sharon 1994).

In seeking to establish ethnographic parallels to such a situation, however, Artzy’s theory
suffers from mixed metaphors. She offers several completely different (to my mind) corollar-
ies to the hypothesized phenomenon. The first are pirates, from the so called “Cilician
pirates” of the first century b.c., to Anglo-American piracy of the eighteenth century, or even
latter-day boat-jackers in Malaysia or off the African coasts. This is a hopelessly wide and
amorphous designation, perhaps because “piracy” is largely in the eyes of the beholder.

If we are to believe Appian’s description that the “Cilician pirates” (Artzy 1997: 6, after
Rauh 1997), had permanent fortified ports and far-flung marketplaces, could and would
mobilize fleets 20–30,000 strong, made and broke treaties, and were ruled by tyrannoi; they
were, for all intents and purposes, independent polities that resorted to priveteering. Indeed,
these descriptions would fit to a tee the appearance and activities of some of the Italian mar-
itime republics in the Middle Ages, or those of the “Robber Barons” of the same times. Had
not the “Cilician pirates” attempted to take the Nostrum out of the Mare . . . and lost, they
might have been quite differently described. Would Sir Francis Drake qualify as a “Sea
Nomad”? Needless to say, people for whom “the boat . . . was not necessarily their home”
and whose “womenfolk, children and others . . . remained behind, continuing their diverse
economic pursuits” (Artzy 1997: 7) can by no stretch of the definition be called “nomads.”

While it may be true that high-seas pirates of the early modern period were in a sense
outcasts from the establishment, they were by no means a nomadic society, or even a society
at all. As Rediker (1987: 285) says “They produced nothing, and had no secure place in the
economic order . . . they were unable to create reliable mechanisms through which they
could either replenish their ranks or mobilize their collective strength.” The facts quoted by
Rediker belie his romantic depiction of eighteenth century piracy. In the “golden age” of
piracy, the single decade between 1716 to 1726, there were, at any point in time, between ten
and twenty pirate ships, manned by some 1,000–2,000 men, plying the Atlantic between the
West Indies and the coast of Africa (Rediker 1987: 256). Though pirate ships sporadically
sailed in consort, and they occasionally vowed to avenge captured comrades, no effective
long-term alliances were ever formed (Rediker 1987: 268). They never, despite the contem-
porary scandalizing reports and the romantic aura of later times, posed a serious threat to

oi.uchicago.edu



592 ILAN SHARON

established maritime powers. (The Royal Navy, whose demobilization after the Spanish
Wars caused this outbreak of piracy in the first place, was never depleted below 10,000 sea-
men [Rediker 1987: 281–82]). Moreover, within that decade, some 4,500–5,000 men are
estimated to have sailed under the skull and crossbones (Rediker 1987: 256) and very few
survived into the second quarter of the century. These two statistics, taken together, mean
that the average life expectancy of a pirate (or at least the length of his career as a renegade)
did not exceed a couple of years.

Things are confused even further when an analogy to desert nomads is made. Despite the
fact that they are dependent upon sedentary states and a market economy for the purchase of
many staples, nomadic pastoralists do form stable societies (Khazanov 1983: 119–23, 148),
perpetuating themselves and their customs. As such, they are ethnic entities. While Hakka-
like “boat people” do exist in the orient (though whether they are “nomads” is debatable), it
has not been shown that such subsistence has ever been , or even could possibly be, practiced
on the Mediterranean.

The most serious defect in Artzy’s intriguing conjecture, however, is its rather slender
empirical basis. While the written records can be read so as to fit the proposed scenario ( just
as they have been interpreted to support ethnic migrations), there is little archaeological evi-
dence in favor of it. We really do not know enough about LB trade mechanisms at all. Over-
seas contacts may have been affected by a class of specialized intermediaries, as Artzy
supposes, but then again they could have been accomplished by “direct marketing” expedi-
tions by the primary producers themselves, or by indirect local barter and short hops up and
down the coast. If “Sea Nomads” ever existed, they left little if any material traces. (The
same is true, I would have to admit, with hypothesized Late Bronze Age pastoralists in the
hill country and that is certainly a glaring defect in my own reconstruction of the establish-
ment of the “Israelites” [Sharon 1994: 126]). The bronze treasure-trove in tiny Tel Nami
(Artzy 1995: 26–29; 1998: 440) is an exciting find, but had it not been for it, there would
have been little to distinguish Nami from dozens of other, large and small, Late Bronze Age
sites. The idea of the innocuous Nami anchorage as a notorious pirate cove is romantic and
novel but is hardly the only (or even the simplest) explanation for the facts.

Artzy confines her discussion to the Bronze Age. It is not clear what relation (if any) her
“Nomads of the Sea” had with the Philistines of the Iron Age. At any rate, this theory does
not offer an explanation as to why these nomads (if indeed they are the same people) sud-
denly established urban centers, some far away from the sea, started to produce a distinctive
material culture, and ceased all visible signs of maritime activity and overseas trade. While
this model makes arguable assumptions about ephemeral archaeological phenomena, it fails
to account for the prima facie empirical evidence.

The “Decentralized Trade” System

A comprehensive economic model, developed from one introduced for Greece (Sheratt
and Sheratt 1991; see Dickinson 1994 for a recent overview of current processual theories to
explain the collapse of Late Bronze civilization in Greece) has been recently “transplanted” to
the Levant by Susan Sheratt (1994, 1998) and embellished by Alexander Bauer (1998).

True-to-form of standard functionalistic critiques of the culture-history genre, Sheratt
censures previous theories as

oi.uchicago.edu



593PHILISTINE BICHROME PAINTED POTTERY

. . . a kind of quasi-politico-military history, elaborated beyond recognition . . . in which
short term events rather than long term processes are seen as the most significant and
decisive factor. Thus, mass invasions of migrants sweeping in . . . are held to account
like deus-ex-machina for a whole package of changes. (Sheratt 1998: 296)

. . . the ostensibly “ethnic” similarities . . . [of] “Sea People” settlements . . . are as likely
to arise directly from this [natural evolution and expansion of international trade] as from
any deep-seated racial or linguistic relationships. (Sheratt 1998: 294)

The Late Bronze Age–Iron Age transformation in the entire Mediterranean is seen by
Sheratt as a transition from a centralized “palatial” monopolist exchange system, which she
thinks was typical of the Bronze Age to “decentralized” private entrepreneurship, character-
istic for the Iron Age in her opinion. This transformation is affected by three complementary
processes, all of which start in the Late Bronze Age and culminate in the Iron Age. The first
is the widening of the exchange market from prestige goods bartered between rulers to com-
modities intended for general consumption. The second is a shift from trade in scarce materi-
als (e.g., gold, silver, tin), high in primary value, to manufactured goods, in which the added-
value component is especially high (e.g., ornate pottery, textiles), the third is a substitution of
the economic center of gravity from inland producers to coastal “gateway communities”
chiefly concerned with distribution.

The motor which jointly runs all of these processes is the growth of a merchant class. The
procurement of exotic goods for the Late Bronze Age elites scattered around the Mediterra-
nean promotes the formation of groups of specialists concerned with moving these goods
around. To supplement their income, sailors add low value, high value-added “trinkets” to
their primary cargo. These cheap products command a wider market than the “primary”
goods, and offer a wider margin of profit for the intermediaries. Increasing concentration in
the moving-around of “secondary” cargoes, and consequently the increase of wealth in the
hands of the merchant class, increasingly undermines the ancien regime of resource-based
aristocracy, and shifts the balance of power from inland centers controlling the resources to
“gateways” concerned with distribution. The final chapter of this struggle is the switch to
commodities self manufactured by the merchant communities because this way they retain
100% of the price, and/or because the original product can no longer be obtained, due to
the collapse of the production system.

It is this final chapter which results in the locally-produced Myc IIIC appearing in Cy-
prus and Philistia. The reasons for why these wares should share an Aegean style are, in
Sheratt’s opinion, peripheral: The consumers are used to viewing this style as “exotic,” and
it is cheaper to manufacture than the hand-made Cypriot-style products.

Now, whatever the merits of this quite elegant theory in the Cypriot sphere, it simply does
not fit the bill for Canaan. What it would have us believe is that what seems to the naked eye
like a catastrophic collapse of the Late Bronze Age importation and a major setback in the
higher apparatus of civilization which would last for centuries, is in reality nothing but an
expansion of international trade (Sheratt 1998: 294; Bauer 1998: 160). While there is some
truth to Muhly’s (1998: 320) claim that the term “dark age” has suffered erosion in the last
decade, does this mean that it was in actuality the opposite?

The facts are that as against tens of thousands of items imported from Cyprus and Greece
to Canaan in the Late Bronze Age, we can, at present, establish perhaps several scores of

oi.uchicago.edu



594 ILAN SHARON

items definitely imported in the Iron Age I period (almost all of them towards the end of the
period, and almost all from non-Philistine sites [see Gilboa 1989: 216–18]). Sheratt (1998:
304–5) counters with some not-quite-testable hypotheses, such as that the trade had switched
from pottery to textiles or that the many metal remelting installations typical of the Iron Age I
in Israel (not only, or even primarily, in Philistine sites—I. S.) and in LC III Cyprus (Sheratt
1998: 300) indicate a trade in scrap-metal (see also Bauer 1998: 160, but would not a simpler
explanation be that recycling was made necessary by the cessation of trade?)

The reasons for perpetuating Aegean prototypes are not valid for Canaan, neither the
local Myc IIIC nor the BPh are any more “industrialized” or easier to manufacture than the
local products (probably quite the opposite). That the merchants provided consumers with
copies of products they were used to is belied by the fact that the range of popular shapes in
both Myc IIIC and BPh groups is quite different than those of Myc IIIB imported in the pre-
vious age. Some (e.g., cooking pots) had never even been seen in Israel before. In any case,
where are these “consumers”? Myc IIIC pottery was never (or almost never, if we are to
believe Finkelstein 1998: 142) used outside the towns where it was manufactured, and while
BPh traveled a little further, it still was primarily a domestic ware.

These shortcomings do not faze Bauer (1998: 150), who judges Sheratt’s version to be
“the most compelling conceptualization of the nature of the Sea Peoples to date.” He proposes
to test this theory with the following hypothesis: “If this is the case, one might expect that . . .
the material culture of these sites would primarily reflect mercantile activities, rather than
being simply “cultural” indicators and in any event would reflect the multi-cultural traditions
of those involved in the maritime trade network.”

The settlements bearing Myc IIIC pottery show no mercantile activities and no involve-
ment in any maritime networks at all. If the purpose of making this pottery was to market
“their own imitation of Mycenaean wares throughout the eastern Mediterranean” (Bauer 1998:
160) then this venture ended in abject failure.

Does this mean that the hypothesis, that the demise of the Bronze Age was hastened by
competition between the established aristocracy and a growing merchant class, has no merit
at all? Not necessarily, but I believe that any valid explanation of the Late Bronze Age col-
lapse will have to include the collapse of the trade networks (and hence of the merchant
class) as well.

The Systemic Approach

The functional view of culture has some obvious drawbacks. First, functional explana-
tions all too easily fall into the trap of naive environmental determinism, and hence into
eighteenth-century style unilinear evolution. Also, there are many aspects of culture that
seem, on the face of it, to offer no functional advantage, and some that are positively
counter-productive. Granted that changing the shape of projectile points may offer some
added efficiency to the Indian’s hunting technique, but what about changes in the headdress
of the chief? In order to rescue processual archaeology from these pitfalls “new archaeol-
ogy” combined functionalism with systems theory (Hodder 1986: 18–33).

Clarke (1968: 43) defined culture as a system of subcultures with interactions between
them. The systemic approach views culture as a structure, composed of smaller units (sub-
cultures). It does not, however, attempt to directly define either the system or its components.
This is typical of the “black box” approach of systems theory.

long

oi.uchicago.edu



595PHILISTINE BICHROME PAINTED POTTERY

Systems theory (like mathematical “theories” in general) is not a theory at all, in the
usual scientific sense, inasmuch as it does not have a subject matter, or a domain. Rather, it
is an abstract vocabulary, or a set of analogies which can be used (in this case) to describe
the way various components might interact to produce a total output in a composite entity
(whether this entity is a car, an institution, or a plant community). Which analogies (if any)
might be usefully invoked in a specific case depends on how we define the “system” being
studied and on the nature of its components. In the case of anthropological or archaeological
use, the system is, of course, the culture. The subsystems can be of several kinds. They can
be groups within the culture (e.g., “young males,” “itinerant craftsmen”) or institutions (“the
court”) or they can be facets of culture (“art,” “mythology”).

Systems theoretical terminology has been used in processual archaeology first of all as
an antidote to the normative assumption of the atomicity of cultures. In the second place it
has been used to explain away apparently non-productive subsystems (e.g. mythology). The
windshield wiper on a car can serve as a crude but effective analogy (I have an aversion to the
system theorists’ fondness of defining concepts of systems theory in system theoretical jar-
gon): no one would be able to guess, from looking at the wiper alone, that it has anything to
do with getting people or commodities from point A to point B. It is only when viewing it in
the holistic context of a working car that it becomes apparent how it does, in fact, aid the
entire system (the car) to achieve its output (move from A to B). Thirdly, systems theory has
been used to explain cultural diversity (i.e., why cultures in similar environmental straits are
not of necessity similar). Any system expends resources in two ways: one is to get its primary
function done (the resources spent in turning the wheels of the car) and the other is the use of
resources to get all secondary functions running smoothly (keep the battery full, the engine
cool, the ball bearings lubricated, etc.) In human organizational systems the first task is often
defined as “work” and the second as “activity.” Now, it seems perfectly sensible that structur-
ally different systems will use different activities to get the same work done.

What, if anything, has the systemic approach contributed to the discussion of the Sea Peo-
ples? Beyond some terminology (“system collapse” is a favorite on both sides of the Medi-
terranean, as are “negative feedback loops” and “interactions”)—not very much. Some of
Renfrew’s models for the demise of the Mycenaean culture are inspired by systems theory or
by catastrophe theory—another mathematical idiom that I do not discuss here. Bunimovitz’s
view of Philistine culture consisting of a Sea Peoples aristocracy and a Canaanite “proletar-
iat” is to a certain extent a systemic view. Note, however, that his view of the nature of each
of these sub-cultures remains strictly normative. The two subsystems must have been inordi-
nately conservative if even after cohabiting for two centuries (the span of time between the
assumed initial disembarkation of the Sea Peoples to the end of Qasile Stratum X) the Qasile
“Philistine” stratum is still assumed to possess a uniquely Aegean culture, at least inasmuch
as anyone making obeisance at a Canaanite-styled sanctuary is assumed to be “Canaanite” by
definition.

Bunimovitz’s approach might be contrasted with a somewhat similar attempt at synthe-
sis of the Cypriot aspects of this process by Iacovou (1998). While not straying far from the
normative view, she points out that the basic happening behind the various epi-phenomena of
the Late Bronze Age to Iron Age transition is the forging of new ethnic identities. The cor-
nerstone of her argument is that “notions like that of a national identity . . . had no meaning
for the . . . [Late Bronze Age] . . . societies. They are the fundamental novelties of the new

oi.uchicago.edu



596 ILAN SHARON

world of the Iron Age.” Of special significance is her observation that even non-Greek groups
(on linguistic and historical evidence) participated in the Aegeanizing koiné.

An Assessment of the Processual Approaches to the Sea Peoples

The advent of “new archaeology” certainly brought a fresh breeze (and quite a few ruf-
fles) to the stagnant pool of archaeological theory in the investigation of the nature of the
Sea Peoples as in other areas of the profession. If we expected, however, that an “explicitly
scientific” archaeology would pay more than lip service to such notions as “theories not
directly grounded in phenomenology have no cognitive meaning” or “a notion which does
not increase the testability of a theory should be ruled out of the field of science,” we were
sadly mistaken. The “nomads of the sea” hypothesis explains away some minor anomalies
but begs the question on the main phenomenological issues, while the “decentralization of
trade networks” theory, in a neo-Platonic twist, would have it that the “truth behind the phe-
nomena” is actually the opposite of what meets the eye. As we have seen in the case of the
“wave theory,” which was widely accepted despite being ill fitted to the facts as then known,
archaeologists are still creatures of the prevailing ideology. Theoretical elegance is a better
key to acceptance then empirical adequacy.

The processual paradigm never took the field in Near Eastern and Mediterranean archaeol-
ogies as completely as it did in the archaeology of northern Europe or North America. Several
explanations have been offered for that (e.g., Trigger 1989: 174). I think the problem is the
basic a-historical and non-humanistic stance of the “new archaeology.” Such a stance is bound
to lack appeal in regions where still (and with good reason) the study of the Bible and/or clas-
sics forms the backdrop for archaeological research.

At any rate, the culture-history, or diffusionistic approach to the settlement of the Sea Peo-
ples is still quite alive today. Lawrence Stager seems well aware that he lacks the home-court
advantage when addressing a congress on “The Archaeology of Society” in Los Angeles in
1993, as his preamble (1995: 332) clearly shows. Nevertheless, he professes his confidence
that the Sea Peoples “established beach-heads all along the shores of the eastern Mediterra-
nean” (Stager 1995: 336) and that “the Sikils, who settled at Dor, also sailed west and gave
their name to Sicily, and the Sherden, who probably established a beachhead in Akko,
bequeathed their name to Sardinia.” Indeed, some of the most militant (in form and content)
formulations of the “invasion theory” are the most recent. Whether it is Stager’s D-Day Sce-
nario quoted above; Ussishkin’s (1998) Monte-Cassino premise, where Megiddo acts as the
bulwark against Sea Peoples’ land offensive, or Bunimovitz’ (1998) Maginot-line hypothesis,
where the distribution of Sea Peoples’ pottery was “contained” by a line of Egyptian forts.

POST-PROCESSUALISM AND THE PHILISTINE QUESTION

Deconstructing the Philistines

The “post-processual” critiques of the now mainstream processual archaeology are not
a monolithic “school.” The ideological background for post-processualism is the so-called
“post-modernism” which, again, is more of an expression of vague discontent than a crystal-
lized world-view. Two main substreams within post-processualism (and within post-modern-

oi.uchicago.edu



597PHILISTINE BICHROME PAINTED POTTERY

ism), however, are the relativistic view of science (or at least the view of archaeology as a
relativistic discipline) and a semiotic view of culture.

Positivistic view of science, the philosophy which processual archaeology attempted to
emulate, argued that science has a privileged position in the quest for truth, by virtue of its
rigid separation between observation and interpretation. Theories are subject to change, but
so long as they are firmly rooted to the immutable database of observed facts, each succeed-
ing theory will be a better approximation of the truth. Relativistic approaches challenge this
view (for the entirety of science or for limited fields within it) by arguing that the observer
cannot be separated from the observation. The observer always operates within an ideologi-
cal or theoretical framework that conditions what he thinks is worthy of observing, which
observations he holds valid, what mental pictures he uses for conceptualizing his observa-
tions, and what vocabulary he is going to use to describe them.

If we hold that all scientific activity is interpretive to a certain degree, then no point of
view is entirely objective, and no cosmology can hold an a priori privileged position. Post-
modernists hold a jaundiced view of the prerogative of science in modern western society.

Perhaps the deadliest weapon in the relativistic arsenal is the technique of deconstruc-
tion, following the history of a theory, and showing how, at every step, proponents of that
theory were swayed by ideological bias (preferably by ideologies the readers no longer
share . . .), and how they fitted their facts around their convictions rather than vice-versa
(e.g., Feyerabend’s 1978: 69–143 treatment of Galileo). It should be stressed that the differ-
ence between the positivist and the relativist frame of mind is not in admitting that scientists
are influenced by the intellectual climate of their times. Most positivists would readily agree
that many (perhaps most) scientists are prejudiced. Still, their position is that “Science” with
a capital S is unbiased. Relativists, on the other hand, argue that bias is not a weakness or a
failing, but goes with the territory. No theory can be free of it.

Though I doubt Esse would have described himself as a post-processualist, his “Scholarly
Ideology and Ceramic Typology” (1991) and “Ceramic Distribution and Ethnicity” (1992)
deconstruct the normative paradigm for “Israelite” ethnicity. One need not, of course, go any
further than this present essay to find a virulent example of deconstruction in the case of
research into the origins of the Sea Peoples; Neil Silberman (1998) furnishes another. Silber-
man’s treatment of the subject parallels the first chapters of his work, though his subdivision
of the era into ideological “schools” is slightly different than mine.

The relativist philosophy operates within an obvious paradox. If no point of view is
privileged, why should the critique enjoy any superiority over what is being criticized? The
honest post-modernist must admit that one is as influenced by one’s own ideology and pre-
conceptions as are any other subjects. While shooting indiscriminately all around, the decon-
structionist cannot avoid hitting a foot.

The Symbolic Approach

Deconstructionism is (as its name implies) a purely negative tool. A positive alternative
to the normative or the processual definitions of “material culture” is a semiotic view. Like
the two others, it really has been around for a long time. Symbolism has been a subject of
research in the social sciences and the humanities (e.g., psychology and literature) almost
since their inception. The meaning of symbols used by ancient societies has, of course, been

oi.uchicago.edu



598 ILAN SHARON

a matter of speculation to archaeologists ever since Hieroglyphs were interpreted as devices
of sorcery. This raises an obvious practical limitation for the use of semiotic approaches in
archaeology: while the symbol itself, or the signifier, is usually obvious enough to the archae-
ologist, its meaning (signification) is often a matter of guesswork. The structural approach to
symbol systems (Levi-Strauss 1963) might offer a way out because its basic premise is that
there is an isomorphism between the structure of symbol systems and the structure of the
social entity which uses it, and that this isomorphism is not dependent on the specific signifi-
cation of these systems. Such an approach is more amenable to archaeological application,
and indeed there is a growing body of such applications (Renfrew 1982: 12–13; Hodder 1986:
34–54) as well as voluble objections.

In the symbolic view a culture is a language by which an individual transmits to him-
or herself and to others an identification with a social group. The most obvious example is
the realm of dress and fashion. We all can, on the evidence of someone’s dress, assess gen-
der, age, social status, religion, and the like. Corollaries from this definition of culture are:

(1) The connection between an individual and his culture is active (Hodder 1982: 12).
The individual does not absorb his culture, but transmits it. Also, this connection is cognitive;
it is a matter of choice. An individual may choose to move from one culture to another. Such
a move will usually be accompanied (if not actually accomplished by) a change in material
culture attributes—such as dress code.

(2) The message that one transmits using material culture is usually quite complex
(Hodder 1982: 75–84); as individuals usually identify themselves with several social group-
ings simultaneously (e.g., a sex group, an age group, an ethnic group, etc.). Still in the facet
of dress, young people (of both sexes) will usually adopt certain fashion attributes which
distinguish them from their elders, while other attributes serve to distinguish males from
females (regardless of age).

(3) Cultures are by nature dynamic, they change constantly, even without any external
stimulus. A social elite (in any definition of the word) will define itself by adopting “stylish”
behavioral or material characteristics. Groups identifying with such an elite will emulate that
attribute. Once that has happened that attribute will be widely distributed but will lose its
property of being a signifier for the elite. That group will then have to adopt a new “fad”
(possibly from a completely different facet) to signify who is “in.” This is exactly the pro-
cess by which slang changes formal languages.

(4) Unlike changes in the symbolic format, changes in the structure of symbolic sys-
tems are evidence of organizational change in society. A structural change in society will
compel individuals to transmit a different set of messages.

(5) An “ethnos” is defined here as a social group which defines itself with a set of symbols
(flag, national cuisine, national dress, etc.).

(6) The model, or analogy, at the base of the semiotic view of cultures is linguistic.
Language is, of course, the most basic and most complex symbolic system known to man.
It is customary therefore to refer to other symbol systems (including material culture) in lin-
guistic terms. Thus one can refer to syntax, composition, or semantic contents of a material
culture. This has actually been quite common in art history for a long time and was used by
Furumark in his analysis of Mycenaean pottery.

It is interesting to note that the symbolic paradigm returns to the normative definition for
the meaning of the spatial and chronological extent of a “material culture.” That extent marks

Long

oi.uchicago.edu



599PHILISTINE BICHROME PAINTED POTTERY

the boundaries of a “social entity,” usually interpretable in ethnic terms (Hodder 1982: 75–
84). There are, however, several important differences between the two paradigms.

A change of material culture is conceived under the symbolic paradigm as a conscious act
of self-determination. This means that immigration, or even conquest, do not in themselves
cause material culture change, unless they are accompanied by conscious effort to transplant
a new social order or a new set of symbolic messages. For example, the emigration of a small
number of English and Dutch colonists wrought far-reaching material changes in North
America. Vast numbers of north, central and east Europeans were then absorbed by that cul-
ture with hardly a trace. This was because the expressed purpose of these immigrants, as
individuals, was to be integrated into the American “melting pot” and not to change it. Sim-
ilarly, one would be hard put to point out specifically “Turkic” material attributes of the Turk-
ish conquest of Palestine, but the British occupation will be easy to identify.

Some years ago (Sharon 1992) I resorted to some elements of semiotic theory in order to
explain away the appearance of the bichrome “Philistine” pottery (as well as the appearance
of “Philistines” as an entity in the biblical narratives) at least a generation later than the pos-
tulated settlement of Sea Peoples in Canaan. I return to this explanation below.

A Comparison of the Different Definitions of Culture

Before coming back to the Philistines let us briefly review the differences between the
different definitions of “culture”:

(1) The functional approach does not see a connection between ethnic identities and
material culture while the normative and the symbolic approaches do.

(2) The normative paradigm sees the connection between the individual and his/her
culture as passive, while the functional and symbolic see it as active.

(3) Under the normative paradigm, the ties between an individual and a culture are
rigid. An observed movement of the culture is taken to mean the physical transhumance of
its participants. Both other paradigms can accept cases where the participants remain static
but change their material culture.

(4) The normative view sees culture as basically conservative (i.e. it tends to resist
change); the functional view sees culture as elastic (it adapts itself to external pressures);
while the symbolic view sees it as dynamic (it changes even without external impulses).

(5) Normative archaeology concentrates its research on the interaction between cultures;
functional archaeology concentrates on the interaction between a culture and its environment;
and systemic archaeology is mainly interested in the interaction of various components within
the culture.

(6) The analogies behind the normative and the functional paradigms are taken from
biology: one from the systemic paradigm in engineering and the one underlying the symbolic
paradigm from linguistics.

(7) The general motivation for human behavior under the normative approach is psycho-
analytical (cultural taboos are located in the deep subconscious); the functional approach
is behavioristic (e.g., Schiffer 1976); and the symbolic approach is related to cognitive
psychology.

(8) The definition of “ethnicity” most often used by the normative paradigm is the gene-
alogical one. This was also the definition used by “classic” nationalism of the nineteenth
century. The definition used by functionalists (if they use any at all) is based on common

oi.uchicago.edu



600 ILAN SHARON

destiny or circumstances and the one favored by the symbolic paradigm is based on self-
determination. In this (as in other things) archaeological theory fits itself into the accepted
ideology of the second half of the twentieth century.

A SYMBOLIC JUSTIFICATION FOR THE ADOPTION OF THE BPh STYLE

A Reexamination of the “Philistine Equation”

As argued above, the definition of Philistine culture as established at the beginning of the
century and consolidated until the 1970s set the “basic equation” as:

prstw (of Egyptian literature) = Philistines (of the Bible) = BPh pottery

Criticisms of this view (Desborough 1964; McClellan 1979; Bunimovitz 1990) tended hith-
erto to concentrate on the right part of this equation—the identity of the BPh pottery with the
Philistines. I suggest leaving this part of the equation intact. I see no reason to doubt that the
Philistines of the Bible, the ones Saul and David fought (historically or figuratively), were
indeed the people who manufactured the BPh pottery. Both chronological and geographical
parameters seem to fit. Moreover, I would tend to dispute the amalgamation of the Myc IIIC
and the BPh styles, and the view of the latter as the “true” Philistine pottery (Bunimovitz
1986), or of both as representing the same phenomenon (Stager 1995), hence the disinclina-
tion to use the term “Philistine Monochrome.” Instead, I intend to take a closer look at the
left identity in that equation and ask, are the Sea Peoples really the same as the Philistines of
the Bible?

So as to remove doubt at the outset, I am not trying to argue here that there is no
connection between BPh pottery or its manufacturers and the Sea Peoples. The question is,
is this connection one of isomorphism, or is it something subtly different? In reviewing
material culture in Early Iron Age Philistia, it seems to me that we are dealing with two
different entities:

Some time after the beginning of the reign of Ramesses III (and after the destruction
of Lachish Stratum VI) Aegean immigrants settle among the Canaanites at some sites in the
southern coastal strip. They manufacture local Myc IIIC pottery. This pottery is exclusively
Aegean in its technique and symbolic contents and is quite limited in its geographical distri-
bution. Except for the occasional stray, it is confined to a few core sites where the new immi-
grants have settled. Even in these sites, regular Canaanite pottery is being manufactured
along with Myc IIIC. Up to this point, the model offered by Bunimovitz seems to be opera-
tive. We have two cultures coexisting. The new immigrants identify themselves as foreigners
(whether as prstw or w·ssw or t·krw or whatever) and they bolster that self-determination by
exact emulation of the Mycenaean pottery. Meanwhile, the Canaanite population goes on
making its own local pottery and displaying its own symbolic array.

At the end of the Twentieth Dynasty Egyptian control of southern Palestine is relin-
quished with it, the Bronze Age social order comes to an end and local populations go into
an acute identity crisis. Out of this period of turmoil emerges a new ethnic identity. It calls
itself (if we are to believe the biblical account) “Philistine.” It is no longer prstw or w·ssw

oi.uchicago.edu



601PHILISTINE BICHROME PAINTED POTTERY

or t·krw nor is it “Canaanite.” It displays its colors (pun intended) by a new bichrome pot-
tery, which retains an Aegean flavor but is equally consciously not Mycenaean. It is much
wider both in its geographic extent and in the semantics of its symbol set. The fact that this
new symbolic array appears at this very point in time is thus neither due to the culmination
of a long process of assimilation, nor to a new “wave” of settlement, nor yet to itinerant pot-
ters from Cyprus or elsewhere setting up shop in Philistia.

The appearance of BPh pottery is not coincidental with the end of Egyptian domination
of Canaan (as it would be under any of the above explanations) but is a response to it (or
both of these happenings are symptoms of one underlying process, which is much the same
thing). In other regions of Canaan, the same crisis promotes the self-determination of other
new ethnic identities, calling themselves (or being called by others) “Israelites” or “Ara-
means.” The conjoint appearance of these ethnic identities is nothing but fortuitous. For this
part of the process I can agree with Iacovou’s (1998: 338) assessment that the new material
dialects constitute a “systematic rejection of the Late Bronze Age” (and hence the insistence
on a “foreign” origin, real or mythic) on the one side, and that on the other “the Philistines
constitute the ‘other’ against which the Israelites came together and asserted their ethnic
identity.” In her recent writings Gilboa (see chapter 9, this volume) argues for a transforma-
tion in Phoenician material culture at the same time (as against the common view of them
as merely “Iron Age Canaanites”). Interestingly enough, that evolution also constitutes the
adoption of a “foreign” (Cypriot) stylistic idiom.

If this, or any similar, reconstruction is at all correct, than the search for a “Philistine”
language, “Philistine” religion (Dothan and Dothan 1992: 156–57), or “non-Philistine Sea
Peoples” (Dothan and Dothan 1992: 153, 186–87, 258) is futile at the outset. The true Sea
Peoples are a phenomenon only one generation long. Aegeans (of Mycenaean, Cypriot, or
other descent) had come into Canaan, together perhaps with immigrants from other places,
seeking their fortune in the cosmopolitan climate of Canaanite city-states at the end of the
Late Bronze Age. It is highly unlikely that they traveled in large hordes, much less in disci-
plined military complements. They more likely came individually or in families, as traders,
sea captains, itinerant craftsmen, mercenaries, or pirates. A good many of them doubtless
tried all of the above at one time or another. By the beginning of the Iron Age, their daughters
or grandsons, most probably by mixed marriage, identified themselves (to their own and to
their neighbors, including the biblical narrator) as a new polity—the Philistines.

REFERENCES

Albright, W. F.
1932 An Anthropoid Clay Coffin from Sahab in Transjordan. American Journal of Archaeology

36: 295–306.
1932–43 The Excavation of Tell Beit Mirsim in Palestine. Annual of the American Schools of Oriental

Research Vols. 12–13, 17, 21–22. New Haven: American Schools of Oriental Research.
1940 From Stone Age to Christianity—Monotheism and the Historical Process. Baltimore: Johns

Hopkins Press.

Artzy, M.
1995 Nami: A Second Millennium International Trading Center in the Mediterranean. Pp. 17–40

in Recent Excavations in Israel: A View to the West, ed. S. Gitin, Archaeological Institutes
of America Colloquia and Conference Papers. Dubuque, Iowa: Kendall/Hunt.

oi.uchicago.edu



602 ILAN SHARON

1997 Nomads of the Sea. Pp. 1–16 in Res Maritima: Cyprus and the Eastern Mediterranean from
Prehistory to Late Antiquity, eds. S. S. Swiney, R. L. Hohlfelder, and H. W. Swiney. Atlanta:
Scholars Press.

1998 Routes, Trade, Boats and ‘Nomads of the Sea.’ Pp. 439–48 in Mediterranean Peoples in Tran-
sition: 13th to Early 10th c. B.C.E., eds. S. Gitin, A. Mazar, and E. Stern. Jerusalem: Israel
Exploration Society.

Asaro, F.; Perlman, I.; and Dothan, M.
1971 An Introductory Study of Mycenean IIIc:1 Ware from Tel Ashdod. Archaeometry 13: 169–75.

Bahn, P. G.
1996 The Cambridge Illustrated History of Archaeology. Cambridge: Cambridge University Press.

Bauer, A. A.
1998 Cities of the Sea: Maritime Trade and the Origin of the Philistine Settlement in the Early Iron

Age Southern Levant. Oxford Journal of Archaeology 17: 149–68.

Benson, J. L.
1961 A Problem in Orientalizing Cretan Birds: Mycenaean or Philistine Prototypes? Journal of Near

Eastern Studies 20: 73–84.

Bernal, M.
1987 Black Athena: The Afroasiatic Roots of Classical Civilization, vol. 1. New Brunswick, NJ:

Rutgers University Press.

Blegen, C. W., and Haley, J.
1928 The Coming of the Greeks. American Journal of Archaeology 32: 141–54.

Bliss, F. J., and Macalister, R. A. S.
1902 Excavations in Palestine during the Years 1898–1900. London: Palestine Exploration Fund.

Braudel, F.
1972 The Mediterranean and the Mediterranean World in the Age of Philip II. [Paris 1949]; trans.

from French. New York: Harper and Row.
1980 History and the Social Sciences—The Longue Dureé. Pp. 25–54 in On History, ed. F. Braudel.

Chicago: University of Chicago Press.

Breasted, J. H.
1906 Ancient Records of Egypt. London: Histories and Mysteries of Man.

Brug, J. F.
1985 A Literary and Archaeological Study of the Philistines. British Archaeological Reports, Inter-

national Series 265. Oxford: British Archaeological Reports.

Bunimovitz, S.
1990 Problems in the “Ethnic” Identification of the “Philistine” Material Culture. Tel Aviv 17: 210–22.
1994 Socio-Political Transformations in the Central Hill Country in the Late Bronze-Iron I Transition.

Pp. 179–202 in From Nomadism to Monarchy: Archaeological and Historical Aspects of Early
Israel, eds. I. Finkelstein and N. Naªaman. Jerusalem: Yad Izhak Ben-Zvi.

1998 Sea Peoples in Cyprus and Israel. Pp. 103–13 in Mediterranean Peoples in Transition: 13th
to Early 10th c. B.C.E., eds. S. Gitin, A. Mazar, and E. Stern. Jerusalem: Israel Exploration
Society.

Catling, H.
1973 The Achaean Settlement of Cyprus. Pp. 34–40 in Acts of the International Archaeological Sym-

posium: The Mycenaeans in the Eastern Mediterranean. Nicosia: Department of Antiquities.

Champollion, J. F.
1836 [–1841] Gramaire égyptienne, ou Principes généraux de l’écriture sacrée égyptienne appliquée à la

representation de la langue parlée. Paris: Firmin Didot Frères.

Childe, V. G.
1925 The Dawn of European Civilization. London: Kegan Paul.
1926 The Aryans: A Study of Indo-European Origins. London: Kegan Paul.
1929 The Danube in Prehistory. Oxford: Oxford University Press.
1964 What Happened in History. 2nd ed. [original 1942]. Harmondsworth: Penguin.

oi.uchicago.edu



603PHILISTINE BICHROME PAINTED POTTERY

1965 Man Makes Himself. 4th ed. [original 1936]. London: Watts.

Clarke, D. L.
1968 Analytical Archaeology. London: Methuen.

Desborough, V. R. d’A.
1964 The Last Mycenaeans and Their Successors: An Archaeological Survey, c. 1200 — c. 1000 b.c.

Oxford: Clarendon Press.

Dever, W. G.
1981 The Impact of “New Archaeology” on Syro-Palestinian Archaeology. Bulletin of the American

Schools of Oriental Research 242: 15–29.

Dickinson, O.
1994 The Aegean Bronze Age. Cambridge: Cambridge University Press.

Dothan, M.
1971 Ashdod II–III: The Second and Third Seasons of Excavations, 1963, 1965; Soundings in 1967.

ºAtiqot (English Series) 9–10. Jerusalem: Department of Antiquities and Museums.

Dothan, M., and Freedman, D. N.
1967 Ashdod I. ºAtiqot 7. Jerusalem: Department of Antiquities and Museums.

Dothan, M., and Porat, Y.
1982 Ashdod IV. ºAtiqot 15. Jerusalem: Department of Antiquities and Museums.

Dothan, T.
1967 The Philistines and Their Material Culture. Jerusalem: Bialik Institute (Hebrew).
1979 Excavations at the Cemetery of Deir el-Balah. Qedem. Monographs of the Institute of

Archaeology 10. Jerusalem: Hebrew University.
1982 The Philistines and Their Material Culture. New Haven: Yale University Press.
1989 The Arrival of the Sea Peoples: Cultural Diversity in Early Iron Age Canaan. Pp. 1–14 in

Recent Excavations in Israel: Studies in Iron Age Archaeology, eds. S. Gitin and W. Dever,
Annual of the American Schools of Oriental Research 49. New Haven: American Schools of
Oriental Research.

1992 Social Dislocation and Cultural Change in the Twelfth Century b.c.e. Pp. 93–98 in The Crisis
Years: The Twelfth Century b.c.—From Beyond the Danube to the Tigris, eds. W. A. Ward and
M. S. Joukowsky. Dubuque, IA: Kendall/Hunt.

1998 Initial Philistine Settlement: From Migration to Coexistence. Pp. 148–61 in Mediterranean Peo-
ples in Transition: 13th to Early 10th c. B.C.E., eds. S. Gitin, A. Mazar, and E. Stern. Jerusalem:
Israel Exploration Society.

Dothan, T., and Dothan, M.
1992 People of the Sea: The Search for the Philistines. New York: Macmillan.

Dothan, T., and Dunayevsky, I.
1978 Qasile, Tel. Pp. 963–68 in Encyclopedia of Archaeological Excavations in the Holy Land,

vol. 4, eds. M. Avi-Yonah and E. Stern. Jerusalem: Massada Press.

Edelman, D.
1991 Toward a Consensus on the Emergence of Israel in Canaan: Introduction. Swedish Journal

of Old Testament 2: 3–6.

Epstein, S. M.
1987 Scholars Will Call It Nonsense—The Structure of Erich von Däniken’s Argument. Expedition

29: 13–18.

Esse, D. L.
1991 The Collared Store Jar—Scholarly Ideology and Ceramic Typology. Swedish Journal of Old

Testament 2: 99–116.
1992 The Collared Pithos at Megiddo: Ceramic Distribution and Ethnicity. Journal of Near Eastern

Studies 51: 81–103.

Feyerabend:
1978 Against Method: Outline of an Anarchistic Theory of Knowledge. 2nd ed. London: Verso.

oi.uchicago.edu



604 ILAN SHARON

Finkelstein, I.
1994 The Emergence of Israel: A Phase in the Cyclic History of Canaan in the Third and Second

Millennia bce. Pp. 150–78 in From Nomadism to Monarchy: Archaeological and Historical
Aspects of Early Israel, eds. I. Finkelstein and N. Naªaman. Jerusalem: Yad Izhak Ben-Zvi.

1995a The Great Transformation: The “Conquest” of the Highland Frontiers and the Rise of Terri-
torial States. Pp. 349–62 in The Archaeology of Society in the Holy Land, ed. T. E. Levy.
London: Leicester University Press.

1995b The Date of the Settlement of the Philistines in Canaan. Tel Aviv 22: 213–39.
1996 The Archaeology of the United Monarchy: An Alternative View. Levant 28: 177–87.
1998 Philistine Chronology: High, Middle or Low? Pp. 140–47 in Mediterranean Peoples in Tran-

sition: 13th to Early 10th c. B.C.E., eds. S. Gitin, A. Mazar, and E. Stern. Jerusalem: Israel
Exploration Society.

Finkelstein, I.; Ussishkin, D.; and Halpern, B., eds.
2000 Megiddo III: The 1992–1996 Seasons, Vols. I–II. Tel Aviv: Emery and Claire Yass Publications

in Archaeology, Institute of Archaeology, Tel Aviv University.

Furtwängler, A., and Loeschcke, G.
1886 Mykenische Vasen: Vorhellenische Thongefässe aus dem Gebiete des Mittelmeeres. Berlin:

Asher.

Furumark, A.
1941a The Mycenaean Pottery, Analysis and Classification. Stockholm: Kungl. Vitterhets Historie och

Antikvitets Akademien.
1941b The Chronology of Mycenaean Pottery. Stockholm: Kungl. Vitterhets Historie och Antikvitets

Akademien.

Gilboa, A.
1989 New Finds at Tel Dor and the Beginnings of Cypro-Geometric Pottery Import to Palestine. Israel

Exploration Journal 39: 204–18.
1999a The View from the East: Tel Dor and the Earliest Cypro-Geometric Exports to the Levant.

Pp. 119–39 in The Historicity of the Cypro-Geometric Horizon, eds. D. Michaelides and
M. Iacovou. Nicosia: University of Cyprus.

1999b The Dynamics of Phoenician Bichrome Style: A View from Tel Dor. Bulletin of the American
Schools of Oriental Research 316: 2–22.

Goldwasser, O.
1984 Hieratic Inscriptions from Tel Seraª in Southern Canaan. Tel Aviv 11: 77–93.

Gunneweg, J.; Perlman, I.; Dothan, T.; and Gitin, S.
1986 On the Origin of Pottery from Tel Miqne/Ekron. Bulletin of the American Schools of Oriental

Research 264: 3–16.

Hall, H. R.
1928 The Civilization of Greece in the Bronze Age: The Rhind Lectures, 1923. London: Methuen.

Hankey, V.
1966 Late Mycenaean Pottery at Beth-Shan. American Journal of Archaeology 70: 169–71.

Harris, M.
1968 The Rise of Anthropological Theory. New York: Crowell.

Hempel, K. G.
1965 Aspects of Scientific Explanation and Other Essays in the Philosophy of Science. New York:

Free Press.

Heurtley, W. A.
1936 The Relationship between “Philistine” and Mycenaean Pottery. Quarterly of the Department

of Antiquities in Palestine 5: 90–110.

Hitzig, F.
1845 Urgeschichte und Mythologie der Philistäer. Leipzig: Weidmann.

Hodder, I.
1982 Symbols in Action. Cambridge: Cambridge University Press.

oi.uchicago.edu



605PHILISTINE BICHROME PAINTED POTTERY

1986 Reading the Past: Current Approaches to Interpretation in Archaeology. Cambridge: Cambridge
University Press.

Hood, S. M. F.
1973 Mycenaean Settlement in Cyprus and the Coming of the Greeks. Pp. 40–50 in Acts of the Inter-

national Archaeological Symposium: The Mycenaeans in the Eastern Mediterranean. Nicosia:
Department of Antiquities.

Humbert, J.-B.
1993 Tel Keisan. Pp. 862–67 in The New Encyclopedia of Archaeological Excavations in the Holy

Land, eds. E. Stern and A. Lewinson-Gilboa. Jerusalem: Israel Exploration Society & Carta;
New York: Simon & Schuster.

Iacovou, M.
1998 Philistia and Cyprus in the Eleventh Century: From Similar Prehistory to Diverse Protohistory.

Pp. 332–44 in Mediterranean Peoples in Transition: 13th to Early 10th c. B.C.E., eds. S. Gitin,
A. Mazar, and E. Stern. Jerusalem: Israel Exploration Society.

Karageorghis, V.
1992 The Crisis Years: Cyprus. Pp. 79–85 in The Crisis Years: The Twelfth Century BC—From

Beyond the Danube to the Tigris, eds. W. A. Ward and M. S. Joukowsky. Dubuque, Iowa:
Kendall/Hunt.

1994 The Prehistory of an Ethnogenesis. Pp. 1–7 in Cyprus in the Eleventh Century BC, ed.
V. Karageorghis. Nicosia: Leventis Foundation.

Khazanov, A. M.
1983 Nomads and the Outside World. Cambridge: Cambridge University Press.

Killebrew, A.
1998 Ceramic Typology and Technology of Late Bronze II and Iron I Assemblages from Tel

Miqne–Ekron: The Transition from Canaanite to Philistine Culture. Pp. 379–405 in Mediter-
ranean Peoples in Transition: 13th to Early 10th c. B.C.E., eds. S. Gitin, A. Mazar, and
E. Stern. Jerusalem: Israel Exploration Society.

Kling, B.
1989 Mycenaean IIIC:1b and Related Pottery in Cyprus. Gothenburg: Paul Åström Förlag.

Knapp, A. B.
1989 Response: Independence, Imperialism, and the Egyptian Factor. Bulletin of the American

Schools of Oriental Research 275: 64–68.
1993 Society and Polity at Bronze Age Pella: An “Annales” Perspective. Sheffield: Sheffield Aca-

demic Press.

Kossina, G.
1911 Die Herkunft der Germanen: Zur Methode der Siedlungsarchäologie. Leipzig: Kabitzsch.

Kuhn, T. S.
1962 The Structure of Scientific Revolutions. Chicago: University of Chicago Press.

Levi-Strauss, C.
1963 Structural Anthropology. New York: Basic Books.

Loud, G.
1939 The Megiddo Ivories. Oriental Institute Publications 52. Chicago: University of Chicago Press.
1948 Megiddo II: Seasons of 1935–1939. Oriental Institute Publications 62. Chicago: University of

Chicago Press.

Macalister, R. A. S.
1912 The Excavation of Gezer, 1902–1905 and 1907–1909. London: Palestine Exploration Fund.

Macdonald, E.; Starkey, J. L.; and Harding, L.
1932 Beth Pelet II. London: British School of Archaeology in Egypt.

Mackenzie, D.
1912 Excavations at Ain Shems (Beth Shemesh). London: Palestine Exploration Fund.

Maier, A. M., and Ehrlich, C.
In press Tel Zafit–1999. Hadashot Arkhe’ologiyot 112.

oi.uchicago.edu



606 ILAN SHARON

Mazar, A.
1980 Excavations at Tell Qasile. Part 1: The Philistine Sanctuary: Architecture and Cult Objects.

Qedem Monographs of the Institute of Archaeology 12. Jerusalem: Hebrew University.
1985a Excavations at Tell Qasile. Part 2: The Philistine Sanctuary: Various Finds, The Pottery,

Conclusions, Appendixes. Qedem Monographs of the Institute of Archaeology 20. Jerusalem:
Hebrew University.

1985b The Emergence of the Philistine Material Culture. Israel Exploration Journal 35: 95–107.
1988 Some Aspects of the Sea People’s Settlement. Pp. 251–60 in Society and Economy in the

Eastern Mediterranean (c. 1500–1000 bce), eds. M. Heltzer and E. Lipinski. Orientalia
Louvaniensia Analecta 23. Leuven: Uitgeverij Peeters.

1990 Archaeology of the Land of the Bible 10,000–586 bce. Anchor Bible Reference Library.
New York: Doubleday.

1993 Beit Shean and the Iron Age: Preliminary Report and Conclusions of the 1990–1991
Excavations. Israel Exploration Journal 43: 201–29.

1994 Beit Shan. Pp. 485, 488–89, 493, 498 in Archaeology in Israel, by S. R. Wolff. American
Journal of Archaeology 98: 481–519.

1997 Iron Age Chronology: A Reply to Israel Finkelstein. Levant 29: 157–67.

Mazar, B.
1951 The Excavations of Tell Qasile, Preliminary Report. Israel Exploration Journal 1: 61–76;

125–40; 194–218.

McClellan, T. L.
1979 Chronology of the “Philistine” Burials at Tell el-Farah (S). Journal of Field Archaeology 6:

57–79.

Muhly, J. D.
1998 Copper, Tin, Silver and Iron: The Search for Metallic Ore as an Incentive for Foreign

Expansion. Pp. 314–29 in Mediterranean Peoples in Transition: 13th to Early 10th c. B.C.E.,
eds. S. Gitin, A. Mazar, and E. Stern. Jerusalem: Israel Exploration Society.

Myers, J. L.
1911 The Dawn of History. London: William and Norgate.

Oren, E. D.
1984 “Governor’s Residencies” in Canaan under the New Kingdom: A Case Study of Egyptian

Administration. Journal of the Society for the Study of Egyptian Antiquities 14: 35–56.
1993a Haror, Tel. Pp. 580–84 in The New Encyclopedia of Archaeological Excavations in the Holy

Land, ed. E. Stern. Jerusalem: Israel Exploration Society & Carta and New York: Simon &
Schuster.

1993b Seraª, Tel. Pp. 1329–35 in The New Encyclopedia of Archaeological Excavations in the Holy
Land, ed. E. Stern. Jerusalem: Israel Exploration Society & Carta and New York: Simon &
Schuster.

Oren, E., ed.
2000 The Sea Peoples and Their World: A Reassessment. Philadelphia: The University Museum,

University of Pennsylvania Press.

Perry, W. J.
1923 The Children of the Sun. London: Methuen.

Petrie, W. F.
1928 Gerar. London: British School of Archaeology in Egypt.

Petrie, W. F., and Tufnell, O.
1930 Beth Pelet I (Tell Fara). London: British School of Archaeology in Egypt.

Phythian-Adams, W. J.
1923 Philistine Origins in the Light of Palestinian Archaeology. Bulletin of the British School of

Archaeology in Jerusalem 3: 20–27.

Popper, K.
1959 The Logic of Scientific Discovery. London: Hutchinson.
1969 Conjectures and Refutations. 3rd ed. London: Routledge and Kegan Paul.

oi.uchicago.edu



607PHILISTINE BICHROME PAINTED POTTERY

Portugali, J.
1994 Theoretical Speculations on the Transition from Nomadism to Monarchy. Pp. 203–217 in

From Nomadism to Monarchy: Archaeological and Historical Aspects of Early Israel, eds.
I. Finkelstein and N. Naªaman. Jerusalem: Yad Izhak Ben-Zvi.

Rauh, N. K.
1997 Who Were the Cilician Pirates? Pp. 263–83 in Res Maritima: Cyprus and the Eastern

Mediterranean from Prehistory to Late Antiquity, eds. S. S. Swiney, R. L. Hohlfelder, and
H. W. Swiney. Atlanta: Scholars Press.

Rediker, M. B.
1987 Between the Devil and the Deep Blue Sea: Merchants, Seamen, Pirates, and the Anglo-American

Maritime World 1700–1750. Cambridge: Cambridge University Press.
Renfrew, C.

1982 Explanation Revisited. Pp. 5–23 in Theory and Explanation in Archaeology, eds. C. Renfrew,
M. J. Rowlands, and B. Abbot-Seagraves. New York: Academic Press.

Rougé, E. de
1867 Extraits d’un mémoire sur les attaques dirigées contre l’Égypte par les peuples de la Méditer-

ranée vers le quatorzième siècle avant notre ère. Revue archéologique 1867: 1–36.
Rowton, M. B.

1976 Chronology. II. Ancient Western Asia. Pp. 193–239 in The Cambridge Ancient History I/1,
3rd ed., eds. I. E. S. Edwards, C. J. Gadd, and N. G. L. Hammond. Cambridge: Cambridge
University Press.

Sandars, N. K.
1963 Later Aegean Bronze Swords. American Journal of Archaeology 67: 140.
1978 The Sea Peoples: Warriors of the Mediterranean: 1250–1150 b.c. London: Thames and Hudson.

Scheepers, A.
1991 Anthroponymes et toponymes du récit d’Ounamon. Pp. 17–84 in Phoenicia and the Bible: Pro-

ceedings of the Conference held at the University of Leuven on the 15th and 16th of March
1990, ed. E. Lipinski. Studia Phoenicia 11. Orientalia Louvaniensia Analecta 44. Leuven:
Peeters.

Schiffer, M. B.
1976 Behavioral Archaeology. New York: Academic Press.

Sharon, I.
1992 Approaches to the Relation Between “Material Culture” and “Culture” and the Question of

“Philistine Ethnicity.” Pp. 8–9 in Eighteenth Archaeological Conference in Israel, Abstracts.
(Tel Aviv University, 12–13 February 1992). Jerusalem: Israel Exploration Society and Israel
Antiquities Authority. (Hebrew).

1994 Demographic Aspects of the Problem of the Israelite Settlement. Pp. 119–34 in Uncovering
Ancient Stones—Essays in Memory of H. Neil Richardson, ed. L. Hopfe. Winona Lake, IN:
Eisenbrauns.

Sheratt, A., and Sheratt, S. E.
1991 From Luxuries to Commodities: The Nature of Mediterranean Bronze Age Trading Systems.

Pp. 351–86 in Bronze Age Trade in the Mediterranean: Papers Presented at the Conference held
in Rewley House, Oxford, in December 1989, ed. N. Gale. Jonsered: Paul Åström Förlag.

Sheratt, S.
1994 Commerce, Iron and Ideology: Metallurgical Innovation in 12th–11th Century Cyprus. Pp. 59–

107 in Proceedings of the International Symposium “Cyprus in the 11th Century B.C.,” ed.
V. Karageorghis. Nicosia: Leventis Foundation.

1998 Sea Peoples and the Economic Structure of the Late Second Millennium in the Eastern Medi-
terranean. Pp. 292–313 in Mediterranean Peoples in Transition: 13th to Early 10th c. B.C.E.,
eds. S. Gitin, A. Mazar, and E. Stern. Jerusalem: Israel Exploration Society.

Singer, I.
1988 The Origins of the Sea Peoples and Their Settlement on the Coast of Canaan. Pp. 239–50

in Society and Economy in the Eastern Mediterranean (c. 1500–1000 bce), eds. M. Heltzer
and E. Lipinski. Orientalia Louvaniensia Analecta 23. Leuven: Uitgeverij Peeters.

oi.uchicago.edu



608 ILAN SHARON

1994 Egyptians, Canaanites and Philistines in the Period of the Emergence of Israel. Pp. 282–338
in From Nomadism to Monarchy: Archaeological Aspects of Early Israel, eds. I. Finkelstein
and N. Na’aman. Jerusalem: Yad Izhak Ben-Zvi.

Smith, G. E.
1923 The Ancient Egyptians and the Origin of Civilization. London: Harper.
1933 The Diffusion of Culture. London: Watts.

Stager, L. E.
1985 Merneptah, Israel, and the Sea Peoples. Eretz-Israel 18: 56*–64*.
1995 The Impact of the Sea Peoples in Canaan. Pp. 332–48 in The Archaeology of Society in the

Holy Land, ed. T. E. Levy. London: Leicester University Press.

Stark, K. B.
1852 Gaza und die Philistäische Küste. Jena: Friedrich Mauke.

Stern, E.
1982 The Material Culture of the Land of the Bible in the Persian Period, 538–332 bc. Warminster:

Aris and Phillips.
1994 Dor—Ruler of the Seas. Jerusalem: Israel Exploration Society.

Trigger, B. G.
1989 A History of Archaeological Thought. Cambridge: Cambridge University Press.

Tufnell, O.
1953 Lachish III: The Iron Age. London: Oxford University Press.

Tufnell, O.; Baden-Powell, D. F. W.; Bate, D. M. A.; Cerny, J.; Diringer, D.; Giles, M.; Helbaek, H.; Isserlin,
B. S. J.; Murray, M. A.; Parker, B.; Porada, E.; Thompson, F. C.; Todd, E.; and Waecher, J.

1958 Lachish IV: The Bronze Age. London: Oxford University Press.

Tufnell, O.; Inge, C. H.; and Harding, L.
1940 Lachish II: The Fosse Temple. London: Oxford University Press.

Ussishkin, D.
1985 Level VII and VI at Tel Lachish and the End of the Late Bronze Age in Canaan. Pp. 213–30

in Palestine in the Bronze and Iron Ages—Papers in Honour of Olga Tufnell, ed. J. N. Tubb.
London: Institute of Archaeology.

1995 The Destruction of Megiddo at the End of the Late Bronze Age and Its Historical Significance.
Tel Aviv 22: 240–67.

1998 The Destruction of Megiddo at the End of the Late Bronze Age and Its Historical Significance.
Pp. 197–219 in Mediterranean Peoples in Transition: 13th to Early 10th c. B.C.E., eds. S. Gitin,
A. Mazar, and E. Stern. Jerusalem: Israel Exploration Society.

Watson, J.; LeBlanc, S. A.; and Redman, C. L.
1971 Explanation in Archaeology. New York: Columbia University Press.

Welch, F. B.
1900 The Influence of the Aegean Civilization on South Palestine. Palestine Exploration Fund Quar-

terly Statement (1900): 342–50.

Wente, E. F., and van Siclen III, C. C.
1976 A Chronology of the New Kingdom. Pp. 217–61 in Studies in Honor of George R. Hughes,

eds. J. H. Johnson and E. F. Wente. Studies in Ancient Oriental Civilization 39. Chicago:
University of Chicago Press.

White, L. A.
1959 The Evolution of Culture. New York: McGraw-Hill.

Wilson, J. A.
1950 Egyptian Historical Texts. Pp. 227–64 in Ancient Near Eastern Texts Relating to the Bible,

ed. J. B. Pritchard. Princeton: Princeton University Press.

Wolff, S. R.
1994 Archaeology in Israel. American Journal of Archaeology 98: 481–519.

oi.uchicago.edu



609PHILISTINE BICHROME PAINTED POTTERY

Wylie, A.
1985 Between Philosophy and Archaeology. American Antiquity 50: 478–90.

Yadin, Y.
1968 And Dan Why Did He Remain in Ships? Australian Journal of Biblical Archaeology 1: 9–23.

Yadin, Y., and Geva, S.
1986 Investigations at Beth-Shean: The Early Iron Age Strata. Qedem Monographs of the Insti-

tute of Archaeology 23. Jerusalem: Hebrew University.

oi.uchicago.edu



610 ILAN SHARON

oi.uchicago.edu



611



THUNDERING HORDES: THE IMAGE OF THE
PERSIAN AND MUSLIM CONQUESTS IN

PALESTINIAN ARCHAEOLOGY



Neil Asher Silberman



Throughout his brief yet highly productive career, Douglas Esse skillfully showed all of
us—friends, colleagues, and students—how powerful the intellectual weapons of gentle com-
mon sense, good-humored skepticism, and unshakable ideological independence can be. Esse
entered the profession at a time when old formulations of biblical archaeology were crumbling.
Yet it was also a time when the critics of the traditional order had not yet formulated a fully
workable alternative. Esse’s aim was not sloganeering or rhetorical one-upmanship; he was
deeply fascinated in understanding the



processes



of ancient history. In his important, path-
breaking on the Early Bronze Age and the “problem” of the Israelite settlement, Esse demon-
strated that it is possible to combine new field data with that collected by earlier generations,
once the older material data is reanalyzed and reshaped by new theoretical frameworks for
understanding social and cultural change (Esse 1991, 1992). As many of the contributors of this
volume acknowledge in connection with their own contributions relating to various historical
and archaeological periods, Esse effectively challenged conventional wisdom about the me-
chanics of cultural transmissions and transitions. He offered us all compelling evidence that
discontinuities observed in the ancient material records need not simply be ascribed to sudden
invasions or migrations of extraregional ethnic groups.

As Esse was well aware, this ethnic hypothesis of culture change has far-reaching modern
implications and subtle social side effects. It has become increasingly clear that in retroject-
ing modern perceptions of ethnicity and nationhood onto the material remains of premodern
societies, archaeological interpretations can sometimes become unwitting instruments of
present-day ethnic politics (see, e.g., the recent essays in Kohl and Fawcett 1995). As potent,
pseudo-scientific metaphor, this kind of archaeological thinking has had far-reaching impli-
cations all over the world for the understanding of every historical period. And even though
the Early Islamic period was not among Esse’s primary areas of interest, I would like to take
this opportunity to examine possible social and political aspects of the study of the Persian
and Muslim conquests of Palestine, which may have some illustrative relevance. For like the
traditional understandings of the Early Bronze Age and the Early Iron Age that Esse worked
so hard to render more complex and sophisticated, the interpretation of the Early Islamic
period can be seen as an aspect of wider intellectual trends. Thus in this paper, in tribute to
Esse’s memory, I attempt to place the history of archaeological research on the “problem” of
the Persian and Muslim conquests into ideological and political perspective—by briefly sur-
veying the evolution and transformation of Western images of the fall of Byzantine rule over
Palestine and of the rise of Islam.



30



oi.uchicago.edu





612



NEIL ASHER SILBERMAN



In recent years, scholars in many branches of the humanities have devoted considerable
attention to studying the mechanics of image making, ideology, rhetoric, and narrative and on
the importance of image making, rhetoric, and narrative as powerful tools for communicating
social understandings (of many, see Berger and Luckmann 1967; Lincoln 1989). The study of
narrative in archaeology, in particular, has proved useful for understanding the implicit mes-
sages and social biases of scholarly work—of both the unwitting and premeditated kinds (Lan-
dau 1991; Trigger 1984; Silberman 1995). Faced with the fragmentary remains of ash layers,
collapsed walls, and diagnostic pottery types, archaeologists have often uncritically applied
modern cognitive categories of gender, ethnicity, and political action to otherwise highly
ambiguous archaeological remains (Leone, Potter, and Shackel 1987; Patterson 1995). As I will
argue below, this has been a particularly influential process in the accepted historiography of
the Early Islamic period, which has been seen since the beginnings of modern archaeology in
Israel as one of the liminal periods of the country’s history, in a sense distinguishing that which
is important and worthy of intensive study from that which is not (Silberman 1989, 1990, 1991).

To be sure, long before archaeology ever entered the picture, ideologically based images
of the Persian and Muslim conquests of the Holy Land were vivid and influential throughout
the Christian world. From the seventh century onward, those images served as powerful soci-
etal narratives (Wilken 1992: chaps. 11, 12). The sudden loss of Byzantine political control
over the Holy Land—in its painful contradiction to the belief in the inevitable triumph of
Christianity—required certain fundamental explanations that did not challenge the Christian
society’s basic ideology. So the earliest and most immediate of the responses, typified by the
works of the monk Antiochus Strategos of Mar Saba about the Persian conquest (Baras 1982:
300–13; Clermont-Ganneau 1898) and the patriarch Sophronius about the Muslim conquest
(Constantelos 1972), devoted little energy to understanding the conquests as symptoms of sec-
ular political change or social interaction as modern anthropologists might do. They framed
the events in purely metaphysical terms.

Like other plagues sent down on the world in other ages—be they Amalekites, Philistines,
or Babylonians—the arrival and victories of the Persians and the Muslims were seen as both
a divine punishment and a challenge to the righteous to regain possession of the Holy Land.
And over the subsequent centuries, as this religious interpretation was ultimately marshaled to
support the conquests of the Crusaders, the earliest European commentaries on the Quran and
speculations on Islamic history maintained the character of polemics. Focusing primarily on
the character of the founder of Islam, they provided vivid details of Muhammed’s idolatry,
black magic, and sexual promiscuity that make the latter-day descriptions in Salman Rushdie’s



Satanic Verses



actually seem quite tame. In fact, medieval Christian polemic in the form of lit-
erary narratives was highly influential in the conduct of East-West interactions both before and
after the Crusader period (Southern 1962; Daniel 1975; B. Smith 1977: 1–15; Rodinson 1987:
11–13). And these early images of Arabs and Islam have struck deep roots in the Western
consciousness. Some of Western society’s most basic stereotypes of Arabs and Muslims—as
still seen in films, cartoons, and popular novels (Terry 1985)—can be traced back to stock
characters in medieval troubadours’ ballads—characters such as the terrifying “Saracen
Giant,” the pitifully emasculated “Defeated Sultan,” and the superficially Christianized (yet
still despised) “Converted Saracen” (Daniel 1984; Metlitzki 1977: chap. 6).

With the Renaissance, however, the focus of Western historiography shifted. As various
European merchant nations sought to negotiate with the rulers of the eastern spice routes,

ONE LONGER

oi.uchicago.edu





613



THUNDERING HORDES



purely legendary elements of early Muslim history began to give way to more factual
accounts. Polemics were less useful than reliable information that might offer commercial
advantage (Rodinson 1987: 23–40). And while most of the added European information per-
tained to



contemporary



Muslim communities and to some extent perpetuated earlier theo-
logical hostility, new historical insights were also developed. Vast numbers of antique
manuscripts were collected, and the pursuit of the historical and linguistic study we now call
“Orientalism” began (Rodinson 1987: 40ff.; Said 1978: 49–73).

Yet changes



within



European society were ultimately even more instrumental than interna-
tional commerce in changing the focus of western historical study. By the time of the Enlight-
enment, older theological and political concepts were being called into question as European
scholars sought to construct a “universal” history in which stages of human progress replaced
earlier biblical concepts (Rossi 1984). Early Islam beckoned to some as an illustration of
primitive religious power, which was preferable in certain respects to more advanced (and
more corrupt) ecclesiastical hierarchies. Among the early works revising the traditional Euro-
pean attitude toward Islam and Islamic history were the anonymous



Mahomet no imposter, or
a Defense of Mahomet (



1720) and Count de Boulanvillier’s



Vie de Mahomed



(1730) which
represented a far wider intellectual trend (Daniel 1960; von Grunebaum 1966). Typical of this
new approach was the history of the early Muslim conquests included in the fifth volume of
Edward



Gibbon’s Decline and Fall of the Roman Empire



published first in 1772. In Gibbon’s
estimation—and in that of the scholars whom he followed—it was only inevitable that the cor-
rupt and decadent Christian Empire of Heraclius would give way to the forces of the vigorous
lawmaker and non-priestly prophet Mohammed, who represented (in true Enlightenment fash-
ion) a Rousseau-like vision of the “Natural Man.”

Propelled by the fire of true faith, rather than wickedness, Mohammed led a charge out of
the desert to conquer the world. This new view of the early Muslims as primitive, alluring, and
ultimately noble was far removed from the image of the horned satyr that Europeans had ini-
tially seen (Porter 1988: 129–32; Lewis 1977). And a pattern was set for a new understanding
of early Islamic history that would continue throughout the nineteenth and twentieth centu-
ries—in the works of Robertson Smith on early Arabian religion (1880, 1885, 1889), and of
Wellhausen on Islamic political structures (1902; English translation 1927). It should be noted
that Julius Wellhausen turned his attention to the historical problem of the early Islamic con-
quest as a continuation of his earlier studies of religious evolution in the biblical period. The
assumption of the early Arabs’ “primitiveness” is implicit in the work of both Wellhausen and
Robertson Smith (Kuper 1988: 83–86). This idealized image of noble primitiveness reached its
most elaborate and pervasive versions in the romantic ethnography of Doughty, Burton, and
T. E. Lawrence (see Said 1978: 226–54) and in the later political romances of John Bagot Glubb
(1963). Yet there was also a clear element of continuity with earlier images. The Western world
still saw the Muslims as basically primitive and antagonistic to settled society (Kuper 1988;
Stocking 1987). And their conquests—whether frightening or alluring—were still seen as the
sudden eruption from the desert of thundering Muslim hordes.

Through the centuries of polemic and contact, the religion and customs of the Arabs were
the main elements of attention, but toward the end of the eighteenth century, a more concrete
racial element entered the historical debate (Stepan 1982; Poliakov 1974). In the dawning age
of the great European empires in Asia and Africa, the tentative probing of individuals gave way
to an imperative to



rule



non-European people—rather than simply convert, exterminate, or

oi.uchicago.edu





614



NEIL ASHER SILBERMAN



conciliate. The expansion of European ambition into the world of Islam required scientific
rationalization, and this was found in the peculiar and unchanging racial characteristics of the
Arabs themselves (Rodinson 1987: 89–90; Said 1978: 206–7). Using such now discredited
indications as brain size, skull features, and genetically inherited “national character” (Gould
1981), European physiologists found an appropriate place for the Arabs in a rigidly hierarchi-
cal scale of human ability. As a particular branch of the Semitic “race” (delineated on both lin-
guistic and anatomical grounds), the Arabs were classified for their violence, asceticism, and
sexual promiscuity—all elements, it should be noted, that were adopted from the earlier
mythic images (Said 1978: 231–34). And as these racial characteristics were seen as biologi-
cally inherited and unchanging, they played a part in historical and linguistic speculations on
the role of the “Semitic” peoples throughout the millennia.

As one influential example of racial determinism, one may note the first chapter of Philip
K. Hitti,



History of the Arabs



(1949), “The Arabs as Semites: Arabia the Cradle of the Semitic
Race.” In it, Hitti notes that Arabian Arabs are “the best representatives of the Semitic family,
biologically, psychologically, socially, and linguistically” and that “ethnic purity is a reward of
the most ungrateful and isolated environment, such as central Arabia affords” (Hitti 1949: 8).
Modern population genetics, however, has shown the concept of ethnic or racial “purity” to be
devoid of any scientific basis (among recent works, dealing with this issue, see Kelves 1985:
129–47 and Gould 1981: 322–33).

Archaeology, in a sense, initially served to illustrate this mistaken understanding. In 1990,
the academic community marked the 100th anniversary of modern archaeology in Palestine,
and although its “founding father,” W. M. F. Petrie, did not contribute directly to the archaeo-
logical study of the Muslim conquests, his intellectual impact was nevertheless profound. As
generations of archaeology students have been taught, at Tell el-Hesi in 1890, Petrie devel-
oped the most basic modern tools of excavation: stratigraphic analysis and pottery typology.
Yet as I have attempted to show elsewhere (Silberman 1993), the way he used those tools was
not ideologically neutral. Petrie’s work must be assessed in the light of his life-long interest in
the question of race and racial warfare as the primary element in human history (Petrie 1906,
1907, 1911). In associating stratigraphic levels with episodes of ethnic dominance, represented
by the appearance of ethnically identified pottery types (arranged rise-floruit-fall cycles), Pet-
rie precisely echoed a popular turn-of-the-century colonialist dogma of the inevitable ascen-
dancy of “superior” or more ‘’vigorous” races over more passive ones. The eugenical theories
of Francis Galton (1869) clearly influenced Petrie, and through him the impact on Palestinian
archaeology was clear (see Silberman 1999, with bibliography). And it was the subtle power
of those conceptual tools, I would argue, that has furthered until fairly recently an antagonistic
and racially based archaeological interpretation of the beginning of the Early Islamic period.

Through the 1920s and 1930s, the first significant archaeological light was shed on the
Persian and Muslim conquests, as excavators encountered destruction levels at Byzantine
churches and Byzantine sites throughout the country. There was, of course, earlier archaeolog-
ical interest. Robinson, Warren, Conder and Kitchener, Clermont-Ganneau, Bliss and Dickie,
and Lawrence and Woolley all noted apparent evidence of the seventh century conquests in
visible archaeological remains. I am more interested here, however, in stratigraphically based
discoveries. And in virtually every case where those levels could be even roughly dated to the
early seventh century, the excavators interpreted them as ample evidence of a bitter racial con-

oi.uchicago.edu





615



THUNDERING HORDES



flict between the Desert and the Sown. A brief review of the material presented in Ovadiah
(1970) indicates the extent of these conclusions. Ovadiah himself goes even further, stating
that “most of the churches in Palestine were destroyed by Chosroes II in the Persian invasion
of 614” (Ovadiah 1970: 218).

Through the matching of military dates with ash layers, toppled frescoes, columns, and
chancel screens, the violent reality of the Persian and Muslim conquests was seemingly con-
firmed. It was clear that some churches and settlements had survived the invasions (Kraemer
1958; Colt 1950; Crowfoot 1941: 161; Kraeling 1938: 68), but the dramatic changes observed
elsewhere provided ample justification for an image of thundering hordes. Of course this potent
archaeological image was not restricted to the seventh century



a.d.



During the 1920s and 1930s,
other archaeologists—applying Petrie’s basic ethnic conceptions on other levels—unhesitat-
ingly found evidence of similar conquests of the country by invading peoples: Hurrians,
Hyksos, Sea Peoples, and Israelites. However, the history of diffusionist racial and ethnic inter-
pretation in the specific case of Palestinian archaeology has not yet been fully studied (see the
suggestive analysis of W. D. Smith 1978).

Yet even as archaeologists in other parts of the world began to recognize the shortcomings
of this fixation with race and racial warfare—at least partially caused by political revulsion
toward the excesses of Nazi racism (Barkan 1988)—and began to suggest new explanations for
sharp discontinuities in material culture, the theories of the thundering hordes in Palestine at
the end of the Byzantine period proved exceedingly difficult to eradicate. Studies of the Roman



limes



in the Negev and southern Transjordan suggested a clear and hostile separation between
settled and nomadic peoples (for conflicting evaluations of recent scholarship, see Parker 1986,
and Mayerson 1986). That hostility, though itself







unexplained, may be the result of the Arabs’
inborn hostility to settled populations or environmental pressure forcing massive migrations
from the Arabian peninsula. In the early twentieth century, environmentalist explanations were
particularly popular (see Huntington 1911, and for a brief survey of similar theories, see the
introduction of Donner 1981). Thus determinist solutions were seen as the cause of the early
seventh-century destructions—no matter how vaguely or imprecisely dated they were.

Similar thinking led some scholars to see marching Persian armies as the culprits in other
destructions, although this contention was supported by only the roughest of chronological
correspondences. Needless to say, ceramic chronology is still not refined enough to distin-
guish 614 from the 630s, but on the basis of the presumed route of the Persian invasion, the
destructions of the church at Shavei-Zion and, more recently, at Nahariya were ascribed to the
earlier event (cf. Prausnitz, Avi-Yonah, and Barag 1967 and Edelstein and Dauphine 1975).
The ascription of the destruction of the acropolis of Avdat to the Persians is even more tenu-
ous (see Negev 1993: 1155, but see Mayerson 1964: 193–96 for another view.) And it is strik-
ing how the basic belief in the sudden, violent invasions was used to support even more
sweeping theories. In the late 1960s, wide-ranging surface surveys, based on only a rough
knowledge of Early Islamic pottery, were combined with the evidence of destruction to sug-
gest that a centuries-long process of “nomadization” began with the Muslim conquest (Tsafrir
1984; Sharon 1976). According to this theory, the fate of the country was sealed with the
arrival of nomads who brought the mentality of the desert and refused to settle in cities—and
with the mass exodus of peaceful (and civilized) Christian peasants who fled in terror from the
thundering hordes (Tsafrir 1984: 74). Such images, I would argue, drew their emotional power

oi.uchicago.edu





616



NEIL ASHER SILBERMAN



as much from turn-of-the-century racial dogma and traditional negative western images of
Arabs as from textual sources or archaeological evidence alone.

Of course that was not the final version of the story. In recent years, a more sophisticated
and careful analysis of early seventh-century levels has profoundly challenged the notion of
general chaos and widespread destruction, replacing it with a more complex conception of
cultural reorganization (for a wide-ranging review of the latest evidence, see Whitcomb
1995). Recent finds at the churches of Umm er-Rasas and at other sites in Jordan have under-
lined the persistence of the architectural and religious aspects of what had long been identified
as “Byzantine” (i.e., pre-Islamic) material culture far beyond its traditional chronological
boundaries (Piccirillo 1988; Schick 1987, 1988). The recognition of the uninterrupted urban
stratigraphy of Jerash, Umm al-Jimal, and a number of other sites on both sides of the Jordan
has cast doubt on the geographical extent of violent destruction connected with the Muslim
conquest (Zayadine 1986; Sauer 1982; DeVries 1981). Recent reconsideration of the sequence
of pottery forms at Pella, with its clear



a.d.



747 earthquake destruction level (Walmsley 1988;
McNicoll and Walmsley 1982); from Khirbat al-Mafjar (Whitcomb 1988); and from the City
of David in Jerusalem (Magness 1992) have offered new correctives to traditional pottery
chronology.

The vibrant urbanism at sites such as Tiberias (Harrison 1992), Ramla (Rosen-Ayalon and
Eitan 1969), and Jerusalem (Ben-Dov 1975), and in the reconstructed baths at Hammat Gader
(Green and Tsafrir 1982: 94–96; Hasson 1982) has further challenged earlier theories about
nomadic loathing of cities and has replaced them with far more wide-ranging hypotheses
about the gradual transition of the classical



polls



to the Islamic



madina



(Kennedy 1985a,
1985b). Moreover, the recognition that elaborately constructed, isolated sites such as Khirbat
al-Mafjar and the “desert castles” of Jordan may have served as administrative centers for a
new system of intensive irrigation agriculture (Conrad 1981) suggests that the transformation
from “Byzantine” to “Islamic” periods was characterized not so much by ethnic migration as
by far-reaching economic and technological change.

At the same time, a more sophisticated understanding of the close interrelation of settled
and pastoral peoples has also undercut the basic premises of the sharp Desert/Sown dichotomy.
Anthropological studies of modern nomadic lifeways throughout the Middle East have
suggested to archaeologists working in a number of historical periods that settled and pastoral
peoples are actually components of a single, dimorphic community (among the many ancient
and modern studies, cf. Rowton 1977; Marx 1978; Dever 1980; and Finkelstein 1988). That
nomads—or “Saracens”—were an element



internal



to Byzantine society in Palestine is becom-
ing increasingly clear in the extensive Negev surveys of Steven Rosen (1987); the studies of
E. B. Banning in Jordan (1986); and the discovery of the still enigmatic cultic center near Sde
Boker (Rosen-Ayalon and Nevo 1982; Nevo n.d.). These finds suggest that pastoralists were
an important component of Palestinian Byzantine society, and that some of the most far-
reaching changes of the early seventh century might have been less the results of mass invasion
than internal social change. An additional factor in this reassessment has been a widespread
refutation of earlier theories of climatic change or population pressure moving outward from
the Arabian peninsula to more fertile regions (summarized in the introduction of Donner
1981)—which was long regarded as the



real



impetus for the Muslim conquests.

ONE SHORTY

oi.uchicago.edu





617



THUNDERING HORDES



In fact, as early as 1964, Philip Mayerson suggested that the earliest Muslim attacks on
southern Palestine were the result of cooperation between an outside Islamic elite and the
indigenous pastoralists of the Negev and Sinai (Mayerson 1964). Other scholars—Crone and
Cook (1977), Wansbrough (1978), Bashir (1984), Sharon (1988), and Koren and Nevo (1991)
subsequently went much further in dismissing the historical reliability of the Islamic tradi-
tions, proposing quite different (and completely internal) reconstructions of the events (but see
the cautionary note against extreme revisionism in Whitcomb 1995: 500–501). In this connec-
tion, it is interesting to note how far Sharon’s historical thinking has come since his earlier
article (Sharon 1976). Less far-reaching in its speculative reconstruction, but more influential
within the discipline is Fred Donner’s recent analysis of Islamic textual sources for the Early
Islamic conquests (1981). In his section dealing with Syria, he demonstrates the complexity of
the developments, beginning with an initial confederation of an Arabian elite with indigenous
pastoral tribes, a period of brief but intense conflicts with Byzantine forces (relatively far from
urban centers), and a relatively peaceful incorporation of cities into the structure of the
expanding Islamic state (see also Hill 1971). Migration or massive population change, Donner
argues, played no significant role in the conquest. And in light of the emerging evidence of
continuing urban life throughout the country, the evidence of the extent of violent outside
intervention becomes more and more restricted. The complex data of historical sociology has
come to replace the once-vivid image of the thundering hordes.

Certainly no one can argue that these new approaches to understanding an admittedly
important period of cultural transition are quite different from those employed by Western
scholars who studied the Persian or Muslim conquest in the Middle Ages, in the Renaissance,
or even in the early days of archaeology. We can now see that many of the changes observed
in the archaeological record were not solely due to sudden, external invasions. Complex fac-
tors such as long-range economic trends, superpower conflict, internal social tensions, and the
political strategies of ethnic minorities must now be taken into account. But can it be just a
coincidence, however, that these new areas of historical interest are also of particular concern
to late twentieth century industrialized Western society? This possible correlation between
modern perception and historical interpretation is precisely the theme that I want to stress.
Because the archaeology of every period is more than a purely objective analysis of fossilized
material culture. Its larger social significance may lay in its ability to craft



modern



historical
metaphors and myths of creation—myths in the older, religious sense of acceptable explana-
tion by vivid analogy (Doty 1986).

There is, of course, a wider context in which these new interpretations must be seen. Over
the last few decades, the human and natural landscapes of the Middle East have undergone
unprecedented changes. With the incorporation of Middle Eastern peoples and economies into
the Western industrial world system, cities have grown up, fellahin and bedouin alike have been
transformed into workers, and the sharp distinction between the Desert and the Sown has been
significantly blurred. For centuries the scholars of the West turned toward the East with the
overtly hostile perspective of their societies, concentrating on fundamental distinctions between
settled and pastoral peoples, between Christians and Muslims, between Indo-Europeans and
Semites. Yet those distinctions have now lost much of their power in an era when the represen-
tatives of Western nations have sought—to greater or lesser degrees—to integrate the peoples

oi.uchicago.edu





618



NEIL ASHER SILBERMAN



of the Middle East into a single, developing world economy. And I would argue that this mod-
ern world view has influenced—if not completely determined—the way archaeological finds
are now interpreted as evidence of complex social interaction and economic development
(Miller 1980; Patterson 1987).

The Muslim and Persian conquests have a particularly potent metaphorical significance in
an era when Middle Eastern conflicts among Arabs, Persians, Israelis, Europeans, and Amer-
icans slowly give way to the beginning of an era of coexistence, if not peace. As citizens of
the end of the twentieth century and the beginning of the twenty-first we cannot help but
reflect the perceptions of our own society. The modern scholarly recognition that there is some
sort of a “problem” with our understanding of the Persian and Muslim conquests should itself
be placed in historical context. For as in the Middle Ages, the Renaissance, the Enlighten-
ment, and the age of the great European empires, our current interpretations of the seventh
century Muslim and Persian conquests cannot help but reflect our society’s evolving relation-
ship to contemporary Middle Eastern peoples—and to the many-faceted reality of social
change.

ACKNOWLEDGMENT

The ideas and general themes expressed in this paper were originally presented at the ses-
sion, “The Problem of the Sassinid (Persian) and Moslem Conquests of Palestine,” at the
Annual Meeting of the American Schools of Oriental Research in Anaheim, California, on
November 20, 1989. Subsequent to that session, I benefited greatly from Doug Esse’s specific
reactions to that paper and from his constructive comments relating to the general problems of
ethnicity and cultural change.

REFERENCES



Banning, E. B.
1986 Peasants, Pastoralists, and Pax Romana.



Bulletin of the American Schools of Oriental Research



261: 25–50.

Baras, Z.
1982 The Persian conquest and the end of Byzantine rule. In



The Land of Israel from the Destruction
of the Second Temple to the Muslim Conquest,



vol. 1, eds. Z. Baras, S. Safrai, Y. Tsafrir, and
M. Stern. Jerusalem: Ben-Zvi Institute (Hebrew).

Barkan, E.
1988 Mobilizing Scientists against Nazi Racism, 1933–1939. Pp. 180–205 in



Bones, Bodies, Behavior:
Essays on Biological Anthropology,



ed. George W. Stocking. Madison: University of Wisconsin
Press.

Bashir, S.
1984



Mugaddimah fi al-Tarikh al-Akhar.



Jerusalem: Israel Publication Services.

Ben-Dov, M.
1975 The Area South of the Temple Mount in the Early Islamic Period. Pp. 97–101 in



Jerusalem
Revealed,



ed. Y. Yadin. Jerusalem: Israel Exploration Society.

Berger, P., and Luckmann, T.
1967



The Social Construction of Reality.



Garden City, NY: Doubleday.

Clermont-Ganneau, C.
1898 The Taking of Jerusalem by the Persians,



a.d.



614.



Palestine Exploration Fund Quarterly State-
ment



(1898): 36–55.

oi.uchicago.edu





619



THUNDERING HORDES



Colt, H. D., ed.
1950



Excavations at Nessana,



Vol. 1. Princeton: Princeton University Press.

Conrad, K. I.
1981 The Qusur of Medieval Islam: Some Implications for the Social History of the Near East.



al-Abhath



29: 7–23.
Constantelos, D.

1972 The Moslem Conquests of the Near East as Revealed in the Greek Sources of the VIIth and VIIIth
Centuries.



Byzantion



42: 325–57.
Crone, P., and Cook, M.

1977



Hagarism: The Making of the Islamic World.



Cambridge: Cambridge University Press.
Crowfoot, J. W.

1941



Early Churches in Palestine.



London: Oxford University Press.
Daniel, N.

1960



Islam and the West: The Making of an Image.



Edinburgh: Edinburgh University Press.
1975



The Arabs in Mediaeval Europe



. London: Longman.
1984



Heroes and Saracens.



Edinburgh: Edinburgh University Press.
Dever, W. G.

1980 New Vistas on the EB IV (MB I) Horizon in Syria-Palestine.



Bulletin of the American Schools
of Oriental Research



237: 35–64.
DeVries, B.

1981 The Umm al-Jimal Project, 1972–77.



Bulletin of the American Schools of Oriental Research



244: 53–72.
Donner, F. M.

1981



The Early Islamic Conquests.



Princeton: Princeton University Press.
Doty, W. G.

1986



Mythography: The Study of Myths and Rituals.



University, AL: University of Alabama.
Edelstein, G., and Dauphine, C.

1975 A Byzantine Church at Nahariya.



Qadmoniot



8: 128–32 (Hebrew).
Esse, D. L.

1991



Subsistence, Trade, and Social Change in Early Bronze Age Palestine.



Studies in Ancient Oriental
Civilization 50. Chicago: Oriental Institute Press.

1992 The Collared Pithos at Megiddo: Ceramic Distribution and Ethnicity.



Journal of Near Eastern
Studies



51: 81–103.
Finkelstein, I.

1988



The Archaeology of the Israelite Settlement.



Jerusalem: Israel Exploration Society.
Galton, F.

1869



Hereditary Genius: An Inquiry into Its Laws and Consequences.



London: Macmillan.
Glubb, J. B.

1963



The Great Arab Conquests.



London: Hodder and Stoughton.
Gould, S. J.

1981



The Mismeasure of Man.



New York: Norton.
Green, J., and Tsafrir, Y.

1982 Greek Inscriptions from Hammat Gader.



Israel Exploration Journal



32: 77–96.
Grunebaum, G. E. von

1966 Islam: The Problem of Changing Perspective. In



The Transformation of the Roman World,



ed. L. White. Los Angeles: University of California Press.
Harrison, T. P.

1992 The Early Umayyad Settlement at Tabariyah: A Case of Yet Another Misr?



Journal of Near
Eastern Studies



51: 51–59.
Hasson, I.

1982 Remarques sur l’inscription de l’époque de Muº



a



wiya à



Ó



ammat Gader.



Israel Exploration
Journal



32: 97–101.

oi.uchicago.edu





620



NEIL ASHER SILBERMAN



Hill, D. R.
1971



The Termination of Hostilities in the Early Arab Conquests,



a.d.



634–656.



London: Luzac.

Hitti, P. K.
1949



History of the Arabs.



London: Macmillan.

Huntington, E.
1911



Palestine and Its Transformation.



New Haven: Yale University Press.

Kelves, D. V.
1985



In the Name of Eugenics.



Berkeley: University of California Press.

Kennedy, H.
1985a From Polis to Madina: Urban Change in Late Antique and Early Islamic Syria.



Past and Present



106: 3–27.
1985b The Last Century of Byzantine Rule in Syria: A Reinterpretation.



Byzantinische Forschungen



10: 141–83.

Kohl, P., and Fawcett, C.
1995



Nationalism, Politics, and the Practice of Archaeology



. Cambridge: Cambridge University
Press.

Koren, J., and Nevo, Y.
1991 Methodological Approaches to Islamic Studies



.







Islam



68: 87–107.

Kraeling, C. H.
1938



Gerasa, City of the Decapolis



. New Haven: Yale University Press.

Kraemer, C. J.
1958



Excavations at Nessana: Non-literary Papyri 3



. Princeton: Princeton University Press.

Kuper, A.
1988



The Invention of Primitive Society



. London: Routledge.

Landau, M.
1991



Narratives of Human Evolution



. New Haven: Yale University Press.

Leone, M.; Potter, P.; and Shackel, P.
1987 Toward a Critical Archaeology.



Current Anthropology



28: 283–302.

Lewis, B.
1977 Gibbon on Muhammad. Pp. 61–74 in



Edward Gibbon and the Decline and Fall of the Roman
Empire,



eds. G. W. Bowersock, J. Clive, and S. Graubard. Cambridge, MA: Harvard University
Press.

Lincoln, B.
1989



Discourse and the Construction of Society



. New York: Oxford University Press.
McNicoll, A., and Walmsley, A. G.

1982 Pella/Fahl during the Early Islamic Period. Pp. 339–45 in



Studies in the History and Archaeol-
ogy of Jordan



I, ed. A. Hadidi. Amman: Department of Antiquities.
Magness, J.

1992 A Reexamination of the Archaeological Evidence for the Sasanian Persian Destruction of the
Tyropean Valley.



Bulletin of the American Schools of Oriental Research



287: 67–74.
Marx, E.

1978 Ecology and Politics of Middle Eastern Pastoralists. Pp. 41–74 in



The Nomadic Alternative:
Modes and Models of Interaction in the African-Asian Deserts and Steppes



, ed. W. Weissleder.
The Hague: Mouton.

Mayerson, P.
1964 The First Muslim Attacks on Southern Palestine (



a.d.



633–634).



Transactions and Proceedings
of the American Philological Association



95: 155–99.
1986 The Saracens and the Limes.



Bulletin of the American Schools of Oriental Research



262: 35–47.
Metlitzki, D.

1977



The Matter of Araby in Medieval England.



New Haven: Yale University Press.

Miller, D.
1980 Archaeology and Development.



Current Anthropology



21: 709–26.

oi.uchicago.edu





621



THUNDERING HORDES



Negev, A.
1993 Eboda. Pp. 1155–65 in



The New Encyclopedia of Archaeological Excavations in the Holy Land



,
vol. 3, Jerusalem: Israel Exploration Society & Carta and New York: Simon & Schuster.

Nevo, Y.
n.d.



Sde Boqer and the Central Negev, 7th–8th Century



a.d.



Jerusalem: Israel Publication Services.
Ovadiah, A.

1970



Corpus of Byzantine Churches in the Holy Land



. Bonn: P. Hanstein.
Parker, S. T.

1986



Romans and Saracens: A History of the Arabian Frontier



. American Schools of Oriental
Research Dissertation Series 6. Winona Lake, IN: Eisenbrauns.

Patterson, T. C.
1987 Development, Ecology, and Marginal Utility in Anthropology.



Dialectical Anthropology



15:
15–31.

1995



Toward a Social History of Archaeology in the United States.



Fort Worth: Harcourt-Brace.

Petrie, W. M. F.
1906 Migrations.



Journal of the Royal Anthropological Institute



36: 189–232.
1907



Janus in Modern Life



. New York: Putnam’s.
1911



The Revolutions of Civilization



. London: Harper.

Piccirillo, M.
1988 The Mosaics at Um er-Rasas in Jordan.



Biblical Archaeology



51: 208–13; 227–31.

Poliakov, L.
1974



The Aryan Myth: A History of Racist and Nationalist Ideas in Europe.



New York: Basic Books.

Porter, R.
1988



Edward Gibbon: Making History.



London: Weidenfeld and Nicholson.

Prausnitz, M.; Avi-Yonah, M.; and Barag, D.
1967



Excavations at Shavei-Zion: The Early Christian Church



. Rome: Centro per le Antichita e Storia
dell’Arte del Vicino Oriente.

Robertson Smith, W.
1880 Animal Worship and the Animal Tribes among Arabs and the Old Testament.



Journal of Phi-
lology



9: 75–100.
1885



Kinship and Marriage in Early Arabia.



Cambridge: Cambridge University Press.
1889



Lectures on the Religion of the Early Semites.



Edinburgh: Adam and Charles Black.

Rodinson, M.
1987



Europe and the Mystique of Islam.



Near Eastern Studies 4. Seattle: University of Washington
Press.

Rosen, S. A.
1987 Byzantine Nomadism in the Negev: Results from the Emergency Survey.



Journal of Field
Archaeology



14: 29–41.

Rosen-Ayalon, M., and Eitan, E.
1969



Ramla Excavations



. Jerusalem: Israel Museum.

Rosen-Ayalon, M., and Nevo, Y. D.
1982



The Early Arab Period in the Negev



. Jerusalem: Ben-Zvi Institute.

Rossi, P.
1984



The Dark Abyss of Time: The History of the Earth and the History of Nations from Hooke to
Vico



. Chicago: University of Chicago Press.

Rowton, M. B.
1977 Dimorphic Structure and the Parasocial Element.



Journal of Near Eastern Studies



36: 181–98.

Said, E.
1978



Orientalism.



New York: Pantheon Books.

oi.uchicago.edu





622



NEIL ASHER SILBERMAN



Sauer, J. A.
1982 The Pottery of Jordan in the Early Islamic Period. Pp. 329–37 in



Studies in the History and
Archaeology of Jordan



I, ed. A. Hadidi. Amman: Department of Antiquities.

Schick, R.
1987 The Fate of the Christians in Palestine during the Byzantine-Umayyad Transition. Ph.D. disser-

tation, University of Chicago.
1988 Christian Life in Palestine during the Early Islamic Period.



Biblical Archaeology



51: 218–21;
239–40.

Sharon, M.
1976 Processes of Destruction and Nomadization in Palestine under Islamic Rule (633–1517). Pp. 9–

32 in



Notes and Studies on the History of the Holy Land under Islamic Rule,



ed. M. Sharon.
Jerusalem: Ben-Zvi Institute (Hebrew).

1988 The Birth of Islam in the Holy Land. Pp. 225–35 in



The Holy Land in History and Thought,
ed. M. Sharon. Leiden: E. J. Brill.

Silberman, N. A.
1989 Between Past and Present. New York: Henry Holt.
1990 The Politics of the Past: Archaeology and Nationalism in the Eastern Mediterranean. Mediterra-

nean Quarterly 1: 99–110.
1991 Desolation and Restoration: The Impact of a Biblical Concept on Near Eastern Archaeology.

Biblical Archaeology 54: 76–86.
1993 Petrie and the Founding Fathers. In Biblical Archaeology Today 1990, Proceedings of the 2nd

International Congress of Biblical Archaeology, eds. A. Biran and J. Aviram. Jerusalem: Israel
Exploration Society.

1995 Promised Lands and Chosen Peoples: The Politics and Poetics of Archaeological Narrative. In
Nationalism, Politics, and the Practice of Archaeology, eds. P. Kohl and C. Fawcett. Cambridge:
Cambridge University Press.

1999 Petrie’s Head. In Assembling the Past: Studies in the Professionalization of Archaeology, eds. A.
Kehoe and M. B. Emmerichs. Albuquerque: University of New Mexico Press.

Smith, B.
1977 Islam in English Literature. Delmar, NY: Caravan Books.

Smith, W. D.
1978 The Social and Political Origins of German Diffusionist Ethnology. Journal of the History of the

Behavioral Sciences 14: 103–12.

Southern, R. W.
1962 Western Views of Islam in the Middle Ages. Cambridge, MA: Harvard University Press.

Stepan, N.
1982 The Idea of Race in Science: Great Britain 1800–1960. Hamden, CT: Archon Books.

Stocking, G. W., Jr.
1987 Victorian Anthropology. New York: Free Press.

Terry, J. J.
1985 Mistaken Identity: Arab Stereotypes in Popular Writing. Washington, D.C.: American–Arab

Affairs Council.

Trigger, B.
1984 Alternative Archaeologies: Nationalist, Colonialist, Imperialist. Man 19: 355–70.

Tsafrir, Y.
1984 The Arab Conquest and the Process of Population Decline in Eretz Israel. Cathedra 32: 69–74

(Hebrew).

Walmsley, A. G.
1988 Convergence of Literary and Historical Evidence. Mediterranean Archaeology 1: 142–59.

Wansbrough, J.
1978 The Sectarian Milieu. Cambridge: Cambridge University Press.

oi.uchicago.edu



623THUNDERING HORDES

Wellhausen, J.
1902 Das arabische Reich und sein Sturz. Berlin: G. Reiner.
1927 The Arab Kingdom and Its Fall. Trans. Margaret G. Weir. Calcutta: University of Calcutta.

Whitcomb, D.
1988 Khirbat al-Mafjar Reconsidered: The Ceramic Evidence. Bulletin of the American Schools of

Oriental Research 271: 51–67.
1995 Islam and the Socio-cultural Transition of Palestine. In The Archaeology of Society in the Holy

Land, ed. T. E. Levy. London: Leicester University Press.

Wilken, R. L.
1992 The Land Called Holy. New Haven: Yale University Press.

Zayadine, F., ed.
1986 Jerash International Project 1981–1983. Amman: Department of Antiquities.

oi.uchicago.edu



624 NEIL ASHER SILBERMAN

oi.uchicago.edu



625



31



PORT POWER IN THE EARLY AND THE
MIDDLE BRONZE AGE: THE ORGANIZATION

OF MARITIME TRADE AND HINTERLAND
PRODUCTION



Lawrence E. Stager



Coastal societies in the Levant developed a distinctive set of characteristics when linked to
long-distance maritime commerce. Integration with the interior regions, the secondary net-
works oriented east-west along the natural drainage, wadi, or transport systems, was essential
for the international trade network to have anything worthwhile and profitable to offer at the
end of the line. These arteries of raw materials and manufactured products supplied the heart
of the system, the eastern Mediterranean seaports, with a lifeline that connected hinterland
with overseas emporia. By integrating diverse domains into the “port power” economic sys-
tem, import-export merchants could reap sizable profits and exercise more economic power
than the harbor princes who protected them or the rulers of the interior whose authority and
power were largely circumscribed by territorial limits.

The port power model I propose for the Early Bronze Age Levant (EB I–III) may apply
equally well or better to other parts of the Mediterranean in the second and first millennia



b.c.



I have borrowed heavily from a model developed by Bennet Bronson (1977) for Southeast
Asia, which was then applied to that region in greater detail by S. R. Hall (1985). From the out-
set it should be said that “no model completely and perfectly fits real life” (Braudel 1984: 63).
It is a heuristic device for organizing data into an intelligible whole. It must be in constant
interaction with empirical data, reinterpreted according to new information, and discarded
when anomalies can no longer be incorporated. It is suggestive and, at its best, predictive, but
never sacrosanct.

I

In Bronze Age Palestine we find at the base of the port power network local village clus-
ters (“communities of exchange”) in the highlands, which periodically converged on a
regional market. The few highland marketing centers (for example, Shechem, Jerusalem, and
Hebron in the Middle Bronze Age [MB]) were then integrated with intermediate markets in
the foot hills and the interior coastal plain. These, in turn, were connected with a major sea-
port, the highest scale entrepot, which then integrated the intermediate and local spheres of
commerce with international trade realms represented by seafaring merchants (figs. 31.1, 31.2
for EB; fig. 31.3 for MB II).

Thus we might find highlanders responding to coastal or regional demands by supplying
wool, meat, and dairy products from the pastoral sector, grapes (wine) and olives (oil) from



oi.uchicago.edu





626



LAWRENCE E. STAGER



FIGURE 31.1.







Map of port power networks in the Early Bronze Age (note: The Egyptian sites el-Dabºa and
Maskhuta belong to the Middle Bronze Age).

oi.uchicago.edu

[image: image192.tif]



627



THE ORGANIZATION OF MARITIME TRADE AND HINTERLAND PRODUCTION



FIGURE 31.2. Map of port power networks in the Early Bronze Age.



the horticultural sector, stone from the hard limestone mountains, timber and charcoal from
the oak and pine forests, and resin (or “turpentine”) from the terebinth and pine tree. These are
products within the catchments of the east-west wadi networks (for highland settlement pat-
terns, see Finkelstein and Gophna 1993; Harrison 1997). And, of course, there was serious
production of olive oil and wine, beginning in the latter half of the fourth millennium, in the
lowlands as well. Farther afield there was salt, sulfur, and bitumen from the Dead Sea, some

oi.uchicago.edu

[image: image193.tif]



628



LAWRENCE E. STAGER



FIGURE 31.3. Map of port power networks in Middle Bronze II.



of which was transported by donkey to western ports. Among the most precious products of
Palestine in great demand overseas was copper, extracted from the mines in Wadi Feinan near
the Arabah during Chalcolithic and EB times as well as from the Sinai during the EB II
period. This archipelago of more distant trade is dealt with below.

I am assuming that power was exercised through economic ties and that this integrated,
hierarchical network can and often did develop outside of or beyond a state exercising coercive

oi.uchicago.edu

[image: image194.tif]



629



THE ORGANIZATION OF MARITIME TRADE AND HINTERLAND PRODUCTION



political-military force over the hinterland. In fact, direct control over mobile populations
such as that represented by the pastoralist sector or heavy-handed measures taken against
other hinterland sectors could be counterproductive as well as expensive.

Asymmetrical relations of economic advantage can be seen almost from the start because
of the system of exchange which operated from the larger underdeveloped hinterland to the
major commercial center, the Mediterranean seaport, the command center of information and
decision making, which translated into real economic power. Power in this decentralized sys-
tem was not necessarily wielded by an overarching political (state) or military force. Relations
between the commercial port were less coercive than if the commercial center and the centers
of production and wholesaling were integrated parts of a single state system. The effectiveness
of port power accrued from the capability of the economic system to penetrate diverse ethnic,
cultural, and political boundaries and to interconnect them with a minimal use of force. This
seems to be especially applicable to the ecological diversity and cultural mosaic of the Levant
throughout the EB, in contrast to the ecological homogeneity and tendencies toward autocracy
of Egypt. The port power model which may explain the network of production and sale of
commodities at one node of the international transaction, such as the Levant, is indifferent to,
and can easily accommodate very different economic systems at the other end, such as Egypt,
where royal authority at the port of entry dominated distribution during periods when kingship
was paramount and trade monopolies prevailed (for a superb study of this complementariness
between Egypt and Palestine, see Butzer 1997).

In the Levant more modest means than outright political control or dominance could be
used to manipulate dependent economies. Implicit in the hierarchy of trade and market
exchange were chains of subordination along the network, involving at points inland and
upland patron-client and trade-partner relationships, and the ultimate advantage over produc-
ers in the highland “communities of exchange,” where non-market or bartering economies can
be expected to dominate trade. In this decentralized system of market exchange, it would be
unlikely for indigenous export-import merchants to involve themselves in direct contact with
the hinterland producers. And if foreign merchants were to invade this area, or even that of
intermediate market centers, it would be a sign that the ultimate consumers had taken rather
tight control of the international network as well. There are signs of this during the late EB I
period, when Egyptian trading posts and mercantile colonies occupied important nodes along
the overland trade routes of southern Palestine (Stager 1992).

Usually, during times of less extreme international asymmetries, foreign merchants, if
present at all, were confined to the ports themselves, as we find in EB II with the Egyptian
mercantile colony at Byblos and perhaps at Ashkelon, when indigenous Levantines were in
control of the ports of power. The enduring sources of power, then, were not those exercised
by political or military means from the ports, whether controlled by indigenous or foreign
sources, but by those import-export merchants, usually an oligarchy, who exercised indirect
economic power through the integrated and hierarchical system of market exchange. In other
words, the Phoenician model of trade operative in the central and western Mediterranean dur-
ing the first millennium



b.c.



is also applicable to the eastern Mediterranean system of the third
and second millennia, except that the geography of trade was much more limited in scope dur-
ing the earlier period.

oi.uchicago.edu





630



LAWRENCE E. STAGER



From the intensive surveys made by Ram Gophna in the southern Levant, it is clear that
sites there proliferate along the wadi basins during EB. Valerie Fargo (1979) was the first to
recognize that the pattern of settlement along these transport and drainage systems was con-
gruent with a dendritic form of marketing system known from the works of E. A. J. Johnson
(1970) and Carol Smith (1976). By ranking sites by size Fargo was able to suggest a regional
or catchment center for each of the wadi systems in EB III and hypothesize that there might
even be a seaport at Ashkelon.

Examples of the EB III inland centers which dominated their respective wadi systems and
“catchment” basins include Tell el-Hesi in the Wadi Hesi/Shiqma basin; Tel Erani (now
known to have been more than 25 ha in EB III) in the Lachish basin to the north; and in the
basin even farther north, the heavily fortified city of Yarmuth, with its impressive palace and
temple complex (Miroschedji 1993). These places would then qualify as intermediate trade
centers in the model I have outlined above (fig. 31.2).

What was missing from this network was the port itself from which, for example, olive oil
and wine were shipped to more distant places (Stager 1985, 1992). Fortunately vestiges of the
EB seaport have been discovered by the Leon Levy Expedition to Ashkelon, a site located right
on the sea and between the major wadi systems we have been discussing: the Wadi Hesi basin
to the south and Na



˙



al Lachish basin to the north. Already known as “Ashkelon” in Egyptian
texts from the early second millennium



b.c.



, the place name is derived from the same West
Semitic root as “shekel” and probably means the “place of weighing”—very relevant nomen-
clature for a seaport (Stager 1991).

Large quantities of residual pottery indicate substantial occupation over much of Tel Ash-
kelon from EB I–EB III (ca. 3500–2250



b.c.



) before Canaanite or Amorite builders dug it up
and threw it into deep fills in and behind the MB IIA city gate and ramparts. The same range
of the EB residual pottery has been found in later fill deposits (the occupation layers have not
yet been reached) near the center of Tel Ashkelon. Oil jars of metallic combed ware and oil
separator vats (indicating production at the site) predominated in EB II–III.

The early EB I ceramic evidence suggests that there was occupation at Tel Ashkelon con-
temporary with the extensive so-called trough settlement(s) a kilometer to the north. Botanical
evidence from there indicates that olive orchards were being grown in the vicinity, and cedar of
Lebanon (



Cedrus libani



) was being shipped to Ashkelon as early as the mid-fourth millennium
(Gophna and Liphschitz 1996; Stager 1993). Kay Prag had already recognized the likelihood of
maritime trade between Byblos and Egypt at this and an even earlier period (Prag 1986). In his
analysis of the interaction between Mesopotamia (Uruk IVa) and Egypt (Naqada II), beginning
ca. 3500



b.c.



, P. R. S. Moorey (1990) proposed active sea lanes between coastal Syria (the Amuq
Plain) and the Egyptian Delta, as part of the larger “Uruk network,” or what Philip Curtin (1984)
might call a “trade diaspora.” Ashkelon would have been a probable intermediate port of call at
that time.

Firm evidence for shipping between Egypt and Palestine later in the EB I comes from the
royal cemetery of Abydos. Just 100 m north of the tomb of Aha (first king of the First
Dynasty), Tomb U-j, belonging to an earlier king of Dynasty 0, whose name is not known, was
filled with mortuary offerings, some time ca. 3150



b.c.



(according to radiocarbon determina-
tions). At least 207 wine jars were stacked in two of the chambers of the multichambered
tomb. Impressions of another 150 wine jars were left in the floor of an adjacent chamber. It is
estimated that as many as 700 wine jars once filled these rooms. Forty-seven of the jars

long

oi.uchicago.edu





631



THE ORGANIZATION OF MARITIME TRADE AND HINTERLAND PRODUCTION



contained grape pips, others, completely preserved grapes. Neutron activation analysis of the
two types of wine jars indicates that they were imported from Palestine (McGovern and Har-
bottle 1997; McGovern 1998), where the grape had been domesticated and grown long before
it was transplanted in Egypt (Stager 1985, 1992). Such a cargo of wine destined for the tomb
of one ruler is not likely to have reached Egypt from Palestine via the EB I overland caravan
route, later known as the “Ways of Horus.” It is most likely that this was just one of many car-
goes of wine shipped from Palestine, probably from the port of Ashkelon, and destined for the
elite of Egypt (Stager 1985). At that time when wine was still a scarce commodity in Egypt,
the Egyptian rulers as far away as Abydos had probably begun to exercise rather strict control
over imports shipped into the Delta seaport and siphoned those luxury items into their coffers.
By the First Dynasty in Egypt and EB II in Palestine, maritime trade had triumphed over over-
land trade (see below). As I suggested above, port power can be exercised in various ways.
For the Egyptian rulers it must have been a much easier task for them to control and regulate
large-scale maritime trade entering a single port in the Delta than small-scale overland opera-
tions entering at a number of points.

I now examine the port power model against a more distant network than the east-west
wadi systems and also examine a trade archipelago, which penetrated deep into the Arabah
and Sinai desert in search of copper.

II

During the floruit of early urban settlement in EB II Palestine and of early dynastic rule in
Egypt, new settlements sprang up in the Sinai and Negev deserts in response to developments,
particularly the demand for copper, in the core areas of Palestine and Egypt.

Because of the excavations of Ruth Amiran at Arad and the survey and excavations of
Itzak Beit-Arieh in the Sinai, it is possible to identify certain nodes in the economic network
that linked Arad with the Sinai, over a distance of more than 300 km, and ultimately with
other parts of Palestine and even Egypt (Stager 1992).

It is usually argued on the basis of some shared material culture that the actual miners and
producers of Sinai copper were “Canaanite colonists,” who settled the Sinai during the EB II
period in order to exploit the copper resources there. I have argued elsewhere that this was the
same indigenous population that had always lived in the Sinai (Stager 1992). Although they
might have adapted some aspects of their culture from southern Palestine (such as the so-
called Arad house), their arid-zone settlements, in both layout and orientation, resemble mod-
ern Bedouin encampments in the Sinai and elsewhere. In addition to their mining activities,
the EB II Sinai tribesmen continued to hunt gazelle and mountain goat and to herd flocks of
sheep and small desert goats as they had been doing in earlier eras.

It seems preferable to view these people not as outsiders who colonized the desert and
adapted very quickly to its constraints, but as part of the indigenous population who adjusted
its strategies to the needs and economic demands of core areas. In profit-motivated trade, it
was not that profitable to control the actual production; profit lay in the price differences
between resource area and the marketing center and most of all between markets very far apart.

During EB II, when large quantities of Levantine pottery containing olive oil, wine, and
tree resins were reaching Egypt during the First Dynasty, turquoise and especially copper were

oi.uchicago.edu





632



LAWRENCE E. STAGER



in greater demand there than ever before. Nowhere is this increase in wealth and status among
the early pharaohs better exemplified than in the turquoise and gold bracelets from Djer’s
tomb at Abydos and in the huge cache of copper objects from his mastaba at Saqqara. Most
likely this copper came from Wadi Feinan and the Sinai; however, there is no evidence that it
was being transported directly to Egypt across the Sinai or by way of the Red Sea.

Trade with Egypt was probably not being carried on directly between the Egyptian court
and the desert miners. The Sinai producers had trade links to the north. Arad served as a “gate-
way” to the Sinai and to the central and northern areas of Palestine, where cereals, olive oil,
wine, and wood could be traded for the resources of the Sinai and the Red Sea.

All of this changed, however, during the Old Kingdom. The economic position of Arad as
a gateway to the south was dealt a fatal blow from which the city never recovered. Lucrative
trade with the Sinai came to an abrupt end when the Egyptians took control of the copper and
turquoise mines during the Third Dynasty. By then the Egyptian state had intervened and
taken over direct control of production, depriving both the Sinai tribesmen and the Levantines
of their profitable roles as long-distance merchants.

Why did the Egyptians not do this earlier? I think it may be due to the fact that the trans-
port system had changed and there was now an opportunity to link the Sinai and its copper to
Egypt via the Red Sea. Egyptian personnel stationed at the mines included scribes, stone cut-
ters, scorpion charmers, coppersmiths, interpreters, and naval officers among others (Stager
1992). Why else were these Egyptian mining expeditions under the supervision of the navy?

Before this time, copper was being extracted from Wadi Feinan and the Sinai and trans-
ported by donkey caravans to intermediate or “wholesale” markets such as Arad. That transac-
tion involved the ultimate producers, the Sinai tribesmen, and donkey caravaneers who served
as the initial set of risk-taking entrepreneurs. The goods might then be traded to an intermediate
center such as Arad. Although the merchants of Arad could have transported the copper directly
overland to Egypt via the Ways of Horus, a route in use during the EB I, Eliezer Oren’s survey
along the north Sinai coast indicates that it had ceased to be used as a caravan route during the
EB II and EB III. Probably the overland route, with its small donkey caravans in the EB I, could
not compete with maritime commerce that began in mid-to-late EB, as the volume, variety, and
price of goods delivered by the large merchantmen on the Mediterranean route known as the
Byblos Run transformed international trade. Only ships carrying timber, oil, copper, and wine
could meet the ever-growing demand of the powerful elite of pharaonic Egypt. These Byblos
ships, named after the most important EB port, transported goods, animals, and people between
Egypt and the Levant. To build and outfit such a fleet of Mediterranean vessels required an out-
lay of capital that only large companies, merchant houses, or the state could provide.

We still have not solved the question of what happened to Feinan and Sinai copper once it
reached Arad but was not transported overland to Egypt. I would suggest that it was probably
sent to Ashkelon from where it was then shipped to Egypt. In addition to copper there were
many other trade commodities exported to Egypt through this port. Unlike Arad, which col-
lapsed after it no longer had access to copper once the Egyptians took over the Sinai mines,
Ashkelon continued through the EB III period and Old Kingdom Egypt because it was more
diversified and not restricted to a single commodity and resource area for its prosperity. The
nodes along the exchange system varied with the changing conditions, although the seaport
remained the same from the EB I to EB III periods. It was more directly linked to the maritime
network with Egypt, which thrived during the First through the Sixth Dynasties. With the First

long

oi.uchicago.edu





633



THE ORGANIZATION OF MARITIME TRADE AND HINTERLAND PRODUCTION



Intermediate Period, which coincides with EB IV/MB I in Palestine, the seaport at Ashkelon
was abandoned and not reoccupied until the resumption of urban life there during the MB IIA
period, when once again a strong Middle Kingdom arose in Egypt; this provides another
example of complementariness between Palestine and Egypt (Butzer 1997).

Thus we should not confuse the small-scale donkey caravaneers of the EB I—whose links
to resource-producing peripheries were steadily attenuated by Egyptian trading posts and mer-
cantile colonies, which penetrated far into the interior of south Palestine and balanced the
“extra military expense with savings on the otherwise excessive exactions of local traders
(Bronson 1978: 46)—with the large-scale seafaring merchants of the EB I and II periods, who
inaugurated a pattern of maritime trade that was to endure for millennia.

With the development of the Palestinian system of city-states in the EB II and the advent
of shipping as the primary means of transport in conducting long-distance trade, the situation
changed dramatically. Shipping transformed trade among the indigenous entrepreneurs of the
Levant, as they themselves were gradually transformed from a “vagabond trading class largely
isolated from local cultures and polities” to a group of “larger and smaller magnates under the
protection of politically competitive harbor princes” (cf. Geertz 1989), engaged in shipping
bulk produce of oil, wine, timber, and resin from the more distant interior highlands as well as
grain, oil, and wine from the foothills and coastal plain. Through connections with Sinai
copper, they held the keys to power during EB II.

In the seaports of the eastern Mediterranean, we can imagine that import-export merchants
sat in their counting houses, enriched by profitable trade from abroad. These shipping and
trader barons had acquired the keys to long-distance trade. In most cases only they possessed
a communication system rapid enough and extensive enough to relate supply to demand, as
they interposed themselves between finished products and marketing them in distant places
(Braudel 1982: 404). They could strive for and maintain monopolies over lucrative commod-
ities not so much by force of arms but by their control of information and its flow.

Price differentials between far-flung markets and ports must have been quite sizable for
certain products. As Fernand Braudel expressed it for Medieval Europe, “Long-distance trade
certainly made super-profits: it was after all based on the price differences between two mar-
kets very far apart, with supply and demand in complete ignorance of each other and brought
into contact only by the activities of the middleman” (Braudel 1982: 405).

Since we have no economic texts for third millennium Palestine and Egypt, we can only
speculate about the price copper brought when it eventually reached the court of the pharaohs,
having passed from the producers to the caravaneers through the middlemen at Arad (perhaps
being processed somewhere in between), transported to Ashkelon, where the import-export
merchants then sent it on its way to Egypt by ship.

The port power model that I have outlined for the third millennium could be applied to
many other periods with some modification, of course, for different cultural, social, and politi-
cal milieus.

III

After nearly three centuries of rural life, cities were once again founded in Palestine, then
properly called “Canaan,” during the latter half of the twentieth century



b.c.



Its founders came
from a West Semitic background, probably part of the “Amorite” culture that dominated Syria

oi.uchicago.edu





634



LAWRENCE E. STAGER



after 2000



b.c.



During the latter half of the twentieth century, they settled first along the coast
of Canaan and began to establish seaports from Akko to Ashkelon. By about 1800



b.c.



, Ash-
kelon, the largest seaport in Canaan, supported a population of around 12,000–15,000 inhabi-
tants within its ramparts. These fortifications, punctuated by elaborate city gates with arched
entryways, extended for 2.2 km to enclose a settlement of 50–60 hectares.

The MB IIA kingdoms of Canaan and Syria were of sufficient strength to attract the atten-
tion of Egyptian foreign intelligence during the nineteenth century



b.c.



, when the Middle
Kingdom was at the height of its power (for salient features of the MB IIA culture in the
Levant, see Dever 1993). In the Egyptian Execration Texts, dating to the Twelfth Dynasty
(from roughly the reigns of Sesostris II through Amenemhet III), pharaohs cursed, among
many others, two rulers of Ashkelon in the earlier series (the Berlin bowls; Sethe 1926) and a
third ruler in the later series (the Brussels figurines; Posener 1940). The notion of W. F.
Albright (1960: 82–85) that these texts reflect an evolution in Canaanite society from the MB I
to MB IIA periods, from nomadism to urbanism, from sheiks to kings, is untenable. The sea-
port of Ashkelon was not occupied at all during the EB IV/MB I period and had become the
largest city in Canaan early in the MB IIA period: thus if any of the rulers of Ashkelon were
“kings,” then all three were.

Contemporary with the MB IIA Levant and Middle Kingdom Egypt was the Old Assyrian
trading colony located at Kanesh in central Anatolia. Two items from the vicinity of that
emporium reached Ashkelon: a ceramic shoe rhyton and an Old Cappadocian cylinder seal, of
a type best known from Karum Kanesh II. From a rich array of cuneiform texts found in the
merchant quarters of Kanesh, scholars have been able to put together a fairly detailed portrait
of the merchants living there and their family-run firms headquartered in Ashur, some five to
six weeks away by donkey caravan (Veenhof 1972; Larsen 1976). These merchants were oper-
ating in a competitive, entrepreneurial environment, not in the Polanyianna world of “market-
less trade” (Adams 1974; Holladay 1997; cf. Polanyi 1977). Some of their dealings in textiles
and metals reaped profits of 100% or more. The Old Assyrian colony in Kanesh is the best
documented of what were surely many more “trade diasporas,” where merchants and invest-
ment firms operated for profit, in the ancient Near East and in the eastern Mediterranean
(Kuhrt 1995: 90–95). There may be even older examples going back into the third and fourth
millennia, such as the Uruk colonies and their trade diaspora, as well as other networks out-
lined above.

In her doctoral dissertation Susan Cohen (2000) has carefully periodized the various
phases of the MB IIA culture in Canaan, which first appeared along the coast and gradually
penetrated eastward along wadi systems into the interior foothills and highlands. By the sev-
enteenth century



b.c.



Canaan reached its zenith as an economic and political power (Dever
1987; Ilan 1998; Oren 1997). During that time masses of Canaanites moved from southern Pal-
estine into the Delta, from which their leaders known as the “Hyksos” (“foreign rulers”) domi-
nated much of Egypt between ca. 1640 and 1540



b.c.



(Redmount 1989; Bietak 1996, 1997).
With a steady growth in overall population, the foothills and highlands of Canaan became

densely settled. The highland population increased tenfold from the MB I through MB II peri-
ods (Finkelstein 1993; 1998: 355), as a hierarchy of settlement, from fortified centers such as
Jerusalem and Shechem to villages and hamlets, shaped the interior. Israel Finkelstein made
the insightful observation that this “impressive population growth [led to] an expansion into
the inhospitable regions which were conducive only for horticultural activity. . . . Demand for

oi.uchicago.edu





635



THE ORGANIZATION OF MARITIME TRADE AND HINTERLAND PRODUCTION



these commodities in the sophisticated lowland centers must have played a role in the pro-
cess” (Finkelstein 1998: 360–61). To this I should add a further stimulus: the demand for wine
and olive oil from abroad, especially from Egypt.

By the late twelfth or early thirteenth Dynasty (1786–1640



b.c.



) jars and boxes bearing
clay sealings stamped with Egyptian scarabs were being sent to Ashkelon from Egypt (tenta-
tive dating provided by Lanny Bell, who is studying some three dozen sealings discovered in
MB IIA Ashkelon). At the same time and somewhat later, large consignments of Canaanite
wine and oil were being shipped to the prosperous seaport of Avaris (Tell el-Dabºa) in the
Delta, where enclaves of Canaanites resided (Bietak 1996: 21–48; Holladay 1997). The ship-
ping containers, commonly known as “Canaanite jars,” are amphoras with a capacity of
approximately 25–30 liters. One of the amphoras that reached Avaris is made of clays from the
Levantine coast. Stamped on its handle is the name of a Canaanite or Amorite municipal offi-
cial (



˙·



ty-º



; commonly translated “mayor”), named Shimu (Bietak 1996: 60–62, fig. 51, pl.
25A–B). The amphora with everted, folded rim is far and away the most common type found
in MB IIA Ashkelon as well as in other parts of the Levant. From Avaris several Canaanite jars
of this type were transshipped some 150 km up the Nile to Lisht, the Middle Kingdom capital
(Arnold, Arnold, and Allen 1995).

Manfred Bietak, who has so successfully directed the excavations of Tell el-Dabºa, has
extrapolated from the imported amphoras to estimate that about two million Canaanite jars,
containing olive oil and wine, arrived at the seaport of Avaris during the MB II period.
Although the overland route via the “Ways of Horus” was in use then (Oren 1997), clearly
shipments of this magnitude must have arrived by sea. With this volume of exports from the
Levant, it is no wonder that port power played a dominant role in the configuration of settle-
ment patterns and economic networks from the lowlands to the uplands of Canaan. The
demand for highland produce—timber, resin, wine, and oil—along the coastal Levant had
never been greater.

In another paper I will try to demonstrate how the port power model helps in understand-
ing Phoenician colonization throughout the Mediterranean, where Phoenician seafarers went
in search of precious metals and other resources but settled mainly on the seacoasts. Only very
rarely do we find traces of Phoenician material culture beyond the seacoast, leading into the
hinterland to the mining and other resource areas. They realized that port power—control over
export abroad—was sufficient for the economic goals they had. In fact it was only during the
Industrial Revolution, when machines made control of production profitable, that we find mer-
chant capitalists venturing into that sector of the economy in a big way.

In these early mercantile ventures, we see noncoercive power exercised in varied and sub-
tle ways. So much so that Fernand Braudel’s remarks about a much later period resonate with
the more distant past:



Capitalism inserts itself into the chain leading from production to wholesale trade, not
seeking to take over entire responsibility for them, but to occupy the strategic points
controlling the key sectors of accumulation. (Braudel 1984: 3: 65)



An expanded version of this study, including Phoenician colonization in the Mediterra-
nean, will appear in a collection of my essays entitled



A Heap of Broken Images: Explorations
in Biblical Archaeology,



to be published by John Knox/Westminster Press.

oi.uchicago.edu





636



LAWRENCE E. STAGER



REFERENCES



Adams, R. M.
1974 Anthropological Perspectives on Ancient Trade.



Current Anthropology



15: 239–58.

Albright, W. F.
1960



The Archaeology of Palestine



. Baltimore: Penguin Books.

Arnold, D.; Arnold, F.; and Allen, S.
1995 Canaanite Imports at Lisht, the Middle Kingdom Capital of Egypt.



Ägypten und Levante



5: 13–32.

Bietak, M.
1996



Avaris: The Capital of the Hyksos: Recent Excavations at Tell el-Dabºa



. London: British
Museum Press.

1997 Avaris, Capital of the Hyksos Kingdom: New Results of Excavations. Pp. 87–139 in



The Hyk-
sos: New Historical and Archaeological Perspectives



, ed. E. Oren. University Museum Sympo-
sium Series 8. Philadelphia: University Museum, University of Pennsylvania.

Braudel, F.
1982



The Wheels of Commerce.



Civilization and Capitalism 15th–18th Century, Vol. 2. New York:
Harper and Row.

1984



The Perspective of the World



. Civilization and Capitalism 15th–18th Century, Vol. 3. New York:
Harper and Row.

Bronson, B.
1977 Exchange at the Upstream and Downstream Ends: Notes toward a Functional Model of the

Coastal State in Southeast Asia.



Michigan Papers on South and Southeast Asia



13: 39–52.

Butzer, K.
1997 Sociopolitical Discontinuity in the Near East ca. 2200



b.c.e.



: Scenarios from Palestine and Egypt.
Pp. 245–26







in



Third Millennium



b.c.



Climate Change and Old World Collapse



, eds. H. N.
Dalfes, G. Kukla, and H. Weiss. NATO ASI Series I: Global Environmental Change, vol. 49.
Berlin-Heidelberg: Springer-Verlag.

Cohen, S. L.
2000 Canaanites, Chronology, and Connections: The Relationship of Middle Bronze Age IIA Canaan

to Middle Kingdom Egypt. Ph.D. dissertation, Harvard University.

Curtin, P.
1984



Cross-Cultural Trade in World History



. Cambridge: Cambridge University Press.

Dever, W. G.
1987 The Middle Bronze Age: The Zenith of the Urban Canaanite Era.



Biblical Archaeology



50: 148–
77.

1993 The Rise of Complexity in the Land of Israel in the Early Second Millennium



b.c.



Pp. 98–109
in



Biblical Archaeology Today II: The Proceedings of the Second International Congress of Bib-
lical Archaeology, Pre-Congress Symposium: Population, Production and Power



,







eds. A. Biran
and J. Aviram. Jerusalem: Israel Exploration Society.

Fargo, V.
1979 Settlement in Southern Palestine during Early Bronze III



.



Ph.D. dissertation, University of
Chicago.

Finkelstein, I.
1993 The Socio-political Organization of the Central Hill Country in the 2nd Millennium



b.c.



Pp.
110–31 in



Biblical Archaeology Today II: The Proceedings of the Second International Con-
gress of Biblical Archaeology, Pre-Congress Symposium: Population, Production and Power



,
eds. A. Biran and J. Aviram. Jerusalem: Israel Exploration Society.

1998 The Great Transformation: The “Conquest”of the Highland Frontiers and the Rise of Territorial
States. Pp. 349–65 in



The Archaeology of Society in the Holy Land



, ed.







T. E. Levy. London and
Washington: Leicester University Press.



short

oi.uchicago.edu





637



THE ORGANIZATION OF MARITIME TRADE AND HINTERLAND PRODUCTION



Finkelstein, I., and Gophna, R.
1993 Settlement, Demographic and Economic Patterns in the Highlands of Palestine in the Chalco-

lithic and Early Bronze Periods and the Beginning of Urbanism.



Bulletin of the American
Schools of Oriental Research



289: 1–22.

Geertz, C.
1989 Review of



Southeast Asia in the Age of Commerce, 1450–1680



by A. Reid.



The New York
Review of Books



, February 16, 1989, 28–29.

Gophna, R., and Liphschitz, N.
1996 The Ashkelon Trough Settlements in the Early Bronze Age I: New Evidence of Maritime Trade.



Tel Aviv



23: 143–53.

Hall, S. R.
1985



Maritime Trade and State Development in Early Southeast Asia



. Honolulu: University of
Hawaii Press.

Harrison, T.
1997 Shifting Patterns of Settlement in the Highlands of Central Jordan during Early Bronze Age.



Bulletin of the American Schools of Oriental Research



306: 1–37.

Holladay, J.
1997 The Eastern Nile Delta During the Hyksos and Pre-Hyksos Periods: Toward a Systematic/

Socioeconomic Understanding. Pp. 183–252 in



The Hyksos: New Historical and Archaeologi-
cal Perspectives



, ed. E. Oren. University Museum Symposium Series 8. Philadelphia: Univer-
sity Museum, University of Pennsylvania.

Ilan, D.
1998 The Dawn of Internationalism—The Middle Bronze Age. Pp. 297–319 in



The Archaeology of
Society in the Holy Land



, ed. T. E. Levy. London and Washington: Leicester University.

Johnson, E. A. J.
1970



The Organization of Space in Developing Countries



. Cambridge, MA: Harvard University Press.

Kuhrt, A.
1995



The Ancient Near East: c. 3000–330 BC



. 2 vols. London and New York: Routledge.

Larsen, M. T.
1976



The Old Assyrian City-State and Its Colonies.



Mesopotamia







4. Copenhagen: Akademisk Forlag.

McGovern, P.
1998 Wine for Eternity.



Archaeology



51: 28–34.

McGovern, P., et al.
1997 The Beginnings of Winemaking and Viniculture in the Ancient Near East and Egypt.



Expedition



39: 3–21.

Miroschedji, P. de
1993 Jarmuth, Tel. Pp. 661–65 in



The New Encyclopedia of Archaeological Excavations in the Holy
Land



,







vol. 2, ed. E. Stern. Jerusalem: Israel Exploration Society & Carta and New York: Simon
& Schuster.

Moorey, P. R. S.
1990 From Gulf to Delta in the Fourth Millennium



bc



: The Syrian Connection.



Eretz-Israel



21 (Ruth
Amiran Volume): 62*–69*.

Oren, E.
1997 The “Kingdom of Sharuhen” and the Hyksos Kingdom



.



Pp. 253–83 in



The Hyksos: New Histori-
cal and Archaeological Perspectives



, ed. E. Oren. University Museum Symposium Series 8.
Philadelphia: University Museum, University of Pennsylvania.

Polanyi, K.
1977



The Livelihood of Man.



New York: Academic Press.

Posener, G.
1940



Princes et pays d’Asie et de Nubie



. Brussels: Fondation Éyptologique Reine Élisabeth.

oi.uchicago.edu





638



LAWRENCE E. STAGER



Prag, K.
1986 Byblos and Egypt in the Fourth Millennium BC.



Levant



18: 59–71.

Redmount, C.
1989 On an Egyptian/Asiatic Frontier: An Archaeological History of the Wadi







Tumilat. Ph.D. disser-
tation, University of Chicago.

Sethe, K.
1926



Die Ächtung feindlicher Fürsten, Völker und Dinge auf altägyptischen Tongefässscherben des
Mittleren Reiches.



Abhandlungen der Preussischen Akademie der Wissenschaften, philologische-
historische Klasse, No. 5. Berlin: Verlag der Akademie der Wissenschaften.

Smith, C. A., ed.
1976



Regional Analysis



, 2 Vols. NewYork: Academic Press.

Stager, L. E.
1985 The First Fruits of Civilization. Pp. 171–88 in



Palestine in the Bronze and Iron Ages



(Olga
Tufnell Volume), ed. J. N. Tubb. London: Institute of Archaeology.

1991



Ashkelon Discovered



. Washington, DC: Biblical Archaeology Society.
1992 The Periodization of Palestine from Neolithic through Early Bronze Times. Pp. 22–41 Vol. 1 and

Pp. 17–60 in Vol. 2, in



Chronologies in Old World Archaeology



, 3rd ed., ed. R. Ehrich. Chicago:
University of Chicago Press.

1993 Ashkelon



.



Pp.







103–12 in



The New Encyclopedia of Archaeological Excavations in the Holy
Land



,







vol. 1, ed.







E. Stern. Jerusalem: Israel Exploration Society & Carta and New York: Simon
& Schuster.

Veenhof, K. R.
1972



Aspects of Old Assyrian Trade and Its Terminology.



Leiden: E. J. Brill.

oi.uchicago.edu



639



PRODUCTION CENTERS OF COLLARED-RIM
PITHOI FROM SITES IN THE CARMEL COAST

AND RAMAT MENASHE REGIONS



Anat Cohen-Weinberger and Samuel R. Wolff



INTRODUCTION

Much attention has been centered in past decades of research on the typology, function,
provenance, chronology, technology and ethnic association of Iron Age I pithoi, especially of
the so-called collared-rim variety (e.g., Ibrahim 1978; Finkelstein 1988: 251–62; Biran 1989;
London 1989a; Zertal 1991; Esse 1992; Artzy 1994; Cohen-Weinberger and Goren 1996;
Clark and London 2000: 104–06 Wengrow 1996). Petrography and neutron activation analyses
on collared-rim pithoi from Shiloh, Sasa, Tel Dan, Tel Mevorakh, Tell Qasile, and Tell Keisan
have contributed to these debated issues (Courtois 1980; Yellin and Gunneweg 1989; Glass et
al. 1993; Cohen-Weinberger and Goren 1996). The present study concentrates on the collared-
rim pithoi type from several sites in an area bordered by the ºIron Ridge (Wadi ºAra) to the
south (El-A



˙



wat), Wadi el-Mil



˙



to the north (ºEn



Ó



agit), the Carmel coast (Tel Dor and Tel
Mevorakh) and ridge (Ramat Hanadiv) to the west and the Jezreel Valley to the east (Tell Qiri,



Tel Qashish, ºAfula) (fig. 32.1).



1



We also include six pithoi from Tel Dor with “wavy-band”
decoration (see Gilboa, chap. 9, this volume), previously termed “Phoenician” or “Tyrian”
pithoi (e.g., Biran 1989; Finkelstein 1988). Thus, the collared-rim pithoi analyzed here come
from a relatively wide geographic area. Most scholars would agree that the sites dealt with in
this study, lying inside the area described above, are located outside the area traditionally asso-
ciated with the Israelites; if anything, these sites are within the Sea Peoples and/or Canaanite
spheres of influence (e.g., Mazar 1994: 41–45; Singer 1994: 318–22). The samples date from
the end of the thirteenth/beginning of the twelfth century



b.c.



to the end of the eleventh cen-
tury



b.c.



(table 32.1). Our aim is to identify the place(s) of manufacture of the pithoi by means
of petrographic analysis and to shed light on some technological aspects of these pithoi. The
results contribute to a better understanding of economic activity and trade patterns in the early
Iron Age.

METHOD

The petrographic method identifies the clay and temper (non-plastic components) minerals.
Thin sections of the pithoi samples were examined under a petrographic (polarizing) microscope.
The samples were divided into petrographic families. A petrographic family encompasses ves-
sels that share similar petrographic affinities in both clay and temper. The classification is an
independent technical criterion for a comparative assortment of ceramic assemblages.



1. Two samples were taken from Beit Shean which is located outside the area that we are dealing with.



32



oi.uchicago.edu





640



ANAT COHEN-WEINBERGER AND SAMUEL R. WOLFF



FIGURE 32.1. Map showing sites with collared-rim pithoi.

oi.uchicago.edu

[image: image195.tif]



641



PRODUCTION CENTERS OF COLLARED-RIM PITHOI



For this study, ninety-four pithoi from seven sites were petrographically examined (table
32.1). Fourteen pithoi from ºEn



Ó



agit were examined with a magnifying glass (marked by *
in table 32.2). Their petrographic families were determined by comparing them with the pet-
rographically analyzed pithoi.

GEOLOGICAL SETTING OF THE SITES

The area under investigation is characterized by several different geological units. The
Carmel and Umm el-Fahm regions are anticlines characterized by Cenomanian and Turonian
carbonatic rocks (limestone and dolomite) accompanied by some volcanic intercalations
(basalts and tuffs; Picard 1956; Sass 1968: 115–16). Senonian rocks are exposed in the flanks
of these structures along Wadi ºIron and Wadi el-Mil



˙



. Ramat Menashe is a syncline between
the Carmel and Umm el-Fahm anticlines and is characterized by Eocene chalk and rendzina
soils, which is generally developed on chalky rocks. There are some Pleistocene basalt hills in
the eastern part of Ramat Menashe and in the Jezreel Valley (Picard 1956; Sneh, Bartov, and
Rosensaft 1998). The Carmel coast is characterized by a Pleistocene



kurkar



ridge (Picard
1956; Ravikovitch 1970). The coastal sand was mostly derived from the Nile, but some com-
ponents were derived from the valleys that are draining the Carmel and Ramat Menashe (i.e.,
Na



˙



al Meºarot and Na



˙



al Daliya). Quartz is the main component of the Israel coastal sand.
The carbonatic percentage in the sand between Hadera to Atlit is usually low, around 8%,
whereas that between Atlit and Haifa is exceptionally rich, comprising up to 90% of the sand
(Nir 1989: 14–15).

one shorter



Table 32.1.



Chronology of the Sites (According to the Excavators) and Number of Samples



Site Date
Number of pithoi

examined



ºEn



Ó



agit Mid–eleventh century



b.c.



(Wolff 1998: 453) 32+14*

El A



˙



wat Thirteenth–twelfth century



b.c.



(Zertal 1996: 5, 39) 6

Tel Dor Late twelfth or early eleventh century



b.c.



(“Sikil city” =
destruction phase), post “Sikil city” until ca. 980



b.c.



(Gilboa
1998: 413)

20

Tel
Mevorakh

Late eleventh/early tenth century



b.c.



(Stern 1978: 66,70) 1

Tel Beit
Shean

Late thirteenth–beginning of the twelfth century



b.c.



(A. Mazar excavations, unpublished)
2

Tell Qiri Eleventh century



b.c.



(Ben-Tor 1992: 1396) 5

Tel Qashish Twelfth century



b.c.



(R. Bonfil pers. comm.) 2

ºAfula Twelfth century



b.c.



(Dothan 1955: 51) 1

Ramat
Hanadiv

Unstratified (Wolff 2000) 11



* These samples were analyzed by magnifying glass.

oi.uchicago.edu





642



ANAT COHEN-WEINBERGER AND SAMUEL R. WOLFF



RESULTS



Petrographic Families (table 32.3)

Family A



The clay of this family is carbonatic with some foraminifera that are often silicified, and
about 5% silty quartz. Some clay lumps that were not mixed well in the matrix appear in the
Tel Mevorakh sample (table 32.3:73). Some vessels are made of noncarbonatic clay. The
non-plastic components of this family (10% of the paste) are mainly coarse chalk (3–4 mm),
basalt, and flint (fig. 32.2). The basalt fragments appear in various degrees of weathering.
Some of them are vesicular basalts having an opaque appearance and inclusions of chlorite.

The silty quartz points to an eolian material. The foraminiferous carbonatic clay and the
chalk fragments suggest a provenance rich with rendzina soil. The chalk fragments are prob-
ably remnants of the source material. In some samples the carbonate of the clay decomposed
due to a high firing temperature. That some vessels are made of noncarbonatic clay suggests
that the rendzina soil source was highly weathered. The silicified foraminiferas indicate an
Eocene chalk source. Eocene formations are exposed in Ramat Menashe (Sneh, Bartov, and
Rosensaft 1998), and in some parts of this area rendzina soils were developed (Ravikovitch
1970). The appearance of basalt fragments suggests an area of contact between the exposed
basalt and rendzina soils



.



The same basalts are well known in vessels from the Carmel coast
(e.g., Tel Nami; E. Marcus pers. comm.). Basalt is exposed in the eastern part of Ramat Mena-
she, in the ºIron and South Carmel ridges, and in the Jezreel Valley (Sneh, Bartov, and Rosen-
saft 1998). Pithoi belonging to this family are found at ºEn



Ó



agit, El A



˙



wat, Tel Dor, Tell Qiri,
Tel Qashish, and Ramat Hanadiv. Thus, the petrographic results strongly suggest that the raw
material of this family was derived from Ramat Menashe and the Carmel or Umm el-Fahm



Table 32.2.



Distribution of the Petrographic Families According to Sites



Petrographic
family/site A B C D E F G A/E



ºAfula — — — 1 — — — —

Tel Beit Shean — — — 1 1 — — —

ºEn Hagit 11+7* 11+2* — 7+4* 2 1 — 1*

El Ahwat 2 1 — 3 — — — —

Tel Dor 2 12 4 1 — — 1 —

Tel Mevorakh 1 — — — — — — —

Tel Qashish 1 1 — — — — — —

Tell Qiri 1 — — 3 1 — — —

Ramat Hanadiv 1 4 — 5 1 — — —



*These samples were analyzed by magnifying glass.

oi.uchicago.edu





643



PRODUCTION CENTERS OF COLLARED-RIM PITHOI



basalts. The distance from each of these sites to the clay and the non-plastic components as
well indicates that they were made somewhere in the vicinity of these sites.



Family B1



The clay of this family is carbonatic with some foraminifera, often silicified, with up to
5% siltic quartz grains. In some samples the clay was mixed with terra rossa soil appearing in
the form of mud balls. The non-plastic components (20% of the paste) consist of mainly fine
quartz grains (0.5 mm) and coarse chalk fragments (3 mm), some fine limestone fragments,
chert and, rarely, feldspar (fig. 32.3).

The clay is the same as the clay of family A and is most probably rendzina soil. Soils in
different degrees of weathering were used for making the pithoi as indicated by the range of
carbonatic to noncarbonatic matrixes. In some samples the carbonate of the matrix decom-
posed due to the firing in a high temperature. Pithoi of this family were found at ºEn



Ó



agit, Tel
Dor, El A



˙



wat, Tel Qashish, and Ramat Hanadiv.



2



For the same reasons mentioned above, the
source of the clay for this family is most likely Ramat Menashe. The source of the non-plastic
components is in the Mediterranean coast. Thus, the production center could have been in
Ramat Menashe or in a coastal site.



2. For the Iron Age pottery from Ramat Hanadiv see S. Wolff (2000).



FIGURE 32.2. Photomicrograph of pithos from ºEn



Ó



agit (table 32.3:6), family A. Chalk and basalt fragments
embedded in rendzina soil. Ck = chalk; Bt = basalt; Fa = foraminifera. XPL. Scale (white bar) = 0.5 mm.

oi.uchicago.edu





644



ANAT COHEN-WEINBERGER AND SAMUEL R. WOLFF



The introduction of terra rossa soil into the clay of this family may be linked to its high iron
content, which enabled the clay to achieve a sufficient degree of sintering, even at relatively
low firing temperature, due to its fluxing activity (Edwards and Segnit 1982). Traditional pot-
ters in the Hebron hills claim that adding terra rossa to the paste reduces the plasticity of the
clay and prevents the collapse of the vessel when wet or its excessive shrinkage during drying
(Krispil 1987). Terra rossa soil was also added to the paste of some Iron I pithoi in the Galilee
(Cohen-Weinberger and Goren 1996: 78, 81).



Family B2



The clay of this family is carbonatic with up to 5% siltic quartz grains. The non-plastic
components (about 15% of the paste) consist mainly of fine quartz grains and coarse limestone
fragments, some feldspar and, rarely, basalt fragments.

The silty quartz points to an eolian source material. The non-plastic components indicate
a coastal source. The basalt fragments were probably washed down to the coastal region via
one of the Carmel wadis. Only one pithos from Tel Qashish belongs to this family. The pro-
duction center could have been Ramat Menashe or the coastal Carmel region, like family B1.



Family C



The matrix consists of ferruginous clay. The non-plastic components of this family, which
comprise about 20% of the paste, are characterized by well-sorted inclusions of generally
rounded quartz and limestone (fig. 32.3), as well as some chert particles. Feldspar, orthopyrox-
ene, and hornblende appear in small quantities.



Amphiroa



sp. algae appear as well.



FIGURE 32.3. Photomicrograph of pithos from ºEn



Ó



agit (table 32.3:20), family B. Coarse chalk and fine
quartz embedded in rendzina soil. Ck = chalk; Qz = quartz; Fa = foraminifera. CPL. Scale (white bar) = 0.5 mm.

oi.uchicago.edu





645



PRODUCTION CENTERS OF COLLARED-RIM PITHOI



The petrographic appearance of the matrix is very similar to that of the



hamra



soils. This
is in accordance with the appearance of most of the non-plastic components, especially with
the



Amphiroa



sp., a



fossile directeur



of the Quaternary coast. The calcareous components sug-
gest a moderately weathered soil, which is typical of the



hamra



along the northern coast of
Israel and Lebanon. Pithoi of this family were found at Tel Dor and related to the type with the
wavy-band decoration. Some pithoi of this type from the Upper Galilee were made from the
same raw material (Cohen-Weinberger and Goren 1996: 78–79).



Family D



This family is characterized by the use of diversified shale fragments, some of which are
ferruginous, while others tend to be more clayey. Some ferruginous oolites also appear in the
clay. The matrix contains up to 5% silt. The non-plastic components (up to 5%) including
mainly quartz and siltstone and some limestone and sandstone (fig. 32.4).

A large bank of comparative data (Goren 1996: 49) shows that the lower formations of the
Lower Cretaceous lithological section were the sources of both clay and non-plastic compo-
nents. These formations outcrop extensively between the southern part of the Dead Sea and
Wadi Zarqa in Transjordan as well as in the Upper Galilee and southern Lebanon. Smaller out-
crops appear in eastern Samaria, in Wadi Mali



˙



and Wadi Farºah (Mimran 1969; Shaliv 1972;
Bender 1974). Pithoi of this family were found at ºEn



Ó



agit, El A



˙



wat, Tel Dor, Tell Qiri, Tel
Beit Shean, ºAfula, and Ramat Hanadiv. Similarly, pithoi belonging to this family were found
at Shiloh (Glass et al. 1993: 276–77, 279–82, n. 4), and at Iron I sites in the southern Samaria



FIGURE 32.4. Photomicrograph of a “wavy-band” pithos from Dor (table 32.3:63), family C. Rounded quartz
and limestone embedded in ferruginous clay. Ls = limestone; Qz = quartz. CPL. Scale (white bar) = 0.5 mm.

oi.uchicago.edu





646



ANAT COHEN-WEINBERGER AND SAMUEL R. WOLFF



survey (N. Shacharon unpublished data). The Lower Cretaceous lithological section was also
the source of some Iron I Age “Galilean” pithoi from Dan and Horvat Avot (unpublished, ana-
lyzed by Cohen-Weinberger).



Family E1



Both carbonatic and non-carbonatic clays occur in this family. The non-plastic compo-
nents (15% of the paste) include mainly rounded limestone and alkali-olivine basalt with alter-
ations to iddingsite (fig. 32.5).

The roundness of non-plastic components indicates the use of wadi sand as temper. A wadi
that drains a region of basalts and carbonatic rocks might be the source of this family. The raw
material of this family was utilized throughout the centuries at sites in the Jezreel Valley
(Goren 1991: 129–30; Cohen-Weinberger 1998: 58–59). The Jordan River, Qishon River, and
wadis that cut the Carmel ridge are also possible sources for the wadi sand. Pithoi of this
family were found at ºEn



Ó



agit, Beit Shean, and Ramat Hanadiv. One pithos from Tell Keisan,
petrographically examined and described by Courtois (1980: 356, table 1), most probably
belongs to this family as well.



Family E2



This family is similar to family E1. The clay is carbonatic with foraminifera. The non-
plastic components include rounded basalt and limestone fragments and also quartz grains. It
is the presence of these quartz grains that distinguish this subfamily from Family E1.



FIGURE 32.5. Photomicrograph of pithos from El A



˙



wat (table 32.3:71), family D. Ferruginous shales and silt-
stone fragment. Ss = siltstone; S = shales. PPL. Scale (white bar) = 0.5 mm.

oi.uchicago.edu





647



PRODUCTION CENTERS OF COLLARED-RIM PITHOI



The source of the quartz is the Nubian sandstone that drained into the Jordan River. Only
one pithos, from Tell Qiri, belongs to this family.



Family F



The clay is carbonatic with foraminifera. The non-plastic components are diversified and
(



⁄



15% of the paste) contain limestone,



kurkar



, quartzolite, several types of feldspar, red
algaes, miliolids, and vesicular basalt with inclusions of chlorite (figs. 32.6, 32.7). The chlo-
rite is an alteration of pyroxene or olivine, and it has a fibrous appearance bearing anomalous
colors.

The source of the non-plastic components is mixed. The origin of the



kurkar



, feldspar, red
algae, and miliolid is coastal. The origin of the basalt and the quartzolite is in the vicinity of
Zikhron Yaºaqov and Shefiyah (E. Sass, pers. comm.). A single pithos of this family was
found at ºEn



Ó



agit. The petrographic results suggest that this family was produced in this
region, probably not far from ºEn



Ó



agit.



Family G



This family is characterized by a micaceous clay. The non-plastic components are mainly
quartz (0.5 mm), elongated biotite (up to 1.5 mm), and minerals of the pyroxene group
(fig. 32.8).

This mineralogical assemblage points to Cyprus or the Aegean zone as the source. Only
one “wavy-band” decorated pithos from Tel Dor belongs to this family.



FIGURE 32.6. Photomicrograph of pithos from ºEn



Ó



agit (table 32.3:4), family E1. Rounded basalt and lime-
stone. Bt = basalt; Ls = limestone. CPL. Scale (white bar) = 0.5 mm.

oi.uchicago.edu





648



ANAT COHEN-WEINBERGER AND SAMUEL R. WOLFF



FIGURE 32.7. Photomicrograph of pithos from ºEn



Ó



agit (table 32.3:31), family F. Coastal non-plastic compo-
nents embedded in carbonatic clay. Mc = microcline; Md = miliolid; Kr = kurkar; Fa = foraminifera. CPL. Scale
(white bar) = 0.5 mm.



FIGURE 32.8. Photomicrograph of pithos from ºEn



Ó



agit (table 32.3:31), family F. As =



Amphiroa



sp. Algae;
Qz = quartz. CPL. Scale (white bar) = 0.2 mm.

oi.uchicago.edu





649



PRODUCTION CENTERS OF COLLARED-RIM PITHOI



DISCUSSION

Our main results, which serve as a basis for the following discussion, are as follows. Every
site includes pithoi made of several different raw materials (tables 32.2 and 33.3). Conversely,
almost every petrographic family is represented at a variety of sites. Collared rim pithoi made
of Families A, B, and D were found at sites located in the northern and southern borders of
Ramat Menashe, Carmel Ridge South, Jezreel Valley, and coastal sites. One pithos from Tel
Beit Shean belongs to family D (table 32.3:74), along with thirty pithoi from Shiloh (Glass
et al. 1993) and a considerable number collected from sites in the southern Samaria survey
(Nurit Shacharon, pers. comm.). This petrographic family was used to manufacture pithoi
found in the Upper Galilee as well (e.g., at Dan, unpublished). “Wavy-band” pithoi made of
family C were found at the coastal site of Tel Dor (table 32.3:61–66) and in the Upper Galilee
(Cohen-Weinberger and Goren 1996). Pithoi belonging to family E were found at ºEn



Ó



agit,
Tel Beit Shean, Tell Qiri, and Ramat Hanadiv. Only one pithos of family F was found at ºEn



Ó



agit, and one pithos of family G at Tel Dor. The later is related to the wavy-band decorated
pithoi and is the only pithos from those examined that was found to have been imported from
overseas, most probably from Cyprus.

Among the sampled collared-rim pithoi from ºEn



Ó



agit (table 32.2), there was a prefer-
ence for the raw material of family A (rendzina + basalt: 40%) over family B (rendzina +
coastal sand: 28%) and family D (Lower Cretaceous clay: 23%). On the other hand, at Tel
Dor a greater number of collared-rim pithoi belonged to family B (80%) over family A
(12.5%). These results may imply that the pithoi belonging to family B were manufactured at
coastal site, perhaps Dor itself, or another site near the coast.



FIGURE 32.9. Photomicrograph of a “wavy-band” pithos from Dor (table 32.3:66), family G. Clinopyroxene
(Cpx) embedded in micaceous clay. CPL. Scale (white bar) = 0.2 mm.

oi.uchicago.edu





650



ANAT COHEN-WEINBERGER AND SAMUEL R. WOLFF



Table 32.3.



Inventory and Results of the Petrographic Examined Pithoi



No Site Basket Locus Level Petrographic family



1 ºEn



Ó



agit 105/11 101 A

2 ºEn



Ó



agit 512/1 207 D

3 ºEn



Ó



agit 534/5 218 A

4 ºEn



Ó



agit 534/10 214 E1

5 ºEn



Ó



agit 554 224/5 B1

6 ºEn



Ó



agit 551 223 A

7 ºEn



Ó



agit 572/6+7 230 B1

8 ºEn



Ó



agit 581/1 224/5 A

9 ºEn



Ó



agit 584/5 231 A

10 ºEn



Ó



agit 585 224/5 A

11 ºEn



Ó



agit 587/2 231 D

12 ºEn



Ó



agit 587/15 231 D

13 ºEn



Ó



agit 591/13 231 A

14 ºEn



Ó



agit 601/7 234 B1

15 ºEn



Ó



agit 607/1 232 B1

16 ºEn



Ó



agit 612/4 232 A

17 ºEn



Ó



agit 618/10 237 B1

18 ºEn



Ó



agit 632/2+3 242 A

19 ºEn



Ó



agit 643/2 243 D

20 ºEn



Ó



agit 686 253 B1

21 ºEn



Ó



agit 689/1 253 A

22 ºEn



Ó



agit 695/6 253 A

23 ºEn



Ó



agit 711 224/5 B1

24 ºEn



Ó



agit 719/1 230 B1

25 ºEn



Ó



agit 740/1 250 B1

26 ºEn



Ó



agit 804/7 403 D

27 ºEn



Ó



agit 815/2 403 D

28 ºEn



Óagit 872/3 422 D

29 ºEn Óagit 872/5 422 B1

30 ºEn Óagit 882/4 422 E1
31 ºEn Óagit 883/7 412 F

32 ºEn Óagit 912/6 424 B1

33 ºEn Óagit* 652/1 250 A

34 ºEn Óagit* 640/4 248 B1

35 ºEn Óagit* 909/4 415 D

36 ºEn Óagit* 856/6 415 A/E

37 ºEn Óagit* 876/2 417 D

38 ºEn Óagit* 801/1 401 A

39 ºEn Óagit* 803/4 402 A

40 ºEn Óagit* 911/3 415 A

41 ºEn Óagit* 678/1 242 A

42 ºEn Óagit* 830/2 403 D

43 ºEn Óagit* 613/4 237 A

44 ºEn Óagit* 587/12 231 B1

45 ºEn Óagit* 591/16+22+4 231 A

46 ºEn Óagit* 872/2 422 D

47 Tel Dor — 12000 Post Destruction B1

oi.uchicago.edu



651PRODUCTION CENTERS OF COLLARED-RIM PITHOI

Table 32.3. Inventory and Results of the Petrographic Examined Pithoi (Cont.)

48 Tel Dor — 12003 Post Destruction B1

49 Tel Dor — 3269 Post Destruction B1

50 Tel Dor — 3097 Post Destruction B1

51 Tel Dor — 3097 Post Destruction B1

52 Tel Dor — 2298 Post Destruction B1

53 Tel Dor — 3256 Post Destruction A

54 Tel Dor — 2694 Post Destruction B1

55 Tel Dor 91156 9902 Post Destruction B1

56 Tel Dor — 9903 Post Destruction B1

57 Tel Dor — 3245 Post Destruction B1

58 Tel Dor 182033 18263 Destruction B1

59 Tel Dor 181189 18033 Destruction D

60 Tel Dor 181974 18237 Destruction B1

61 Tel Dor (“wavy band” pithoi) — 9829 Destruction A

62 Tel Dor (“wavy band” pithoi) — 9829 Destruction C

63 Tel Dor (“wavy band” pithoi) — 9985 Destruction C

64 Tel Dor (“wavy band” pithoi) — 7926 Destruction C

65 Tel Dor (“wavy band” pithoi) — 9813 Destruction C

66 Tel Dor (“wavy band” pithoi) — 7914 Destruction G

67 El A˙wat — 34273/32 A

68 El A˙wat — ? D

69 El A˙wat — 34275 B1

70 El A˙wat — 34246 D

71 El A˙wat — 33076 D

72 El A˙wat — 34248 A

73 Tel Mevorakh 971 6204 A

74 Tel Beit Shean 187221/80 D

75 Tel Beit Shean 184237 18411 E1

76 Tell Qiri 662/1 690 D

77 Tell Qiri 506/17 579 D

78 Tell Qiri 3034 1817 D

79 Tell Qiri 2052/1 1011 A

80 Tell Qiri 2271/6 690 E2

81 ºAfula Dothan 1955: fig.16:4 — D

82 Tel Qashish 2466/7 374 B2

83 Tel Qashish 2466/6 374 A

84 Ramat Hanadiv 3642 546 B1

85 Ramat Hanadiv 6161/1 582 D

86 Ramat Hanadiv 9138 943 D

87 Ramat Hanadiv 6164 681 D

88 Ramat Hanadiv 3698 524 D

89 Ramat Hanadiv 3649 501 E1

90 Ramat Hanadiv 9006/1 A904 A

91 Ramat Hanadiv 3941/1 — B1

92 Ramat Hanadiv 6714/1 — B1

93 Ramat Hanadiv 3130/4 311 D

94 Ramat Hanadiv 3189 325 B1

* These samples were analyzed by magnifying glass.

oi.uchicago.edu



652 ANAT COHEN-WEINBERGER AND SAMUEL R. WOLFF

For the present discussion it is important to describe the petrographic families of the
Samaria region (Nurit Shacharon, pers. comm. and Glass et al. 1993). The examined pithoi
from the southern Samaria survey and Shiloh excavation belong to three different petrographic
families: one made of the Lower Cretaceous formations (family D); one made of Mo˛a clay
and tempered by rhombic-shaped dolomite sand of Aminadav Formation (for a detailed
description of this family, see Goren 1995); and one made of Taqiye Marl Formation.

Pithoi made of the Lower Cretaceous formations (family D) were made in the northern
Samaria region (in Wadi Farºah and Wadi Mali˙), or in Transjordan. These pithoi were exported
to the southern part of the Samaria hills, and to the more distant regions of Ramat Menashe,
Carmel Ridge, Jezreel Valley, and the Carmel coast.

Pithoi made of a coastal raw material (families B and C) were mainly found at coastal
sites. Among the pithoi from the Jezreel Valley that were sampled, only one, from Tel Qashish
(table 32.3:82), was made of family B2. For comparison, petrographic results of pottery from
the southern Samaria survey and Shiloh do not indicate the presence of pithoi made of coastal
components in the region of Samaria.

Most of the southern Samaria survey sites, as well as Shiloh, contain pithoi belonging to
at least two petrographic families. Thus, it is significant that pithoi of the second and third fam-
ilies (those made of Taqiye and Mo˛a formations) are not found outside of the Samaria region.
Two possible interpretations of these observations are offered here. (1) If the provenance of
family D pithoi was Samaria, then we would expect to find pithoi made of all three families at
sites investigated in this study. Since this is not the case, the production center of family D
pithoi may be sought in Transjordan. (2) Pithoi were manufactured in several central work-
shops in the Samaria region, as suggested previously by Glass et al. (1993: 279). Thus, one
might assume that the inhabitants of the coast and the northern valleys were for some reason
in contact with the central workshop in northeastern Samaria and not with workshops in other
parts of Samaria.

The Relationship Between Typology and the Petrographic Families

Several typological subdivisions of collared-rim pithoi have been suggested in the past
(e.g., Sinclair 1960: 16–18; Kelso 1968: 63; Callaway 1969: 8–9). Finkelstein (1988: 276–77)
concluded that these typological divisions are not chronologically significant. The collared-
rim pithoi from Shiloh were divided into two different types, with each one assigned, on the
basis of petrographic analysis, its own production center, as described above (Glass et al.
1993: 279–80).

At ºEn Óagit the picture is more complicated. Most of the collared-rim pithoi belong to
one type, a short outfolded rim with short upright neck. Typologically, this type is very similar
to Group A from Shiloh. Vessels and sherds of this subtype grouped with families A, B1, D,
E1, and F, suggesting a multiplicity of sources. A less common type at ºEn Óagit (only three
examples) has a rim that is similar to the first type but with a long upright neck (table 32.3:26,
28, 30). The findspots of these sherds may suggest that this type is earlier than the first type
(Wolff forthcoming). When analyzed, two of these sherds grouped with family D and one with
family E1, suggesting dual sources for this type.

The collared-rim pithoi from Ramat Hanadiv are very similar in shape to the majority from
nearby ºEn Óagit. The collared-rim pithoi from Tel Dor, most belonging to family B, exhibit
a diversity in rim morphology.

oi.uchicago.edu



653PRODUCTION CENTERS OF COLLARED-RIM PITHOI

The following conclusions can be drawn:

1. Of the six analyzed sherds of “wavy-band” pithoi from Tel Dor, three sources were
determined: Cyprus (family G), the region of Ramat Menashe (family A), and the
northern Phoenician coast (family C). Previously examined pithoi of this type from the
Upper Galilee add at least two more petrographic families: one made of the Lower
Cretaceous formations (family D), and one made of Cenomanian formations, par-
ticularly Deir Hana Formation, of the Upper Galilee (Cohen-Weinberger and Goren
1996: 78–79).

2. The most typical “central hill country” type of collared-rim pithos was manufactured
in a variety of localities: the central hill country, Ramat Menashe, Galilee (Cohen-
Weinberger and Goren 1996: 79–80), Carmel coast, Jezreel Valley, Transjordan, and
Jordan Valley.

3. One sherd from ºEn Óagit with a double ridge (table 32.3:19) is associated with family
D (Wadi Farºah or Transjordan source). A similar vessel was found at Tell el-ºUmeiri,
Jordan (Herr, Geraty, and Younker 1991: 162, fig. 6:3–4). The distribution and petro-
graphic analysis suggest a Transjordanian source for this subtype.

Relations Between Chronology and Petrography

It seems that the same raw materials were employed by the potters during the thirteenth,
twelfth, and eleventh century pithoi. The earliest samples analyzed here, from El A˙wat and Tel
Beit Shean, grouped with families A, B, D, and E1. The same can be said for the latest samples
(postdestruction levels from Tel Dor, ºEn Óagit, Tel Mevorakh). The sources of families C and
G ceased to be exploited after the destruction phase at Tel Dor (late twelfth or early eleventh
century b.c.). This is not so much a chronological phenomenon but rather a typological one;
that is, these sources were utilized for the production of “wavy-band” pithoi, whose floruit
is earlier than that of collared-rim pithoi.

The earliest pithoi analyzed here, both of the long-necked and short-necked variety, grouped
with several families. The majority, however, grouped with families D and E1, whose prove-
niences are found in the eastern part of Cisjordan or even in Transjordan. Thus, the contention
that the earliest pithoi are coastal in origin (Artzy 1994: 138) is put into question. The analysis
of the pithoi from Tel Nami will be crucial in determining the reliability of this observation.

Petrographic Analysis and Neutron Activation Analysis

The pithos from Tel Mevorakh (table 32.3:73) was previously analyzed by means of neu-
tron activation analysis by Yellin and Perlman (1978). They compared the elemental compo-
nents of this pithos to those of some presumed local pottery (two bichrome and two coarse
ware sherds) from Tel Mevorakh, a presumption that needs to be proven analytically, and to a
reference collection from Tel Megadim (Ariel et al. 1985: 150–51; Yellin and Perlman 1978:
90, 94, table 7) that contains samples from several periods and several vessel types (unpub-
lished list in Tel Megadim archives). They concluded that its chemical composition was local
to the coastal region of Tel Mevorakh (Yellin and Gunneweg 1989: 139). According to our
petrographic analysis, the sample that they defined as being local to Tel Mevorakh belongs to
our family A and therefore seems to have been produced in the Ramat Menashe region, which

oi.uchicago.edu



654 ANAT COHEN-WEINBERGER AND SAMUEL R. WOLFF

is not so far from Tel Mevorakh but has a petrographic fingerprint that clearly distinguishes
itself from the Carmel coast.3

Petrography and Trade

Previous studies of Iron Age I collared-rim pithoi have also suggested considerable vari-
ability in the production localities in each examined assemblage. As mentioned above, petro-
graphic analysis of the collared-rim pithoi from Shiloh indicates that they were produced in
two different, relatively distant, regional workshops (Glass et al. 1993). Petrographic analysis
of “Galilean” and “Tyrian” pithoi from Sasa indicates the existence of two separate, contem-
poraneous production centers for each of them (Cohen-Weinberger and Goren 1996). Neutron
activation analysis of the collared-rim pithoi from Tel Dan indicates that nearly one-third of
the analyzed samples were locally made, whereas the rest were imported from seven different
geographical regions (Yellin and Gunneweg 1989). At a manufacturing site, one would expect
to find pithoi made of the same raw material, which is not the case. In our opinion these results
imply that pithoi were not manufactured at any of the sites that were sampled until now (with
the possible exception of Tel Dor; see above).

The heterogeneity of sources causes us to reject a model according to which the collared-
rim pithoi were manufactured by itinerant potters (London 1989a: 44) and were marketed in
nearby settlements, as ethnoarchaeological research conducted in Cyprus, Crete, and Jordan
suggests (London 1989b: 68–69; Voyatzoglou 1974). For example, it makes no sense that a
vessel produced in northeastern Samaria was shipped to the coastal region (where identical
vessels were also manufactured) when the same itinerant potters could have produced them
there as well. Rather, they were manufactured in independent, specialized potters’ workshops.
The heterogeneity of sources have been interpreted by some as an indication that the pithoi
were used as vases-récipients, that is, pottery whose value was in the product it contained and
not in the vessel itself (e.g., Artzy 1994; Wengrow 1996). The extraordinary weight of pithoi
filled with produce leads us to reject this theory. Rather, collared-rim pithoi were used first and
foremost as vases-marchandises (pottery traded not for the value of its contents but for its own
value). A pithos, in comparison to present-day concepts, was an expensive item since it de-
manded high technical skill to produce. We offer two explanations regarding the heterogeneity
of sources at one site. (1) Purchasers of pithoi sought the maximum combination of high qual-
ity and reasonable price. Because of the price and quality, which are subject to fluctuation, the
consumer changed preferences according to the situation and times. Shifts in preferences are
reflected in the heterogeneity of sources. (2) Pithoi had a long life (London 1989a: 44), so much
so that an assemblage of pithoi could have been purchased during a stretch of several years.

The results of our petrographic analysis strengthen the hypothesis that there is no connec-
tion between the place where pithoi were manufactured and a specific ethnic group (as argued
by earlier researchers), rather that they were manufactured and used by a variety of ethnic
groups.

3. J. Yellin (Hebrew University) conducted neutron activation analysis on the same collared-rim pithoi from
ºEn Óagit that are described above, but the results are still unavailable. A comparision of his results to those
reported in this paper will provide complementary information.

oi.uchicago.edu



655PRODUCTION CENTERS OF COLLARED-RIM PITHOI

ACKNOWLEDGMENTS

We are grateful to Nurit Shacharon for providing us with the unpublished data of her south
Samaria survey petrographic analysis; Yuval Goren, Eitan Sass, and Ezra Marcus for their
respective comments on the petrographic analysis; and Ruhama Bonfil, Ayelet Gilboa, Yizhar
Hirschfeld, Amihai Mazar, and Adam Zertal for providing Cohen-Weinberger with samples
from their respective excavations. We gratefully acknowledge the financial support provided
by the National Center for Collaboration between Natural Science and Archaeology (admin-
istered by the Weizmann Institute). The petrographic analysis was conducted in the laboratory
of the Israel Antiquities Authority, Jerusalem. The map (fig. 32.1) was prepared by Leticia
Barda, Israel Antiquities Authority.

REFERENCES

Ariel, D. T.; Sharon, I; Gunneweg, J.; and Perlman, I.
1985 A Group of Stamped Hellenistic Storage Jar Handles from Dor. Israel Exploration Journal 35:

135–52.

Artzy, M.
1994 Incense, Camels and Collared-Rim Jars: Desert Trade Routes and Maritime Outlets in the Sec-

ond Millennium. Oxford Journal of Archaeology 13: 121–47.

Bender, F.
1974 The Geology of Jordan. Berlin and Stuttgart: Gebrüder Borntraeger.

Ben-Tor, A.
1992 Qiri, Tel. Pp. 1228–29 in The New Encyclopedia of Archaeological Excavations in the Holy

Land, vol. 4, ed. E. Stern. Jerusalem: Israel Exploration Society & Carta and New York: Simon
& Schuster.

Biran, A.
1989 The Collared-Rim Jars and the Settlement of the Tribe of Dan. Pp. 71–96 in Recent Excavations

in Israel: Studies in Iron Age Archaeology, eds. S. Gitin and W. G. Dever. Annual of the Ameri-
can Schools of Oriental Research 49. Winona Lake, IN: Eisenbrauns.

Callaway, J. A.
1969 The 1966 ºAi (et-Tell) Excavations. Bulletin of the American Schools of Oriental Research 196:

2–16.

Clark, D. R., and London, G. A.
2000 Investigating Ancient Ceramic Traditions on Both Sides of the Jordan. Pp. 100–110 in The

Archaeology of Jordan and Beyond: Essays in Honor of James A. Sauer, eds. L. E. Stager, J. A.
Greene, and M. D. Coogan. Winona Lake, IN: Eisenbrauns.

Cohen-Weinberger, A.
1998 The Typology and Petrography of the Egyptian Pottery from Tel Beth Shean in the Light of the

Renewed Excavation. M.A. thesis, Department of Archaeology, Hebrew University (Hebrew).

Cohen-Weinberger, A., and Goren, Y.
1996 Petrographic Analysis of Iron Age I Pithoi from Tel Sasa. ºAtiqot 28: 77–83.

Courtois, L.
1980 Examen pétrographique et caractérisation matérielle de quelques céramiques de Tell Keisan.

Pp. 353–60 in Tell Keisan (1971–1976): une cité phénicienne en Galilée, eds. J. Briend and
J. B. Humbert. Orbis Biblicus et Orientalis. Fribourg: Editions universitaires.

Dothan, M.
1955 The Excavations at ºAfula. ºAtiqot 1: 19–70.

Edwards, W. I., and Segnit, E. R.
1982 Pottery Technology of the Chalcolithic Site of Teleilat Ghassul (Jordan). Archaeometry 26: 69–77.

oi.uchicago.edu



656 ANAT COHEN-WEINBERGER AND SAMUEL R. WOLFF

Esse, D. L.
1992 The Collared Pithos at Megiddo: Ceramic Distribution and Ethnicity. Journal of Near Eastern

Studies 51: 81–103.

Finkelstein, I.
1988 The Archaeology of the Israelite Settlement. Jerusalem: Israel Exploration Society.

Gilboa, A.
1998 Iron Age I Pottery Evolution at Dor: Regional Contexts and the Cypriot Connection. Pp. 413–25

in Mediterranean Peoples in Transition: Thirteenth to Early Tenth Centuries b.c.e., eds. S. Gitin,
A. Mazar, and E. Stern. Jerusalem: Israel Exploration Society.

Glass, J.; Goren, Y; Bunimovitz, S.; and Finkelstein, I.
1993 Petrographic Analysis of Middle Bronze Age III, Late Bronze Age and Iron Age I Pottery

Assemblages. Pp. 271–86 in Shiloh. The Archaeology of a Biblical Site, ed. I. Finkelstein. Tel
Aviv: Tel Aviv University.

Goren, Y.
1991 Petrographic Examination of the Ceramic Assemblage from Tel ºAmal. ºAtiqot 20: 129–30.
1995 Shrines and Ceramics in Chalcolithic Israel: The View Through the Petrographic Microscope.

Archaeometry 37: 287–305.
1996 The Southern Levant in the Early Bronze Age IV: The Petrographic Perspective. Bulletin of the

American Schools of Oriental Research 303: 34–72.

Herr, L. G.; Geraty, L. T.; and Younker, R.
1991 Madaba Plains Project: The 1989 Excavations at Tell el ºUmeiri and Vicinity. Annual of the

Department of Antiquities of Jordan 35: 155–80.

Ibrahim, M. M.
1978 The Collared-Rim Jar of the Early Iron Age. Pp. 117–26 in Archaeology in the Levant: Essays

for Kathleen Kenyon, eds. P. R. Moorey and P. J. Parr. Warminster: Aris and Phillips.

Kelso, J. L.
1968 The Excavation of Bethel (1934–1960). Annual of the American Schools of Oriental Research

39. Cambridge, MA: American Schools of Oriental Research.

Krispil, N.
1987 As Clay in the Hands of the Potter. Pp. 43–55 in Craftsmen in Mount Hebron. Tel Aviv: Kibbutz

Movement. (Hebrew).

London, G.
1989a A Comparison of Two Contemporaneous Lifestyles of the Late Second Millennium b.c.e. Bul-

letin of the American Schools of Oriental Research 273: 37–55.
1989b On Fig Leaves, Itinerant Potters and Pottery Production Locations in Cyprus. Pp. 65–80 in

Cross-Craft and Cross-Cultural Interactions in Ceramics, eds. P. E. McGovern and M. D. Notis,
Ceramic and Civilization, vol. 4. Westerville, OH: American Ceramic Society.

Mazar, A.
1994 The 11th Century b.c.e. in the Land of Israel. Pp. 39–57 in Proceedings of the International

Symposium on Cyprus in the 11th Century b.c.e., Nicosia, 30–31 October 1993, ed. V. Kara-
georghis. Nicosia: Leventis Foundation.

Mimran, Y.
1969 The Geology of the Wadi Malih Area. M.A. thesis, Department of Geology, Hebrew University,

Jerusalem (Hebrew).

Nir, J.
1989 Sedimentological Aspects of the Israel and Sinai Mediterranean Coasts. Jerusalem: Geological

Survey of Israel (Hebrew).

Picard, L.
1956 Zikhron Yaºakov, 1:100,000. Geological Map of Israel. Series C: Northern Part. Jerusalem:

Ministry of Development and Geological Survey.

Ravikovitch, S.
1970 Soil Map 1:250,000. Rehovot: Hebrew University Faculty of Agriculture.

oi.uchicago.edu



657PRODUCTION CENTERS OF COLLARED-RIM PITHOI

Sass, E.
1968 Geology of the Umm El Fahm Area, Northern Israel. Israel Journal of Earth Sciences 17: 115–30.

Shaliv, G.
1972 The Geology of the Wadi Farºah Area. M.A. thesis, Department of Geology, Hebrew University,

Jerusalem (Hebrew).

Sinclair, L. A.
1960 An Archaeological Study of Gibeah (Tell el-Fûl). Annual of the American Schools of Oriental

Research 34–35:1–52. New Haven: American Schools of Oriental Research.

Singer, I.
1994 Egyptians, Canaanites, and Philistines in the Period of the Emergence of Israel. Pp. 282–338

in From Nomadism to Monarch: Archaeological and Historical Aspects of Early Israel, ed.
I. Finkelstein and N. Naªaman. Jerusalem: Yad Izhak Ben-Zvi and Israel Exploration Society.

Sneh, A.; Bartov, Y.; and Rosensaft, M.
1998 Geological Map of Israel 1:200,000. Sheet 1. Jerusalem: Geological Survey of Israel.

Stern, E.
1978 Excavations at Tel Mevorakh, From the Iron Age to the Roman Period. Qedem 9. Jerusalem:

Hebrew University.

Voyatzoglou, M.
1974 The Jar Makers of Thrapsano in Crete. Expedition 16: 18–24.

Wengrow, D.
1996 Egyptian Taskmasters and Heavy Burdens: Highlands Exploration and the Collared Rim Pithos

of the Bronze/Iron Age Levant. Oxford Journal of Archaeology 15: 307–26.

Wolff, S. R.
1998 An Iron Age I Site at ºEn Óagit (Northern Ramat Menashe). Pp. 449–54 in Mediterranean Peoples

in Transition: Thirteenth to Early Tenth Centuries b.c.e., eds. S. Gitin, A. Mazar, and E. Stern.
Jerusalem: Israel Exploration Society.

2000 Iron Age Pottery. Pp. 529–36 in Ramat Hanadiv Excavations: Final Report of the 1984–1998
Seasons, Y. Hirschfeld. Jerusalem: Israel Exploration Society.

Forthcoming ºEn Óagit. A Middle Bronze Age III and Iron Age I Village in Northern Ramat Menashe: Final
Report. Jerusalem: Israel Antiquities Authority.

Yellin, J., and Gunneweg, J.
1989 Instrumental Neutron Activation Analysis and the Origin of Iron Age I Collared-rim Jars and

Pithoi from Tel Dan. Annual of the American Schools of Oriental Research 49: 133–41.

Yellin, J., and Perlman, I.
1978 Provenance of Iron Age Pottery from Tel Mevorakh. Pp. 86–94 in Excavations at Tel Mevorakh,

From the Iron Age to the Roman Period, E. Stern. Qedem 9. Jerusalem: Hebrew University.

Zertal, A.
1991 Israel Enters Canaan. Biblical Archaeology Review 17: 28–49.
1996 El-A˙wat, A Fortified Site of the Sea Peoples near the ºIron Valley. A Preliminary Report of

Three Seasons of Excavations 1993–1995. Haifa: University of Haifa (Hebrew).

oi.uchicago.edu



658 ANAT COHEN-WEINBERGER AND SAMUEL R. WOLFF

oi.uchicago.edu



659



THE EARLY BRONZE AGE IA OF
SOUTHWESTERN CANAAN



Yuval Yekutieli



The following paper focuses on a certain time and space within the Early Bronze (EB) I
framework—its early part—the EB IA of southwestern Canaan. It describes its attributes,
reconstructs some of its socioeconomic aspects, and discusses its contribution to several
related issues.

The most accepted chronological scheme for southwestern Canaanite EB I argues for its
division into two phases: early—EB IA, and late—EB IB (Amiran and Gophna 1992; Stager
1992: 28). For a summary of researchers’ opinions see table 33.1. Since S. Yeivin’s (1961)
excavations at Erani, the EB IB of southwestern Canaan became relatively well documented
and defined (Gophna 1990b). Its remains were recorded in many sites within the research area
(fig. 33.1) such as ºEn Besor, Tel



Ó



alif Terrace, Maªa



˙



az, Erani, and Afridar (see also Porat
1992: map 1), and it has been found that its most characteristic feature is an abundance of
Egyptian, or Egyptian-inspired finds. However, the phase preceding it, the EB IA, is poorly
understood, and needs further clarification (Dessel 1991: 93).

THE ENVIRONMENT

The area dealt with in this paper (fig. 33.2) is bordered on the south by Nahal Besor (Wadi
Ghazzeh) and its large tributary Na



˙



al Beersheva. The valleys separating the southern She-
phelah (the inland plain) from the Hebron mountains border it on the east. Its northern limit is
the course of Na



˙



al Lachish to its point of debouchement. This region is considered, in this
paper, as a single unit, defined as “southwestern Canaan” (others who have treated more or
less the same geographical entity as a distinguishable unit in the EB I include Porat 1992;
Kempinski 1992; Brandl 1992). Three geographical subregions are located within the research
area (fig. 33.3):



“Pleshet” (biblical Philistia) is a lowland coastal plain covered by sand dunes near the
shore and, farther to the east, by



hamra



and



qurqar



(calcareous sandstone) soils. The
“northern Negev” is notable for its plains and low hilly areas covered by loessy soils.
The “southern Shephelah” is one of rounded, chalky limestone hills covered by a harder
top layer, known locally as Nari. (Dan 1988: 51)



The soils in the research area are fertile, and at present wheat is grown in this region by the
dry fall method. Irrigation enables the cultivation of additional crops including vegetables and
fruit (Dan 1988: 56). A transition between two major natural vegetation zones occurs within
the limits of the research area. The northeast lies within the Mediterranean zone, while the
southern part is included in the more arid Irano-Turonian zone (Danin 1988: 59). Rainfall
occurs in southwestern Canaan from October to May (Zangvil 1988: 45). Average rainfall in



33



oi.uchicago.edu





660



YUVAL YEKUTIELI



FIGURE 33.1. Map of EB I sites mentioned in the text.

oi.uchicago.edu

[image: image196.tif]



661



THE EARLY BRONZE IA OF SOUTHWESTERN CANAAN



the northern part of the research area is 400 mm per year, and in its southern part less than 200
mm. The fluctuations of the 200 mm isohyte, the so-called aridity line, passes within the
research area’s limits (Meigs 1953). An area with less than that average is usually defined as
a desert zone (Stern 1988: 15). Recent interdisciplinary research concerning the ancient cli-
mate of Canaan, based on different methods and environmental aspects, has determined that
in the fourth millennium



b.c.



, which includes the latter half of the Chalcolithic period and
the EB I, the climate of Canaan was moister than today (Copeland and Vita-Finzi 1978;
Horowitz 1974: 407–11; 1979; Neev and Emery 1967; Govrin 1990: 108; Frumkin et al. 1991;



FIGURE 33.2. The research area of Southwestern Canaan.

FIGURE 33.3. Geographical subregions within the research area of Southwestern Canaan.

oi.uchicago.edu

[image: image197.tif][image: image198.tif]



662



YUVAL YEKUTIELI



Goldberg and Rosen 1987; Goldberg 1983: 147). The evidence further shows that at the end
of the fourth millennium



b.c.



(end of EB I), a decrease in rainfall occurred, which brought
about climatic conditions close to those prevailing today (Govrin 1990: 108; Frumkin et al.
1991; Neev and Emery 1967; Baruch 1986).

THE SITES

In the research area eight EB IA sites are presently known: Nizzanim, Afridar (within the
modern town of Ashkelon), Erani, Gat-Guvrin, the Northwestern settlement of Lachish, Tel



Ó



alif Terrace, Taur-Ikhbeineh, and Site H (fig. 33.4). For the purposes of the research, the
known sites were approached in the following way:

1. Sites that were excavated in the past were analyzed using the data published in the ex-
cavation reports: Lachish (Tufnell et al. 1958), Site H (Macdonald 1932; Gophna 1976,
1990a), and Tel



Ó



alif Terrace 101 (Dessel 1991).



1



Where possible unpublished material
was also studied.



2



2. At sites known from surveys only, trial excavations were initiated: Nizzanim (Yekutieli
and Gophna 1994) and Taur-Ikhbeineh (Oren and Yekutieli 1992).

3. The finds from sites excavated but not published were processed, researched, and
prepared for publication. These include the “Silo” site at Tel



Ó



alif Terrace (Alon and



1. The author wishes to thank J. P. Dessel and the Lahav Research Project for allowing use of drawings of finds
from Tel



Ó



alif Terrace Site 101.

2. The author wishes to thank Ram Gophna for making material from his excavations at Site H available for
this study, and for allowing use of drawings of some finds from that site.



Table 33.1.



Terminology, Inner Division, and Absolute Chronology of EB I



DATE



b.c.



2800 2900 3000 3100 3200 3300 3400 3500

Wright 1937 EB II EB IB EB IA Late Chal.

Wright 1958 EB II EB IC EB IB EB IA Ghassulian

Albright 1960 EB II EB I Late Chal.

Kenyon 1960 EB II EB I Proto Urban Ghassul

Hennessy 1967 EB II EB I Proto Urban

Lapp 1968: 70 EB IC EB IB EB IA Late Chal.

de Vaux 1970 EB II EB IB EB IA Ch. Superior

Amiran 1970: 78 EB II EB IB EB IA Late Chal.

Callaway1972 EB II IC IB IA Late Chal.

Gilead 1993 EB II Late EB I Middle EBI Early EB I

Amiran and
Gophna 1992

Late Middle Early
EB II EB IB EB IB EB IB EB IA

2800 2900 3000 3100 3200 3300 3400 3500

Egypt Early Dynastic Period Dyn. 0 \ Naq. 3 2D 2C 2B 2A

oi.uchicago.edu





663



THE EARLY BRONZE IA OF SOUTHWESTERN CANAAN



Yekutieli 1995), and the renewed excavations at Tel Erani



3



(Area DII: Kempinski and
Gilead 1991).

4. The unpublished material from Gat-Guvrin (Perrot 1961) was not available for study,
and the EB IA site of Afridar was only revealed and excavated just as this paper was
being written. Therefore, these sites were not investigated in this study.

Most of the excavations treated here are of limited size. For the set task of defining a
regional culture, data from large-scale excavations at major sites would have been preferable.
However, even in the absence of such material, it was possible, from the available evidence,
to identify such a chronological, regional entity and further distinguish within it two sub-
phases—early and late EB IA (EB IA1 and EB IA2).



Features of the EB IA Settlements



A typical EB IA settlement in the research area is a small, sparsely populated, unfortified
village. Within its confines, between the dwelling units, are many open courtyards, probably
reserved for household activities. Proximity to water resources and arable land seem to have
been the attractions for settlers; none of the settlements seem to have been located for strategic
purposes. The size of a typical EB IA settlement in the research area is between 5 to 20
dunams. After consulting the relevant literature on ancient demography (Johnson 1973: 51–
70; Broshi and Gophna 1984: 147–57; Rollefson and Kohler-Rollefson 1989; Marfoe 1980:
320; Gophna and Portugali 1988: 11–28; Hassan 1980), a coefficient of fifteen people per
dunam was chosen for estimating the EB IA sites’ populations. Accordingly the largest site’s
population is calculated to be of approximately 300 people. Within the settlements two main
types of dwellings are encountered: caves and free-standing structures. The first type is con-
fined to the hilly, rocky region in the east of the research area (Lachish, Tel



Ó



alif Terrace). In



3. The author wishes to thank Aharon Kempinski and Isaac Gilead for the opportunity to study the pottery of
their excavation at Erani.

FIGURE 33.4. EB IA sites in the research area of Southwestern Canaan.

oi.uchicago.edu

[image: image199.tif]



664



YUVAL YEKUTIELI



that region people learned to cope with and utilize their environment. Irregular-shaped dwell-
ing caves are carved out of the layers of soft limestone, leaving the harder Nari layer above as
a roof. Similar practices are known from the local Chalcolithic culture as well (Govrin 1987;
Alon 1991) and therefore imply a continuity of traditions. Free-standing structures, built from
either stone, mudbrick, or a combination of the two, appear all over the research area. Such
structures were excavated at Site H, Tel



Ó



alif Terrace, Nizzanim, Afridar, and some remains
hint at the existence of free-standing structures at Lachish as well. Rectangular broadroom
plans (Site H) link them to local (i.e., southern Canaanite) Chalcolithic traditions (Porath 1985:
14); rectilinear structures with rounded corners at Site H imply northern late EB IA influences
(Braun 1985, 1989a); and curvilinear structures at Afridar (Braun and Gophna 1994) suggest
connections with northern EB IA.

Features within the sites included many storage and refuse pits, silos, and unidentified instal-
lations. The recurrence of installations for grain storage is especially noteworthy. A pottery kiln
was excavated at Lachish Northwestern settlement, at Cave 1525 (Tufnell et al. 1958: fig. 11).



Stratigraphy of the Sites



The majority of the sites mentioned above had more than a single EB IA stratum. A cor-
relation, made through the use of similar attributes of the material culture between the several
sites, has proven the existence of two phases of southwestern Canaanite EB IA—henceforth
labeled EB IA1 (the earlier) and EB IA2 (the later). The two-tiered, chronological structure of
EB IA was first attested at Nizzanim during the analysis of the pottery assemblage of that
period. The pottery of Stratum V was noted to have different attributes from the pottery of
Strata IV and III (Yekutieli and Gophna 1994). Macdonald’s (1932) notes accompanying the
published pottery of Site H, indicating the elevation in which it was found, enabled a recon-
struction of the site finds’ stratigraphical order with a reasonable degree of assurance. The
organization of pottery vessels according to their elevations revealed, at Site H, the same
attributes which differentiate between the early and late EB IA pottery at Nizzanim, in the
same sequence (Yekutieli 1992). A final confirmation of these observations came from Tel



Ó



alif Terrace Site 101. A rearrangement of the many pottery drawings in Dessel’s (1991) the-
sis, according to their strata, produced yet again the same pattern (although with local varia-
tions), in the same sequence (Yekutieli 1992). Following the newly revealed pattern, it could
be determined that EB IA1 remains are found at Nizzanim Stratum V, Tel



Ó



alif Terrace Site
101 Stratum V, the Tel



Ó



alif Terrace “Silo” site Stratum III, the lower levels of Site H, and in
certain caves at the Northwestern settlement of Lachish. The EB IA2 material culture is found
at Nizzanim Strata IV–III, Tel



Ó



alif Terrace Site 101 Stratum IV (and perhaps Stratum III as
well), at Site H in the upper levels and in the dwellings at Taur-Ikhbeineh Stratum IV, at Tel
Erani (Kempinski and Gilead’s excavation) Layer D, and certain caves of the Northwestern
settlement of Lachish (table 33.3).

It is important to note that—excluding Gat-Guvrin and Afridar, from which data are
unavailable yet—only at one site in the research area, Tel



Ó



alif Terrace, were remains of the
Chalcolithic period encountered (“Silo” site Stratum IV). There, an EB IA1 stratum is super-
imposed directly on the Chalcolithic stratum (Alon and Yekutieli 1995), thus ruling out the
suggestion of a Chalcolithic–EB IA overlap (see table 33.2 for various opinions about the
chalcolithic–EB I transition). Another conclusion from this observation is that all other sites
were newly established in EB IA.

oi.uchicago.edu





665



THE EARLY BRONZE IA OF SOUTHWESTERN CANAAN



.



a.



Yekutieli and Gophna 1994



b.



Dessel 1991



c.



Alon and Yekutieli n.d.



d.



Macdonald 1932



e.



Tufnell et al. 1958



Table 33.2.



Chalcolithic–Early Bronze I: Nature and Degree of Continuity



Long time gap Sequential character Overlap



Elliott 1978 Wright 1937 Kenyon 1960

Gilead 1993 Albright 1954, 1957 Contenson 1961

Lapp 1970 Hennessy 1967

Amiran 1985 Miroschedji 1971

Joffe 1993 Chapman 1990



Table 33.3.



Synchronological Sequence of Strata at Some Fourth Millennium



b.c.



Sites



EB II EB IB EB IA Chalcolithic

3000 3050 3300 3400 3450 3650 3900

Egyptian
terminology

Dynasty 0
/Naqada 3

Naqada
2d 2c 2b 2a 1

Nizzanim



a



III IV V

Tel



Ó



alif 101



b



II III IV V

Tel



Ó



alif “Silo”



c



I–II III IV

Site H



d



Late Early

Lachish NW 1520
settlement



e



1509 1517, 1540

1534 1528, 1550

Taur–Ikhbeineh



f



II III IV V?



14



C: 3370

Erani DII



g



C



14



C: 3331
D

Erani



h



V, VI–XI



14



C: 3307
(Str. XI)

XII

Maadi



i



Settlement

Dynasty 0
/Naqada 3

Naqada

0 / 3 2d 2c 2b 2a 1

3000 3050 3300 3400 3450 3650 3900



f.



Oren and Yekutieli 1992



g.



Kempinski and Gilead 1991



h.



Yeivin 1961



i.



Rizkana and Seeher 1988, 1989

oi.uchicago.edu





666



YUVAL YEKUTIELI



FIGURE 33.5. Pottery forms of EB IAl.



Type Site Stratum/Locus Reference



1. Storage jar Lachish NW settlement Cave 1503 Tufnell et al. 1958: pl. 56:26
2. Storage jar Lachish NW settlement Cave 1517 Tufnell et al. 1958: fig. 7:5
3. Storage jar Tel



Ó



alif “Silo” site Stratum III Yekutieli 1992: pl. 22:5
4. Holemouth jar Tel



Ó



alif “Silo” site Stratum III Yekutieli 1992: pl. 22:10
5. Holemouth jar Tel



Ó



alif “Silo” site Stratum III Yekutieli 1992: pl. 22:11
6. Bowl Nizzanim Stratum V Yekutieli 1992: pl. 36:10
7. Bowl Tel



Ó



alif Site 101 Stratum V Dessel 1991: pl. 40:9
8. Bowl Tel



Ó



alif Site 101 Stratum V Dessel 1991: pl. 41:11

oi.uchicago.edu

[image: image200.tif]



667



THE EARLY BRONZE IA OF SOUTHWESTERN CANAAN



SMALL FINDS

The small finds at the sites include vessels and tools made of pottery, flint, and a variety of
stones, copper, and organic materials. Four aspects of these finds are important to discuss:
form, material and its provenance, technology and manufacture tradition, and function.

POTTERY (For a detailed study of this pottery see Yekutieli 2000.)



Form



The repertoire of ceramic forms of both EB IA phases distinguished in the research area is
presented in figures 33.5 and 33.6. Forms of EB IA1 (fig. 33.5) show a great resemblance to
local Chalcolithic forms. These include V-shaped bowls, sometimes with a delicate carination
on the rim’s edge, thin-rimmed holemouth jars, and high-necked storage jars, sometimes with
an indented rim. Chalcolithic forms missing from the EB IA1 assemblage include churns and
perhaps cornets as well. The new introductions that appear in the EB IA1 are ledge handles and
small, crude, globular bowls. It is worth noting that the ledge handles appear usually on stor-
age jars, below their widest part, and have many indentations. The EB IA2 forms (fig. 33.6)
are different. Innovations include small loop-handled storage jars, vessels with folded and
thumb-indented rims, and globular or drop-shaped juglets.



Material and Its Provenance



Petrographic studies concerning the ultimate origin of pottery from the research area were
conducted by Yuval Goren and Naomi Porat, who describe in detail their results and conclu-
sions (Goren 1987, 1991a; Porat 1989). They indicate that both local (southwestern Canaanite)
and imported wares were recognized. The imports originated in the Judaean mountains area and
Egypt. Porat (1992: 433–35) has shown and stressed the fact that within southwestern Canaan-
ite EB I (in general), one encounters pottery vessels made of local clays shaped in Egyptian
forms or made with Egyptian techniques (= Manufacture Traditions, see below); they are the
so-called Egyptianized vessels (term coined by Brandl 1989). Within the time span dealt with
in this article, her observation is relevant mainly in regard to the EB IA2 pottery. EB IA1 pottery
seems much more unaffected by any Egyptian influence. Among the ware types found in the
research area and worthy of note is the straw tempered ware. Its use is confined mainly to the
EB IA1 globular bowls. It was so common in the EB IA1 strata of Tel



Ó



alif Terrace, Lachish
Northwestern settlement, and Site H, that it is useful as a



fossil directeur



of that period.



Technology and Manufacture Traditions



Five well-defined pottery traditions, manifested in manufacturing techniques and surface
treatment, were noted in southwestern Canaan during the EB IA period:

1. Chalcolithic tradition: Distinguished by the appearance of Chalcolithic ceramic
manufacturing techniques such as indented rims, wheel-made V-shaped bowls, and
“streaky wash” (Hennessy 1969), a burnish of widely spaced horizontal lines made
when the clay was still soft (Goren 1987: 13). It is important to note that some
Chalcolithic traditions ceased, such as painting on vessels’ surfaces.

oi.uchicago.edu





668



YUVAL YEKUTIELI



FIGURE 33.6. Pottery forms of EB IA2.

oi.uchicago.edu

[image: image201.tif]



669



THE EARLY BRONZE IA OF SOUTHWESTERN CANAAN



2. Red on White painting tradition: Irregular red patterns or red vertical stripes painted on
top of a white wash.

3. Folded and thumb-indented rims tradition.
4. Red slipped or red slipped and burnished tradition.
5. Egyptian tradition: Distinguished by use of silty sediments with a small amount of

temper, organic tempers, and firing temperatures of over 800 degrees C (Porat 1989:
table 9.7b), and sometimes by a treatment that leaves straw impressions on the vessel’s
surface, as in contemporary Egyptian wares.

Tradition number 1 was attested mainly in EB IA1 contexts, while numbers 3–5 were
encountered in EB IA2 associations. Tradition number 2 began in EB IA1 as irregular red pat-
terns on top of a white wash, but in EB IA2 the irregular patterns were replaced by red vertical
stripes on top of the same white wash. Pottery traditions 1–3 are local southwestern Canaanite.
Numbers 1 and perhaps even 3 have a Chalcolithic origin, while tradition 2 is an EB IA inno-
vation. An Egyptian influence, often regarded as a major factor in southern Canaanite EB I
pottery (Porat 1989; Brandl 1992), is discerned in pottery traditions 4 and 5. A few spatial dis-
tinctions exist as well; tradition 2 is more common in the hilly parts of the research area, while
tradition 3 occurs more in the plains and on the coast. Pottery traditions 4 and 5 intensify as
one moves southwest within the research area (toward Egypt).



Function



At the functional level, different sizes were noted for storage vessels as well as cooking
and serving vessels. Of special note are two kinds of storage jars, ledge handled and loop han-
dled, which were used as transport vessels along the trade route to Egypt, as finds from south-
western Canaan, north Sinai, and Lower Egypt (see below) attest.

STONE INDUSTRY

The chipped stone industry (fig. 33.7) shows mainly local, and a few Egyptian-inspired,
traditions. Some of the local traditions indicate continuity with the Chalcolithic period. This
influence is found in the inclusion of fan scrapers, backed sickle blades, as well as a relatively
high percentage of retouched bladelts (Gilead and Marder in Oren and Yekutieli 1992; Gilead
1984; Roshwalb 1981: 278; S. A. Rosen 1989: 214).



FIGURE 33.6.



Type Site Stratum/Level Reference



1. Storage jar Site H Upper level Macdonald 1932: pl. 40:67
2. Storage jar Taur-Ikhbeineh Stratum IV Oren and Yekutieli 1992: fig. 10:13
3. Storage jar Taur-Ikhbeineh Stratum IV Oren and Yekutieli 1992: fig. 11:6
4. Storage jar Site H Dwelling 1 Roshwalb 1981: fig. H7:1
5. Storage jar Tel



Ó



alif Site 101 Stratum IV Dessel 1991: pl. 24:8
6. Storage jar Taur-Ikhbeineh Survey Unpublished
7. Holemouth jar Nizzanim Stratum IV Yekutieli and Gophna 1994: fig. 11:14
8. Bowl Nizzanim Stratum IV Yekutieli and Gophna 1994: fig. 11:17
9. Bowl Taur-Ikhbeineh — Oren and Yekutieli 1992: fig. 9:4

10. Bowl Lachish NW settlement Cave 1503 Tufnell 1958: pl. 56:28
11. Juglet Site H Dwelling 1 Roshwalb 1981: fig. H6:14

oi.uchicago.edu





670



YUVAL YEKUTIELI



FIGURE 33.7. Chipped stone industry, stone mace-head, and basalt bowls.

oi.uchicago.edu

[image: image202.tif]



671



THE EARLY BRONZE IA OF SOUTHWESTERN CANAAN



Other EB IA developments are the “Canaanean” blade (S. A. Rosen 1989) and a regional
version of a “toothed” sickle blade (Friedmann and Gophna 1990). The Egyptian-inspired tra-
dition is manifested in tools such as twisted bladelets (Roshwalb 1981: 278).

The raw material for the manufacture of simple, ad hoc tools was usually collected in the
immediate vicinity of each site (Gilead and Marder in Oren and Yekutieli 1992; Kempinski
and Gilead 1991), while more “elaborate” tools, such as fan scrapers and Canaanean sickle
blades, were probably imported to the sites from certain manufacturing centers (S. A. Rosen
1983, 1989).



4



The ground stone tools and vessels repertoire (fig. 33.8) includes examples fashioned of
basalt (e.g., Braun 1990) and limestone. Bowls, mortars, and grinding stones were made from
this array of stone types and were encountered at all the sites within the research area. Pear-
shaped mace-heads of limestone were recovered at Nizzanim (Yekutieli and Gophna 1994)
and at Lachish (Tufnell et al. 1958: 71). Grooved stones, probably used as arrow straighteners,
were encountered at Lachish (Tufnell et al. 1958 pl. 21:4) and at Site H (Macdonald 1932: pl.
XXVI: 52); and basalt or phosphorite spindle whorls were associated with Nizzanim
(Yekutieli and Gophna 1994), Lachish (Tufnell et al. 1958: 71), Erani, Tel



Ó



alif Terrace, and
Site H.

It is important to note that basalt bowls and spindle whorls, similar to those from the
research area which proved to be Canaanite in origin (Porat and Seeher 1988; Rizkana and
Seeher 1988: 53, 57), were found at Maadi. The phosphorite spindle whorl (from Nizzanim)
is worth noting, too, since our only knowledge of spindle whorls in stone resembling basalt is
confined to Chalcolithic contexts (Goren 1989, 1991b). This new information gives the impli-
cation of a continuing tradition. The shapes of basalt bowls (fig. 33.7:7, 9), the mace-head (fig.
33.7:8), and spindle whorls indicate a continuation of Chalcolithic forms as well. Most of the
stone tools and vessels (chipped and ground stone) were used for domestic activities. Only a
minority of them, such as the pear-shaped mace-heads and the fan scrapers, is considered by
some scholars to have had ceremonial or cult functions (Beebe 1989: 294; McConaughy 1979:
304; S. A. Rosen 1989: 202). Simple tools, such as grinding stones, mortars, and ad hoc flint
tools, were produced in a way indicating a domestic, nonspecialized mode of production.



4. Circumstantial evidence reaffirming this theory includes a cache of eight Canaanean blades found in Nizza-
nim—probably still in their original packing, as imported to the site (Yekutieli and Gophna 1994).

FIGURE 33.7.



Type Site Stratum/Locus Reference



1. Blade Nizzanim Stratum III Yekutieli 1992: pl. 10:1
2. Sickle blade Nizzanim Stratum III Yekutieli 1992: pl. 10:2
3. Backed sickle blade Nizzanim Stratum III Yekutieli 1992: pl. 10:3 fig. 11:6
4. Sickle blade Nizzanim Stratum III Yekutieli and Gophna 1994: fig. 14:5
5. Tabular fan scraper Nizzanim Stratum V Yekutieli 1992: pl. 10:4
6. Arrow straightener Lachish NW settlement Area 1500 Tufnell et al. 1958: pl. 21:4
7. Basalt bowl Nizzanim Stratum IV Yekutieli 1992: pl. 10:8
8. Mace-head Nizzanim Survey Yekutieli 1992: pl. 10:5
9. Basalt bowl Nizzanim Survey Yekutieli 1992: pl. 10:9

oi.uchicago.edu





672



YUVAL YEKUTIELI



FIGURE 33.8. Mud stoppers, clay stamps, stone figurines, and clay figurine.

oi.uchicago.edu

[image: image203.tif]



673



THE EARLY BRONZE IA OF SOUTHWESTERN CANAAN



However, the quality of the raw materials and the manufacturing skills attested in the more
“elaborate” tool types (e.g., fan scrapers, Canaanean blades, basalt bowls, spindle whorls, and
mace heads) indicate the existence of specialized manufacturers and even manufacturing cen-
ters (S. A. Rosen 1989: 217; Braun 1990: 95). The wide distribution of those items, within the
research area and beyond, indicates the existence of an efficient distribution system. It also
emphasizes the value attributed to these Canaanite products “abroad” as far as Lower Egypt.

COPPER INDUSTRY

The common copper tool in the research area was the square-sectioned awl found at Nizza-
nim (Yekutieli and Gophna 1994) and at Site H (Macdonald 1932: pls. 24:47, 26:56, 28:2–3).
Parallels are known at Maadi (together with specimens with rounded sections; Rizkana and See-
her 1989: 14, pl. 3). A fishing hook was recovered at Site H (Macdonald 1932: 13, pl. 28:8) and
is paralleled at Maadi (Rizkana and Seeher 1989: 14, pl. 3:1–5). Until some years ago, some
scholars estimated that fourth millennium southern Canaanite copper originated somewhere in
the north (i.e., Armenia; Key 1980: 243). In recent years copper mines of this period were
located much closer in the cliffs of the Arava segment of the Great Rift Valley. These mines are
presently considered to be the main sources of copper in both the Chalcolithic and EB I periods
(Rothenberg 1978; 1985: 124; Raikes 1980: 55; Hauptman, Weisberger, and Knauf 1985; Sha-
lev and Northover 1987: 362, 366; Goren 1989; Ilan and Sebanne 1989), together with a less
possible source in the southern Sinai (Beit Arieh 1977, 1980, 1983; Ilan and Sebanne 1989). It
should be noted that an analysis of the copper artifacts, and copper ore lumps from Maadi, indi-
cates that there is a great probability that they, too, originated in the Arava (Rizkana and Seeher
1989: 17, 78). Ilan and Sebbane (1989) noted that contrary to the rule in the Chalcolithic period,
in the EB I the manufacture of copper tools moved out of the settlements into the mining areas.
A comparison between the Chalcolithic and the EB IA copper objects also reveals a change in
forms and in manufacturing techniques (Ilan and Sebbane 1989), as well as a complete cessation
of artistic endeavors in the EB I copper industry. All the known EB IA copper objects belong
to the domestic tool kit. Thus a shifting balance in the “category” of copper tools is discern-
ible—from items that primarily express power and ideology (in the Chalcolithic period) to those
of necessity and daily use (in the EB IA). Since copper tools were probably imported into the
research area, such a shift in consumption patterns is not just an internal southwestern Canaanite
affair but has to do with a change on a larger scale than that of the research area.

ADDITIONAL FINDS

Some few remains of ornaments (especially beads) and domestic industries associated
with organic materials (i.e., weaving and basketry) indicate certain additional skills practiced



FIGURE 33.8.



Type Site Stratum Reference



1. Mud stopper Taur-Ikhbeineh Stratum V Yekutieli 1992: pl. 12:6
2. Mud stopper Site H — Yekutieli 1992: pl. 12:5
3. Clay stamp Tel



Ó



alif Terrace “Shaar-hakibbutz” — Yekutieli 1992: pl. 12:4
4. Stone figurine Tel



Ó



alif Terrace Site 101 Stratum IV Seger 1989: fig. 58:3;
Dessel 1991: fig. 13:4

5. Stone figurine ? Nizzanim Stratum III Yekutieli 1992: pl. 13:3
6. Clay figurine Site H — Gophna 1990a: fig 3:7

oi.uchicago.edu





674



YUVAL YEKUTIELI



during the EB IA period. The evidence for the latter group is, however, mostly secondary,
(e.g., mat and cloth impressions on clay objects) and does not allow for any elaborating
research, although they must have been major industries which are, unfortunately, archaeolog-
ically invisible (cf. Stager 1990).

In summarizing the small finds category, a clear distinction can be made between two lev-
els of the EB IA industries—that of the household level and that of the specialist. Simple
industries not requiring special skills include the production of ad hoc flint tools, some grind-
ing stones, and a few types of pottery vessels. Perhaps basketry and simple weaving belong to
the first level, too. These kinds of industries were probably practiced at the sites within the
research area. The copper industry, special chipped and ground stone industries, and certain
types of pottery, on the other hand, belong to a specialized industry level. Presumably special-
ized industries were located at a few, as yet unknown manufacturing centers (Braun 1990: 95;
S. A. Rosen 1989: 217; Ilan and Sebbane 1989) probably near the sources of raw materials
and thus out of the research area. Southwestern Canaan was one of a few markets consuming
these industries’ products and probably was not the main market.

CULT AND RELIGION

No cult places have been identified in the research area, although a few cult objects have
been found. Two kinds of figurines, closely related to local prototypes, are dated to the EB IA
(fig. 33.8:4–6). One found at Tel



Ó



alif Terrace, fashioned from stone, is best defined as a
highly schematic evolution of a Chalcolithic violin figurine (Seger 1988: fig. 58:3; Dessel
1991: fig. 13).

A piece, probably of another figurine of the same kind, was found at Nizzanim (Yekutieli
and Gophna 1994). The second kind, of which an example was found at Site H (Gophna
1989: 40), is made of pottery and is slightly more naturalistic, with breasts indicated and
punctations on the pubic region emphasizing its feminine gender.



5



Cemeteries of the EB IA
period have not yet been found within the research area. The only burials excavated were in-
trasite, child inhumations, either directly in pits in the ground or within whole or portions of
pottery vessels, below living floors (Nizzanim: Yekutieli and Gophna 1994; Afridar: Braun
and Gophna 1994). This practice is also known in the north in the EB I at Tel Teo (Eisenberg
1989: 38, fig. 8), Kabri (Kempinski and Niemeier 1991: 76), and Beit Yera



˙



(Maisler, Steke-
lis, and Avi-Yonah 1952: 229). Intrasite child inhumations are known in the research area also
in periods earlier than the EB IA at Neolithic Qatif (Epstein 1984: 210) and at the Chalcolithic
sites of Shiqmim, Gilat (D. Alon pers. comm.), Neve Noy (Eldar and Baumgarten 1985: 138),
Bir Safadi (Perrot 1959: 141–2), Grar (Gilead 1989: 383), and Tell Abu-Matar (Perrot 1955:
13–174). Such practices are not confined to Canaan; they are well documented throughout the
Levant in the fifth and fourth millennia



b.c.



on the Syro-Lebanese coast



6



at Sidon-Dakerman
(Saidah 1979: 42, figs. 14–15), Byblos (Dunand 1973: 246–65), and Ras Shamra (Courtois



5. See Gophna’s (1989: 43 n. 26) suggestion for an alternate date for a similar looking figurine from Tel



Ó



alif
Terrace (Dessel 1988: 13).

6. Braun (1989a; 1989b; 1991) has indicated some sort of connection between this culture and the northern
Canaanite early EBI.



short

oi.uchicago.edu





675



THE EARLY BRONZE IA OF SOUTHWESTERN CANAAN



1962), and farther north in Mesopotamia and Anatolia, and south in Egypt (Hole 1989; Riz-
kana and Seeher 1989: 67, 8). Thus, it is an old and well-rooted tradition in the Levant, mani-
fested in the research area as well.

AGRICULTURE ANIMAL HUSBANDRY

At Nizzanim (analyzed by J. Klenck) and the Tel



Ó



alif Terrace “Silo” site (analyzed by L.
Kolska Horwitz and R. Rabinovitch Goren), the following animal species (listed in descend-
ing quantitative order) were found: sheep and goat, cattle, pig, and equid. In Nizzanim’s
assemblage the presence of large quantities of fish bones indicates this coastal site’s likely
dependence on marine resources (Yekutieli and Gophna 1994). At Tel



Ó



alif Terrace deer was
additionally identified. Donkey remains were recorded at Site H (Gophna 1976), and fish
bones were found at Erani in Stratum D (A. M. Rosen 1991: 199). Recent research indicates
that in the EB I, sheep and goats were milked much more intensively than in the Chalcolithic
period (Smith and Horwitz 1984; Horwitz and Tchernov 1989: 286). A number of representa-
tions of donkeys in the EB I (Macalister 1912: I: 9; III: pl. 75:18; Kaplan 1993: 521; Eitan
1969: 51; Arnon 1984: 100; Gophna 1974: pl. 13:19), together with remains of these animals,
both at Canaanite (Nizzanim, Tel



Ó



alif Terrace, Site H) and Egyptian sites (Maadi: Rizkana
and Seeher 1989) suggest their use as pack animals both in Canaan and along trade routes to
Egypt. It is highly plausible that cattle were used as plow animals already in the EB I (Amiran
1992). According to this evidence it is safe to assume that the “secondary products revolution”
(Sherratt 1981, 1983) was completed in southwestern Canaan by the EB IA. Products derived
from nondomesticated animals were also found to be a part of the material culture at the EB
IA sites, although these are probably trade items not necessarily gathered by the southwestern
Canaanites themselves. Shells from the Mediterranean were found at all sites under study, as
were shells from the Red Sea at Lachish (Baden-Powell 1958). Shells and specialized bones
from the Nile were found at Site H (



Aspatharia



Nile shells and spikes of Nile catfish; Mac-
donald 1932: 14, pls. 23:33, 34; 26: 61). Also recovered were ostrich eggs, which were used
as storage vessels. Some of these eggs both from the research area and Egypt, had incised dec-
orations (Taur-Ikhbeineh: Oren and Yekutieli 1992; Site H: Macdonald 1932: pls. 25:60,
28:13; Maadi: Rizkana and Seeher 1989: 19–20, pl. 5:1, 2, 4).



Field Agriculture



“Mediterranean agriculture” (Ben-Tor 1990: 5) was already practiced in southwest Canaan
in the EB IA. Its main markers—olives and grapes—were a part of the diet of the EB IA peo-
ples of Nizzanim (Liphschitz 1989), while remains of different grains and cereals were found
at other sites in the research area (Hubbard 1981; A. M. Rosen 1991: 202; Oren and Yekutieli
1992). Some scholars noted a change in agricultural patterns, from pastoral to field oriented,
between the Chalcolithic and the EB I periods (Gophna 1974; Hanbury-Tenison 1986: 88). In
trying to explain this change, they suggested technological progress in the EB I involving the
introduction of a better plow and new irrigation methods (Miroschedji 1971: 128), a change in
social values (Hanbury-Tenison 1986: 88), or a mere preference for field agriculture in the
EB I (Gophna 1974). An interpretation of such a change should take into consideration too the

oi.uchicago.edu





676



YUVAL YEKUTIELI

abandonment of the Chalcolithic settlements of the northern Negev steppes7 and the demo-
graphic shift that followed this event in the EB I, toward the Mediterranean zone of Canaan
(Finkelstein and Gophna 1993: 9). Since agricultural patterns depend on an environment’s
characteristics, it is but logical to expect a preference of a pastoral mode in the semiarid
Negev steppes and of field agriculture in the Mediterranean zone. Thus the change might have
happened due to spatial reasons as well.

TRADE

From the EB IA onward, stable, permanent, and planned trade activities are discernible in
southwestern Canaan. The intensification of trade, through the different subphases of the EB I,
seems to be the major stimulator of socioeconomic development of southern Canaan during that
period. The main trade routes crossed or criss-crossed the Levant connecting sources of raw
materials to the major centers of the day—Mesopotamia and Egypt—and perhaps even con-
nected the two through some intermediaries (Kantor 1965; Prag 1986; Moorey 1990; Von der
Way 1987; Algaze 1989). Southwestern Canaan is located at the junction of three trade routes,
leading southwest to Egypt, east to the Arava copper mines and the Dead Sea bitumen sources
(Nissenbaum, Serban, and Amiran 1984), and north to northern Canaan/Syria. Southwestern
Canaan in this scenario is more a corridor than a destination in itself. Remains of traded goods,
some originating from the Dead Sea-Arava area, and of their storage and packing facilities have
been found at sites in the research area and along the trade routes all the way to Egypt. Dead
Sea (most probably) bitumen is associated with Nizzanim (Yekutieli and Gophna 1994), Tel
Óalif Terrace, Lachish (Tufnell et al. 1958: 71), Afridar Area G (Braun and Gophna pers.
comm.), Site H (Gophna 1976), and Maadi (Rizkana and Seeher 1989: 71, 72). Arava (most
probably) copper ores, either for copper manufacture or to be used (when ground) as cosmetics,
were found at north Sinai EB I sites (Oren 1973: 205, although which phase of EB I is not indi-
cated), and at Maadi in Egypt (Rizkana and Seeher 1989: 17, 78). A phosphorite spindle whorl,
probably originating around the Jordan Valley (Goren 1989), was found in Nizzanim (Yekutieli
and Gophna 1994). In addition to raw materials, complete vessels and tools were traded as well.
Canaanite copper, flint, wood, and basalt vessels and tools were recorded at Maadi (Rizkana
and Seeher 1988; 1989: 14, 24, 25, pl. 3; Porat and Seeher 1988). Other, archaeologically
invisible goods considered to have been transported from Canaan into Egypt include olive oil
(Lev-Yadun and Gophna 1992) and wine (Ward 1991; Lev-Yadun and Gophna 1992). Archae-
ological evidence for Egyptian items that were probably exchanged for those goods include
Nile catfish (Synodontis schall) spikes, Aspatharia shells (apparently prized for their mother of
pearl interiors), as well as a few alabaster mace-heads that were found in an EB IA tomb at Bâb
edh-Dhr⺠(Beebe 1989: 293) and probably reached that destination via southwestern Canaan.
The traded items listed above suggest a flow of many “necessity” items (flint, basalt, wood, and
copper tools, bitumen, copper ores, oil, and wine) from Canaan to Egypt, contra some, seem-
ingly more “luxurious” items (mace-heads, shells, and spikes) flowing in the opposite direction.
This observation might be wrong since there is a whole range of necessity items invisible in the

7. Any discussion of reasons for this abandonment are beyond the scope of this paper.

sho rt

oi.uchicago.edu



677THE EARLY BRONZE IA OF SOUTHWESTERN CANAAN

archaeological record which might have been traded too, in either direction. These include
grains, textiles, leather products, dried fish, and dates. Contemporary records of trade transac-
tions in Mesopotamia testify that the main volumes traded from Mesopotamia toward its
periphery were primarily these products, all of which are almost untraceable today (Algaze
1989: 573; Edens 1992: 127). It is highly plausible that such items also moved on the EB IA
trade routes in whatever direction, thus making the present picture only fragmentary.

The transport vessels are an important testimony to trade as well. Canaanite EB IA pottery
vessels have been found in Lower Egypt (Maadi: Rizkana and Seeher 1989; Heliopolis:
Debono and Mortensen 1988), and Egyptian vessels have been found in the research area
(Taur-Ikhbeineh: Oren and Yekutieli 1992; Site H: Gophna 1992).

Of special note are stamps, and stamped unbaked clay lids and bullae, which were used to
close and seal ceramic vessels and other containers. A net-patterned seal (fig. 33.8:3) and sev-
eral stamped bullae were found at Tel Óalif Terrace (Alon and Yekutieli 1995; Seger 1989;
Dessel 1991: 80). Similar parallels in Jordan, and in the Jordan Valley (Helms 1987: 59–60;
Glueck 1951: pl. 84:10; Mabry 1989: fig. 15, 14:5), hint at their connection with that region.
Other, locally manufactured mud bullae and stoppers (fig. 33.8:1–2) were found at Erani
(Kempinski and Gilead 1991: 187, figs. 14–16), Site H (Yekutieli 1992: fig. 12), and Taur-Ikh-
beineh (Oren and Yekutieli 1992). Stamped items as such suggest a beginning of formal con-
trol on storage and exchange transactions (Tosi 1984: 45) during the EB IA.

Internal Canaanite trade should not be dismissed—it was responsible for the dissemination
of such items as fish (A. M. Rosen 1991: 199), flint tools (S. A. Rosen 1989: 203), pottery ves-
sels (from the Judaean mountains into the research area),8 copper tools (Ilan and Sebbane
1989), and basalt bowls (Braun 1990: 95).

The intense trade suggests the development of a social group of specialized traders, who
might have been “freelance middlemen” (Harrison 1993), independent entrepreneurs (Stager
1992), agents of local rulers, temples, and the like. Whatever their definition, the important
point is that the intensification of trade increased social complexity in southwestern Canaan
during the EB IA period.

CHRONOLOGY

Beginning of EB IA

On the basis of radiocarbon determinations Levy (1992) has estimated that the collapse of
the Beersheva Valley Chalcolithic system occurred around 3700 b.c. radiocarbon dates from
the Chalcolithic Na˙al Mishmar Cave cover a range of 4243–3389 b.c. (five samples; Gilead
1993: table 1; Levy 1992: fig. 4), thus perhaps indicating even a somewhat later date for the
end of the Chalcolithic period. Having established that the EB IA follows the Chalcolithic
period and does not overlap with it (see above), it may be safely stated that a date around 3600
b.c. for its inception is indicated. As has been shown above (description of pottery, copper,
basalt, flint, shells, fish spikes, raw materials analysis, etc.), there are many similarities

8. In this way northern vessels such as the “Gray burnished ware” bowls might have reached southern sites
such as Alaªyik (Pritchard 1958[Pritchard 1958 not in Refs]) and Palmachim (Gophna 1974).

oi.uchicago.edu



678 YUVAL YEKUTIELI

between the material culture of southwestern Canaan during the EB IA and that of Maadi, indi-
cating at least a certain period of contemporaneity. Maadi is considered by many scholars to
have been deserted by Naqada 2c at the latest (Rizkana and Seeher 1989: 81; Von Ulrich
1994), that is, 3400 b.c. (Von der Way 1993: fig. 21). Radiocarbon dates from Maadi range
slightly earlier, from 3985 to 3515 b.c. (seven samples; Rizkana and Seeher 1989: 82). Since
Maadi and the southwestern Canaanite EB IA sites were contemporaneous, at least in part, it
can be safely suggested that the EB IA had begun around 3600 b.c.

End of EB IA

A radiocarbon sample from Taur-Ikhbeineh Strata III–IV, which represents the end of the
EB IA2, indicates a date of 3370 b.c. (Oren and Yekutieli 1992). Another, from Stratum C at
Tel Erani, which fits according to its material culture remains to the very beginning of the EB
IB, and covers an EB IA2 layer (Stratum D), is dated to 3331 b.c. (Kempinski and Gilead 1991:
171). A third radiocarbon date of 3307 b.c. (Gilead 1993) is derived from Stratum XI of
Yeivin’s excavations at Erani. This level has a pottery assemblage, as far as can be identified
from Yeivin’s (1961) plates, similar to that of Stratum C of the Kempinski and Gilead’s (1991)
excavation—Early EB IB. Therefore the end of the EB IA period can be determined at circa
3350 b.c.

A REVISION OF THE CHALCOLITHIC–EB IA TRANSITION
IN SOUTHWESTERN CANAAN

Settlement patterns in southwestern Canaan during the fourth millennium b.c. show
extreme contrast between the Chalcolithic period and the EB IA. More than 120 Chalcolithic
sites, of different sizes, were scattered in the northern Negev plains (Gazit 1986: 61), in con-
trast to only one site (Site H) that existed during EB IA in the same area. Besides the drastic
change in quantity, magnitude, and location of sites, other changes were noted as well, such as
the lack of any artistic endeavors, and presentations of power and ideology, in EB IA, com-
pared with the Chalcolithic period; thus a major collapse is attested. A phenomenon as such,
of a flourishing society reaching a peak never achieved before in its region, and then collaps-
ing in a way that combines a desertion of sites, a decline in wealth, technology, art, and power
is not unique. It has been noted in the archaeological records of many places around the world.
Such collapses have marked the end of the Mycenaean culture of Greece (Renfrew 1979), the
Maya culture of Central America (Adams 1973), the Pre-Pottery Neolithic B culture of the
Jordan Valley (Kohler-Rollefson 1988), and the EB II–III Canaanite culture (Dever 1989), to
mention but a few. Collapses as such led Renfrew (1978) to use the “catastrophe theory”
(Thom 1975), and Dever (1989) to use the “systems theory” to explain some of these events
noted in the archaeological record.

The catastrophe theory predicts that a cultural change will follow a system’s collapse
(Renfrew 1978). The EB IA material culture, and especially the pottery assemblage, indeed
reflects such predicted cultural change. Analysis of the pottery assemblage even makes it pos-
sible to be more precise and detect its inner fluctuations during the process. At first (EB IA1:
fig. 33.5) the ceramic repertoire is mainly a direct continuation of southern Chalcolithic tradi-
tions, while later (EB IA2: fig. 33.6) new traditions, which fully represent the cultural change,
join in.

oi.uchicago.edu



679THE EARLY BRONZE IA OF SOUTHWESTERN CANAAN

It is therefore suggested that the Chalcolithic population of the northern Negev sites did
not disappear after the collapse, but rather dispersed throughout Canaan. It is likely that the
resulting turmoil may have caused further population movements within the region of the
Canaanite cultural milieu, from which the EB IA culture, described above, emerged.

This emerging EB IA culture was distinguished not only in the research area, but at north-
ern Canaan as well (Braun 1989a), and at both places it combines local indigenous and new
traditions. In southwestern Canaan the indigenous components of the new culture appear as an
evolutionary continuation of the local Chalcolithic culture. These include dwelling methods,
burial customs, and types of pottery vessels, chipped stone tools, ground stone vessels, copper
tools, and cult figurines. The innovations of southwestern Canaanite EB IA culture include
new economic strategies (i.e., emphasis on trade), new technologies (e.g., copper metallurgy),
new settlements’ locations (from the eight EB IA sites in the research area, six were estab-
lished where no earlier Chalcolithic sites existed before; only at Tel Óalif Terrace and Gat-
Guvrin were Chalcolithic remains found as well) and some foreign (to southwestern Canaan)
architectural ideas (i.e., structures with rounded corners at Site H). The testimonies of evolu-
tionary continuation on the one hand, and the stratigraphy at Tel Óalif Terrace on the other
hand, preclude the possibility of either a significant time gap, or an overlap between the Chal-
colithic and the EB IA periods in the research area (see table 33.2 for various opinions on this
issue).

A REEVALUATION OF THE EGYPTIAN ROLE IN

SOUTHWESTERN CANAAN IN THE EB IA

Some scholars emphasize the Egyptian role and impact on the culture of southwestern
Canaan in the EB I regardless of its internal subphases. In the EB IA there were indeed trade
contacts, exchange of goods, and some cultural imitation on both sides of the Canaan-Egypt
trade route, but an Egyptian colony in southwestern Canaan during the EB IA as suggested by
Brandl (1992: 442, 447) is not evident. The intensive Egyptian activity occurred during the
late EB IB. The magnitude of Egyptian involvement in southwestern Canaan in the EB IA was
smaller than in the late EB IB in terms of quantity of Egyptian imports, adaptation of Egyptian
technological traditions, and spatial distribution of both. The zone of Egyptian influence in the
late EB IB, according to the above-mentioned parameters, stretched farther northeast into
Canaan than that of the EB IA. A second observation in regard to Egypto-Canaanite relations
within the EB IA is that they were minor in the EB IA1 and intensified, again in terms of quan-
tity of Egyptian imports, adaptation of Egyptian technological traditions, and spatial distribu-
tion of both, in the EB IA2 period.

Reconstruction of the EB IA System of Southwestern Canaan

Viewing the ancient landscape as sets of cores and peripheries (Algaze 1989) sharpens and
clarifies the differences between the Chalcolithic and the EB IA of southwestern Canaan. Dur-
ing the Chalcolithic period the regional core was at the northern Negev—expressing itself in
wealth, technology, art, power, and religion and exerting gravitational forces that affect a wide
peripheral zone (i.e., see Gilat sanctuary’s centrality and influence in the southern Levant:
Alon and Levy 1990). After the collapse of this system in a negative-feedback loop (Dever

oi.uchicago.edu



680 YUVAL YEKUTIELI

1989), the regional core shifted, in the EB IA, toward the Nile Delta, changing southwestern
Canaan’s status from core to periphery. In this manner southwest Canaan found itself in the
EB IA period at a trade route’s junction on the periphery of a strengthening Egyptian core, on
the eve of its turning into a pristine nation-state. This situation together with the intensifying
trade, channeled through the research area, triggered a positive-feedback loop (Dever 1989).
It activated a flow of new ideas and innovations into southwestern Canaan which, as a direct
consequence, aroused its social complexity and economic prosperity through the EB IA to
EB IB. This scenario is not unique. A parallel of a more or less similar situation—a peripheral
society channeling foreign resource input into a large core, and benefiting from it—is Baha-
rein’s (Dilmun) case in the Mesopotamian Early Dynastic period (Edens 1992).

The described process, initiated in EB IA in southwestern Canaan, culminated in an early
form of urbanization as soon as early EB IB (Erani Layer C: Kempinski and Gilead 1991),
earlier than elsewhere in the southern Levant.

ACKNOWLEDGMENTS

This paper is based on the author’s M.A. work, presented to Tel Aviv University and
guided by Ram Gophna (Tel Aviv University) and Eliezer Oren (Ben Gurion University). Spe-
cial thanks to Eliot Braun who read the manuscript and made many useful comments, and to
Patrice Kaminski who arranged the figure drawings.

REFERENCES

Adams, R. C. W.
1973 The Collapse of Maya Civilization: A Review of Previous Theories. Pp. 21–34 in The Classic

Maya Collapse, ed. T. P. Culbert. Albuquerque: University of New Mexico Press.

Aharoni, Y.
1982 The Archaeology of the Land of Israel. Philadelphia: Westminster.

Albright, W. F.
1954 The Archaeology of Palestine. Harmondsworth: Penguin.
1957 From Stone Age to Christianity. Baltimore: Johns Hopkins Press.
1960 The Archaeology of Palestine. Baltimore: Penguin.

Algaze, G.
1989 The Uruk Expansion: Cross Cultural Exchange in Early Mesopotamian Civilization. Current

Anthropology 30: 571–608.

Alon, D.
1991 Horvat Abu Hof. Daroma 1: 2–13 (Hebrew).

Alon, D., and Levy, T. E.
1990 The Gilat Sanctuary: Its Centrality and Influence in the Southern Levant during the Late 5th–

Early 4th Millennium b.c.e. Eretz-Israel 21: 23–36 (Hebrew).

Alon, D., and Yekutieli, Y.
1995 The Tel Óalif Terrace “Silo Site” and its Implications for the Early Bronze Age I. ºAtiqot 27:

149–89.

Amiran, R.
1970 The Beginnings of Urbanization in Canaan. Pp. 83–100 in Near Eastern Archaeology in the Twen-

tieth Century: Essays in Honor of Nelson Glueck, ed. A. J. Sanders. Garden City, NY: Doubleday.
1985 The Transition from the Chalcolithic to the Early Bronze Age. Pp. 108–12 in Biblical Archaeol-

ogy Today, ed. J. Amitai. Jerusalem: Israel Exploration Society.

oi.uchicago.edu



681THE EARLY BRONZE IA OF SOUTHWESTERN CANAAN

1992 An Egyptian Artist from Megiddo and a Local Canaanite Artist from Tell Farºah (North) — Both
from the Early Bronze I. Unpublished paper presented at the Eighteenth Archaeological Confer-
ence in Israel.

Amiran, R., and Gophna, R.
1992 The Correlation between Lower Egypt and Southern Canaan during the EB1 Period. Pp. 357–60

in The Nile Delta in Transition: 4th–3rd Millennium B.C., ed. Edwin C. M. Van den Brink. Tel
Aviv: Edwin C. M. Van den Brink.

Amiran, R.; Paran, U.; Shiloh, Y.; Brown, R.; Tsafrir, Y.; and Ben-Tor, A.
1978 Early Arad: The Chalcolithic Settlement and Early Bronze City I. First–Fifth Seasons of Exca-

vations 1962–1966. Jerusalem: Israel Exploration Society.

Arnon, C.
1984 Tel Qishyon—1981. Excavations and Surveys in Israel 1: 100.

Baden-Powell, D. F. W.
1958 Shells. Pp. 323–24 in Lachish IV: The Bronze Age, by O. Tufnell; Baden-Powell, D. F. W.; Bate,

D. M. A.; Cerny, J.; Diringer, D.; Giles, M.; Helbaek, H.; Isserlin, B. S. J.; Murray, M. A.;
Parker, B.; Porada, E.; Thompson, F. C.; Todd, E.; and Waecher, J. London: Oxford University
Press.

Baruch, U.
1986 The Late Holocene Vegetational History of Lake Kinneret (Sea of Galilee), Israel. Paléorient 12:

37–48.

Beebe, H. K.
1989 The Mace Heads from Bâb edh Dhrâº. Pp. 289–94 in Bâb edh Dhrâº: Excavations in the Ceme-

tery Directed by Paul W. Lapp (1965–67), eds. R. T. Schaub and W. E. Rast. Winona Lake, IN:
Eisenbrauns.

Beit Arieh, I.
1977 Southern Sinai in the Early Bronze Age Period. Ph.D. dissertation, Tel Aviv, Tel Aviv University

(Hebrew).
1980 A Chalcolithic Site near Serabit al-Khadim. Tel Aviv 7: 45–64.
1983 Central Southern Sinai in the Early Bronze Age II and Its Relationship with Palestine. Levant 15:

39–48.

Beit Arieh, I., and Gophna, R.
1977 A Note on a Chalcolithic Site in Wadi Araba. Tel Aviv 4 : 105–9.

Ben-Tor, A.
1990 The Trade Relations of Palestine in the Early Bronze Age. Pp. 3–20 in Commerce in Palestine

throughout the Ages, eds. B. Z. Kedar, T. Dothan, and S. Safrai. Jerusalem: Yad Izhak Ben Zvi
and the Israel Exploration Society (Hebrew).

Brandl, B.
1989 Observations on the Early Bronze Age Strata of Tel Erani. Pp. 357–88 in L’urbanization, ed.

P. de Miroschedji, British Archaeological Reports, International Series 527, pt. 2. Oxford: Brit-
ish Archaeological Reports.

1992 Evidence for Egyptian Colonization in the Southern Coastal Plain and Lowlands of Canaan dur-
ing the EB1 Period. Pp. 441–77 in The Nile Delta in Transition: 4th–3rd Millennium B.C., ed.
Edwin C. M. Van den Brink. Tel Aviv: Edwin C. M. Van den Brink.

Braun, E.
1985 En Shadud: Salvage Excavations at a Farming Community in the Jezreel Valley, Israel. British

Archaeological Reports, International Series 249. Oxford: British Archaeological Reports.
1989a The Problem of the Apsidal House: New Aspects of Early Bronze I Domestic Architecture in

Israel, Jordan, and Lebanon. Palestine Exploration Quarterly 121: 1–43.
1989b The Transition from the Chalcolithic to the Early Bronze Age in Northern Israel and Jordan: Is

There A Missing Link? Pp. 7–27 in L’urbanization, ed. P. de Miroschedji, Pierre, British
Archaeological Reports, International Series 527, pt. 1. Oxford: British Archaeological Reports.

1990 Basalt Bowls of the EB I Horizon in the Southern Levant. Paléorient 16: 87–96.

oi.uchicago.edu



682 YUVAL YEKUTIELI

Braun, E., and Gophna, R.
1994 Ashkelon (Afridar). In Archaeology in Israel, by Samuel R. Wolff. American Journal of Archae-

ology 98: 486–87.

Broshi, M., and Gophna, R.
1984 The Settlements and Population of Eretz Israel in the Early Bronze Age II–III. Eretz-Israel 17:

147–57. (Hebrew).

Callaway, A. J.
1972 The Early Bronze Age Sanctuary at ºAi (et-Tell). London: Quaritch.

Chapman, L. R.
1990 In Pursuit of the Early Bronze Age: Review article of L’urbanisation, ed. P. de Miroschedji.

British Archaeological Reports, International Series 527, pts. 1–2. Bulletin of the Anglo-Israel
Archaeological Society 10: 73–83.

Contenson, H. de
1961 Remarques sur la Chalcolithique de Beer-sheba: Étude Typologique. Israel Exploration Journal

6: 163–79, 226–38.

Copeland, L., and Vita-Finzi, C.
1978 Archaeological Dating of Geological Deposition in Jordan. Levant 10: 10–25.

Courtois, J.-C.
1962 Contribution à l’étude des Niveaux II et III de Ras Shamra. Pp. 329–414 in Ugaritica IV, ed.

F. T. C. Schaeffer. Paris: P. Geuthner.

Dan, Y.
1988 The Soils of the Southern Shefelah. Pp. 50–59 in Man and Environment in the Southern

Shefelah, eds. E. Stern and D. Urman. Givºataim: Massada (Hebrew).

Danin, A.
1988 Vegetation in the Southern Shefelah. Pp. 59–64 in Man and Environment in the Southern

Shefelah, eds. E. Stern and D. Urman. Givºataim: Massada (Hebrew).

Debono, F., and Mortensen, B.
1988 The Predynastic Cemetery at Heliopolis: Season March–September. Mainz am Rhein: von

Zabern.

Dessel, J. P.
1988 An Iron Age Figurine from Tel Halif. Bulletin of the American Schools of Oriental Research

269: 59–64.
1991 Ceramic Production and Social Complexity in Fourth Millennium Canaan: A Case Study from

the Halif Terrace. Ph.D. dissertation, University of Arizona.

de Vaux, R.
1970 Palestine in the Early Bronze Age. Pp. 208–37 in Ancient History, vol. 1, part 2, eds. I. E. S.

Edwards, C. J. Gadd, and N. G. L. Hammond. Cambridge: Cambridge University Press.

Dever, W. G.
1989 The Collapse of the Urban Early Bronze Age in Palestine: Toward a Systemic Analysis. Pp. 225–

46 in L’urbanisation, ed. P. de Miroschedji. British Archaeological Reports, International Series
527, pt. 2. Oxford: British Archaeological Reports.

Dunand, M.
1973 Fouilles de Byblos, Tome V. L’architecture, les tombes, le matériel domestique, des origines

néolithiques à l’avènement urbain (Texte et Planches). Paris: A. Maisonneuve.

Edens, C.
1992 Dynamics of Trade in the Ancient Mesopotamian “World System.” American Anthropologist 94:

118–39.

Eisenberg, E.
1989 The Chalcolithic and Early Bronze I Occupations at Tel Teo. Pp. 29–40 in L’urbanisation,

ed. P. de Miroschedji. British Archaeological Reports, International Series 527, pt. 1. Oxford:
British Archaeological Reports.

oi.uchicago.edu



683THE EARLY BRONZE IA OF SOUTHWESTERN CANAAN

Eitan, A.
1969 Excavations at the Foot of Tel Ras Haªayin. ºAtiqot 5: 49–68

Eldar, I., and Baumgarten, Y.
1985 Neve Noy, a Chalcolithic Site of the Beer-Sheba Culture. Biblical Archaeologist 48: 134–39.

Elliott, C.
1978 The Ghassulian Culture in Palestine: Origins, Influences and Abandonment. Levant 10: 37–54.

Epstein, C.
1984 A Pottery Neolithic Site Near Tel Qatif. Israel Exploration Journal 34: 209–19.

Finkelstein, I., and Gophna, R.
1993 Settlement, Demographic, and Economic Patterns in the Highlands of Palestine in the Chalco-

lithic and Early Bronze Periods and the Beginning of Urbanism. Bulletin of the American
Schools of Oriental Research 289: 1–22.

Friedmann, E., and Gophna, R.
1990 Early Bronze Sickle Blades from Site H in Nahal Besor. Tel Aviv 17: 87–90.

Frumkin, A; Magaritz, M.; Carmi, I.; and Zak, I.
1991 The Holocene Climatic Record of the Salt Caves of Mount Sedom, Israel. The Holocene 1:

191–200.

Gazit, D.
1986 HaBesor Region. Jerusalem: Society for Protection of Nature (Hebrew).

Gilead, I.
1984 The Microendscraper: A New Tool Type of Chalcolithic Period. Tel Aviv 11: 3–10.
1989 Grar: A Chalcolithic Site in Nahal Grar, Northern Negev, Israel. Journal Field Archaeology 16:

377–94.
1993 Sociopolitical Organization in the Northern Negev at the End of the Chalcolithic Period. Pp. 82–

97 in Biblical Archaeology Today 1990, Pre-Congress Symposium: Population. Production and
Power, Proceedings of the Second International Congress on Biblical Archaeology, eds. A. Biran
and J. Aviram. Jerusalem: Israel Exploration Society.

Gitin, S.; Biran, A.; Eitan, A.; and Aviram, A.
1982 Excavations and Surveys in Israel 1982, I. Jerusalem: Israel Antiquities Department and

Museums.

Glueck, N.
1951 Explorations in Eastern Palestine IV, Parts 1 and 2. Annual of the American Schools of Oriental

Research 25–28.

Goldberg, P.
1983 The Research of Climatic Changes through the Eyes of a Geologist. Catedra 28: 146–47

(Hebrew).

Goldberg, P., and Rosen, A. M.
1987 Early Holocene Paleoenvironments of Israel. Pp. 23–33 in Shiqmim I. Studies Concerning

Chalcolithic Societies in the Northern Negev Desert, Israel (1982–1984), ed. T. E. Levy.
British Archaeological Reports, International Series 356, pt. 1. Oxford: British Archaeological
Reports.

Gophna, R.
1974 The Settlement of the Coastal Plain of Eretz Israel during the Early Bronze Age. Ph.D. disserta-

tion, Tel Aviv University (Hebrew).
1976 Notes and News: Besor Bridge (“Site H”). Israel Exploration Journal 26: 199.
1989 The Early Bronze I Settlement at ºEn Besor Oasis. Eretz-Israel 20: 37–40 (Hebrew).
1990a The Early Bronze I Settlement at ºEn Besor Oasis. Israel Exploration Journal 40: 1–11.
1990b The Egyptian Pottery of ºEn Besor. Tel Aviv 17: 144–62.
1992 The Contacts Between the ºEn Besor Oasis, Southern Canaan, and Egypt during the Late Pre-

dynastic and the Threshold of the First Dynasty: A Further Assessment. Pp. 385–94 in The Nile
Delta in Transition: 4th–3rd Millennium B.C., ed. Edwin C. M. van den Brink. Tel Aviv: Edwin
C. M. van den Brink.

oi.uchicago.edu



684 YUVAL YEKUTIELI

Gophna, R., and Portugali, J.
1988 Settlement and Demographic Process in Israel’s Coastal Plain from the Chalcolithic to the Mid-

dle Bronze Age. Bulletin of the American Schools of Oriental Research 269: 11–28.

Goren, Y.
1987 Petrographical Analysis of the Pottery of the EB I “Silo” Site, Near Kibbutz Lahav. Unpublished

report (Hebrew).
1989 The Contacts between Jordanian and Canaanite Sites in the Chalcolithic Period According to

the Petrographic Research of Pottery, Stone and Metal Tools and Vessels. Mitekufat Haeven
22: 82–84.

1991a The Beginning of Pottery in Eretz-Israel. The Technology and Typology of Pottery Assemblages
from the Proto-Historic Periods of Israel (6th–4th Millennium bce). Ph.D. dissertation, Hebrew
University, Jerusalem (Hebrew).

1991b Phosphorite Vessels of the 4th Millennium b.c.e. in the Southern Levant: New Data, New Inter-
pretations. Journal of the Israel Prehistoric Society 24: 102–11.

Govrin, Y.
1987 Horvat Hor: A Dwelling Cave from the Chalcolithic Period in the Northern Negev. Mitekufat

Haeven 20: 119–27.
1990 The Question of the Origin of Early Arad’s Population. Eretz-Israel 21: 107–10 (Hebrew).

Hanbury-Tenison, J. W.
1986 The Late Chalcolithic to Early Bronze Age I Transition in Palestine and Transjordan. British

Archaeological Reports, International Series 311. Oxford: British Archaeological Reports.

Harrison, P. T.
1993 Economics with an Entrepreneurial Spirit: Early Bronze Trade with Late Predynastic Egypt.

Biblical Archaeologist 56: 81–93.

Hassan, A. F.
1980 Prehistoric Settlements along the Main Nile. In The Sahara and the Nile, eds. A. J. Martin Wil-

liams and H. Faure. Rotterdam: Balkema.

Hauptmann, A.; Weisberger, G.; and Knauf, E. A.
1985 Archäometallurgische und Bergbauarchäologische Untersuchungen im Gebiet von Fenan, Wadi

Arabah (Jordanian). Der Anschnitt 37: 163–95.

Helms, W. S.
1987 Jawa, Tell Um Hammad and the EB I/Late Chalcolithic Landscape. Levant 19: 49–81.

Hennessy, J. B.
1967 The Foreign Relations of Palestine during the Early Bronze Age. London: Quaritch.
1969 Preliminary Report on a First Season of Excavations at Teleilat Ghassul. Levant 1: 1–24.

Hole, F.
1989 Patterns of Burial in the Fifth Millennium. Pp. 149–80 in Upon This Foundation, The ºUbaid

Reconsidered: Proceedings from the ºUbaid Symposium, Elsinore, May 30th–June 1st 1988, eds.
F. E. Henrikson and I. Thuesen. Copenhagen: University of Copenhagen Press.

Horowitz, A.
1974 Preliminary Palynological Indications as to the Climate of Israel during the Last 6000 years.

Paléorient 2: 407–14.
1979 The Quaternary of Israel. New York: Academic Press.

Horwitz, L. K., and Tchernov, E.
1989 Animal Exploitation in the Early Bronze Age of the Southern Levant: An Overview. Pp. 279–313

in L’urbanization, ed. P. de Miroschedji. British Archaeological Reports, International Series 527,
pt. 2. Oxford: British Archaeological Reports.

Hubbard, R. L. N. B.
1981 Appendix 5: Plant Evidence. Pp. 455–57 in Protohistory in Wadi Ghazzeh: A Typological and

Technological Study Based on the Macdonald Excavations, by A. Roshwalb. Ph.D. dissertation,
University of London.

short

oi.uchicago.edu



685THE EARLY BRONZE IA OF SOUTHWESTERN CANAAN

Ilan, O., and Sebbane, M.
1989 Copper Metallurgy, Trade and the Urbanization of Southern Canaan in the Chalcolithic and

Early Bronze Age. Pp. 139–62 in L’urbanisation, ed. P. de Miroschedji. British Archaeological
Reports, International Series 527, pt. 1. Oxford: British Archaeological Reports.

Joffe, A. H.
1993 Settlement and Society in the Early Bronze I and II of the Southern Levant: Complementarity and

Contradiction in a Small-Scale Complex Society. Monographs in Mediterranean Archaeology 4.
Sheffield: Academic Press.

Johnson, G. A.
1973 Local Exchange and Early State Development in Southwestern Iran. Anthropological Paper

Number 51. Ann Arbor: Museum of Anthropology, University of Michigan.

Kantor, H. J.
1965 The Relative Chronology of Egypt and Its Foreign Correlations before the Late Bronze Age.

Pp.1–46 in Chronologies in Old World Archaeology, rev. ed., ed. W. R. Ehrich. Chicago: Univer-
sity of Chicago Press.

Kaplan, J.
1978 Givºatayim. P. 520 in The New Encyclopaedia of Archaeological Excavations in the Holy

Land, ed. E. Stern. Jerusalem: Israel Exploration Society & Carta and New York: Simon &
Schuster.

Kempinski, A.
1992 Reflections on the Role of the Egyptians in the Shefelah of Palestine—in the Light of Recent

Soundings at Tel Erani. Pp. 419–26 in The Nile Delta in Transition: 4th–3rd Millennium B.C.,
ed. Edwin C. M. Van den Brink. Tel Aviv: Edwin C. M. Van den Brink.

Kempinski, A., and Gilead, I.
1991 New Excavations at Tel Erani: A Preliminary Report of the 1985–1988 Seasons. Tel Aviv 18:

164–91.

Kempinski, A., and Niemeier, W. D.
1991 Kabri 1990. Excavations and Surveys in Israel 10: 76–77.

Kenyon, K.
1960 Archaeology in the Holy Land. London: Ernest Benn.

Key, C. A.
1980 Trace Element Composition of the Copper and Copper Alloy Artifacts of the Nahal Mishmar

Hoard. Pp. 239–43 in The Cave of the Treasure, P. Bar Adon. Jerusalem: Israel Exploration
Society.

Kohler-Rollefson, I.
1988 The Aftermath of the Levantine Neolithic Revolution in the Light of Ecological and Ethno-

graphic Evidence. Paléorient 14: 87–93.

Lapp, P. W.
1968 Bâb edh Dhr⺠Tomb A76 and Early Bronze I in Palestine. Bulletin of the American Schools of

Oriental Research 189: 12–41.
1970 Palestine in the Early Brronze Age. Pp. 101–31 in Near Eastern Archaeology in the Twentieth

Century: Essays in Honor of Nelson Glueck, ed. A. J. Sanders. Garden City, NY: Doubleday.

Levy, T. E.
1992 Radiocarbon Chronology of the Beersheva Culture and Predynastic Egypt. Pp. 345–56 in The

Nile Delta in Transition: 4th–3rd Millennium B.C., ed. Edwin C. M. Van den Brink. Tel Aviv:
Edwin C. M. Van den Brink.

Lev-Yadun, S., and Gophna, R.
1992 What Plant Products were Exported from Canaan into Egypt in the Early Bronze Age I: Botan-

ical and Cultural Remarks on the Paper by William A. Ward. Bulletin of the American Schools
of Oriental Research 287: 89–90.

oi.uchicago.edu



686 YUVAL YEKUTIELI

Liphschitz, N.
1989 Dendroarchaeological Investigations in Israel: Nizzanim. Mimeographed Report No. 185. Tel

Aviv: Institute of Archaeology, Tel Aviv University (Hebrew).
Mabry, J.

1989 Investigations at Tell el-Handaquq, Jordan (1987–88). Annual of the Department of Antiquities
of Jordan 33: 59–96.

Macalister, R. A. S.
1912 The Excavation of Gezer I–III. London: John Murray.

Macdonald, E.
1932 Beth-Pelet II: Prehistoric Fara. London: British School of Archaeology in Egypt.

Maisler, B.; Stekelis, M.; and Avi-Yonah, M.
1952 Excavations at Beth Yerah. Israel Exploration Journal 2: 218–29.

Marfoe, L.
1980 Review of The Rise of an Urban Culture, A. Kempinski (Jerusalem: Israel Ethnographic So-

ciety, 1978) and Early Arad, R. Amiran et al. (Jerusalem: Israel Exploration Society, 1978).
Journal of Near Eastern Studies 39: 315–22.

McConaughy, M.
1979 Formal and Functional Analyses of the Chipped Stone Tools from Bâb edh-Dhrâº, Jordan. Ph.D.

dissertation, University of Michigan.
Meigs, P.

1953 Design and Use of Homoclimatic Maps; Dry Climates of Israel as Example. Desert Research
Symposium: Jerusalem, 1953: 99–111. Jerusalem: Research Council of Israel.

Miroschedji, P. de
1971 L’ époque pré-urbaine en Palestine. Paris: Gabalda.

Moorey, R. S. P.
1990 From Gulf to Delta in the Fourth Millennium: The Syrian Connection. Eretz-Israel 21: 62*–69*.

Neev, D., and Emery, O. K.
1967 The Dead Sea, Depositional Processes and Enviroments of Evaporites. Geological Survey of

Israel Bulletin 41: 1–147.
Nissenbaum, A.; Serban, A.; and Amiran, R.

1984 Dead Sea Asphalt from the Excavations in Tel Arad and Small Tell Malhata. Paléorient 10:
157–61.

Oren, E. D.
1973 The Overland Route Between Egypt and Canaan in the Early Bronze Age, Preliminary Report.

Israel Exploration Journal 23: 198–205.
1989 Early Bronze Age Settlement in Northern Sinai: A Model for Egypto-Canaanite Interconnections.

Pp. 389–405 in L’urbanisation, ed. P. de Miroschedji. British Archaeological Reports, Interna-
tional Series 527, pt. 1. Oxford: British Archaeological Reports.

Oren, E. D., and Yekutieli, Y.
1992 Taur Ikhbeineh—Earliest Evidence for Egyptian Interconnections. Pp. 361–84 in The Nile

Delta in Transition: 4th–3rd Millennium B.C., ed. Edwin C. M. Van den Brink. Tel Aviv: Edwin
C. M. Van den Brink.

Perrot, J.
1955 The Excavations at Tel Abu Matar near Beersheba. Israel Exploration Journal 5: 17–40, 73–84,

167–89.
1959 Bir es Safadi. Israel Exploration Journal 9: 141–42.
1961 Notes and News: Gat Guvrin. Israel Exploration Journal 11: 76.

Porat, N.
1989 Composition of Pottery—Application to the Study of Interrelations between Canaan and Egypt

during the 3rd Millennium. Ph.D. dissertation, Hebrew University, Jerusalem.
1992 An Egyptian Colony in Southern Palestine during the Late Predynastic/Early Dynastic Period.

Pp. 433–40 in The Nile Delta in Transition: 4th–3rd Millennium B.C., ed. Edwin C. M. Van den
Brink. Tel Aviv: Edwin C. M. Van den Brink.

oi.uchicago.edu



687THE EARLY BRONZE IA OF SOUTHWESTERN CANAAN

Porat, N., and Seeher, J.
1988 Petrographic Analysis of Pottery and Basalt from Predynastic Maadi. Mitteilungen des Deutschen

Archäologischen Instituts Abteilung Kairo 44: 215–28.
Porath, Y.

1985 A Chalcolithic Building at Fasaªel. ºAtiqot 17: 1–19.
Prag, K.

1986 Byblos and Egypt in the Fourth Millennium b.c. Levant 18: 59–74.
Raikes, T. D.

1980 Notes on Some Neolithic and Later Sites in the Wadi Araba and Dead Sea Valley. Levant 12:
40–60.

Renfrew, A. C.
1978 Trajectory Discontinuity and Morphogenesis. American Antiquity 43: 203–22.
1979 Systems Collapse as Social Transformation: Catastrophe and Anastrophe in Early Societies.

Pp. 481–506 in Transformations: Mathematical Approaches to Culture Change, eds. A. C.
Renfrew and L. K. Cooke. New York: Academic Press.

Rizkana, I., and Seeher, J.
1988 Maadi II, The Lithic Industries of the Predynastic Settlement. Mainz: von Zabern.
1989 Maadi III, The Non Lithic Small Finds and the Structural Remains of the Predynastic Settle-

ment. Mainz: von Zabern.
Rollefson, O. G., and Kohler-Rollefson, I.

1989 The Collapse of Early Neolithic Settlements in the Southern Levant. Pp. 73–90 in People and
Culture in Change, ed. I. Hershkovitz. British Archaeological Reports, International Series 508,
pt. 1. Oxford: British Archaeological Reports.

Rosen, A. M.
1991 Early Bronze Age Tel Erani: An Environmental Perspective. Tel Aviv 18: 192–204.

Rosen, S. A.
1983 Tabular Scraper Trade: A Model of Material Culture Dispersion. Bulletin of the American

Schools of Oriental Research 249: 79–86.
1989 The Analysis of Early Bronze Age Chipped Stone Industries. Pp. 199–222 in L’urbanisation, ed.

P. de Miroschedji. British Archaeological Reports, International Series 527, pt. 1. Oxford: Brit-
ish Archaeological Reports.

Roshwalb, A.
1981 Protohistory in Wadi Ghazzeh: A Typological and Technological Study Based on the Macdonald

Excavations. Ph.D. dissertation, University of London.
Rothenberg, B.

1978 Excavation at Timna Site 39. Pp. 1–23 in Chalcolithic Copper Smelting, eds. B. Rothenberg,
R. F. Tylecote, and P. J. Boydell. London: Institute for Archaeo-Metallurgical Studies.

1985 Copper Smelting Furnaces in the Arabah, Israel: The Archaeological Evidence. Pp. 123–35 in
Furnaces and Smelting Technology in Antiquity, eds. P. T. Craddock and M. J. Hughes. London:
Institute for Archaeo-Metallurgical Studies.

Saidah, R.
1979 Fouilles de Sidon-Dakerman: l’agglomeration Chalcolithique. Berytus 27: 29–55.

Seger, J. D.
1988–89 Tel Óalif–1987. Excavations and Surveys in Israel 7–8: 69–71.
1989–90 Tel Óalif–1989. Excavations and Surveys in Israel 9: 67–68.

Shalev, S., and Northover, J. P.
1987 Chalcolithic Metal and Metalworking from Shiqmim. Pp. 357–71 in Shiqmim I. Studies Con-

cerning Chalcolithic Societies in the Northern Negev Desert, Israel (1982–1984), ed. T. E. Levy.
British Archaeological Reports, International Series 356, pt. 1. Oxford: British Archaeological
Reports.

Sherratt, A.
1981 Plough and Pastoralism: Aspects of the Secondary Products Revolution. Pp. 261–305 in Pattern

of Past: Studies in Honour of David Clarke, eds. I. Hodder, G. Issac, and N. Hammond. Cam-
bridge: Cambridge University Press.

oi.uchicago.edu



688 YUVAL YEKUTIELI

1983 The Secondary Exploitation of Animals in the Old World. World Archaeology 15: 90–104.

Smith, P., and Horwitz, L. K.
1984 Radiographic Evidence for Changing Patterns of Animal Exploitation in the Southern Levant.

Journal of Archaeological Science 11: 467–75.

Stager, L. E.
1990 Painted Pottery and its Relationship to the Weaving Crafts in Canaan During the Early Bronze

Age I. Eretz-Israel 21: 83*–88*.

Stern, E.
1988 The Southern Shefelah—Boundaries and Geographical Defenition. Pp. 12–20 in Man and Envi-

ronment in the Southern Shefelah, eds. E. Stern, and D. Urman. Givºataim: Massada (Hebrew).

Thom, R.
1975 Structural Stability and Morphogenesis. Reading: Benjamin Cummings.

Tosi, M.
1984 The Notion of Craft Specialization and Its Representation in the Archaeological Record of Early

States in the Turanian Basin. Pp. 22–52 in Marxist Perspectives in Archaeology, ed. M. Spriggs.
Cambridge: Cambridge University Press.

Tufnell, O.; Baden-Powell, D. F. W.; Bate, D. M. A.; Cerny, J.; Diringer, D.; Giles, M.; Helbaek, H.; Isserlin,
B. S. J.; Murray, M. A.; Parker, B.; Porada, E.; Thompson, F. C.; Todd, E.; and Waecher, J.

1958 Lachish IV. London: Oxford University Press.

Von der Way, T.
1987 Tell el Farain—Buto 2. Bericht. Mitteilungen des Deutschen Archäologischen Instituts, Abtei-

lung Kairo 43: 241–57.
1993 Untersuchungen zur Spätvor- und Frühgeschichte Unterägyptens. Heidelberg: Heidelberger

Orientverlag.

Von Ulrich, H.
1994 Bemerkungen zur Chronologie der Beziehungen Ägyptens zu Südkanaan in Spätprädynastischer

Zeit. Mitteilungen des Deutschen Archäologischen Instituts, Abteilung Kairo 50: 107–13.

Ward, W. A.
1991 Early Contacts Between Egypt, Canaan and Sinai: Remarks on the Paper by Amnon Ben-Tor.

Bulletin of the American Schools of Oriental Research 281: 11–26.

Wright, G. E.
1937 The Pottery of Palestine from the Earliest Times to the End of the Early Bronze Age. New

Haven: American Schools of Oriental Research.
1958 The Problem of the Transition between the Chalcolithic and Bronze Ages. Eretz-Israel 5: 37*–

45*.

Yeivin, S.
1961 First Preliminary Report on the Excavations at Tel “Gat” (Tell Sheykh Ahmed el-Areyny). Sea-

sons 1956–1958. Jerusalem: Tel Gat Expedition (Hebrew).

Yekutieli, Y.
1992 The Early Bronze IA of Southwestern Canaan—Settlement, Economy and Society. M.A. thesis,

Tel Aviv University (Hebrew).

Yekutieli, Y., and Gophna, R.
1994 The Excavations at an Early Bronze Age Site near Nizzanim. Tel Aviv 21: 162–86.

Zangvil, A.
1988 The Climate of the Southern Shefelah. Pp. 42–49 in Man and Environment in the Southern

Shefelah, eds. E. Stern and D. Urman. Givºataim: Massada (Hebrew).

oi.uchicago.edu



689



WEDGE- AND CIRCLE-IMPRESSED POTTERY:
AN ARABIAN CONNECTION



Jeffrey R. Zorn



In his 1982



The Material Culture of the Land of the Bible in the Persian Period 538–



332



b.c.



,



Stern summarized what was then known about a class of decorated pottery character-
ized by rows of impressed wedges and/or rows of impressed circles which was generally dated
to the Persian period (1982: 133–36).



1



This is an extremely valuable resource as Stern was able
to examine and briefly describe specimens from excavations that were then, and the majority
of which still are, unpublished, as well as those published up to that time. Since Stern’s work,
other examples have come to light, including one piece from the Persian period at Tel Poleg
(Herzog, Rapp, and Negbi 1989: fig. 35:2.16); four examples from Tell el-Fûl from Stratum
IIIB deposits of the sixth century from Cistern 1 (N. L. Lapp 1981: pl. 65:5–7, 13); eight pieces
from late sixth to early fifth century contexts at Shechem (N. L. Lapp 1985: fig. 6:1–8); three
pieces from the fifth century at Qadum are listed, of which two pieces are illustrated (Stern and
Magen 1984: 10–14, figs. 4, 9.1); six pieces from Ramat Ra



˙



el not published in the two later
formal site reports (Aharoni 1956: pl. 13B); five pieces from pre-Herodian fills in Jerusalem
(Tushingham 1985: fig. 15:15–16, 18–19, 21, p. 35); five pieces from Hellenistic and Roman
strata at Tell Anafa (Berlin 1997: pl. 71:565–569); two examples from fifth to fourth century
contexts at Gezer (Gitin 1990: pl. 28:27, 29, pp. 230–31); one example from an uncertain con-
text at Ashdod (Dothan and Porath 1982: pl. 25:4); one piece from a sixth to fifth century con-
text at Jabel Nimra, near Hebron (Hizmi and Shabtai 1994: 78, pl. 2:10); one piece from a
generally Persian fill at Tell el-Sumeiriya (Feig 1988–1989: pl. 13:6); seven pieces, of which
four are illustrated, are known from Jericho (Kenyon and Holland 1982: figs. 210: 12, 19, 212:
13–14). Unfortunately the Jericho pieces come from erosional washes, destruction debris
exposed to erosion and later pitting, and an uncertain phase of burnt material (see the convo-
luted presentations in Kenyon 1981: 17, 111–13, 524–25). Kenyon and Holland (1982: 455–
56) correctly note that erosion has likely removed occupation levels of the Late Iron Age and
possibly the Persian period from the mound. So far only one, perhaps two, examples are pub-
lished from Jordan, from Tell el-ºUmeiri, in contexts described as Late Iron II or Early Persian
(Lawlor 1991: 27, fig. 3.12.34; Herr 1991: 241–42; Lawlor 1997: 46, fig. 3.22.12; Herr 1997:
245). Unpublished fragments include: four examples from Tell Erani; one piece from Pisgat
Zeªev D; an unspecified number of sherds from the City of David in Jerusalem; fragments of
at least four vessels from a cave near the Holyland Hotel in Jerusalem (Ben-Arieh 2000, figs.



1. Other brief treatments are those of Wampler (1940) and P. W. Lapp (1970: 185–86).



34



oi.uchicago.edu





690



JEFFREY R. ZORN



8–9);



2



and four pieces from recent excavations on Mt. Scopus, Area E. Finally, the example
cited by Stern from Macalister and Duncan (1926) is in Annual of the Palestine Exploration
Fund Annual, volume 4 (pl. 19:3), not volume 5.

Since Stern’s publication, some additional information on this decorative style has come
to light. It is also possible to make a few observations on its geographic distribution.

The most common types of pottery on which this decoration occurs are deep kraters
(Wampler 1947: pl. 67:1510) and holemouth jars (Wampler 1947: pl. 7:96). These can either
have no handles or two–four handles which can be either horizontal or vertical (vertical handles
usually, though not always, reach the rim). Fewer examples are known from deep bowls, and
fewer still from jars or pyxis forms (Stern 1982: 133).

The decoration itself most often takes the form of impressed wedges, usually linked to each
other at their bases, arranged in one or more rows along the rim of the vessel (Aharoni 1964:
fig. 13:9; Pritchard 1964: fig. 48:17), below the rim (Wampler 1947: pl. 66:1497), at the base
of the neck (Tufnell 1953: pl. 91:405), on the shoulder (Wampler 1947: pl. 67:1521) or some
combination of the above; rarely they occur on a raised band on the neck of a krater (Wampler
1947: pl. 20:349). Sometimes the wedges are smaller and more widely spaced (Sellers et al.
1968: fig. 20:6). One example bears a line of inverted “U”s (Stern and Magen 1984: fig. 9:1).
At least one example has two bands of incised wedges along with a row of impressed wedges
(Wampler 1947: pl. 66:1495; possibly Crowfoot, Crowfoot, and Kenyon 1957: fig. 12:17). The
pyxis form seems to be the only vessel type ever to be almost completely covered with wedges
(Pritchard 1964: fig. 33:13; Lamon and Shipton 1939: pl. 43:6; Mazar and Dunayevsky 1964:
pl. 28: top center). Rows of circles are less common and usually accompany rows of wedges
(Wampler 1947: pl. 67:1510); only rarely do they occur on their own.



3



Occasionally other
stamped designs accompany the wedges, either rosettes or squares within squares (Stern 1982:
fig. 218; Mazar and Dunayevsky 1964: pl. 28: bottom right) or lozenges with checkerboard pat-
terns (Wampler 1947: pl. 20:348). An example from Samaria has a raised modeled ram’s head
(Crowfoot, Crowfoot, and Kenyon 1957: fig. 32:9A–B). There is one example of the use of
drops of clay set in rows (Stern 1982: fig. 227; Mazar and Dunayevsky 1964: pl. 28: top left).
A holemouth form with painted human, animal, and floral decoration, as well as a one-word
graffitto, was found at Qadum (Stern and Magen 1984: fig. 4). In his corpus Stern also includes
vessels stamped with floral motifs, but his example from Mevorakh contains neither wedges
nor circles, and so should be excluded from this body of material (Stern 1978: fig. 8:21, pl. 26:5).

Excavations at the site of Qasr al-Hamr



a



ª at Taym



a



ª in northwestern Arabia have produced
seven examples of wedge decoration on kraters or deep bowls, four of which have been illus-
trated (fig. 34.1 = Abu-Duruk 1986: 86:47; fig. 34.2 = Bawden, Edens, and Miller 1980: pl. 64:
15; figs. 34.3 and 34.4 = Abu-Duruk and Murad 1985: pls. 59:1, 60:2; also Abu-Duruk and



2. I am grateful to Sam Wolff (Israel Antiquities Authority) for the information on the Tell Erani pieces, to
Yonatan Nadleman (Israel Antiquities Authority) for reference to the Pisgat Zeªev example, to Sam Wolff
and Alon DeGroot (Israel Antiquities Authority) for the City of David reference, to Sarah Ben-Arieh (Israel
Antiquities Authority) for the pieces from the Jerusalem cave, and to Sam Wolff, David Amit, Jon Seligman,
and Irina Zilberbed for the Mt. Scopus specimens. I thank all of these colleagues for permission to cite these
pieces here in advance of their publications of this material.

3. Most often they appear alone on handles (Macalister 1912: pl. 182:8; Sellers et al. 1968: fig. 20:1), or on
sherds too small to be certain that the lack of wedges is not accidental (Sellers et al. 1968: fig 20: 6; though
see N. L. Lapp 1981: pl. 65:13).

oi.uchicago.edu





691



WEDGE- AND CIRCLE-IMPRESSED POTTERY: AN ARABIAN CONNECTION



FIGURE 34.1. Krater from Taym



a



ª. Scale 1:2.

FIGURE 34.2. Bowl from Taym



a



ª. Scale 1:4.

FIGURE 34.3. Krater from Taym



a



ª. Scale not given.

FIGURE 34.4. Bowl from Taym



a



ª. Scale not given.

oi.uchicago.edu

[image: image204.tif][image: image205.tif][image: image206.tif][image: image207.tif]



692



JEFFREY R. ZORN



Murad 1986: pl. 54). Figure 34.1, a krater, is decorated first with a row of circles below its rim,
followed by a row of small wedges, a row of rosettes, a second row of circles, and finally by a
row of wedges. This piece thus contains almost all the decorative devices found in this style.
All seven pieces were found in what appears to be a palace/temple complex on the small mound
of Qasr al-Hamr



a



ª. After the first season of excavation it was believed to be a single-period
occupation site which dated to the ten years (ca. 550–540



b.c.



) when Nabonidus, the last of the
Neo-Babylonian kings, made his base at Taym



a



ª (Abu-Duruk 1986: 54, 96; Abu-Duruk and
Murad 1985: 61–64). Remains of a stone table and a stela decorated with Babylonian motifs
were found in this complex (Abu-Duruk 1986: 56–66, figs. 7–8, pls. XLIX–L; Bawden, Edens,
and Miller 1980: 83–84, pl. 69). However, subsequent excavation, discussion, and evaluation
have determined that there are two periods of occupation at the site and that the period from
which the wedge-decorated vessels originate (the earliest occupation phase) begins in the
sixth century and probably extends to the end of the fifth (Parr 1989: 53–61). Another wedge-
decorated deep bowl, published without a context, is known from al-Hijr, also in northwest
Arabia (fig. 34.5 = Ibrahim and Al-Talhi 1989: pl. 14:5)

In his summary Stern traces the discussion of the dating of wedge-decorated vessels (1982:
135). After an initial assignment to the Hellenistic period, the dating of this decorative style
has gradually crept backward into the Persian period generally, and since Stern’s work, to the
end of the sixth–beginning of the fifth centuries.



4



Aharoni published six fragments of wedge-
impressed pottery from Ramat Ra



˙



el which are said to come from the floor of the courtyard
of the Stratum VA citadel and its surrounding casemate wall (1956: pl. 13B:142–43). If so,
these pieces would be the earliest examples yet excavated, dating no later than the beginning
of the sixth century. However, no subsequent report provided details as to their precise find-
spots, making an evaluation of their archaeological contexts impossible. Unfortunately the
new evidence from Taym



a



ª and al-Hijr does not allow for any more precise dating than that
arrived at for the specimens from ancient Israel.

It has been assumed that the original inspiration for wedge-decorated pottery was Meso-
potamian cuneiform (Stern 1982: 136; Wampler 1940: 15).



5



Although wedge-decorated pot-



4. Yadin et al. (1961: pls. CLXXIII:3, CCCL:15) show a fragment of a vessel of uncertain type decorated with
incised wedges and impressed circles from Stratum X of the tenth century



b.c.



This may represent the ear-
liest example of this general style of decoration in the Iron Age. However, the 400+ year gap between this
example and those pieces with impressed wedges of the sixth and fifth centuries makes a direct connection
difficult to accept.

5. Stern (1982: 136) suggests that another possible origin for wedge decoration derives from motifs on Assyr-
ian and Persian metal vessels, but he provides no examples. Zertal (1989) discusses the nature and dating of
late Iron Age bowls whose bases on the interior are covered with impressed wedges bounded by an incised



FIGURE 34.5. Bowl from al-Hijr. Scale not given.



one long

oi.uchicago.edu

[image: image208.tif]



693



WEDGE- AND CIRCLE-IMPRESSED POTTERY: AN ARABIAN CONNECTION



tery is known from Mesopotamia from as early as the Hassuna period (Zertal 1989: 81), its use
as an external decorative device is relatively rare and does not seem to occur on large kraters
or holemouth jars. A few shallow bowls are known from Nippur, but they could be either Neo-
Babylonian or Persian (McCown and Haines 1967: 71, pl. 103:16). A bowl from Nimrud is
from a Hellenistic grave (Mallowan 1966: fig. 295). If cuneiform, as observed by potters in
sixth century Israel or northern Arabia, did provide the inspiration for external wedge-dec-
oration, it is just as possible that Israel or northern Arabia was the original source for vessels so
decorated, and that the style spread east to Mesopotamia, rather than from the east to the west.

It is not surprising that identical decorative styles on similar vessel types should be found
in both ancient Israel and northern Arabia. Contacts between the two areas would have begun
to intensify with rising Assyrian intervention in the west. Tiglath Pileser III and Sargon II
made efforts to control the movement of Arab herdsman and caravans along the periphery of
their empire and used Arab leaders to control the border with Egypt, where Sargon established
a trading center (Ephºal 1982: 83–100, 108). Sargon also settled some Arabs in Samaria
(Ephºal 1982: 105–08)



.



Sennacherib campaigned in northern Arabia (Ephºal 1982: 118–23).
Assyrian military and political intervention brought at least some of the tribes of northern Ara-
bia within the Assyrian orbit certainly by the reign of Esarhaddon who relied on them to sup-
ply his army with water when he invaded Egypt in 675



b.c.



(Saggs 1984: 107–08).
Assurbanipal (Ephºal 1982: 114), and later Nebuchadnezzar of Babylon (Roux 1980: 349),
both campaigned in northern Arabia. Nabonidus, the last Babylonian ruler, moved his court to
Taym



a



ª and used it for ten years as a base of operations in Arabia, reaching even as far as
Medina in the south (Roux 1980; Ephºal 1982: 179–82). Assyrian forts and administrative
centers dot the area from the south coastal zone around Gaza, extending inland across the
northern Negev to Edom (Finkelstein 1995: 147), and Assyrian Palace Ware is common at
most sites in the same region (Finkelstein 1995). Assyrian campaigns and building activities
seem aimed at linking northwestern Arabia with southern coastal Palestine.

One of the reasons for this interest in Arabia might have been a desire to control the spice
trade routes which threaded through the region. A number of biblical texts, mostly from con-
texts of the seventh and sixth centuries, refer to Arab traders and their activities, including the
trade in frankincense (1 Kgs 10:1–13 and 2 Chron 9:1–12; Isa 21:13, 60: 6; Jer 6:20; Ezek
27:22, 38:13; Job 6:20; Ps 72:10). The increasing number of small, cuboid limestone incense
altars (Stern 1982: 182–95; Gitin 1992: 46; Stern 1973: 52–53) found in ancient Israel from
the seventh century onward attest to the importance of spices, and the trade which brought
them, in this region. An increase in the relative quantity of camel bones from Tell Jemmeh in
the seventh century and the recovery of south Arabian inscriptions from southern Israel and
Edom may also be connected with the rising importance of this trade (Finkelstein 1995: 148;
Shiloh 1987).

Although Stern notes that this type of decoration is known throughout ancient Israel and
is not limited to Judah and Samaria, as had been thought earlier, there is some patterning to its



circle. He suggests that the inspiration for the wedges derived from the experience of Mesopotamian depor-
tees exiled to the region of Samaria in the eighth–seventh centuries



b.c.



with the cuneiform writing system.
What is unclear is why these exiles would adopt a decorative scheme in their new settlements which was rela-
tively rare in their homeland. Perhaps the use of wedges in these bowls was not decorative but served some
now-obscure utilitarian role today (see London 1992).

oi.uchicago.edu





694



JEFFREY R. ZORN



distribution. The total number of published examples comes to approximately 160. Of this
total about 130 come from the area of Judah (including Jericho), while the remainder come
from the north (including Gezer).



6



Of the 130 from Judah, sixty-two come from Tell en-Nas-
beh, with most of the rest coming from Jerusalem, Ramat Ra



˙



el, En-Gedi, Tell el-Fûl, Beth
Zur, and Jericho, mainly from sites that continued to be occupied after the fall of Jerusalem,
or that were resettled soon after the return from the Exile.



7



The northern sites that have
yielded the largest number of examples are Shechem and Samaria. It thus seems that this form
of decoration was most popular in the south, showing up at many relatively small sites, while
in the north it is most often found at major population centers.

If the Ramat Ra



˙



el sherds do date to the seventh–early sixth centuries, the rise and spread
of the wedge- and circle-impressed pottery may be connected with the suggested increase in
Mesopotamian interest in exerting some control over Arabian trade, which began under the
Assyrians and continued under the Babylonians and Persians. At this stage of research, how-
ever, it remains uncertain if the perceived overlap of the generally southern range of this form
of decorated pottery and the distribution of Assyrian forts, pottery, and south Arabian inscrip-
tions is fortuitous, or in some way related. The two specimens from ºUmeiri may be a fore-
runner of future discoveries in Jordan that will provide the crucial link between northern
Arabia and southern Israel.

The high incidence of such decorated vessels at Tell en-Nasbeh/Mizpah should not be sur-
prising. The floruit of Mizpah’s development, when it served as the administrative center for
Judah under the Babylonians, was in the wake of the Babylonian conquest in 586



b.c.



, and it
continued as a major administrative center down into the fifth century.



8



One would naturally
expect such an important southern center to have some connection with the rising Arabian
trade. A fragment of one of the limestone cuboid incense altars was found at Tell en-Nasbeh
in a wall of Stratum 2 of the Babylonian to Persian period and helps underscore the impor-
tance of the spice trade in ancient Judah (McCown 1947: 236–37, fig. 61A, pl. 84:14).

Archaeological exploration of northern Arabia is still modest compared with the intensive
research carried out in Israel and, to a lesser extent, in Jordan. For this reason it is not certain
if the relative paucity of wedge-impressed pottery in northern Arabia is due to the vagaries of
excavation, or if it truly reflects a concentration of such material in ancient Israel. It is also



6. See Stern 1982: 133 and n. 1 (above) for the distribution of this material in Israel.

7. Stern 1982: 135, refers to some 140 vessels decorated with wedges and/or circles at Tell en-Nasbeh; however,
the records in the Badé Institute of Biblical Archaeology in Berkeley, California, contain references to only
approximately seventy-five vessels with this style of decoration, not all of which were published in
Wampler’s 1947 report. See Neh 3:1ff. and 7:6ff. for a list of those who helped rebuild the walls of Jerusalem
and those towns settled after the return from the Exile. Ezek 47:10 suggests that En-Gedi was settled during
this time as well.

8. The two largest clusters of wedge-decorated sherds at Tell en-Nasbeh are found in the southwest side of the
tell, from AD20 southwestward to AG17 (25 total), and in the intergate area from X12 south to AA24 (eleven
total). These two areas contain extensive building remains from Stratum 2, which this author has suggested
was constructed in the Babylonian period and continued down to the end of the fifth century



b.c.



See Zorn
(1993a: 163–85; 1993b: 1098–1102) for a discussion of the stratigraphy and architecture of Stratum 2. See
2 Kgs 25:23ff. and Jer 40:6ff. for Mizpah in the Babylonian period. See Neh 3:7, 15, 19 for Mizpah in the Per-
sian period.

oi.uchicago.edu





695



WEDGE- AND CIRCLE-IMPRESSED POTTERY: AN ARABIAN CONNECTION



unclear if the vessels themselves, or their contents, were the object of the trade. Until a source
analysis is performed on the Arabian and Israelite material, it will be impossible to say in
which direction the trade in these vessels flowed, or if these vessels were locally produced and
this form of decoration was common to both regions.

ACKNOWLEDGMENTS

In 1978 Douglas Esse wrote a seminar paper on “Settlement Patterns in the Persian
Period.” In an appendix to this paper he briefly discussed “The Stamped Impressed Ware”
which is the subject of the present article. Though the paper was never published it showed a
clear grasp of the issues involving this decorative style and contained useful insights. It is fit-
ting that in a volume honoring Doug Esse’s memory there be a piece which hearkens back to
his formative years as a student. Much of this paper was prepared during my time at the
Albright Institute of Archaeological Research, Jerusalem, as a National Endowment for the
Humanities Fellow. I would like to thank Sam Wolff for inviting me to contribute this article,
for helpful comments during its preparation, and for bringing several wedge-impressed spec-
imens to my attention. Any errors or omissions are mine alone.

REFERENCES



Abu-Duruk, H. I.
1986



Introduction to the Archaeology of Tayma



. Riyadh: Department of Antiquities and Museums.
Abu-Duruk, H., and Murad, A. J.

1985 Preliminary Report on Qasr al-Hamra Excavations; Tayma: The Second Season 1404/1984.



Atlal



9: 55–64.
1986 Preliminary Report on Qasr al-Hamra Excavations; Tayma: Third Season 1405/1985.



Atlal



10:
29–35.

1988 Preliminary Report on Qasr al-Hamra Excavations and Explorations in Taymaª during the
Fourth (Last) Season 1406/1986.



Atlal



11: 29–36.
Aharoni, Y.

1956 Excavations at Ramath Ra



˙



el, 1954: Preliminary Report.



Israel Exploration Journal



6: 102–11,
137–57.

1964



Excavations at Ramat Ra



˙



el: Seasons 1961 and 1962.



Serie Archeologica 6. Rome: Centro di
Studi Semitici.

Bawden, G.; Edens, C.; and Miller, R.
1980 The Archaeological Resources of Ancient Taym



a



: Preliminary Excavations at Taym



a



.



Atlal



4:
69–106.

Ben-Arieh, S.
2000 Salvage Excavations near the Holyland Hotel, Jerusalem.



ºAtiqot



40: 1–24.
Berlin, A.

1997 The Plain Wares. Pp. 1–211 in



Tel Anafa II,



1:



The Hellenistic and Roman Pottery



, ed. S. C. Her-
bert. Ann Arbor: Kelsey Museum, University of Michigan.

Crowfoot, J. W.; Crowfoot, G. M.; and Kenyon, K. M.
1957



The Objects from Samaria.



London: Palestine Exploration Fund.
Dothan, M., and Porath, Y.

1982 Ashdod IV.



ºAtiqot



15. Jerusalem: Department of Antiquities and Museums.
Ephºal, I.

1982



The Ancient Arabs



. Leiden: E. J. Brill.
Feig, N.

1988–1989 Wine Presses from the Hellenistic Period at Tell el-Sumeiriya.



Israel-ºAm Weªaretz



5–6: 73–91
(Hebrew).

oi.uchicago.edu





696



JEFFREY R. ZORN



Finkelstein, I.
1995



Living on the Fringe: The Archaeology and History of the Negev, Sinai and Neighboring
Regions in the Bronze and Iron Ages



. Monographs in Mediterranean Archaeology 6. Sheffield:
Sheffield Academic Press.

Gitin, S.
1990



Gezer III



. Jersualem: Hebrew Union College, Jewish Institute of Religion.
1992 New Incense Altars from Ekron: Context, Typology and Function.



Eretz-Israel



23: 43*–49*



.



Herr, L. G.
1991 Pottery Typology and Chronology. Pp. 232–45 in



Madaba Plains Project 2: The 1987 Season
at Tell el-ºUmeiri and Vicinity and Subsequent Studies



, eds. L. G. Herr, L. T. Geraty, Ø. S.
LaBianca, and R. W. Younker. Berrien Springs: Andrews University Press.

1997 The Pottery. Pp. 228–49 in



Madaba Plains Project 3: The 1989 Season at Tell el-ºUmeiri and
Vicinity and Subsequent Studies



, eds. L. G. Herr, L. T. Geraty, Ø. S. LaBianca, R. W. Younker,
and D. R. Clark. Berrien Springs: Andrews University Press.

Herzog, Z.; Rapp, G. Jr.; and Negbi, O., eds.
1989



Excavations at Tel Michal, Israel



. Tel Aviv University, Publications of the Institute of Archaeol-
ogy 8. Minneapolis/Tel Aviv: University of Minneapolis Press/Tel Aviv University, Sonia, and
Marco Nadler Institute of Archaeology.

Hizmi, H., and Shabtai, Z.
1994 A Public Building from the Persian Period at Jabel Nimra. Pp. 65–86 in



Judea and Samaria
Research Studies



3, ed. Z. H. Erlich and Y. Eshel. Kedumim-Ariºel: Research Institute of the
College of Judea and Samaria.

Ibrahim, M. A., and Al-Talhi, D.
1989 Preliminary Report on the Excavation at al-Hijr, Second Season.



Atlal



12: 21–29.

Kenyon, K. M.
1981



Excavations at Jericho



, Vol. Three.



The Architecture and Stratigraphy of the Tell:







Text and
Plates.



Oxford: Oxford University Press.

Kenyon, K. M., and Holland, T. A.
1982



Excavations at Jericho



, Vol. Four.



The Pottery Type Series and Other Finds.



Oxford: Oxford
University Press.

Lamon, R. S., and Shipton, G. M.
1939



Megiddo I



. Oriental Institute Publications 42. Chicago: University of Chicago Press.

Lapp, N. L.
1981



The Third Campaign at Tell el-Fûl: The Excavations of 1964.



Annual of the American Schools
of Oriental Research 45. Cambridge: American Schools of Oriental Research.

1985 The Stratum V Pottery from Balâah (Shechem).



Bulletin of the American Schools of Oriental
Research



257: 19–43.

Lapp, P. W.
1970 The Pottery of Palestine in the Persian Period. Pp. 179–97 in



Archäologie und Altes Testament:
Festschrift für Kurt Galling



, eds. A. Kuschke and E. Kutsch. Tübingen: Mohr.

Lawlor, J. I.
1991 Field A: The Ammonite Citadel. Pp. 15–52 in



Madaba Plains Project 2: The 1987 Season at Tell
el-ºUmeiri and Vicinity and Subsequent Studies



, eds. L. G. Herr, L. T. Geraty, Ø. S. LaBianca,
and R. W. Younker. Berrien Springs: Andrews University Press.

1997 Field A: The Ammonite Citadel. Pp. 21–52 in



Madaba Plains Project 3: The 1989 Season at Tell
el-ºUmeiri and Vicinity and Subsequent Studies



, eds. L. G. Herr, L. T. Geraty, Ø. S. LaBianca,
R. W. Younker, and D. R. Clark. Berrien Springs: Andrews University Press.

1992 Reply to A. Zertal’s “The Wedge-Shaped Decorated Bowl and the Origins of the Samaritans.”



Bulletin of the American Schools of Oriental Research



286: 89–90.

London, G.
1992 Reply to A. Zertal’s “The Wedge-Shaped Decorated Bowl and the Origins of the Samaritans.”



Bulletin of the American Schools of Oriental Research



286: 89–90.

oi.uchicago.edu





697



WEDGE- AND CIRCLE-IMPRESSED POTTERY: AN ARABIAN CONNECTION



Macalister, R. A. S.
1912



The Excavation of Gezer



, Vol. 3. London: Palestine Exploration Fund.

Macalister, R. A. S., and Duncan, J. G.
1926



Excavations on the Hill of Ophel, Jerusalem.



Palestine Exploration Fund Annual 4. London:
Palestine Exploration Fund.

Mallowan, M. E. L.
1966



Nimrud and Its Remains,



Vol. 1. London: Collins.

McCown, C. C.
1947



Tell en-Nasbeh I: Archaeological and Historical Results



. Berkeley: Palestine Institute.

McCown, D. E., and Haines, R. C.
1967



Nippur I.



Oriental Institute Publications 78. Chicago: University of Chicago Press.

Mazar, B., and Dunayevsky, I.
1964 En-Gedi: The Third Season of Excavations: Preliminary Report.



Israel Exploration Journal



14:
121–30.

Parr, P. J.
1989 Aspects of the Archaeology of North-West Arabia in the First Millennium



b.c.



Pp. 39–66 in



L’Arabie préislamique et son environment historique et culture:



Actes du Colloque de Stras-
bourg, 24–27 juin 1987, ed. T. Fahd. Leiden: E. J. Brill.

Pritchard, J. B.
1964



Winery, Defenses and Soundings at Gibeon



. Philadelphia: University Museum.

Roux, G.
1980



Ancient Iraq



, 2nd ed. New York: Pelican.

Saggs, H. W. F.
1984



The Might That Was Assyria



. London: Sidgwick & Jackson.

Sellers, O. R.; Funk, R. W.; McKenzie, J. L.; Lapp, P.; and Lapp, N.
1968



The 1957 Excavations at Beth-Zur



. Annual of the American Schools of Oriental Research 38.
New Haven: American Schools of Oriental Research.

Shiloh, Y.
1987 South Arabian Inscriptions from the City of David, Jerusalem.



Palestine Exploration Quarterly



119: 9–18.

Sinclair, L. A.
1960



An Archaeological Study of Gibeah (Tell el Fûl)



Part 1. Pp. 1–52 in Annual of the American
Schools of Oriental Research 34–35. New Haven: American Schools of Oriental Research.

Stern, E.
1973 Limestone Incense Altars. Pp. 52–53 in



Beer-Sheba I



, ed. Y. Aharoni. Tel Aviv: Institute of
Archaeology.

1978



Excavations at Tel Mevorakh (1973–1976



), Part 1. Qedem 9. Jerusalem: Hebrew University.
1982



The Material Culture of the Land of the Bible in the Persian Period 538–332



b.c.



Warminster:
Aris & Phillips.

Stern, E., and Magen, Y.
1984 A Pottery Group of the Persian Period from Qadum in Samaria.



Bulletin of the American Schools
of Oriental Research



253: 9–27.

Tufnell, O.
1953



Lachish III: The Iron Age



. London: Oxford University Press.

Tushingham, A. D.
1985



Excavations in Jerusalem 1961–1967



, Vol. 1. Toronto: Royal Ontario Museum.

Wampler, J. C.
1940 Triangular Impressed Design in Palestinian Pottery.



Bulletin of the American Schools of Oriental
Research



80: 13–16.
1947



Tell en-Nasbeh II: The Pottery



. Berkeley: Palestine Institute.

oi.uchicago.edu





698



JEFFREY R. ZORN



Yadin, Y.; Aharoni, Y.; Amiran, R.; Dothan, T.; Dunayevsky, I.; and Perrot, J.
1961



Hazor III–IV



,



Plates



. Jerusalem: Magnes.

Zertal, A.
1989 The Wedge-shaped Decorated Bowl and the Origin of the Samaritans.



Bulletin of the American
Schools of Oriental Research



276: 77–84.

Zorn, J. R.
1993a Tell en-Nasbeh: A Re-evaluation of the Architecture and Stratigraphy of the Early Bronze Age,

Iron Age and Later Periods. Ph.D. dissertation, University of California at Berkeley.
1993b Nasbeh, Tell en-. Pp. 1098–1102 in



The New Encyclopedia of Archaeological Excavations in the
Holy Land



, vol. 3, ed. E. Stern. Jerusalem: Israel Exploration Society & Carta and New York:
Simon & Schuster.

oi.uchicago.edu



699



PUBLICATIONS OF DOUGLAS L. ESSE



1981 Report of the George A. Barton Fellow at the Albright Institute, 1979–80.



American Schools of
Oriental Research Newsletter



, pp. 3–4.
1983 Harran in the Near East.



Oriental Institute News & Notes,



April.
Khirbet Kerak Publication Project. Pp. 36–38 in



Oriental Institute Annual Report



.
1984 A Chronological Mirage: Reflections on Early Bronze IC in Palestine.



Journal of Near Eastern
Studies



43: 317–30.
Harran: City of Abraham and the Moon God.



Bulletin of the Society for Mesopotamian Studies



8: 5–13.
Khirbet Kerak Project. Pp. 35–36 in



Oriental Institute Annual Report



.
1986 Levantine Trade in the Early Bronze Age: From Pots to People. Pp. 327–39 in



Proceedings of
the 24th International Archaeometry Symposium



, eds. J. Olin and M. J. Blackman. Washing-
ton, DC: Smithsonian (with Philip Hopke).

1987 Review of



The Early Bronze Age Citadel and Lower City at Ai (et-Tell)



by Joseph Callaway and
Kermit Schoonover.



Journal of Near Eastern Studies



46: 151–53.
Notes and News: Ashkelon, 1985–1986 (with Lawrence E. Stager),



Israel Exploration Journal



37: 68–72.
1987–1988 An Early Bronze Age Cemetery in the Northern Jordan Valley, Israel. Pp. 33–35 in



Oriental
Institute Annual Report



.
1988 Review Symposium on



Shiqmim I: Studies Concerning Chalcolithic Societies in the Northern
Negev Desert, Israel,



by Thomas Levy.



Mitekufat Haeven: Journal of the Israel Prehistoric
Society



21: 139–44.
Obituary for Geoffrey Shipton (1910–1987) (with Robert Braidwood).



Bulletin of the American
Schools of Oriental Research



272: 1–2.
Review of



Material Culture of the Land of the Bible in the Persian Period 538–332 B.C.,



by
Ephraim Stern.



Journal of Near Eastern Studies



47: 190–91.
Review of



Shechem I: The Middle Bronze IIB Pottery,



by Dan P. Cole.



Journal of Near Eastern
Studies



47: 283–85.
Review of



The Archaeology of the Israelite Settlement,



by Israel Finkelstein.



Biblical Archaeology
Review



14: 6–12.
From Israel to Turkey: Spring 1988. Pp. 1–5 in



Oriental Institute News & Notes,



September–
October.

1988–1989 Tell Yaqush. Pp. 27–31 in



Oriental Institute Annual Report



.
1989 Secondary State Formation and Collapse in Early Bronze Age Palestine. Pp. 81–96 in



L’urban-
isation en Palestine à l’âge du bronze ancien; Actes du Colloque d’Emmaus (Centre National
de la Recherche Scientifique),



ed. P. de Miroschedji. British Archaeological Reports,







Interna-
tional Series 527 (2 vols.). Oxford: British Archaeological Reports.
Village Potters in Early Bronze Age Palestine: A Case Study. Pp. 77–92 in



Essays in Ancient
Civilization Presented to Helene J. Kantor,



eds



.



A. Leonard, Jr. and B. Beyer Williams. Studies
in Ancient Oriental Civilization 47. Chicago: Oriental Institute, University of Chicago.
Notes from an Early Bronze Age Village



.



Pp. 1–3 in



Oriental Institute News & Notes,



November–December.
1990 Early Bronze Age Cylinder Seal Impressions from Beth-Yerah.



Eretz-Israel



21: 27*–34*.
News and Notes: Yaqush 1989.



Israel Exploration Journal



40: 222–23.
1991



Subsistence, Trade, and Social Changes in Early Bronze Age Palestine.



Studies in Ancient Ori-
ental Civilization 50. Chicago: Oriental Institute Press.
The Collared Store Jar: Scholarly Ideology and Ceramic Typology.



Swedish Journal of Old Tes-
tament



2: 99–116.
Review of



The Late Bronze and Early Iron Ages of Central Jordan: The Baqºah Valley Project,
1977–1981



by Patrick McGovern.



Journal of Near Eastern Studies



50: 219–21.



oi.uchicago.edu





700



PUBLICATIONS OF DOUGLAS L. ESSE

1992 The Collared Pithos at Megiddo: Ceramic Distribution and Ethnicity.



Journal of Near Eastern Stud-
ies



51: 81–103.
Ashkelon. Pp. 487–90 in



The Anchor Bible Dictionary



, Vol. 1, ed. D. N. Freedman. New York:
Doubleday.
Review



of Byblos in the Third Millennium B.C



., by Muntaha Saghieh.



Journal of Near Eastern
Studies



51: 141–43.
Review of Michael Avi-Yonah Memorial Volume,



Eretz-Israel



Vol. 19.



Journal of Near Eastern
Studies



51: 218–20.
1993 Yaqush. Pp. 1502–4 in



The New Encyclopedia of Archaeological Excavations in the Holy Land



,
Vol. 4, ed. E. Stern. Jerusalem: Israel Exploration Society & Carta and New York: Simon &
Schuster.

oi.uchicago.edu





701



LIST OF CONTRIBUTORS



Algaze, Guillermo
Department of Anthropology
University of California, San Diego
9500 Gilman Drive
San Diego, CA 92093
galgaze@weber.ucsd.edu

Alon, David†
Kibbutz Mishmar Hanegev
Israel

Amiran, Ruth
Israel Museum
91710 Jerusalem, Israel

Ben-Tor, Amnon
Institute of Archaeology
Hebrew University
Mt. Scopus
91905 Jerusalem, Israel
bentor@h2.hum.huji.ac.il

Braun, Eliot
Israel Antiquities Authority
P.O.B. 586
91004 Jerusalem, Israel
eliot@israntique.org.il

Carmi, Israel
Weizmann Institute
76100 Rehovot, Israel

Cohen-Weinberger, Anat
Israel Antiquities Authority
P.O.B. 586
91004 Jerusalem, Israel
cohen@israntique.org.il

Dessel, J. P.
Department of History
915 Volunteer Blvd.
6th Floor, Dunford Hall
University of Tennessee
Knoxville, TN 37996-0450
jdessel@utk.edu

Dever, William G.
Department of Near Eastern Archaeology and
Anthropology
University of Arizona
Tucson, Arizona 85721
wdever@u.arizona.edu

Fessler, Daniel
Department of Anthropology
University of California, Los Angeles
Los Angeles, CA 90095-1553
dfessler@anthro.ucla.edu

Gal, Zvi
Israel Antiquities Authority
P.O.B. 586
91004 Jerusalem, Israel
tsvika@israntique.org.il

Garfinkel, Yosef
Institute of Archaeology
Hebrew University
Mt. Scopus
91905 Jerusalem, Israel
garfinkel@h2.hum.huji.ac.il

Gilboa, Ayelet
Israel Antiquities Authority
P.O.B. 586
91004 Jerusalem, Israel

and

Institute of Archaeology
Hebrew University
Mt. Scopus
91905 Jerusalem, Israel
gilboaa@internet-zahav.net

Gophna, Ram
Department of Archaeology and Ancient Near
Eastern Civilizations
Tel Aviv University
P.O.B. 39040
69978 Tel Aviv, Israel
gophna@post.tau.ac.il

Goren, Yuval
Department of Archaeology and Ancient Near
Eastern Civilizations
Tel Aviv University
P.O.B. 39040
69978 Tel Aviv, Israel
ygoren@post.tau.ac.il

Gorny, Ronald
University of Chicago
Advisor in the College
5454 South Shore Dr.
Chicago, IL 60615
rlg2@midway.uchicago.edu

oi.uchicago.edu





702



PUBLICATIONS OF DOUGLAS L. ESSE

Greenberg, Raphael
Department of Archaeology and Ancient Near
Eastern Civilizations
Tel Aviv University
P.O.B. 39040
69978 Tel Aviv, Israel
grafi@post.tau.ac.il

Hallote, Rachel S.
56 Irving Pl.
New Rochelle, NY 10801
rhallote@mindspring.com

Harrison, Timothy P.
Department of Near and Middle Eastern
Civilizations
University of Toronto
4 Bancroft Ave.
Toronto, Ont. M5S 1C1
Canada
tim.harrison@utoronto.ca

Herr, Larry G.
Canadian University College
235 College Avenue
College Heights, Alberta T4L 2E5
Canada
lherr@cauc.ab.ca

Hesse, Brian
Department of Anthropology
The University of Alabama at Birmingham
1212 University Boulevard, Room 338
Birmingham, AL 35294-3350
Pbhesse@bellsouth.net

Ilan, David
Department of Archaeology and Ancient Near
Eastern Civilizations
Tel Aviv University
P.O.B. 39040
69978 Tel Aviv, Israel
davilan@post.tau.ac.il

Ilan, Ornit
Israel Museum
91710 Jerusalem, Israel
ornitil@imj.org.il

Joffe, Alexander H.
56 Irving Pl.
New Rochelle, NY 10801
alexjoffe@mindspring.com

Kansa, Eric C.
Harvard University
Department of Anthropology
Peabody Museum
11 Divinity Ave.
Cambridge, MA 02138
Ekansa@fas.harvard.edu



Killebrew, Ann E.
Classics and Ancient Mediterranean Studies
108 Weaver Building
Penn State University
University Park, PA 16802
aek11@psu.edu

Kolska-Horwitz, Liora
Department of Evolution, Systematics and
Ecology
Hebrew University- Givat Ram
91904







Jerusalem, Israel

Lass, Egon H. E.
Israel Antiquities Authority
P.O.B. 586
91004 Jerusalem, Israel
egon@israntique.org.il

Levy, Thomas E.
Department of Anthropology
9500 Gilman Dr.
University of California, San Diego
La Jolla, CA 92093
tlevy@ucsd.edu

Mazar, Amihai
Institute of Archaeology
Hebrew University
Mt. Scopus
91905 Jerusalem, Israel
mazar@h2.hum.huji.ac.il

Milevski, Ianir
Israel Antiquities Authority
P.O.B. 586
91004 Jerusalem, Israel
ianir@israntique.org.il

Miroschedji, Pierre de
Maison de l’Archéologie et de l’Ethnologie
CNRS UMR 7041
Boite 14
21, allée de l’Université
92023 Nanterre Cedex France
miroschedji@mae.u-parisio.fr

Raban, Avner
Recanati Center for Maritime Studies
University of Haifa
31905 Mount Carmel, Haifa, Israel
rabana@research.haifa.ac.il

Rast, Walter E.
Department of Theology
Valparaiso University
Valparaiso, IN 46383
wrast@exodus.valpo.edu

LIST OF CONTRIBUTORS

oi.uchicago.edu





703



Rosen, Arlene M.
Institute of Archaeology
University College London
31-34 Gordon Square
London WC1E 6BT
United Kingdom
a.rosen@ucl.ac.uk

Rosen, Steven A.
Archaeology Division
Ben Gurion University of the Negev
P.O.B. 653
84105 Beersheva, Israel
rosen@bgumail.bgu.ac.il

Saltz, Daniella
4137 Hardwoods Dr.
West Bloomfield, MI 48323
dsaltz@aol.com

Segal, Dror
Museum of Local and Mediterranean
Archaeology
19150 Gan Hashlosha, Israel

Sharon, Ilan
Institute of Archaeology
Hebrew University
Mt. Scopus
91905 Jerusalem, Israel
sharon@h2.hum.huji.ac.il

Silberman, Neil Asher
Clos de Pommiers 3
B-1380 Ohain
Belgium
neil.silberman@ename974.org

Stager, Lawrence E.
Harvard Semitic Museum
6 Divinity Ave.
Cambridge, MA 02138
stager@fas.harvard.edu

van den Brink, Edwin C. M.
Israel Antiquities Authority
P.O.B. 586
91004 Jerusalem, Israel
edwin@israntique.org.il

Wapnish, Paula
Department of Anthropology
University of Alabama-Birmingham
338 Ullman
Birmingham, AL 35294
pbhesse@bellsouth.net

Wolff, Samuel R.
Israel Antiquities Authority
P.O.B. 586
91004 Jerusalem, Israel
sam@israntique.org.il

Yekutieli, Yuval
Archaeology Division
Ben Gurion University of the Negev
P.O.B. 653
Beersheva, Israel
yuvaly@bgumail.bgu.ac.il

Zorn, Jeffrey R.
Department of Near Eastern Studies
Cornell University
Ithaca, NY 14853
jrz@cornell.edu

LIST OF CONTRIBUTORS

oi.uchicago.edu





704



PUBLICATIONS OF DOUGLAS L. ESSE

oi.uchicago.edu