AMERICAN JOURNAL
OF ARCHAEOLOGY
THE JOURNAL OF THE ARCHAEOLOGICAL INSTITUTE OF AMERICA
Volume 111
●
No. 4
October 2007
AMERICAN JOURNAL
OF ARCHAEOLOGY
THE JOURNAL OF THE ARCHAEOLOGICAL INSTITUTE OF AMERICA
EDITORS
Naomi J. Norman, University of Georgia
Editor-in-Chief
Madeleine J. Donachie
Vanessa Lord
Managing Editor
Assistant Editor
John G. Younger
Elizabeth Bartman
Editor, Book Reviews, University of Kansas
Editor, Museum Reviews
ADVISORY BOARD
Susan E. Alcock
Sarah P. Morris
Brown University
University of California at Los Angeles
John Bodel
Robin Osborne
Brown University
Cambridge University
Larissa Bonfante
Jeremy Rutter
New York University
Dartmouth College
John F. Cherry
Michele Renee Salzman
Brown University
University of California at Riverside
Jack L. Davis
Guy D.R. Sanders
University of Cincinnati
American School of Classical Studies at Athens
Janet DeLaine
Andrew Stewart
Oxford University
University of California at Berkeley
Natalie Boymel Kampen
Columbia University
University of Manitoba
Claire L. Lyons
Cheryl A. Ward
Getty Research Institute
Florida State University
Andrew M.T. Moore
Katherine Welch
Rochester Institute of Technology
Ian Morris
Stanford University
Lea Stirling
Institute of Fine Arts, New York University
Greg Woolf
University of St. Andrews
Jenifer Neils, ex officio
Case Western Reserve University
EDITORIAL ASSISTANTS
Kathryn Armstrong Peck, Kimberly A. Berry, Julia Gaviria, Deborah Griesmer, Benjamin Safdie
THE AMERICAN JOURNAL OF ARCHAEOLOGY, the journal of the Archaeological Institute of America,
was founded in 1885; the second series was begun in 1897. Indices have been published for volumes 1–11
(1885–1896), for the second series, volumes 1–10 (1897–1906) and volumes 11–70 (1907–1966). The
Journal is indexed in ABS International Guide to Classical Studies, American Humanites Index, Anthropological
Literature: An Index to Periodical Articles and Essays, Art Index, Arts & Humanities Citation Index, Avery Index
to Architectural Periodicals, Book Review Index, Current Contents, Humanities Index, International Bibliography of
Periodical Literature in the Humanities and Social Sciences (IBZ), and Wilson Web.
MANUSCRIPTS and all communications for the editors should be addressed to Professor Naomi J. Norman,
AJA Editor-in-Chief, Department of Classics, Park Hall, University of Georgia, Athens, GA 30602-6203, fax
706-542-8503, email [email protected]. The American Journal of Archaeology is devoted to the art and
archaeology of ancient Europe and the Mediterranean world, including the Near East and Egypt, from
prehistoric to Late Antique times. The attention of contributors is directed to “Editorial Policy, Instructions
for Contributors, and Abbreviations,” AJA 111 (2007) 3–34. Guidelines for AJA authors can also be found
on the AJA Web site (www.ajaonline.org). Contributors are requested to include abstracts summarizing
the main points and principal conclusions of their articles. Manuscripts may be submitted electronically
via the AJA Web site; hard-copy articles, including photocopies of figures, should be submitted in triplicate
and addressed to the Editor-in-Chief; original photographs, drawings, and plans should not be sent unless
requested by the editors. In order to facilitate the peer-review process, all submissions should be prepared
in such a way as to maintain anonymity of the author. As the official journal of the Archaeological Institute
of America, the AJA will not serve for the announcement or initial scholarly presentation of any object
in a private or public collection acquired after 30 December 1973, unless its existence was documented
before that date or it was legally exported from the country of origin. An exception may be made if, in the
view of the Editor-in-Chief, the aim of the publication is to emphasize the loss of archaeological context.
Reviews of exhibitions, catalogues, or publications that do not follow these guidelines should state that the
exhibition or publication in question includes material without known archaeological findspot.
BOOKS FOR REVIEW should be sent to incoming AJA Book Review Editors Professors Rebecca Schindler and
Pedar Foss, Department of Classical Studies, DePauw University, Greencastle, IN 46135-0037, tel. 765-6584760, fax 765-658-4764, email [email protected]. The following are excluded from review and should
not be sent: offprints; reeditions, except those with significant changes; journal volumes, except the first in
a new series; monographs of very small size and scope; and books dealing with New World archaeology.
THE AMERICAN JOURNAL OF ARCHAEOLOGY (ISSN 0002-9114) is published four times a year in January, April, July, and October by the Archaeological Institute of America, located at Boston University, 656
Beacon Street, Boston, MA 02215-2006, tel. 617-353-9361, fax 617-353-6550, email [email protected]. Subscriptions to the AJA may be addressed to the Institute headquarters. A subscription request form can also be
downloaded from www.ajaonline.org. An annual subscription is $75 (international, $95); the institutional
rate is $250 (international, $290). Membership in the AIA, including a subscription to the AJA, is $125 per
year. Student membership is $73; proof of full-time student status is required. International subscriptions
and memberships must be paid in U.S. dollars, by a check drawn on a bank in the U.S., by money order,
or by credit card. Subscriptions are due 30 days prior to issue date. No replacement for nonreceipt of any
issue of the AJA will be honored after 90 days (180 days for international subscriptions) from the date
of issuance of the fascicle in question. When corresponding about memberships or subscriptions, always
give the account number as shown on the mailing label or invoice. A microfilm edition of the AJA, beginning with volume 53 (1949), is issued after the completion of each printed volume. Subscriptions to the
microfilm edition, which are available only to subscribers to the printed AJA, should be sent to National
Archive Publishing Company, 300 North Zeeb Road, Ann Arbor, MI 48106. For permission to photocopy
or reuse portions of the AJA for republication, use in course readers, or for electronic use, please contact
Copyright Clearance Center (www.copyright.com). Please see the AJA Web site for back issue ordering
information. Back issues from 2002–present are available as nonprintable pdfs on the AJA Web site. Back
issues from 1885–2001 are available on JSTOR as printable pdfs to JSTOR subscribers; individuals without
authorized full-text access to JSTOR who locate an AJA article through search engines or linking partners
may purchase the article with credit card. Exchanged periodicals and correspondence relating to exchanges
should be directed to the AIA. Periodicals postage paid at Boston, Massachusetts, and additional mailing
offices. Postmaster: send address changes to the American Journal of Archaeology, Archaeological Institute
of America, located at Boston University, 656 Beacon Street, Boston, MA 02215-2006.
The opinions expressed in the articles and book reviews published in the American Journal of Archaeology
are those of the authors and not of the editors or of the Archaeological Institute of America.
Copyright © 2007 by the Archaeological Institute of America
The American Journal of Archaeology is composed in ITC New Baskerville
at the offices of the Archaeological Institute of America, located at Boston University.
The paper in this journal is acid-free and meets the guidelines for permanence and durability of the
Committee on Production Guidelines for Book Longevity of the Council on Library Resources.
The Square Temple at Tell Asmar and the
Construction of Early Dynastic Mesopotamia,
ca. 2900–2350 B.C.E.
JEAN M. EVANS
Abstract
During the 1930s, the excavations of the Iraq Expedition of the Oriental Institute uncovered extensive
remains at the Diyala sites of Tell Asmar, Khafajah, and
Tell Agrab. The archaeological levels at these sites fell
principally within the earlier third millennium B.C.E.,
which had been an ill-defined time variously referred to
as the prediluvian, Lagash, pre-Sargonid, plano-convex,
early Sumerian, or “early dynastic.” The results of the
Diyala excavations were used to delineate an Early Dynastic period of Mesopotamia, so named because the
earliest historically attested dynasties fell within it.1 In
1935, the Diyala excavators subdivided the Early Dynastic period into ED I, II, and III. The tripartite Early
Dynastic periodization was subsequently applied to
greater Mesopotamia. Ever since, archaeologists have
maintained that the Diyala displays certain regional
characteristics that are not applicable elsewhere. In
particular, the attempts to identify ED II contexts outside the Diyala region have been unsuccessful.
In this article, I review the criteria currently used to
characterize ED II. In the face of criticisms about the
applicability of ED II as defined in the Diyala, some
archaeologists have singled out new diagnostic pottery forms to date contexts outside the Diyala to ED
II, while others have considered Fara-style glyptic as
an ED II chronological marker. As discussed below,
these attempts to bolster the tripartite Early Dynastic
periodization and retain ED II terminology are problematic. I then turn to the Abu Temple sequence at
Tell Asmar, in which the Square Temple building
period formed the basis for the ED II subdivision.2 I
examine in particular detail the Square Temple archi-
* I dedicate this article to my teacher, Donald P. Hansen
(1931–2007), who shared with me his many insights into the
Early Dynastic period. It was his suggestion that ED II should
refer only to an art historical phase in the Diyala, set forth in
publication and elaborated upon in lectures and discussions,
which led me to a close examination of the Diyala contexts
in which the Early Dynastic subdivisions were established. I
am grateful to him for his many comments and for making
available the field records of the Inanna Temple excavations. I
would like to thank Clemens Reichel of the Oriental Institute
of the University of Chicago for providing me with access to
the field notebooks and other unpublished data from the Abu
Temple excavations and for reviewing a draft of this article. I
also thank the staff of the Oriental Institute Museum for assisting in the examination of Diyala objects in storage. I am grate-
ful to the many who read and commented on various drafts
of this article, including Joan Aruz, Kim Benzel, Madeleine
Cody, Paul Collins, Andrew Leung, Oscar Muscarella, Edward
Ochsenschlager, Holly Pittman, Vincent van Exel, and Richard Zettler. I thank Editor-in-Chief Naomi J. Norman and the
anonymous reviewers for the AJA for their helpful comments.
This article is a revision of certain chapters in Evans (2005),
of which parts were also presented in 2006 at the Institute of
Fine Arts, New York University, and at the annual College Art
Association meeting; I am grateful for the many useful suggestions I received at both.
1
Frankfort 1932, 1, 48–9; 1936, 35.
2
In the Diyala excavations, a building period represents a
complete rebuilding of a given structure and can contain multiple occupation levels (Delougaz and Lloyd 1942, 3).
The archaeological subdivision of the Early Dynastic
period of Mesopotamia into ED I, II, and III was established in the 1930s during excavations conducted by the
Iraq Expedition of the Oriental Institute of the University
of Chicago in the Diyala region east of Baghdad. The
Square Temple building period, part of a temple sequence
uncovered at the Diyala site of Tell Asmar, defined the
ED II subdivision along with the Asmar sculpture hoard
and Fara-style glyptic. In particular, the geometric-style
sculpture in the Asmar hoard was considered so significant
that all Diyala temple levels in which such sculpture first
appeared were correlated with the onset of ED II. Subsequently, ED II has proven to be elusive in the archaeological record of greater Mesopotamia. A review of the Square
Temple excavations suggests that the ED II subdivision
is problematic because of the criteria upon which it was
established: sculpture style was given precedence over the
Square Temple material assemblage, which is ED I. It is
therefore concluded here that even within the Diyala region itself, the concept of ED II is largely untenable, and
new interpretations are proposed for the original contexts
upon which the Early Dynastic subdivisions of Mesopotamia were formulated. More generally, and as an assessment of an established chronology, this article addresses
the various issues involved in negotiating the intersection
of stratigraphy, material culture, and dating.*
599
American Journal of Archaeology 111 (2007) 599–632
600
JEAN M. EVANS
tecture, pottery, cylinder seals, and sculpture. Many of
my observations are based on unpublished excavation
records housed at the Oriental Institute. I argue here
that ED II is paradoxically just as elusive in the Square
Temple context upon which it was established as it is
in any other context in Mesopotamia.
To a greater extent than has been previously acknowledged, the Early Dynastic periodization depended on the assumed chronological significance of
a geometric sculpture style and a realistic sculpture style
defined by the Diyala excavators. Hundreds of Early
Dynastic statues were excavated from the Diyala temple
remains. When the tripartite Early Dynastic periodization was formulated, all geometric-style sculpture was
dated to ED II and all realistic-style sculpture to ED
III. Yet the significance accorded to dedicatory sculpture styles obscured its limitations as a chronological
marker. Early Dynastic sculpture was intended for
a conservative temple context that encouraged the
continuation of certain styles and workshop traditions
even after new modes were available. For determining
the chronology of Early Dynastic Mesopotamia, pottery instead of sculpture should be used in association
with other considerations.3
The Tell Asmar sculpture hoard of 12 well-preserved
statues was the prime example of sculpture in the
geometric style, characterized by a concentration of
corporal forms into abstract shapes (fig. 1).4 Geometricstyle sculpture provided the principal criterion both for
the establishment of an ED II subdivision and for determining the relative chronology of the Diyala temples.
Although surviving in lesser quantities, associated pottery and cylinder seals from temple contexts indicate
that all levels in which geometric-style sculpture first
appeared are not contemporary with one another at
the beginning of ED II, as the Diyala excavators maintained. A consideration of the contribution sculpture
styles made to the formation of the Early Dynastic
periodization therefore highlights an additional factor behind our inability to identify ED II, even in the
Diyala excavations.
early dynastic periodization and early
dynastic ii
Although the Sin Temple at Khafajah and the Abu
Temple at Tell Asmar were of foremost importance in
defining the Early Dynastic subdivisions in the Diyala,
3
Porada et al. 1992, 103.
Frankfort 1935a, 55–78; 1935b; 1939; 1943; 1954, 23–31.
5
Frankfort 1935a, 87 n. 19; 1936, vii, 35–59. The Diyala site
of Tell Agrab also yielded Early Dynastic remains, but it was
excavated after the tripartite subdivision was established (see
Delougaz and Lloyd 1942, 218–28).
6
Frankfort 1935a, 7, 79.
4
[AJA 111
it was only after the excavation of the Abu Temple
had been completed in the winter of 1934–1935 that
a specifically tripartite subdivision was established.5
Whereas the Sin Temple building periods were numbered (Sin I–X) and grouped according to similarities
in foundations and plans, the Abu Temple building periods were given names that underscored the distinct
plan of each.6 The excavators believed that the Early
Dynastic Archaic Shrine, Square Temple, and SingleShrine Temple building periods of the Abu Temple
mirrored widespread cultural shifts:
[T]he three phases to which we have given names are
each distinguished by a complete change of plan. This
did not alter the character of the worship to which
the sanctuary was dedicated. . . . But for this very reason one is inclined to attribute the changes of plan
to general causes affecting contemporary civilization
as a whole. And the changes in plan coincide in fact
with the changes in style of a good proportion of the
objects found in the buildings.7
The Archaic Shrine, Square Temple, and Single-Shrine
Temple thus established ED I, II, and III, respectively:
“it seems, then, if we take into account the Archaic
Shrine as well as the Square Temple and the SingleShrine Temple, that the early dynastic period went
through three stages” (table 1).8
According to the excavators, material remains
confirmed the significance of the three Abu Temple
building periods. Of chief importance was the Asmar
hoard, associated with the Square Temple and “carved
in a new archaic style.”9 The sculpture of the Asmar
hoard “had no parallel among the finds from other
sites and suggested a clearly defined line between two
subdivisions of the Early Dynastic period, represented
by these successive versions of the same temple.”10 To
understand the great impression that the Asmar sculpture hoard made upon its discovery in 1934, it must be
remembered that the hoard at that time represented
the oldest monumental stone sculpture ever excavated
in Mesopotamia. Frankfort related the geometric style
of the Asmar sculpture hoard—its “pristine quality, a
vigorous and inventive stylization with obvious traces
of experiment”—to the Square Temple, which represented the most substantial architecture in the Abu
Temple sequence.11 The geometric-style sculpture of
the Asmar hoard thus confirmed the significance of
the Square Temple plan and vice versa.
7
Frankfort 1935a, 79.
Frankfort 1935a, 86.
9
Delougaz and Lloyd 1942, 157; see also Frankfort 1935a,
78, 83–4; Delougaz and Lloyd 1942, 123 n. 84.
10
Delougaz and Lloyd 1942, 157.
11
Frankfort 1943, 1.
8
2007]
THE SQUARE TEMPLE AT TELL ASMAR
601
Fig. 1. The sculpture hoard from the Abu Temple at Tell Asmar (Frankfort 1935a, fig. 63; courtesy the Oriental Institute of the University of Chicago).
The Fara-style cylinder seals in the Square Temple
further reinforced the significance of the Square
Temple plan and the associated Asmar sculpture
hoard. Fara-style glyptic, carved with friezes of humans,
animals, and fantastic creatures engaged in combat,
was so called because it appeared in great quantity at
the site of Fara during excavations conducted by the
Deutsche Orient-Gesellschaft (DOG) from 1902 to
1903.12 The DOG excavations did not provide a precise date for Fara-style glyptic. Instead, the Diyala excavators dated Fara-style glyptic from the site of Fara
to ED II on the basis of the Diyala excavations, and in
a circular manner then cited this glyptic style to confirm the date of the Square Temple. Fara-style glyptic
“clinched the question of date, and the Square Temple
must be assigned to Early Dynastic II.”13
12
Heinrich 1931.
Delougaz and Lloyd 1942, 157. See also Frankfort (1936,
41–2) and the comment on the chronological table at the end
of the report.
14
Delougaz and Lloyd 1942, 125, and the chronological
chart at the end of the volume. Of the Nintu Temple building periods (Nintu I–VII), only Q 45:4 and its immediate sur13
When the relative chronology of the Diyala was established, it was the perceived chronological value of
geometric-style sculpture that ultimately superseded
other dating criteria and became decisive for dating
to the onset of ED II all the levels in which geometric-style sculpture first appeared. For example, Nintu
Temple V at Khafajah should have been dated to ED I
on the basis of architectural criteria, but the Diyala excavators dated it to ED II because it yielded geometricstyle sculpture like that in the Asmar hoard.14 The
Main Level of the Shara Temple at Tell Agrab bore
architectural similarities to Temple Oval I at Khafajah and to the Square Temple, which might have
indicated to the excavators that earlier Shara Temple
levels were ED I.15 Instead, all these levels were dated
to ED II by the geometric-style sculpture in the Shara
roundings were excavated below Nintu VI (Delougaz and
Lloyd 1942, 92). The Nintu Temple cannot be dated by pottery (Delougaz 1952 [B.001.200a, B.416.371, C.504.367]) or
cylinder seals (Frankfort 1955, nos. 277–83, Kh. VIII 16, Kh.
VIII 230), as the few examples retrieved do not have diagnostic value.
15
Delougaz and Lloyd 1942, 249–50, 261–65.
602
JEAN M. EVANS
[AJA 111
Table 1. The Relative Chronology of Diyala Levels Cited in this Article (after Delougaz and Lloyd 1942; Frankfort
1943, 1955; Delougaz 1952).
Tell Asmar
Abu Temple
Khafajah
Sin Temple
Houses
1
IIIb
Single-Shrine
Temple I
IIIa
Tell Agrab
Temple Oval Nintu Temple
III
2
X
3
Latest Building
II
VII
4
Early
Dynastic
II
I
Protoliterate d
Square Temple
I–III
Archaic Shrine
I–IV
Earliest Shrine
IX
5
VIII
6
VII
7
8
9
10
VI
V
11
IV
12
Temple.16 On the basis of pottery, at least part of the
Shara Temple is ED I.17
After Nintu Temple V and the Shara Temple were
correlated with the Square Temple on the basis of
geometric-style sculpture, these three temples were
then correlated with the Temple Oval/Houses/Sin
Temple sequence at Khafajah (see table 1). At Khafajah, the Temple Oval and the Sin Temple were
cleared to their lowest levels before excavation began on domestic structures, dubbed the Houses, that
separated them. Absolute levels were instead used to
correlate the building periods of these three areas. I
have accepted these correlations here.18
The only pottery forms from the Diyala excavations
designated as representative solely of ED II were a
16
Delougaz and Lloyd 1942, 260.
Delougaz 1952, 60.
18
See also Wilson (1985, 50–77; 1986, 64–6), who has demonstrated the validity of these correlations for earlier Sin Temple and Houses levels.
19
Delougaz 1952, 80–6, 141–42. See also two spouted vessel forms (C.546.262, C.556.362)—unillustrated and known
17
Shara Temple
Main Level
VI
I
V
Intermediate
Earlier Building
IV
III
Earliest Remains
II
I
type of fruitstand and a type of pilgrim flask defined
by a few examples from levels 6–4 of the Houses
(= Houses 6–4) (fig. 2).19 The fruitstand—with a tall,
flaring stem surmounted by a bowl—is known from six
examples.20 The pilgrim flask—with a lentoid shape,
narrow neck, and beveled-ledge rim—is known from
four examples.21 Other pottery from ED II Diyala contexts includes forms that span the entire Early Dynastic
period, such as jars with upright handles and certain
small and medium jars, bottles, and spouted vessels.
Other pottery from ED II Diyala contexts first appears during ED I and continues into ED II (e.g., redpainted jars and squat jar lids with solid knob handles),
or first appears in the Diyala during ED II and continues into ED III (e.g., fruitstands and plain stands
by one example each in Grave 89 of Houses 6—described by
Delougaz (1952, 81) as ED II.
20
Delougaz 1952, 85 (C.366.810, C.367.810).
21
Delougaz 1952, 83 (B.806.570, B.807.570). One ED II
type pilgrim flask catalogued by Delougaz (1952 [B.806.570])
as “Houses 4?” is from an unknown context, according to Delougaz et al. 1967, 32.
2007]
THE SQUARE TEMPLE AT TELL ASMAR
from graves). Delougaz suggested that some pottery
forms spanning the Early Dynastic period have certain
decorative elements representative solely of ED II, but
these examples should not be considered distinctly ED
II on the basis that each is singular.22
The Houses 6–4 sequence at Khafajah thus represents the only Diyala context dated to ED II on the basis of ceramics. The founding of the Temple Oval (as
Temple Oval I) and the rebuilding of the Sin Temple
(as Sin VIII) were correlated with level 6 (Houses 6)
of the Houses.23 Sin VIII, however, represents the only
Diyala temple context dated solely to ED II that can
be confirmed ceramically by association with Houses
6–4. Both Temple Oval I and the rebuilding of the
Sin Temple as Sin IX date only in part to ED II; they
continue beyond the ED II Houses 6–4 and into the
time of the ED III Houses 3.
One geometric-style sculpture fragment from Sin
VIII of a male head was cited in order to correlate
the Square Temple (and by association with it, Nintu
Temple V and the Shara Temple) with the onset of
ED II at Khafajah (founding of Temple Oval I/Houses 6/Sin Temple VIII).24 Geometric-style sculpture
therefore was considered so significant that its presence de facto indicated contemporaneity—the onset
of ED II—among all the temple contexts in which it
first appeared. According to Frankfort:
The extraordinary power and inspiration which we shall
recognize upon analyzing the statues of the Tell Asmar
hoard and those of the earlier style at Khafajah can best
be understood as part of the general intensification of
cultural life to which this building activity at the beginning of Early Dynastic II testifies. The statues cannot
belong to a much earlier or a much later age.25
early dynastic ii since the diyala
excavations
Criticisms
Since the Diyala excavations, archaeologists have
maintained that certain regional characteristics in the
Diyala are not applicable elsewhere, particularly with
reference to ED II.26 Although ceramically ill-defined,
22
Delougaz 1952, 141 (C.524.350, incised fish on the shoulder of a fragmentary neckless jar; C.525.362b, checker pattern
in reserved slip on a fragmentary spouted vessel).
23
Delougaz 1940, 138–39; Delougaz and Lloyd 1942, 122–
23.
24
Frankfort 1939, 25–6.
25
Frankfort 1939, 18.
26
Perkins 1953, 49; Abu al-Soof 1967, 133; Gibson 1972, 115
n. 6; Weiss 1975, 435–36; Moorey 1979, 117–19; Moon 1981,
74; Algaze 1983–1984, 155; Collon 1988, 20; Zettler 1989, 385;
Roaf 1990, 79; Porada et al. 1992, 107–8; Kuhrt 1995, 1:27–8;
Matthews 1997, 11, 18.
603
Fig. 2. ED II pottery forms from Houses 6–4 at Khafajah: a,
tall fruitstand (scale 1:10); b, pilgrim flask (scale 1:5) (after
Delougaz 1952, pls. 167 [B.806.570], 174 [C.366.810]; courtesy the Oriental Institute of the University of Chicago).
the Diyala excavators had characterized ED II as the
most important of the Early Dynastic subdivisions, “a
time of extraordinary expansion and creativeness.”27
In contrast, the ceramically well-defined ED I subdivision was characterized in the Diyala as transitional.28
Since the Diyala excavations, the two ED II diagnostic
ceramic forms in Houses 6–4 have not been identified in any subsequent archaeological excavation in
Mesopotamia, and ED I has come to be recognized as
a period of great importance on the basis of ceramics.29 The ED I diagnostic ceramics defined in the Diyala region are represented at sites in both southern
Mesopotamia and the Hamrin.30
Beginning in 1957, soundings made within the Inanna Temple precinct at Nippur provided the first continuous stratigraphic sequence spanning the Middle Uruk
period to the Early Dynastic period.31 By comparison
with the Inanna Temple precinct excavations, Wilson
equated Sin Temple IV at Khafajah, which had been
dated earlier (to Protoliterate d ) by the Diyala excavators, with Early Dynastic I Nippur.32 Three fragmentary sealings from Inanna Temple (IT) IXA, the latest
ED I level of the Inanna Temple at Nippur, belong to
27
Frankfort 1939, 18.
Frankfort 1936, 49–59, 61–73. For the role of glyptic styles
in the characterization of ED I as transitional, see Nissen 2007,
21–2.
29
Abu al-Soof 1967; Hansen 1971, 54; 1975, 159; Amiet
1980, 204–5; Winter 1984, 103; Nissen 1988, 130–35; 2007, 20–
4; Zettler 1989, 385; Porada et al. 1992, 103–7.
30
Porada et al. 1992, 103–7.
31
For the Inanna Temple at Nippur, see Hansen and Dales
1962; Hansen 1963, 1965, 1971, 1975; Wilson 1985, 1986; Porada et al. 1992; Zettler 1992; Gibson et al. 2001.
32
Wilson 1985, 67; 1986, 65–6.
28
604
JEAN M. EVANS
the Fara style of glyptic.33 Hansen thus concluded that
Fara-style glyptic, long dated to ED II on the basis of the
Diyala, dates at least in part to ED I.34 The substantial
ED I remains of IT XI–IX were so dated on the basis of
ceramics, and IT VIIB was dated to ED IIIa on the basis
of ED IIIa Fara tablets.35 The intervening IT VIII did
not yield remains that could be considered distinctly
ED II as defined in the Diyala.36
Hansen was particularly struck by the two phases
so evident in the Inanna Temple rebuildings that fell
within the Early Dynastic period: the nearly identical
plans of the initial Early Dynastic IT XI–IX, the dramatic change of plan between IT IX and IT VIII, and
the close similarities between the latest Early Dynastic IT VIII and IT VII.37 Ultimately, Hansen proposed
that ED II should refer only to an art historical or
transitional phase in the Diyala that is contemporary
with the end of ED I in southern Mesopotamia.38 This
proposal has received a mixed reaction.39 As discussed
below, attempts instead have been made to bolster the
tripartite Early Dynastic periodization by singling out
additional pottery forms or glyptic styles in an effort
to make ED II a more definable and recognizable subdivision of the Early Dynastic period.
Shortly after the discovery of the Asmar sculpture
hoard, the excavation of stone sculpture from level III
of the Eanna precinct at Uruk and an eye inlay from Sin
IV at Khafajah disproved that the hoard represented
the oldest monumental stone sculpture in Mesopotamia.40 A few stone statues—small in scale—were also
excavated from ED I levels of the Diyala temples.41 Excavation of a realistic-style statue from Nintu Temple
VI at Khafajah demonstrated that the realistic style first
appeared before ED III, to which all examples had initially been dated.42 Ultimately, Frankfort stated that the
33
Hansen 1971, 52–3, nos. 8, 12, 13, pls. 20a, 21e–f.
Porada et al. 1968, 303–4; Hansen 1971, 54; Porada et al.
1992, 104. On the basis of IT IXA, Amiet (1980, 204–5) supported an ED I date for his série archaique of Fara-style glyptic;
see also Hrouda 1971, 112–13.
35
Hansen 1971, 48–9; Porada et al. 1992, 107–8; Gibson et
al. 2001, 552–55.
36
Porada et al. 1992, 107–8.
37
Hansen and Dales 1962; Hansen 1963, 153 n. 42; Zettler
1989, 385.
38
Porada et al. 1992, 107–8.
39
Martin (1988, 74–5), Porada (1991, 170, 172), and Matthews (1997, 11, 31) favor retaining ED II on the basis of Farastyle glyptic; Marchetti (2006) retains ED II terminology in his
study of Early Dynastic royal sculpture; Hrouda (1971, 112–
13), Amiet (1980, 204–5), and Kuhrt (1995, 27–8) favor a
lengthier ED I followed by a shorter ED II; Zettler (1989, 385)
supports the elimination of ED II.
40
For Uruk, see Nöldeke and Lenzen 1940, pls. 1, 32 (W
17878); for Khafajah, see Delougaz and Lloyd 1942, 26.
41
Frankfort 1939, nos. 92, 97; 1943, 5, As. 33:631, no. 208.
42
Frankfort 1943, no. 232. Initially, Frankfort (1943, 6) ar34
[AJA 111
geometric style possibly evolved in ED I and characterized the early part of ED II, while the realistic style
appeared during the later part of ED II and continued
throughout ED III.43 This modification is significant,
given the importance of sculpture styles in formulating
a tripartite Early Dynastic periodization, but neither
conclusion led to a reassessment of the periodization
itself. Subsequent studies of the chronology of Early
Dynastic sculpture fail to recognize the circular argumentation required to establish a chronology of sculpture styles in light of a periodization determined in
part by the sculpture styles that are being evaluated.44
Hansen proposed that some sculpture typically considered ED II is instead ED I on the basis of style.45
Attempts to Define Additional ED II Pottery Forms
Subsequent to the Diyala excavations, attempts were
made to identify new pottery forms that could be defined as representative solely of ED II. At the southern
Mesopotamian site of Abu Salabikh, the excavators
maintain that levels distinct from ED I and ED III are
present. ED II dating is used at Abu Salabikh, although
the excavations provide “only a limited idea of what
ED II material looks like.”46 A chronological sequence
proposed for spouted vessels at Abu Salabikh begins
with an earlier ED II type with a convex base that was
found in graves correlated with level II of the 6G 54C
sounding in area E, a context in which ED II pilgrimflask sherds were mistakenly identified.47 The identification of convex-based spouted vessels with ED II has
persisted in the literature. Although the identification
of a rounded or convex base is problematic, the convex base is attested already in ED I (fig. 3).
In the Diyala, at least one convex-based spouted jar
is from an ED I grave at Khafajah, but flat and round-
gued that the Nintu VI statue had been displaced by deeply
dug drains in the Houses level above the Nintu Temple, but see
Delougaz et al. 1967, 16, pl. 14.
43
Frankfort 1954, 28–9; 1955, 2–3.
44
Strommenger 1960; Braun-Holzinger 1977.
45
Hansen 1971, 54; 1975, 159; Porada et al. 1992, 105.
46
Martin et al. 1985, 17 n. 2.
47
Postgate 1977, 291–92; Moon 1987, 128. Moon (1987, 64,
no. 328) subsequently published only one pilgrim-flask sherd
from the 6G 54C sounding, noting that it was the basis for “the
‘superficial impression’ that there were many pieces of pilgrim
flask.” According to Moon (1987, 64), the sherd is not related
to the ED II Diyala pilgrim flask; correcting its findspot to level
I of the sounding, Moon dated the sherd to ED IIIa. For the 6G
54C sounding, see Postgate and Moorey 1976, 156–57; Postgate 1977, 281–82. Convex-based spouted vessels were cited
from Graves 38 and 81 as well as “graves close to Grave 81”
(Postgate 1977, 291). According to Martin et al. (1985, 4), however, Grave 81 is the only grave stratified before level II in area
A. For the correlation of the graves with the area E sounding,
see Martin et al. 1985, 6. Grave 38 is a disturbed grave containing vessels that have been assigned individual dates ranging
2007]
THE SQUARE TEMPLE AT TELL ASMAR
ed bases were sometimes classified together in the Diyala six-digit numerical classification system.48 Other
convex-based spouted jars could be from ED I Diyala
contexts, but not every pottery form or every example
of every pottery form was illustrated or recorded individually. In general, publications do not necessarily
indicate if a base is flat or convex and instead merge
such features into a single category. Numerical classification systems, photographs, and drawings can also
prohibit identifying the features of a base, raising the
likelihood that the chronological range, frequency,
and significance of such a feature cannot be accurately assessed. At the site of Kish, the representative
drawing for type 1a in Algaze’s ceramic typology is a
convex-based spouted vessel, but published drawings
of some type 1a vessels have flat bases, suggesting that
the base of this vessel type is not a defining feature.49
Two spouted vessels from the Kish Y graves at 4–6 m
below plain level, now dated to ED I by Algaze, are
convex-based, to judge by drawings.50 Flat and convexbased spouted vessels were catalogued together at the
site of Ur.51
In subsequent studies, Martin defined ED II contexts at the sites of al-’Ubaid and Fara principally on
the basis of convex-based spouted vessels and other
ceramic forms newly defined as ED II.52 ED II Fara
cannot verify ED II al-’Ubaid, and vice versa, when
from ED II to ED III (Martin et al. 1985, 6, 92; Moon 1987,
178); Grave 81 is now dated to late ED I or ED II (Martin et al.
1985, 150; Moon 1987, 128). Other contexts at Abu Salabikh
were also dated to ED II on the basis of other incorrectly identified pilgrim-flask sherds, including a house in 5I of area A
(Postgate 1984, 108; Moon 1987, 64, no. 329).
48
In the Diyala pottery classification system, a “2” for the
fourth digit indicates a flat or discoid base (Delougaz 1952,
17). Six spouted vessels (Delougaz 1952 [C.525.262c], from
Grave 79 of Houses 8 at Khafajah; see also Delougaz et al.
[1967, 89], in which Grave 79, assigned to “Houses 8?,” could
have been dug from either Houses 8 or 7, both ED I) from
an ED I grave at Khafajah are classified as having a “2” base.
Since the drawing shows a convex base, at least one example
is convex-based. Other spouted vessels from ED I Diyala contexts have a “2” base but are not illustrated (Delougaz 1952
[C.515.262, C.516.262, C.517.262]).
49
Algaze 1983–1984, 156.
50
Algaze 1983–1984, 141–48. For convex-based spouted vessels, see Ash. 1931.215 from Grave 612 at 4 m (Moorey 1978,
microfiche 2:E11) and Ash. 1932.972 from Grave 624 at 6 m
(Moorey 1978, microfiche 2:F09).
51
Woolley 1934, pl. 264 (type 209).
52
Martin 1982, 150–52, 167; 1988, 26, 50.
53
Martin 1982, 160, 166. Porada et al. (1992, 105) doubt the
validity of an ED II grave group. The principal pottery forms
cited by Martin (1982, 166) to distinguish ED II graves were
convex-based spouted jars and round-based bottles. Of the
two spouted vessel types cited by Martin, at least one example
of type LXXVI has a ring base (Hall and Woolley 1927, 203,
Grave C.92). Round-based bottles appear throughout the Early Dynastic period. For the Diyala, see Delougaz 1952, pl. 102,
605
Fig. 3. A convex-based spouted vessel from ED I Houses 8
or 7 at Khafajah (scale 1:5) (after Delougaz 1952, pl. 179
[C.525.262c]; courtesy the Oriental Institute of the University of Chicago).
the evidence from these contexts is itself uncertain.
At al-’Ubaid, Martin ultimately determined that the
evidence for convex-based spouted jars is inconclusive
and noted more generally that ED II, “when judged by
its pottery, does not stand out as a distinct period in its
own right.”53 At the site of Fara, Martin defined ED II
contexts in the deep cut in DE 38/39 and in the two
adjacent squares labeled FG 42/43 excavated by the
University of Pennsylvania for one season in 1931.54
and others, e.g., B.663.540a, B.664.540b; for Abu Salabikh, see
Moon 1987, 64–8; for Kish, see Algaze 1983–1984 (type 10a).
54
Martin 1982, 149–51; 1988, 20–5, 31–3. Martin (1988,
25–6, 33) also dated graves in DE 38/39 and FG 42/43 to ED
II. After conical bowls, medium convex-based jars and small
jars with a string-cut base were described as the most common
forms in the graves (Martin 1982, 150; 1988, 50). Martin (1988,
26, 33) argued that the dimensions of the conical bowls correspond to those in the al-’Ubaid graves, but see Martin 1982,
154. Medium convex-based jars appear throughout the Early Dynastic period (Delougaz 1952 [B.545.520, B.546.520,
B.556.520, B.566.560]; Martin 1982, 156–57; 1988 [Penn Pot
nos. 50, 51, and comparanda]; infra n. 66). At Abu Salabikh,
medium convex-based jars are catalogued as round-based jars,
which “do not, on present evidence, offer much as dating criteria” (Moon 1987, 71). Small jars with string-cut bases first appear at Fara in ED I (Martin 1988 [Penn Pot no. 48, FP 361]);
for ED I Diyala, see Delougaz 1952 (B.515.220, B.545.220c).
Upright-handled jars appear throughout the Early Dynastic
period (Delougaz 1952, pls. 76, 77). See Delougaz et al. (1967,
88) for additional findspot information for C.526.371. A type
of ring base set high such that the rounded base of the pot
protrudes below it is not distinguished in the Diyala classification system, but see Delougaz 1952 [D.515.362, D.525.362,
D.526.371]. At Abu Salabikh, this type of ring base appears in
graves related to the misidentified pilgrim-flask sherds at that
site (Moon 1987, nos. 675 [Grave 80], 677 [Grave 81], 685
[Grave 52]). Similarities between Grave 205 from the 5I house
and Grave 80 were cited to confirm an ED II date for the latter (Martin et al. 1985, 142). See Martin et al. (1985, 110) for
Grave 52; Postgate (1984, 103) for Grave 205; supra n. 47.
606
JEAN M. EVANS
Martin concluded, however, that ED II could not be
distinguished from ED IIIa among the finds from the
DOG excavations at Fara.55 Nevertheless, she upheld
the ED II date for Fara-style glyptic first proposed in
the Diyala excavations.56
The Date of Fara-Style Glyptic
With the University of Pennsylvania excavations
guiding the organization of the DOG material, Martin
undertook a stylistic and chronological classification
of all glyptic from Fara, producing comprehensively
defined categories (fig. 4).57 Some 190 seals and seal
impressions excavated at the site of Fara are carved
in the Fara style, which Martin divided into an earlier
elegant style and a later crossed style.58 Underscoring
a shift from horizontal to vertical compositions, the
elegant style consists of upright, slender figures in
symmetrical combat scenes. Filler motifs typically are
representational. Martin’s argument that the elegant
style developed from glyptic compositions, in which
male figures attack lions or two lions attack a central
prey, is supported by the stratification of Fara-style
glyptic in IT IXA at Nippur one level later than related
IT IXB sealings.59 Martin outlined a development from
the elegant style to the crossed style, in which crossing animals eliminate empty space and render filler
motifs increasingly rare.60 The crossed style is related
to ED IIIa glyptic.
55
Martin 1988, 115, 126–28.
On the basis of the Diyala excavations, Karg (1984) also
dated Fara-style glyptic and earlier glyptic styles to ED II, but
for criticism of his methodology, see Martin 1988, 133–34 n.
29; Porada 1991, 172.
57
Martin 1975, 1988.
58
Martin 1988, 72–4. See also Amiet (1980, 54–6), who
similarly subdivided Fara-style glyptic into an earlier and later
phase.
59
Martin 1988, 71, 73. Such compositions belong to the “late
ED I group” related by Martin (1988, 71, 133–34 n. 29) to glyptic from IT IXB at Nippur, SIS 4–8 at Ur, and the Y sounding
at Kish. The seal impressions from IT IXB (Hansen 1971) confirm an ED I date for seal impressions from SIS 4–8 (Legrain
1936), along with reconstructions of the pottery forms sealed
in the SIS (Zettler 1989). Martin (1988, 134 n. 32) suggested
that IT IXA, which yielded Fara-style seal impressions, could
be dated to ED II on the basis of two convex-based spouted
vessels, but both vessels have parallels in the ED I Houses at
Khafajah; supra n. 33 (for the IT IXA sealings). For the Inanna
Temple spouted vessels, see Hansen 1971, 49 (7N553, 7N558).
Three archaic tablets were also retrieved from IT IXA (Buccellati and Biggs 1969, 5, nos. 1–3, “probably Early Dynastic I”).
60
Martin 1988, 74–5.
61
Martin 1988, 66.
62
Martin 1988, table 13. It could be instead that the I d/e
glyptic comprises disparate but contemporary ED I styles.
56
[AJA 111
The DOG excavations at Fara produced a large
corpus of seal impressions, but few have recorded
findspots. From the available data, Martin nevertheless
was able to establish the general chronological distinction between an earlier elegant style and a later crossed
style. According to the field register, seal impressions
were found in two principal locations: a large dump
area in trenches I d/e yielding more than 800 seal
impressions, and houses with ED IIIa tablets.61 Thirtyfive of the 40 examples of the Fara elegant style with
recorded findspots are from the I d/e trenches, which
Martin characterized as a mixed assemblage of Jamdat Nasr, ED I, and ED II glyptic styles.62 Seven of the
54 crossed-style seals and seal impressions excavated
at Fara have identifiable findspots: all were recovered
with ED IIIa Fara tablets, and one was also associated
with ED IIIa seal impressions.63
Despite this general chronological distinction derived from the admittedly few recorded findspots for
the elegant style and the crossed style, Martin maintained an ED II date for Fara-style glyptic. Citing two
cylinder seals from DE 38/39 and FG 42/43, she associated the elegant style with pottery forms she defined
as ED II.64 Regardless of the issues associated with these
forms, the seals from DE 38/39 and FG 42/43 cannot
date the elegant style to ED II. The seal from FG 42/43
either is a surface find or is from a mixed level.65 The
seal from DE 38/39 has an uncertain findspot and,
as Porada noted, is not carved in the elegant style.66
63
Martin (1988, 75) cites seven crossed-style sealings with
known findspots, but eight are listed in her table 13.
64
Martin 1988, 74–5, nos. 284, 291. Martin (1988, 75) also
cited a sealing from Abu Salabikh, but it is from level II of the
area E sounding, in which ED II pilgrim-flask sherds were mistakenly identified. For the sealing, see Postgate 1977, 298, pl.
34e (AbS 1121).
65
Level 1 of FG 42/43 was defined as the surface to 70 cm
below, and Seal 291 is from 0–10 cm below the surface. Martin (1988, 32) suggested that a solid-footed goblet, from 50–70
cm below the surface, “might just be an earlier shape dug up
by the excavators of the silo, Pit 1, next to FG 43, and then discarded.” If this were correct, all of level 1 is mixed, as Martin
suggested. It seems reasonable to conclude that at least the material near the surface of FG 42/43, including an ED IIIa tablet
at 20 cm below, is out of context.
66
Martin 1988, 25; Porada 1991, 171. Seal 284 from DE
38/39 depicts a chain of kneeling human figures; incongruous with the elegant style are the exclusively human figures,
the lack of verticality, the tête-bêche arrangement, repetitive
design elements, abstract filler motifs, and the large, flat form
of the figures. Seal 284 is from DE 38/39, “found .60 m below surface (and possibly from grave 14),” but Martin (1988,
25) discussed Seal 284 instead with level 1 of DE 38/39, defined as 30–80 cm below the surface. Pottery from level 1 of DE
38/39 consists of a conical bowl, a medium convex-based jar,
and the base of a solid-footed goblet (Martin 1988, 25 [FP 581,
2007]
THE SQUARE TEMPLE AT TELL ASMAR
Citing the Diyala excavations, Martin also dated the
crossed style to ED II despite the examples associated
with ED IIIa at Fara.67
Attempts to define new pottery forms as ED II diagnostic and to uphold an ED II date for Fara-style glyptic
reflect the shortcomings of applying the tripartite ED
periodization developed in the Diyala excavations to
greater Mesopotamia. I conclude from the above review that the new criteria proposed for distinguishing
ED II pottery cannot be accepted. The ED II date of
Fara-style glyptic can be upheld only by accepting as
ED II the Diyala contexts in which this style appears.
Fara-style glyptic cannot be dated to ED II at the site of
Fara, and it is already present at Nippur during ED I.
It is necessary to turn now to a reexamination of ED II
in the Diyala itself. Beginning with the Square Temple,
I argue that the identification of ED II in the Diyala
is paradoxically just as elusive as the identification of
ED II outside the Diyala.
607
a
b
abu temple excavations
The Oriental Institute excavations at Tell Asmar
consisted of five campaigns from 1930 to 1935.68 During the third season of excavation (1932–1933), temple remains were discovered on the northern part of
the tell and dubbed the Abu Temple because an inscribed copper bowl in a hoard from the nearby Earlier Northern Palace was dedicated to the god Abu.69
The excavators defined four building periods in the
Abu Temple, which were further subdivided into occupation levels (table 2). Additional remains in the
Abu Temple sequence were poorly recorded and only
briefly described in publications. About 3 m of occupational debris separated the earliest Abu Temple
building period (Earliest Shrine) from virgin soil. Between the Archaic Shrine and the Square Temple was
a “predecessor” to the Square Temple followed by a
“construction pavement.” An “Intermediate” building or period with two “pavements” was encountered
between the Square Temple and the Single-Shrine
Temple (see table 2).
An elevation taken from the bitumen-paved threshold of Ablution Room D 17:5 put the first occupation
level of the Square Temple (Square Temple I) at
32.30 m. Twenty centimeters above Square Temple I
(at 32.50 m) was a second occupation level (Square
Fig. 4. Glyptic styles at the site of Fara: a, “Early ED I group”;
b, “Late ED I group”; c, Fara elegant style; d, Fara crossed
style; e, “ED IIIa” (a, scale 1:2; b–e, scale 1:1) (Martin 1988,
nos. 130, 197, 250, 365, 438).
FP 582], 176, and microfiche no. 2, appx. 8, 230 [FP 242]).
Martin (1988, 26) concluded that medium convex-based jars
“became popular during ED II,” but Seal 284—if indeed from
level 1 of DE 38/39—and the solid-footed goblet base support
an ED I date for DE 38/39.
67
Martin 1988, 75. See also Martin (1988, 38): “clearly there
is a possibility that the Crossed Style seals continue into ED
IIIa.”
68
Preliminary Abu Temple excavation reports were written
by Frankfort (1934, 1935a, 1936), and the final publication
for architecture was written by Lloyd (Delougaz and Lloyd
1942, 156–217).
69
Delougaz and Lloyd 1942, 159, 298, no. 12; Delougaz et
al. 1967, 184–85.
c
d
e
608
JEAN M. EVANS
Table 2. The Stratigraphic Sequence of the Abu
Temple at Tell Asmar.
Building
Period
Occupation
Level
Other Remains
IV
III
Single-Shrine
Temple
II
I
Intermediate
building/period
III (33.0–33.5 m elev.)
(33.0 m floor elev.)
II (32.5–33.0 m elev.)
(32.5 m floor elev.)
Square Temple
I (31.5–32.5 m elev.)
(32.3 m floor elev.)
construction
pavement (31.8 m
elev.)
predecessor to
Square Temple
(no elev. given)
IV
III
II
Archaic Shrine
I
Earliest Shrine
3 m occupational
debris
(virgin soil)
Temple II), which was described as essentially the same
in plan. The third occupation level, at 33 m (Square
Temple III), represented “a rebuilding of which insufficient traces were left to make a plan.”70 One representative plan was drawn for Square Temple I, II, and
III (fig. 5). Square Temple II and III are only briefly
70
Delougaz and Lloyd 1942, 177.
Delougaz and Lloyd 1942, 209–12.
72
The finds from Archaic Shrine IV are catalogued as
“29.80–31.50 m” (Delougaz and Lloyd 1942, 208), but the elevations of individual occupation levels in the Archaic Shrine
are not published.
73
Frankfort 1939, 1943.
74
One of the sculpture finds at 32.30 m is a plaque fragment
71
[AJA 111
mentioned in the Diyala publications, and few finds
are assigned to these levels.
The Square Temple finds were catalogued by building period and occupation levels, which were assigned
a range of absolute elevations: Square Temple I (ca.
31.50–32.50 m); Square Temple II (ca. 32.50–33.00
m); and Square Temple III (ca. 33.00–33.50 m).71
The range in findspot elevations for items catalogued
as Square Temple I extends 80 cm below the Square
Temple I floor.72 The construction pavement at 31.80
m and the predecessor to the Square Temple fall
within the 80 cm below Square Temple I. Any finds
retrieved from either of these levels would have been
catalogued as Square Temple I. Only the findspot elevations for Abu Temple sculpture are published (table 3).73 Twenty of the 22 sculpture finds catalogued
as Square Temple I—including the Asmar hoard as
well as other sculpture from D 17:6, D 17:9, and E
17:20—have a findspot elevation of 31.85 m, which is
45 cm below the Square Temple I floor.74
The published findspot elevations for sculpture
suggested to me that additional finds catalogued as
Square Temple I in actuality were from below the
Square Temple I floor. Unpublished findspot elevations do indeed suggest a wider distribution for cylinder seals and pottery catalogued as Square Temple I.
The excavation records from the Abu Temple excavations include two field notebooks written by Lloyd
that summarize the excavations by locus. Lloyd’s field
notebooks contain some additional descriptions of object findspots beyond those described in Diyala publications.75 Some additional information is available
on locus cards housed at the Oriental Institute. For
the majority of finds catalogued as Square Temple I,
however, only findspot elevations can indicate a more
precise context.
The elevations of the Square Temple occupation
levels are obviously representative in nature. Lloyd’s
notebooks describe a greater range for any given occupation level and indicate that additional floor levels
were encountered during excavation of the Square
Temple and below it. At the same time, a description
in Lloyd’s notebook of, for example, the clearing of
the Square Temple I floor can be followed through
references to “the 32.3 m floor.” In the notebooks,
(As. 33:102) that joins to fragments from 31.85 m (As. 33:435)
and 32.45 m (As. 33:350); see Frankfort 1939, no. 194.
75
E.g., only four pottery findspots are described by Lloyd:
C.356.010 in the D 17:5 sink (Delougaz and Lloyd 1942, fig.
138); C.357.010b from “a somewhat higher level” (Delougaz
and Lloyd 1942, 179); C.39-.0-- and As. 33:493 in D 17:8 were
found 1 m before the altar (Delougaz and Lloyd 1942, 183,
fig. 145).
2007]
THE SQUARE TEMPLE AT TELL ASMAR
609
Fig. 5. Plan of the Square Temple at Tell Asmar, with the predecessor to the Square Temple indicated in broken lines (Delougaz and Lloyd 1942, pl. 22; courtesy the Oriental Institute of the University of Chicago).
Lloyd often distinguished between the 32.30 m Square
Temple I floor and excavation below it. I have made a
general distinction here between the finds in Square
Temple I and the finds below Square Temple I. Some
findspot elevations can be correlated with specific floor
levels recorded in Lloyd’s notebooks, so I have been
more precise in my observations in these instances.
I should stress here, however, that a comprehensive
revision of the Square Temple excavations is a different topic entirely.
The finds that can be assigned to below Square
Temple I are significant because the building remains
encountered there likely constitute actual occupation
levels, albeit poorly preserved and poorly recorded, in
the Abu Temple sequence. When these remains are
taken into consideration, the Square Temple plan is
no longer the radical innovation that the Diyala excavators believed it to be. Rather, the Abu Temple
plan simply evolved slowly over time from the Archaic
Shrine to the Square Temple. Reinterpreting the transition in the Abu Temple sequence from the Archaic
Shrine to the Square Temple as a gradual evolution
is significant, for it weakens the perceived novelty of
the Square Temple plan, which necessitated the ED
II subdivision.
Square Temple and Its Predecessor
The plan of the Square Temple consists of a central space surrounded by rooms. Three cellas were
identified (Shrines I–III), each with an entrance in a
–
–
–
–
–
–
Square Temple II (32.50)
32.45
Square Temple I (32.30)
32.00
31.85
31.20
b
a
–
32.90
–
12, 14,
33, 61
–
–
–
–
–
–
–
–
–
194
–
–
–
–
–
–
155
66
–
–
63
93, 94,
95
–
–
1–11, 15,
16, 96, 194
–
–
194
–
–
–
–
–
–
–
–
–
–
–
–
–
97,
As. 33:631b
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
14, 29, 67,
68, 99, 169,
171, 172,
175
–
–
–
–
–
–
–
–
–
13
–
–
–
–
–
–
The sculpture is cited by the catalogue numbers in Frankfort 1939, 1943. Frankfort 1943, nos. 256, 257, 323 do not appear here because no findspot elevations were recorded
As. 33:631 is not catalogued in these volumes and therefore is referred to by its excavation number
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
Square Temple III (33.00)
–
–
–
–
–
–
Single-Shrine foundations,
bottom (33.50)
–
–
E 17:20
–
_
–
–
179
199
E 17:12
–
–
–
–
–
–
E 17:11
–
–
–
–
–
D 17:15
33.75
–
–
–
62
D 17:12
–
_
–
–
–
D 17:9
–
–
–
–
–
D 17:8
98
–
–
–
D 17:7
34.00
–
_
_
D 17:6
–
–
–
a
–
178, 186,
199
34.50
177, 180
D17.2
Single-Shrine Temple I
(34.40)
–
D 17:1
35.00
Elevation (m)
Table 3. Findspot Elevations and Loci for Sculpture from the Abu Temple at Tell Asmar.
610
JEAN M. EVANS
[AJA 111
2007]
THE SQUARE TEMPLE AT TELL ASMAR
long wall providing a bent-axis approach to the altar,
set against a short wall. According to Frankfort, the
Square Temple existed “for a very short period” as a
single cella accessed from a central court, and Lloyd
noted that the Square Temple “had an irregularly
planned and less pretentious predecessor.”76 Little
information is available regarding these architectural
remains—the “predecessor”—below Square Temple I.
No elevation is given, and Lloyd was uncertain about
the location of its exterior walls. The predecessor walls,
preserved to a maximum height of 30 cm, had been
cut down for the Square Temple foundations. The
predecessor plan is included on the Square Temple
plan (see figs. 5, 6).
Frankfort, who also wrote the preliminary excavation reports, initially had an understanding of the Abu
Temple sequence that differed from that of Lloyd.
Frankfort viewed both the predecessor and the construction pavement as Square Temple occupation levels. He noted that “five successive floor levels indicate
four reconstructions, none of them very thorough . . .
the Square Temple had been in use continuously.”77
These five successive floor levels correspond to levels
designated by Lloyd as the predecessor, the construction pavement, and Square Temple I, II, and III.78
In contrast to Frankfort, Lloyd designated Square
Temple I “the main and probably the earliest genuine occupational level” of the Square Temple—“a
complete rebuilding” of a “new temple” with a “new
plan.”79 Lloyd also suggested some continuity with the
predecessor, noting alternately that “the form of the
Square Temple was already foreshadowed” and “important elements of the Square Temple were already
embodied” in the predecessor.80
Two levels between Square Temple I and Archaic
Shrine IV on the section through the Abu Temple can
be identified as the construction pavement at 31.80
m and the predecessor below it (fig. 7).81 According
to Lloyd, the construction pavement represented
76
Frankfort 1935a, 13; Delougaz and Lloyd 1942, 172.
Frankfort 1939, 3–4; see also Tunca 1984, 22.
78
Frankfort (1943, 5), citing Delougaz and Lloyd (1942),
later conceded that the Asmar hoard must “be assigned to the
earliest occupation level proper of the temple.”
79
Delougaz and Lloyd 1942, 172–73, 177.
80
Delougaz and Lloyd 1942, 170, 172.
81
On the Abu Temple section, Archaic Shrine IV has four
occupation levels, but only three (levels A–C) are discussed
in Delougaz and Lloyd 1942. Delougaz and Lloyd (1942, fig.
132) confirm that the predecessor directly followed Archaic
Shrine IV because a predecessor wall is directly above the Archaic Shrine IV altar.
82
Delougaz and Lloyd 1942, 176–77.
83
Delougaz and Lloyd 1942, fig. 131. Tunca (1984, 14–26)
77
611
the surface to which the predecessor had been leveled; builders then used the level to lay the Square
Temple foundations.82 On the section through the
Abu Temple, however, only one wall of the predecessor is shown as level with the construction pavement.
While inconsistencies prohibit taking the section as
the final word, a photograph showing the predecessor
altar preserved almost to the waist level of a workman
standing before it also indicates that the predecessor
was not leveled at the construction pavement.83 If the
construction pavement were an actual occupation
level, as Tunca has argued, this would mitigate the
impression of a total rebuilding that is implied by the
otherwise singular example in the Diyala of an intermediary construction phase.84
Even with only the meager recorded data, the predecessor is clearly related to the Square Temple (see
fig. 6). Both consist of rooms grouped around a central
space, with D 17:2 and Cella D 17:1 of the predecessor rebuilt as D 17:7 and Cella D 17:8 of the Square
Temple. These central spaces have a circular mudbrick
structure and are accessed on the north.85 The fragmentary west wall of E 16:40 of the Square Temple is
similarly positioned at the level of the predecessor.
Both cellas have an entrance in the long east wall. The
predecessor altar and the D 17:8 Square Temple altar
have a vertical, bitumen-lined chase in the east face;
directly below it, a bitumen-lined trough or basin was
set into the floor of the predecessor.86 Lloyd suggested
that the predecessor wall had been incorporated into
the west wall of D 17:8 of the Square Temple, just as
the predecessor altar had been incorporated into the
D 17:8 altar of the Square Temple.87
During the Square Temple building period, D 17:9
and E 16:40 were regarded by Lloyd as a new, northward extension of the Abu Temple.88 According to the
predecessor plan, however, walls were excavated in the
area corresponding to E 16:40 of the Square Temple.
Lloyd also described a smaller altar beneath the D 17:9
dismisses the Abu Temple section as schematic in conception
and ultimately unreliable.
84
Tunca 1984, 21.
85
Delougaz and Lloyd 1942, 172. In a version of the Abu
Temple section published in a preliminary report (Frankfort
1936, fig. 2), the mudbrick structure is at the construction
pavement.
86
Delougaz and Lloyd 1942, 172.
87
Delougaz and Lloyd 1942, 174.
88
Delougaz and Lloyd 1942, 173. The north wall of D 17:9
was “founded only a few centimeters beneath the Level I
pavement” (Delougaz and Lloyd 1942, 191), but in another
passage, the limits of the predecessor were “not altogether certain” (Delougaz and Lloyd 1942, 172).
612
JEAN M. EVANS
[AJA 111
Fig. 6. Plan of the Square Temple predecessor (solid lines), with the Square Temple plan in the background (cross-hatched
lines) (adapted from Delougaz and Lloyd 1942, pl. 22; courtesy the Oriental Institute of the University of Chicago).
altar of the Square Temple that does not appear on
any plan, and a circular hearth full of ashes seemingly
on the same floor level was excavated some distance
before it.89 The smaller altar appears in a photograph (fig. 8) and corresponds to its description as
“hardly more than a pedestal.”90 According to Lloyd’s
field notebook, the smaller altar seems to belong to
a floor at 32.16 m. The Square Temple I floor, therefore, was not the earliest floor level in the D 17:9 area.
Certainly, the D 17:9 smaller altar anticipates the D
89
Delougaz and Lloyd 1942, 191.
Delougaz and Lloyd 1942, 191, fig. 130, with the caption
stating that the smaller altar (a) rests on the Square Temple
floor (b). However, the identification of b as the Square Temple floor is incorrect—the smaller altar belongs to a 32.16 m
90
17:9 Square Temple altar. Although the smaller altar
is at a higher level than the predecessor, the D 17:9
smaller altar also would have coexisted at least briefly
with the predecessor altar in D 17:8, given that this
altar was standing until it was incorporated into the
Square Temple I altar. The D 17:9 smaller altar and the
D 17:8 predecessor altar, in rooms arranged around a
central court, suggest an intermediary phase between
the one altar in Archaic Shrine IVC and the three altars in the Square Temple.91
floor.
91
Lloyd (Delougaz and Lloyd 1942, 191) suggested the
smaller altar in D 17:9 was a survival of an altar near the entrance to Archaic Shrine IVA–B, but the latter is slightly east of
and aligned differently than the smaller altar.
2007]
THE SQUARE TEMPLE AT TELL ASMAR
613
Fig. 7. Section through the Abu Temple at Tell Asmar, taken through D 17:8, D 17:7, E 17:20. The Asmar hoard is indicated
by an X in the D 17:8 area (modified from Delougaz and Lloyd 1942, pl. 24a; courtesy the Oriental Institute of the University
of Chicago).
The predecessor is also related to Archaic Shrine
IVC (fig. 9), for some of the predecessor walls were
built in roughly the same position as the earlier Archaic Shrine IVC walls.92 For example, predecessor
walls correspond to the long exterior walls on the north
and south of Archaic Shrine IVC. However, entrances
through the predecessor walls now provide access to
what on the Archaic Shrine IVC plan had been the
long area of D 17:15 and the partially excavated area
north of Cella D 17:10. Walls in the E 16 area northeast
of Archaic Shrine IVC proper continue on the predecessor plan, as does D 17:11 of Archaic Shrine IVC,
which is rebuilt as E 17:3. Two small rooms (E 17:1, E
17:2) on the predecessor plan recall the trapezoidal
form characteristic of certain rooms appearing during all the Archaic Shrine occupation levels. These
trapezoidal rooms represent an irregular partitioning of space that is no longer present in the Square
Temple.
A similar observation can be made regarding the
overall form of the Square Temple plan. As the dimensions of the Abu Temple expanded, the space
surrounding the Archaic Shrine hypothetically would
92
For a discussion of Archaic Shrine IV, see Tunca 1984,
18–21.
have been appropriated for the temple proper. Although Lloyd stated that the exterior walls of the predecessor were uncertain, the entrances to the north
and south from D 17:2 of the predecessor indicate
that the expansion culminating in Square Temple I
has begun. At the same time, the walls in E 16 and the
D 17:11 room adjacent to Archaic Shrine IVC proper
continue on the predecessor plan; the Abu Temple
likely had not yet fully expanded into this area. The
regularity in shape that gives the Square Temple its
name has not yet been obtained at the time of the predecessor, and the predecessor therefore also retains
strong ties to the Archaic Shrine.
Square Temple Pottery
Although a preliminary excavation report attributed
most of the pottery from the Square Temple to “the
last phase of its occupation,” all Square Temple pottery was catalogued as Square Temple I in Pre-Sargonid
Temples.93 In Pottery from the Diyala,94 published 10
years later, occupation levels are not listed for Square
Temple pottery. According to unpublished findspot
elevations, the Square Temple pottery can be assigned
93
94
Frankfort 1936, 45; Delougaz and Lloyd 1942.
Delougaz 1952.
614
JEAN M. EVANS
[AJA 111
on the basis of findspot elevations to below Square
Temple I, to Square Temple I itself, and to Square
Temple II (table 4). All identifiable illustrations of
Square Temple pottery appear in figure 10.95
Only a small quantity of pottery was catalogued from
the Square Temple. With the exception of stands, no
vessels are complete, and the majority are sherds. It is
certainly possible that some sherds are residual. Square
Temple Scarlet Ware sherds were described as surviv-
als from ED I, although Delougaz argued on the basis
of two vessels lacking context but bearing elaborate
painted designs that Scarlet Ware continued into ED
II.96 Additional pottery was retrieved during excavation of the Square Temple, but it was not recorded.
For example, “a large earthenware storage jar with
horizontal flutings” in D 17:6 and “fragments of a potstand or brazier pierced with window-like apertures” in
E 17:20 were not catalogued.97 It is not clear to which
occupation level these vessels belonged.
It was noted in a preliminary excavation report that
the majority of Square Temple pottery consisted of
solid-footed goblets and reserved-slip ware, but only
one solid-footed goblet (As. NR:6) and two vessels with
reserved-slip (As. 33:424, As. 33:425) were catalogued
from the Square Temple.98 The solid-footed goblet has
neither a findspot nor a complete excavation number.
Delougaz accepted the attribution of the solid-footed
goblet to the Square Temple and described its lower
part as “more regular in shape and of better workmanship” than ED I solid-footed goblets, perhaps to
explain why it is the only solid-footed goblet from the
Diyala excavations cataloged in an ED II context.99 The
solid-footed goblet is a hallmark of ED I, concentrated
in the middle of the period and less abundant in later
ED I levels.100
Seven sherds have findspot elevations that are below
the Square Temple I floor. Four sherds have incised
decoration; four-lugged vessels with incised decoration are characteristic of ED I.101 One sherd with an
imitation rope handle and ridges (As. 33:406) is from
a type of red-painted jar known from two examples
from ED I Diyala contexts.102 According to Delougaz,
the continuation of this vessel type into ED II “is indicated by sherds,” of which the Square Temple sherd is
the only catalogued example.103 One sherd described
as red and another as purple are from painted wares.
Unlike red-painted jars with plastic ornamentation,
plain red-painted jars first appear in the Diyala during ED I and continue into ED II with two examples
from Houses 6 at Khafajah.104
95
Pottery from the following loci is not considered here because the loci are outside the Square Temple proper: D 17:12,
D 17:13, and D 17:16. Frankfort 1936, pl. 1, nos. 3 (conical
bowl), 5 (fragmentary spouted vessel), 10 (sherd), 12 (painted
sherd), and 13 (painted sherd) cannot be identified.
96
Delougaz 1952, 69–72, 80. With the exception of one Scarlet Ware sherd from Houses 4 (Delougaz 1952 [Kh. III 592]),
no other Scarlet Ware in Delougaz (1952) can be securely attributed to an ED II context.
97
Delougaz and Lloyd 1942, 180, 192.
98
Frankfort 1935a, 18, 81; see also Frankfort 1936, 46, 56–7.
On the basis of this pottery, Wilson (1985, 78) concluded that
at least part of either the predecessor or the Square Temple is
ED I; see also Porada et al. 1992, 105.
99
Delougaz 1952, 82. One solid-footed goblet (Delougaz
1952 [B.087.700]) from a grave in “Houses 7?” is from either
Houses 7 or 6; see Delougaz et al. 1967, 91. An asterisk following its findspot indicates “some doubt as to the exact shape of
the specimen” (Delougaz 1952, 153).
100
Delougaz 1952, 56–7, 136; Hansen 1965, 209. Solid-footed
goblets are concentrated in Sin V–VI, Houses 10, and Archaic
Shrine III in the Diyala and in IT XI at Nippur (Wilson 1986,
63, fig. 11[1]).
101
Delougaz 1952, 53–5.
102
Delougaz 1952, 72, 80 (D.514.370b, Archaic Shrine III,
Houses 9 or 8).
103
Delougaz 1952, 80.
104
Delougaz 1952, 72, 80 (C.514.370a, C.515.270).
Fig. 8. The D 17:9 area of the Abu Temple at Tell Asmar.
The smaller altar (a) is set on a floor (b) at 32.16 m; the Asmar hoard was excavated in area c. The altar (d ) belongs
to Archaic Shrine IV (after Delougaz and Lloyd 1942, fig.
130; courtesy the Oriental Institute of the University of
Chicago).
2007]
THE SQUARE TEMPLE AT TELL ASMAR
615
Fig. 9. Plan of the Square Temple predecessor (solid lines) superimposed on the plan of Archaic Shrine IVC (cross-hatched
lines) (adapted from Delougaz and Lloyd 1942, pl. 21b; courtesy the Oriental Institute of the University of Chicago).
A fragmentary conical bowl (As. 33:543), a fragmentary jar (As. 33:542), and a sherd (As. 33:549) from E
16:40 with no extant illustrations can be further defined by their classification numbers. Given that the
latter two have ED III and later parallels and that E
16:40 was partially disturbed by a pit, these vessels are
likely out of context.105
Other pottery (13 examples) can be assigned to
Square Temple I. Two spouted vessels with reserved
slip (As. 33:424, As. 33:425) are fragmentary. Spouted
vessels with carination between the shoulder and the
body, a rimless neck, and reserved-slip decoration are
characteristic of ED I–II Diyala.106 The reserved slip
decoration on one spouted vessel is specified, con-
sisting of a checker or gridlike pattern on its shoulder, which Delougaz suggested might indicate an ED
II date.107 No additional examples of this motif have
been identified. One sherd (As. 33:238) has a transversely pierced beak-lug attached to a notched ridge
and is from the shoulder of a four-lugged neckless jar
characteristic of ED I.108 Painted wares include a Scarlet Ware sherd; two sherds with unspecified painted
decoration; and a red sherd. No illustrations are available for a conical bowl (As. 33:366) and a sherd with
a plaited handle (As. 33:139).
A variety of stands can also be assigned to Square
Temple I on the basis of findspot elevations.109 In the
Diyala, such stands are restricted to temple contexts.110
105
Delougaz 1952, 100 (A.645.720, Houses 2; A.654.720,
Houses 2; B.656.720, Old Babylonian).
106
Delougaz 1952, 52–3, 80–1, 91–3, 143; see also Zettler
1989, 381, 384.
107
Delougaz 1952, 141.
108
Delougaz 1952, 53–5, 136. Four-lugged neckless jars are
typically incised, but see Delougaz 1952 (C.534.313, Archaic
Shrine II; C.533.313, Sin V, with the lugs set laterally).
109
Two fragmentary Square Temple stands (As. 33:493, As.
33:494) have fenestration comparable to two stands or braziers (Delougaz 1952 [C.236.010]) from the ED I ceramic assemblage of the Hill B sounding at Tell Agrab, but the latter
have more slender proportions.
110
A type of plain stand appears in ED II–III graves at Khafajah (Delougaz 1952, 81 [B.356.000, B.357.000, C.356.000,
C.357.000]), and one plain stone stand is from a grave in the
ED I Houses at Khafajah (Delougaz et al. 1967, 84–6 [Grave
72, Kh. V 244]). At Kish, stands with excised decoration were
common in the Early Houses stratum of the Y sounding, but
in the associated Y graves, plain cylindrical stands of sheet
copper were instead predominant; see Moorey 1966, 35, 41.
Only plain stands are found in the Early Dynastic graves at Abu
Salabikh (Postgate 1980, 73). Such variations are likely due
to context, not chronology, and stands from temple contexts
therefore should not be correlated with stands from funerary
contexts.
616
JEAN M. EVANS
[AJA 111
Table 4. Findspots of Square Temple Pottery According to Findspot Elevations.
Oriental Institute Records
Excavation
No.
Description
Oriental Institute Publications
Locus/
Context
Elev.
(m)
Frankfort
1935a
Frankfort
1936
Delougaz and
Lloyd 1942
Delougaz
1952
Below Square Temple I (below 32.30 m)
As. 33:376
sherd, incised
D 17:6
31.85
–
–
–
–
As. 33:377
sherd, incised
D 17:6
31.85
–
–
–
–
As. 33:364
sherd, incised
D 17:6
32.00
–
–
–
–
As. 33:404
sherd, incised
D 17:7
32.00
–
–
–
–
As. 33:405
sherd, red
D 17:7
32.00
–
–
–
–
As. 33:406
sherd, applied dec.
D 17:7
32.00
–
pl. 1
rope/loop
design
_
As. 33:407
sherd, purple
D 17:7
32.00
–
–
–
–
Square Temple I (floor at 32.30 m)
As. 33:362
stand, mended
D 17:5
32.30
fig. 17
pl. 1
C.356.010
pl. 68f
As. 33:238
sherd, applied dec.
D 17:6
32.30
–
pl. 1
pierced lug
–
As. 33:674
stand, beveled rim
D 17:6
32.30
fig. 17
pl. 1
C.355.010
pl. 68g
As. 33:366
bowl, conical
D 17:8
32.30
–
–
C.082.200
C.082.200
As. 33:494
stand, ridges, fenestrated
D 17:9
32.30
fig. 16
pl. 1
C.35-.0–
pl. 68d
As. 33:119
sherd, painted dec., scarlet
E 17:20
32.30
–
pl. 1
–
pl.136r
As. 33:139
handle, plaited
E 17:20
32.30
–
–
plaited handle
–
As. 33:424 a
jar with spout
E 17:20
32.30
fig. 17
pl. 1
C.525.362
pl. 67f
As. 33:425
jar with spout
E 17:20
32.30
fig. 17
–
C.526.262c
pl. 67h
As. 33:116 a
stand
D 17:6
32.45
fig. 16?
–
C.357.010b
C.357.010b
As. 33:344
sherd, painted dec.
D 17:9
32.45
–
–
–
–
As. 33:345
sherd, painted dec.
D 17:9
32.45
–
–
–
–
As. 33:117
sherd, red
E 17:20
32.45
–
–
–
–
Square Temple II (floor at 32.50 m)
As. 33:149
stand, notched ridges
D 17:5
32.50
fig. 17
pl. 1
C.357.010b
pl. 68b
As. 33:81
sherd, painted dec., red
D 17:7
32.50
–
–
–
–
As. 33:82
sherd, painted dec., dark red
D 17:7
32.50
–
–
–
–
As. 33:493
stand, incised ridges
D 17:8
32.50
–
pl. 1
stand frag.
–
As. 33:495
stand, fenestrated
D 17:8
32.50
fig. 16
pl. 1
C.39-.0–
pl. 68e
As. 33:499
colander
D 17:8
32.50
fig. 16
–
–
–
sherd, painted dec., red stripe E 17:20
32.50
–
–
–
–
31.85
–
–
C.082.210
C.082.210
As. 33:71
E 16:40 (pit)
As. 33:543
bowl
E 16:40
2007]
THE SQUARE TEMPLE AT TELL ASMAR
617
Table 4 (continued).
Oriental Institute Records
Oriental Institute Publications
Excavation
No.
Description
Locus/
Context
Elev.
(m)
Frankfort
1935a
Frankfort
1936
Delougaz and
Lloyd 1942
Delougaz
1952
As. 33:541
jar
E 16:40
31.85
–
–
–
–
As. 33:542
jar, pointed base
E 16:40
31.85
–
–
B.756.720
B.756.720
As. 33:549
sherd
E 16:40
31.85
–
–
A.645.720
A.645.720
As. 33.531
sherd, painted dec.
D 17:8
–
–
pl. 1
painted sherd
pl. 136s
As. 34.24
sherd, painted
D 17:8
31.50
–
pl. 1
–
pl. 136q
vessel
–
–
fig. 17
pl. 1
–
pl. 70j
Wall
Unknown loci
As. NR:6
a
Catalogued in Delougaz (1952), with an asterisk after the locus, indicating “that there is some doubt as to the exact shape of the
specimen concerned” (Delougaz 1952, 153)
The Diyala excavators considered the decorative elements on stands to be diagnostic of Early Dynastic subdivisions: stands with incised or excised decoration and
features such as rims and lugs were dated to ED I, and
plain stands were dated to ED II.111 The chronological significance of the decorative elements on stands,
however, is ambiguous in the Diyala, for such a distinction is discernible only by comparing Archaic Shrine
IV stands to Square Temple I stands.112 A full range of
excised, incised, and plain stands is already present
in ED I levels in the Inanna Temple at Nippur.113 The
decorative elements on stands therefore should not be
considered diagnostic of Early Dynastic subdivisions.
To judge by catalogued examples, the archaeological basis for the perceived continuation into ED II Diyala of characteristically ED I pottery—the solid-footed
goblet, four-lugged neckless jar, red-painted jar with
imitation rope handles, and Scarlet Ware—is largely
contingent on the appearance of these forms in the
Square Temple. Perhaps the Square Temple pottery
had to be ED II because it was from the same context
that necessitated the ED II subdivision. It therefore
would have seemed a foregone conclusion to the
Diyala excavators that in the ED II Square Temple,
111
Delougaz 1952, 55–6, 81, 141.
Decorated stands were found in Sin VI (C.357.010) and
Archaic Shrine IV (As. 33:675, C.3-.0–, C.3–.063, C.35-.0–b,
C.357.073). Plain stands were found in Square Temple I (C.35.0--a, C.355.010, C.356.010, C.357.010b, C.39-.0-) and in the
Shara Temple (C.357.010a, C.358.010). One of the Shara
Temple stands (C.358.010) has a vertical band rim but is plain,
and a plain stand (C.355.010) from Square Temple I has a
beveled rim, suggesting that the addition of rims cannot be
112
with its geometric-style sculpture and Fara-style cylinder seals, characteristically ED I pottery forms simply
continued or survived. Regardless of the small quantity recorded, the date of the Square Temple pottery
should not be determined by working back from the
assumed ED II date of the geometric-style sculpture
and Fara-style cylinder seals with which the Square
Temple pottery was found. It does not seem reasonable to continue to maintain an ED II date for Square
Temple pottery when it comprises an ED I assemblage.
The same thinking can be applied to the date of the
Square Temple itself. Therefore, Square Temple I
and the remains below it are dated here to ED I on
the basis of pottery.
Both stands and painted sherds can be assigned to
Square Temple II on the basis of findspot elevations.
Only 20 cm separate Square Temple I and II. It may
be that given the representative nature of the Abu
Temple findspot elevations, a true distinction between
finds in Square Temple I and II cannot be made. The
few stands and painted sherds correlated with Square
Temple II would not alter the ED I date of Square
Temple I. So little pottery can be attributed to Square
Temple II that it seems preferable to leave open the
considered ED II even according to the dating established by
the Diyala excavators. Stands with lugs appear only in Archaic
Shrine IV (C.357.073, C.3--.063).
113
7NP126/IT XI (excised); 7NP67/IT IXB (incised);
7NP90/IT X or IXB (plain). Plain stands (7NP38, 7N514,
7N515, 7N517) continue into IT VIII, which also yielded a
stand (7NP49) decorated with clay strips applied in a wavy
pattern.
618
JEAN M. EVANS
[AJA 111
Fig. 10. Pottery from the Square Temple and below at Tell Asmar: a, As. 34:24; b, As. 33:119; c, As. 33:531; d, As. 33:238; e, As.
33:406; f, As. NR:6; g, As. 33:424; h, As. 33:425; i, As. 33:499; j, As. 33:674; k, As. 33:362; l, As. 33:116 (?); m, As. 33:149; n, As. 33:495;
o, As. 33:494; p, As. 33:493 (stands, scale 1:10; all other pottery, scale 1:5) (after Frankfort 1935a, fig. 16; 1936, pl. 1, nos. 6, 7, 18;
Delougaz 1952, pls. 67 [f, h], 68 [b, d–g], 70 [ j], 136 [q–s]; courtesy the Oriental Institute of the University of Chicago).
2007]
THE SQUARE TEMPLE AT TELL ASMAR
dating of both Square Temple II and Square Temple
III, to which no pottery can be assigned.
Square Temple Cylinder Seals
In the Diyala publications, the Square Temple
seals were catalogued, with few exceptions, as Square
Temple I. Some cylinder seals instead have findspot
elevations corresponding to below Square Temple I,
and others have findspot elevations corresponding
to Square Temple II (table 5). All extant illustrations
of cylinder seals from the Square Temple appear in
figure 11.
A cylinder seal (As. 34:44) in the ED I brocade style
known principally from the Diyala was found in the
exterior west wall of the Square Temple, which may
have incorporated the predecessor wall.114 Other seals
from either the walls or foundations of the Square
Temple include one carved in the glazed steatite style
(As. 34:30) and another (As. 34:39) with a banqueting
scene, an inverted eagle with outstretched wings, and
star and crescent filler motifs.115
Five cylinder seals (As. 33:248, As. 33:254, As. 33:270,
As. 33:454, As. 33:456) between the altar and the long
west wall of D 17:8 are from a hoard of objects including stamp seals and beads.116 According to Lloyd’s field
notebook, this hoard lay some 10 cm above a pavement
at 31.74 m and underneath multiple replasterings of
the D 17:8 altar, which, according to the findspot elevation, would refer to the earlier predecessor altar that
114
The earliest examples of the brocade style, bearing
compositions of balanced, allover patterning related to the
glazed steatite glyptic style, appear in Sin V (Frankfort 1955,
nos. 220–22, 224) and Archaic Shrine III (Frankfort 1955, no.
447). The full brocade style then appears in Sin VI (Frankfort
1955, no. 229) and Archaic Shrine IV (Frankfort 1955, no.
450); see also Wilson 1985, 69, 86.
115
The seal in the glazed steatite style is carved from shell
and has parallels in Sin IV and in the Earliest Shrine; see
Frankfort 1955, nos. 97, 441; Pittman 1994, 108–13, 222–23.
Banqueting scenes with antithetic seated figures first appear
in SIS 4-5 at Ur (Legrain 1936, nos. 373, 377, 381).
116
Lloyd (Delougaz and Lloyd 1942, 181–82) suggested
these objects fell from the altar and were covered by a Square
Temple II replastering, but Frankfort (1939, 4–5) described
the hoard underneath Square Temple I plaster. According
to Lloyd’s field notebook, two groups of objects were found
between the altar and long west wall of D 17:8—one including stamp seals and beads above the 31.74 m pavement and
the other including a bronze mirror and beads at a 31.16 m
pavement.
117
Moortgat-Correns 1959, 346–49; accepted by Porada
et al. 1968, 304 n. 4; Hansen 1971, 50; Behm-Blancke 1979,
56; Amiet 1980, 50–1; Porada 1980, 7; Wilson 1986, 65; Martin 1988, 133 n. 16; Pittman 1994, 63; Matthews 1997, 85 n.
125. Although defined principally by examples in the Shara
619
survived to be incorporated into the Square Temple
altar. Two of the cylinder seals (As. 33:248, As. 33:254)
belong to the so-called Jamdat Nasr–style cylinder seals
that appear in the Diyala throughout the Early Dynastic period. Another cylinder seal (As. 33:454) belongs
to the Fara elegant style; the empty space, lack of symmetry, and use of only three figures are unusual. Another seal (As. 33:270) is carved in a soft style with a
ram, bull, and antelope in file and a recumbent animal
in the upper register. Moortgat-Correns first suggested
an ED I date for this glyptic style, which is concentrated in the Earlier Building second occupation of the
Shara Temple at Tell Agrab.117 As noted above, most
excavated levels of the Shara Temple were originally
dated to ED II principally because of geometric-style
sculpture like that in the Asmar hoard.118 However, an
ED I wing-lugged jar containing a jewelry hoard was
buried from the Earlier Building second occupation
and supports an ED I date for this Shara Temple level
in which the soft glyptic style is concentrated.119 Precursors to the soft style were identified by Wilson in Sin
Temple III, which spans the transition from Jamdat
Nasr to ED I.120 At Khafajah, the softly carved animals
with wishbone-shaped horns on a cylinder seal from
the ED I Houses 7 parallel this style.121
A hoard including six cylinder seals (As. 33:663, As.
33:666, As. 33:677, As. 33:697, As. 33:698, As. 33:699)
was found buried in the upper part of the D 17:8 altar.122 One (As. 33:698) bears an eye motif common
Temple, the softly carved style is known from Fara, Kish, Uruk,
al-Hiba, Nippur, and Ur (Amiet 1980, pls. 53–5, 57; Hansen
1987, pl. 12[2] [al-Hiba]). Karg (1984) dated this glyptic style
to ED II, but for objections, see Martin 1988, 133–34 n. 29; Porada 1991, 172. The findspot elevations of Shara Temple pottery, cylinder seals, and sculpture are published in Frankfort
1943, 1955; Delougaz 1952. These objects are not catalogued
by building periods or occupation levels in any of the Diyala
publications; see also Delougaz and Lloyd 1942. The context
of the Shara Temple finds can be derived by correlating the
object findspot elevations with the elevations of occupation
levels published in Delougaz and Lloyd 1942; see Evans 2005,
119–65.
118
Delougaz and Lloyd 1942, 260.
119
Delougaz 1952 (D.526.373, published as C.526.373a in
Delougaz and Lloyd 1942). For the findspot, see Delougaz
and Lloyd 1942, 253–54, pl. 27B; Delougaz 1952, 57. For winglugged jars, see Delougaz 1952, 57–8; Wilson 1986, fig. 11[8].
120
Wilson 1986, 65.
121
Frankfort 1955, no. 290; see also Amiet 1980, 50.
122
Frankfort 1935a, 24; Delougaz and Lloyd 1942, 183. The
D 17:8 altar incorporated part of the predecessor altar (Delougaz and Lloyd 1942, fig. 131), but the “upper part” is seemingly
that of the Square Temple I portion; the findspot elevation
would, however, suggest the predecessor altar. The altar was
replastered—not rebuilt—for Square Temple II and III.
620
JEAN M. EVANS
[AJA 111
Table 5. Findspots of Square Temple Glyptic.
Oriental Institute Records
Excavation
No.
Description
Oriental Institute Publications
Locus/
Context
Elev.
(m)
Delougaz and
Lloyd 1942
Frankfort
1955
Square Temple walls and foundations
As. 33:532
OI: no description a
D 17:8, wall
–
–
–
As. 34:48
geometric
foundations,
wall
–
seal
452
As. 34:30
glazed steatite style
D 17:8
foundations
–
seal
459
As. 34:25
OI: geometric
D 17:8, wall
31.50
seal
2 rows of oblique lines
slanting in diff. directions
As. 34:39
banquet
D 17:8
foundations
32.00
seal
465
As. 34:44
brocade style
wall
32.00
seal
466
Below the Square Temple (below 32.30 m)
As. 33:418
Fara style
D 17:9
31.85
seal
471
As. 33:269
OI: worn design
D 17:8
32.00
–
–
As. 33:271
OI: uncarved
D 17:8
32.00
bead
–
As. 33:380
Fara style
D 17:8
32.00
seal
464
Underneath replastering related to earlier D 17:8 “predecessor” altar
As. 33:454
Fara style
D 17:8
31.74
seal
463
As. 33:456
OI: no description
D 17:8
31.74
seal
linear design
As. 33:248
animal file
D 17:8
32.00
seal
460
As. 33:254
animal and lines
D 17:8
32.00
seal
461
As. 33:270
animal file
D 17:8
32.00
seal
462
Buried in the D 17:8 altar
As. 33:663
Fara style
D 17:8
32.00
seal
457
As. 33:666
combat
D 17:8
32.00
seal
456
As. 33:677
lions tête-bêche
D 17:8
32.00
seal
458
As. 33:697
OI: unfinished
D 17:8
32.00
unfinished
unfinished design,
traces of horizontal lines
As. 33:698
geometric
D 17:8
32.00
seal
454
As. 33:699
brocade style
D 17:8
32.00
seal
455
Square Temple II (floor at 32.50 m)
As. 33:205
brocade style
D 17:6
32.50
seal
468
As. 33:206
OI: worn away
D 17:6
32.50
seal
no record of worn design
As. 33:202
OI: worn down
D 17:8
32.50
–
–
As. 33:226
OI: no description
D 17:8
32.50
seal
–
As. 33:701
Fara style
D 17:9
32.50
seal
470
2007]
THE SQUARE TEMPLE AT TELL ASMAR
621
Table 5 (continued).
Oriental Institute Records
Oriental Institute Publications
Excavation
No.
Description
Locus/
Context
Elev.
(m)
Delougaz and
Lloyd 1942
Frankfort
1955
As. 33:310
OI: sealing
D 17:7
32.52
–
–
E 17:20
33.00
seal
473
Square Temple III (floor at 33.00 m)
As. 33:151
a
OI: blank
OI = Oriental Institute records
to the Jamdat Nasr–style cylinder seals that appear in
the Diyala throughout the Early Dynastic period, and
another (As. 33:699) is brocade style. Seal As. 33:677
depicts a chain of lions, each closing its jaw over the
leg or tail of an adjacent lion. The delicate carving
and the viewpoint of the lion heads recall the Fara
elegant style, but the broad, flat bodies of the lions
and the horizontal composition of repetitive design
elements differ from it. The filler motif of a crescent
containing dots has a parallel with a seal from an ED
I context at the site of Fara, and the splayed rear legs
and contorted poses of the lions are characteristic of
sealings from the ED I level IXB of the Inanna Temple
at Nippur.123 Seal As. 33:666 is carved with a symmetrical elegant-style composition, but the seal is more
coarsely carved and in higher relief than the elegant
style, and the squat shape of the seal along with the
drilled star motifs and drooping mouths of the lions instead recall seals from earlier Protoliterate contexts.124
The open composition of Seal As. 33:663 resembles
Martin’s earliest subgroup of Fara elegant-style seals.125
The sharply bent arms of the bull-man and the male
figure are elongated in order to accommodate the
open spacing of the composition, similar to one of
the Fara-style sealings from ED I Nippur, which also
combines male and bull-man figures.126
Additional seals can be assigned to below Square
Temple I. One is a fragmentary cylinder seal (As.
33:418) in the Fara elegant style with unusual iconography, and another seal (As. 33:380) is carved in the Fara
elegant style with a dense composition of six figures
forming two symmetrical units of combat. An increase
in the number of figures and the formation of two units
of three figures each is considered late by Martin because these trends approach the crossed style.127
123
Hansen 1971, 52, no. 7, pl. 19h; Martin 1988, no. 206.
Frankfort 1955, 28; Buchanan 1956, 72.
125
Martin 1988, 73.
126
Hansen 1971, 52, no. 8, pl. 20a; Amiet 1980, 204–5, no.
1707.
124
No cylinder seals can be assigned to Square Temple
I on the basis of findspot elevations. A poorly preserved cylinder seal in the brocade style (As. 33:205)
and a Fara crossed-style cylinder seal (As. 33:701) have
findspot elevations corresponding to Square Temple
II. The 32.50 m findspot elevation of the Fara crossedstyle seal is accepted here as a specific reference to
Square Temple II because in his field notebooks, Lloyd
describes a 32.45 m floor in D 17:9 in addition to the
32.30 m Square Temple I floor. Martin compared the
heavily hatched animal bodies to a Fara crossed-style
seal from late in the excavated sequence of the Shara
Temple, suggesting that both may represent a local
Diyala style.128
Given that the remains below Square Temple I were
poorly recorded, it is difficult to consider any one
group of cylinder seals from below Square Temple
I stratigraphically earlier than another. Thus, for example, while the softly carved style of glyptic is stratigraphically earlier than the Fara elegant style in the
Shara Temple sequence, such a distinction cannot be
made in the Square Temple. The Fara crossed-style
seal in Square Temple II is, however, from a later context than the elegant-style seals below Square Temple
I. The findspot elevations for Square Temple cylinder
seals therefore support a chronological distinction between the elegant style and the crossed style. Such a
distinction has already been indicated more generally
by the two broad findspots for the elegant style and
crossed style at the site of Fara.
The Location of the Tell Asmar Sculpture Hoard
The Asmar hoard was excavated between the altar and the long north wall of D 17:9. A boxed area
(see fig. 7) representing the hoard and roughly cor-
127
Martin 1988, 73.
Martin 1988, 77. For the Shara Temple seal, see Frankfort
(1955, no. 875), with a findspot elevation corresponding to
the Main Level second occupation.
128
622
JEAN M. EVANS
[AJA 111
Fig. 11. Modern impressions of cylinder seals from the Square Temple and below at Tell Asmar: a, As. 33:151; b, As. 33:205; c,
As. 33:248; d, As. 33:254; e, As. 33:270; f, As. 33:380; g, As. 33:418; h, As. 33:454; i, As. 33:663; j, As. 33:666; k, As. 33:699; l, As.
33:677; m, As. 33:698; n, As. 33:701; o, As. 34:30; p, As. 34:39; q, As. 34:44; r, As. 34:48 (c, e, f, h–j, l–n, p, scale 1:1; a, b, d, g, k, q,
r, scale 3:5; o, scale 1:3) (after Frankfort 1955, nos. 452, 454–66, 468, 470, 471, 473; courtesy the Oriental Institute of the University of Chicago).
2007]
THE SQUARE TEMPLE AT TELL ASMAR
responding to its known dimensions appears on the
Abu Temple section, connected to the Square Temple
I floor only by dashed lines. Because the Abu Temple
section is through D 17:8/D 17:7/E 17:20 and not
D 17:9, where the hoard was found, the presence of
the hoard on the Abu Temple section confirms the
schematic nature of the section; nevertheless, it also
suggests that the hoard is not directly connected to
the Square Temple I floor. The stratigraphic location
of the cut for the hole containing the Asmar hoard is
not made explicit in the Diyala publications. From the
evidence discussed below, however, it is clear that the
hole containing the Asmar hoard was below not only
the Square Temple I floor but also additional floors
in D 17:9. The burial of the Asmar hoard therefore
should not be associated with the Square Temple
proper but with one of the poorly recorded levels
below Square Temple I. The disassociation of the
Asmar hoard from the Square Temple is noteworthy
because the hoard had confirmed the significance of
the Square Temple plan and necessitated an ED II
subdivision.
Like the majority of sculpture catalogued as Square
Temple I, the findspot elevations for the sculpture
in the hoard are all 31.85 m, some 45 cm below the
Square Temple I floor (32.30 m). Calculations derived
from the dimensions for the hole containing the hoard
confirm its 31.85 m findspot elevation, which must
refer to the top of the hole containing the hoard. According to Lloyd, the hoard was at a “total depth” of
1.25 m below the Square Temple I floor in an 85 x 50
cm oblong hole; after the statues had been placed in
the hole, the hole was packed with “hardened tablet
clay” rolled into balls “for 30 cm beneath the actual
pavement.”129 According to Frankfort, the hole for the
hoard was “about” 60 cm deep.130 If the bottom of the
hole was 1.25 m below Square Temple I, and the hole
was 60 cm deep—including 30 cm of clay packing—
then the beginning of the hole containing the hoard
would have been 65 cm below the Square Temple I
floor at 31.65 m. However, since the 12 statues in the
hoard were stacked in three or four layers of three or
four statues side-by-side, more space likely would have
been needed to accommodate the statues, and thus
129
Delougaz and Lloyd 1942, 189.
Frankfort 1939, 3.
131
Delougaz and Lloyd 1942, figs. 148, 150, 151. The average height of the statues in the Asmar hoard is 42 cm. One
of the statues from the Asmar hoard (Frankfort 1939, no. 9)
now in the Metropolitan Museum of Art (Fletcher Fund, 1940
40.156) is 29 cm tall and has a maximum width of 10 cm from
front to back. Thus, at least 30–40 cm would be needed to accommodate statues stacked in three to four rows.
130
623
the findspot elevation derived from the dimensions
for the hole approaches the recorded findspot elevation for the hoard of 31.85 m.131
Photographs of the hoard in situ also indicate that
the hole containing the hoard is some distance below
the Square Temple I floor.132 The mudbrick structures of a D 17:9 floor are visible in the foreground
of figure 12, which shows the hoard in situ.133 In the
Diyala publications, the mudbrick structures in D 17:9
are assigned to the 32.30 m Square Temple I floor.134
In Lloyd’s notebook, the mudbrick structures are assigned to the 32.45 m floor, which in D 17:9 is a separate floor level corresponding to Square Temple II.
It is unclear which is correct. Near the altar in figure
12, the floor with the mudbrick structures has already
been removed in order to investigate the construction
of the altar.135 The line of the Square Temple floor preserved along the front and side of the altar and along
the west wall of the cella (cf. fig. 13) is particularly
prominent because at this level, the floor was plastered
with greenish “paint” of a consistency resembling gypsum mixed with mud.136 The line of the Square Temple
floor is some distance from the top of the hole, which
still contains all 12 statues in the hoard. In another
photograph (fig. 14), a workman squatting near the
hole, which contains the bottom three statues of the
hoard, gives a good indication of scale: the plaster remains of the Square Temple floor are visible behind
his head. This same mass of accumulated plaster is also
visible in figure 8, where it is higher than the 32.16 m
floor with the smaller altar.
In the entries under D 17:9 in the field notebook,
three floors are identified in the Square Temple
and below: 32.16 m (with the smaller altar), 32.30 m
(Square Temple I), and 32.45 m (Square Temple II).
An additional floor was potentially identified in D 17:9
on 27 January 1934, the day the Asmar hoard was discovered. According to Lloyd’s field notebook:
Began cleaning between shrine and pedestals down to
3230 pavement and a little beneath mid-morning struck
hoard of statues between shrine and north wall. These
were buried in a hole beneath a pavement (?) (31.81)
the spaces between filled with spherical lumps of tablet
clay and covered in with the same material.137
132
Evans 2005, 89–92. See also Marchetti (2006, 27), who
has made similar observations.
133
Cf. Delougaz and Lloyd (1942, fig. 148) with Delougaz
and Lloyd 1942, fig. 151.
134
Delougaz and Lloyd 1942, 191.
135
Delougaz and Lloyd 1942, 189.
136
Delougaz and Lloyd 1942, 181, 187.
137
Lloyd 1933–1934.
624
JEAN M. EVANS
[AJA 111
Fig. 12. The sculpture hoard from the Abu Temple at Tell Asmar during excavation (modified from Delougaz and
Lloyd 1942, fig. 148; courtesy the Oriental Institute of the University of Chicago).
The beginning of the hole that contained the Asmar
hoard therefore was potentially also beneath a 31.81
m floor, which represents another floor in addition
to the 32.16 m floor that would have separated the
Asmar hoard from Square Temple I.
As noted above, the stratigraphic location of the cut
for the hole containing the Asmar hoard is unknown.
It is unlikely that the hole for the hoard was dug from
the Square Temple I floor because the lumps of tablet clay between the hole containing the hoard and
the 31.81 m floor provide a good indication of when
the Asmar hoard was buried. In contrast to the Asmar
hoard, a sculpture hoard in Nintu Temple V at Khafajah was buried in a hole filled with earth. The earth
fill settled and caused a noticeable depression in the
floor that went uncorrected, indicating to the excavators that the hoard had been buried at the end of
Nintu Temple V when the floor was no longer in use.138
138
Delougaz and Lloyd 1942, 94–5, fig. 85.
Martin et al. 1985, 20 (Grave 1, Room 39, level I, area E).
140
Postgate and Moorey 1976, 146; Martin et al. 1985, 20.
In another example, a bitumen-lined depression in the Main
139
More similar to the clay packing of the Asmar hoard
is the clay filling used at Abu Salabikh in association
with an ED IIIa grave.139 The grave had been filled with
earth, which subsequently caused a depression in the
floor that overlay the grave. Because the floor was still
in use, the depression had to be leveled several times
with clay filling.140 The clay packing above the Asmar
hoard was likely intended to prevent a depression
from forming in a floor that was or would be in use,
since clay packing would not settle like earth fill. The
31.85 m findspot elevation for the Asmar hoard—corresponding to the 31.81 m floor in D 17:9—therefore
indicates the top of the clay packing that extended “for
30 cm beneath the actual pavement,”141 and the Asmar
hoard was likely buried at the beginning or during the
use of the 31.81 m floor rather than at its end.
The Diyala excavators left open the question of
when the Asmar hoard was buried in relation to an
Level cella floor of the Shara Temple was later filled in with a
layer of baked brick rather than dirt in order to bring the depression up to floor level (Delougaz and Lloyd 1942, 234).
141
Delougaz and Lloyd 1942, 189.
2007]
THE SQUARE TEMPLE AT TELL ASMAR
occupation level, but other arguments for dating the
burial of the hoard have been formulated by drawing
stylistic comparisons with the remaining sculpture
catalogued as Square Temple I.142 Since all sculpture,
with two exceptions, catalogued as Square Temple I
was found at 31.85 m, any stylistic affinities would associate the hoard with sculpture from below Square
Temple I. Other scholars have suggested on the basis
of the 31.85 m findspot elevation that the statues in the
Asmar hoard originated in a level below Square Temple I.143 For example, Behm-Blancke associated the
hoard with the predecessor and then argued on the
basis of architectural criteria that the predecessor encompassed the ED I–II transition.144 Braun-Holzinger
also suggested that the Asmar hoard might have
originated in the predecessor but considered such
a distinction inconsequential because, in contrast to
Behm-Blancke, she related the predecessor plan to
Square Temple I. This architectural continuity signaled to her that both were ED II.145 These opposing
conclusions demonstrate the subjective nature of such
criteria. The Asmar hoard should be dated instead by
pottery, and the pottery correlated with Square Temple I provides a terminus ante quem of the end of ED
I for the date of the burial of the Asmar hoard.146
postscripts
Fara-Style Glyptic, Geometric-Style Sculpture, and ED II
Khafajah
Having reviewed the Square Temple, I turn now to
Khafajah where, as discussed above, Houses 6–4 and
the corresponding Sin VIII represent the only Diyala
contexts dated solely to ED II on the basis of ceramics;
Sin IX and Temple Oval I continue into ED III and
thus represent mixed assemblages. The principal pottery retrieved from Sin VIII was a ceramic “cult wagon”
142
Frankfort 1939, 4; Delougaz and Lloyd 1942, 191; Strommenger 1960, 10; Braun-Holzinger 1977, 34; Behm-Blancke
1979, 57.
143
Behm-Blancke 1979, 57; Tunca 1984, 192; Marchetti
2006, 27.
144
Behm-Blancke (1979, 57, 62, 64) dated the predecessor essentially to the end of ED I but associated the Asmar
hoard with ED II (“Mesilim-Zeit”), such that his dating of the
predecessor did not affect his dating of the hoard. See BehmBlancke (1979, table 3), with the predecessor and Square Temple I at the beginning of ED II.
145
Braun-Holzinger 1977, 29, 34.
146
Hrouda (1971, 112) argued that some figures in the Asmar hoard hold ED I solid-footed goblets, which has been accepted by Behm-Blancke (1979, 57) and Porada et al. (1992,
105). Braun-Holzinger (1977, 44) argued that the vessel forms
were not distinct enough to be identified as solid-footed
goblets.
147
Delougaz 1952, 85–6, pls. 82, 83 (C.99). Other ceramics
625
Fig. 13. Cella D 17:9 of the Square Temple at Tell Asmar
with the Square Temple floor intact (Delougaz and Lloyd
1942, fig. 146; courtesy the Oriental Institute of the University of Chicago).
in which two upright-handled jars and a short fruitstand were set.147 In a photograph of the crushed cult
wagon in situ, the stem of the fruitstand is visible, but
nothing remains of the upper part; a fragment of an
additional incised base with the beginning of a stem
is visible in the photograph and does not belong to
the cult wagon.148 The additional fragment was not
catalogued, but it seems to belong to a fruitstand. If
Sin VIII were not correlated by absolute levels with
the ED II Houses 6–4, fruitstands would date Sin VIII
to ED II–III.
The Fara elegant style is represented at ED II Khafajah by one cylinder seal from Houses 4; given the
evidence at Nippur, Fara, and Tell Asmar, this seal
should not represent the earliest appearance of the
Fara style at Khafajah.149 Indeed, one crossed-style cylinder seal from Sin VIII indicates that the Fara style is
already present at Khafajah at the onset of ED II.150 One
catalogued as Sin VIII are Kh. IV 391, a basketlike clay vessel,
and E.205.310, a large storage vessel (Delougaz and Lloyd
1942, 143–45); see also Delougaz 1952.
148
Delougaz and Lloyd 1942, 58, fig. 48. According to Delougaz (1952, 86), the fruitstand on the cult wagon was later
restored using the dish from an ED III fruitstand.
149
Frankfort 1955, no. 305. One fragmentary elegant-style
cylinder seal (Frankfort 1955, no. 253) was catalogued as Temple Oval I, but it is from a denuded area ( J 45:4, with the outer
oval wall of Temple Oval I poorly preserved in J 45/46; see Delougaz 1940, pls. 3, 4). An elegant-style seal (Frankfort 1955,
no. 321) catalogued under “Houses 3 or 2” is out of context,
retrieved from square J 43 southwest of the Houses north of
the Temple Oval (Delougaz et al. 1967, pl. 13). Frankfort 1955,
nos. 271 (Temple Oval II) and 282 (Nintu VII) were classified
by Amiet (1980, nos. 873, 883) as his série archaique of the Fara
style, which corresponds to the elegant style.
150
Frankfort 1955, no. 245.
626
JEAN M. EVANS
Fig. 14. The sculpture hoard from the Abu Temple at Tell
Asmar during excavation (modified from Delougaz and
Lloyd 1942, fig. 150; courtesy the Oriental Institute of the
University of Chicago).
crossed-style seal from Sin IX and three from Temple
Oval I are from ED II–III contexts.151 In the Diyala, the
crossed style is also found in ED III contexts.152
Following the chronological distinction between
the elegant style and the crossed style at Fara and in
the Square Temple, the elegant style should precede
the appearance of the crossed style in Sin VIII at Khafajah. However, no examples of Fara-style glyptic were
retrieved from earlier levels. The lack of Fara-style
glyptic in ED I Khafajah levels is likely insignificant,
151
Frankfort 1955, nos. 246, 254, 255, 258. See also no. 247,
a crossed-style seal from Q 42 catalogued as “Sin IX?”
152
Frankfort 1955, nos. 330 (Houses 2), 498 (structure
north of the Earlier Northern Palace). For the findspot of no.
498, see Delougaz et al. 1967, 185, 242.
153
Frankfort 1955, nos. 288–91.
154
Delougaz and Lloyd 1942, 143; Delougaz 1952 (Kh. V
160a, b); Frankfort 1955, nos. 227–37. No other pottery is catalogued as Sin VII. For Sin VI, see Delougaz 1952 (C.802.200,
shallow oval dish; C.545.320, medium-sized jar with ring base;
C.357.010, stand).
155
Frankfort 1939, 25–6, no. 60.
156
Frankfort 1939, 10. Twenty-one examples were catalogued as Sin VIII in Frankfort (1939, 1943), but no. 91 is
[AJA 111
given that few seals in general were retrieved from
these levels. Only four cylinder seals were found in
Houses 9–7, the latest ED I Houses: one seal with a
geometric motif, two brocade-style seals, and a seal
related to the soft glyptic style concentrated in the
Shara Temple and known from one example below
Square Temple I.153 The scant finds in Houses 9–7 are
paralleled by the few finds catalogued under the corresponding Sin VI and VII, which were disturbed by
later foundations. In these levels, brocade-style seals
predominate, and of the few catalogued ceramics, burnished ware, represented by sherds, is characteristic
of ED I.154 The Fara elegant-style seals below Square
Temple I are thus the earliest stratified examples of
Fara-style glyptic in the Diyala.
The Square Temple was correlated with the Khafajah sequence on the basis of one male head from
Sin VIII carved in the geometric style.155 Although
Frankfort characterized the Sin VIII sculpture as “but
few” and “of an inconclusive character,” a great deal
of sculpture (20 illustrated examples) was retrieved.156
The majority of Sin VIII sculpture fragments are heads
from female figures, which played a small role in
defining Early Dynastic sculpture styles because the
Diyala excavators considered their stylistic qualities
to be less distinct than sculpture of male figures.157
Leaving aside the validity of distinguishing sculpture
styles among female figures, it suffices to note that few
male figures—of which only two male heads are illustrated—were retrieved from Sin VIII.158 The second
male head bears elements of the realistic style in the
accurately proportioned skull and deep creases from
the nose to the lips accentuating full cheeks.159 Admittedly fragmentary, the evidence for realistic-style sculpture in Sin VIII is just as compelling as the evidence
for geometric-style sculpture in Sin VIII, suggesting
that at Khafajah, both geometric- and realistic-style
sculpture were present by the onset of ED II.
The realistic-style sculpture associated with the
Single-Shrine Temple at Tell Asmar formed the basis
for Frankfort’s outline of a geometric sculpture style
from a locus outside the temple proper.
157
Frankfort 1935a, 73; 1935b, 121; 1939, 25–6 n. 6, 31.
Frankfort (1939, 31) did, however, describe realistic-style female figures; see also no. 120 (Sin VIII) with soft, full cheeks
and modeled lips.
158
Frankfort 1939, nos. 83, 87, 114, 115, 120, 124, 127, 130,
133, 139, 140, 145, 148, 151a, 151b, and 1943, no. 250 are
female; Frankfort 1939, nos. 58, 60 are male; no. 86 is of indeterminate gender; no. 88 is a couple. Of the unillustrated
sculpture catalogued as Sin VIII, five are fragmentary female
figures, four are fragmentary male figures, and three fragments preserve the feet and base (Frankfort 1943, 41–2).
159
Frankfort 1939, no. 58.
2007]
THE SQUARE TEMPLE AT TELL ASMAR
becoming a realistic sculpture style.160 Only small fragments of sculpture, however, were recovered from the
Single-Shrine Temple proper, and the association of
realistic style with the Single-Shrine Temple relied
instead on sculpture principally from E 17:11, which
defines an area east of and below the Single-Shrine
Temple.161 As with the transition from the Archaic
Shrine to the Square Temple, the transition from the
Square Temple to the Single-Shrine Temple was poorly recorded and poorly understood, and conflicting
accounts are given.162 As with the objects catalogued
as Square Temple I, finds from below Single-Shrine
Temple I were combined with finds in Single-Shrine
Temple I to form a single category.163 Although realistic-style sculpture cannot be associated in the Abu
Temple sequence exclusively with the Single-Shrine
Temple, it is accurate to say that only realistic-style
sculpture was found in the Abu Temple from Square
Temple I onward.164 This perceptible shift in sculpture
styles is unique to the Abu Temple.
In Sin IX, Frankfort distinguished between sculpture from Room Q 42:7 belonging to the first Sin IX
floor and sculpture in Court Q 42:3 belonging to a
later Sin IX floor.165 A greater number of statues carved
in the realistic style were retrieved from the later Q
42:3 context than from the earlier Q 42:7 context,
which signaled to the Diyala excavators a shift in sculpture styles in the Sin Temple. The sculpture from the
earlier Q 42:7 context consists primarily of fragments
from female figures—20 of 31 published examples. In
contrast, primarily male figures—12 of 16 published
examples—were retrieved from the later Q 42:3 con-
160
Frankfort 1935a, 73, 83–4; 1939, 16–17, 29–30.
For the association of realistic-style sculpture with the
Single-Shrine Temple, see Frankfort 1939, 16–17, 29–30,
nos. 62 (35 m/D 17:12), 63 (32.90 m/Square Temple II), 66
(32.30 m/Square Temple I), and 67–8 (33.75 m/E 17:11). A
wall, which does not appear on the plan, abutted the exterior
northeast corner of Single-Shrine Temple I, and E 17:11 was
used to designate the area to the east that it enclosed. The E
17:11 sculpture is from a gypsum-coated pavement more than
half a meter below, which Lloyd suggested corresponded to
one of the two intermediate pavements below Single-Shrine
Temple I (Delougaz and Lloyd 1942, 199). Frankfort (1934,
45; 1935a, 7) instead associated the sculpture with a court or
open space north of the Single-Shrine Temple.
162
Delougaz and Lloyd 1942, 156, 192.
163
Delougaz and Lloyd 1942, 212–13.
164
One geometric-style sculpture (Frankfort 1939, no. 14)
from E 17:11 joins with feet found at 31.85 m.
165
Frankfort 1943, 6; see also Delougaz and Lloyd 1942,
65–6. A comparison of the published findspot elevations supports that the Q 42:7 sculpture and the majority of the Q 42:3
sculpture are from different occupation levels, with the range
in findspot elevations for Q 42:7 sculpture perhaps reflecting
that some was found on “low mud-brick benches” (Delougaz
161
627
text.166 Given that primarily female figures were also
found in Sin VIII, the increase in male figures from the
earlier Q 42:7 context to the later Q 42:3 context could
reflect a shift to principally male rather than female
donors in Sin Temple dedicatory practices. Carved in
both the geometric and the realistic styles, the Sin IX
statues could be understood then as a continuation,
in greater quantity, of the sculpture styles represented
by the two male heads in Sin VIII.167
Sin X, dated by the excavators to ED III, had been
dug illicitly and produced no sculpture during the
controlled excavations.168 At Khafajah, sculpture was
also retrieved from the Temple Oval and the Nintu
Temple. The temple proper of the Temple Oval complex was not preserved, and the surviving sculpture is
fragmentary.169 The Nintu Temple cannot be dated
by either pottery or glyptic.170
Dating Geometric-Style Sculpture
A lengthy period of production can be posited for
the geometric style on the basis of archaeological context, for sculpture in the geometric style continues to
appear in ED III contexts. Perhaps it is more accurate to
speak of various geometric styles in the Early Dynastic
period rather than of one uniform geometric style.
A sculpture hoard from the North Temple at Nippur contains five statues carved in varying styles.171 According to the excavators, the statues in the hoard were
either discarded outside the temple at level III, dated
to ED III, or buried below the floor of the cella when
it was extended in level II, dated to the Akkadian period.172 The most complete of the statues is a standing
and Lloyd 1942, 66).
166
Unillustrated Sin IX statue fragments catalogued from
Q 42:3 and Q 42:7 also reflect this shift in gender. Of the 12
sculpture fragments from Q 42:7, 10 are from female figures
and two are from male figures (Frankfort 1943, 40–1). Of the
24 sculpture fragments from Q 42:3, four are of unidentifiable
gender, eight are from female figures, and 12 are from male
figures (Frankfort 1943, 39–40).
167
The appearance of a tufted skirt among the male figures
in the later Q 42:3 context (Frankfort 1939, nos. 39, 113; 1943,
nos. 252, 253) would then represent a trend within an existing
realistic style rather than a shift from a geometric to a realistic
style. The tufted skirt does not appear among the few male figures in the earlier Q 42:7 context, but this may be insignificant
because among female figures, a tufted garment is already
present in Sin VIII (Frankfort 1943, no. 250) and in Sin IX,
Q 42:7 (Frankfort 1939, nos. 76, 106–8; see also 1943, 23 for
findspot corrections).
168
Delougaz and Lloyd 1942, 2–3, 6.
169
Frankfort 1939; 1943.
170
Supra n. 14.
171
Zettler 1978, 349 n. 12.
172
McCown et al. 1978, 22, 71, pls. 3A, 32.
628
JEAN M. EVANS
male figure about 75 cm high, carved with geometric
forms.173 Although too worn to be completely legible,
an inscription across the back of the figure records a
land-sale contract contemporary with ED IIIa Fara tablets.174 Despite its geometric forms, the male figure has
elongated arms and a deep bend at the pointed elbows,
paralleling other ED III dedicatory sculpture in southern Mesopotamia with so-called realistic forms.175
Largely because of the geometric-style sculpture associated with the Kleiner Antentempel, Tell Chuera in
northern Syria—has been characterized as a site with
ED III–Akkadian pottery but with ED II material remains.176 Two of the geometric-style sculpture fragments
are from the fill of the level 2 cella, which was sealed by
level 1, but the other sculpture fragments were found
strewn over the debris in rooms north of the Kleiner
Antentempel.177 Levels 1–3 of the Kleiner Antentempel
are now dated to the Chuera ID phase, which has most
recently been equated with either ED III or ED III–
Akkadian.178 The progressively later dating of the Kleiner Antentempel area would seem to support the conclusion that the statues are out of context. Yet, the most
recent excavations of the Kleiner Antentempel demonstrated that the temple itself was of short duration,
and no additional sculpture was retrieved.179 If geometric-style sculpture continued throughout the Early Dynastic period, the seeming conflict with an ED III/ED
IIIb–Akkadian context is unfounded. The Tell Chuera
statues have no exact parallels for the narrow platform
upon which the feet are carved, the slender proportions, and the diminutive, relief-carved facial features.
The closest parallel for the severity of the geometric style represented by the Asmar hoard is among the
sculpture from the ED IIIa level VIIb of the Inanna
Temple at Nippur, which produced an assemblage of
sculpture carved in disparate styles.180 The pottery associated with Square Temple I provides a terminus ante
quem of the end of ED I for the burial of the Asmar
hoard. A lengthy span of time therefore exists between
the Asmar hoard and the Inanna Temple sculpture.
Perhaps reasons for the duration of geometric-style
sculpture are related to the temple context. During
the Early Dynastic period, temples were maintained
with the ritual renewal of the cella and its cultic furniture and installations.181 The typologies and styles of
artifacts dedicated to temples were likely maintained
173
McCown et al. 1978 (3N–402).
McCown et al. 1978, 72, no. 1; Gelb et al. 1991, 90–1, no.
25. Perhaps related is a notation on the upper right arm of the
statue of Enmetena of Lagash recording land donated to a
temple (Cooper 1986 [La 5.17]; Hansen 2003, 29–30).
175
Cf. Hansen 1975, figs. II, 31.
176
Zettler 1978, 349; Schwartz 1990, 765; Pruss 2000, 1434.
177
Moortgat 1965, 23; 1967, 14, 21.
174
[AJA 111
in a manner analogous to the maintenance of the
temple itself. Certainly, it is an accepted assumption
that some objects found in temples must have been
preserved in levels later than the one in which they
entered the temple because they were valuable, fulfilled a certain function, or had to remain within the
temple confines once they were dedicated.182 Temple
“heirlooms” therefore existed.
To examine any given temple level only to designate
the style of individual artifacts as either contemporary
with or earlier than the context in which they were
retrieved, without recourse to a greater significance,
seems limited. If temple objects were preserved in levels later than the one in which they entered the temple, the effect that the presence of such heirlooms had
on artistic production should be considered. It seems
reasonable to suppose that, for example, if sculpture is
preserved in later temple levels, the style of the older
sculpture could have influenced the style of the new
sculpture. In other words, it might be desirable to
carve new statues to resemble old statues, for some of
the same reasons that heirlooms existed: the style was
valuable, fulfilled a certain function, or merely represented one of the various styles present because dedicated objects remained within the temple. Thus, rather
than finding disparate sculpture styles in a single archaeological context problematic, it seems more
reasonable to conclude instead that sculpture production in various styles continued throughout the Early
Dynastic period.183 The presence of various geometric styles in the ED IIIa Inanna Temple at Nippur,
in the ED III–Akkadian North Temple at Nippur,
in ED III temples at Khafajah, and in the ED IIIb–
Akkadian Kleiner Antentempel at Tell Chuera should
not be explained only by designating as heirlooms
the sculpture that does not conform to the sequence
established by the Diyala excavators of a geometric
style becoming a realistic style. As a result, no single
sculpture style should be considered contemporary
in all of Mesopotamia, and no single sculpture style
should be considered exclusive to any one single Early
Dynastic subdivision.
conclusions
The Diyala excavators identified three distinct
building periods in the Abu Temple sequence at Tell
178
Pruss 2000, figs. 2, 11; 2004, table 2.
Dohmann-Pfälzner and Pfälzner 1996, 3.
180
Hansen 1975, figs. II, 20–2, 23b.
181
Hansen 2003, 28–9.
182
Frankfort 1939, 16; Delougaz and Lloyd 1942, 4; Hansen
2003, 29.
183
Hansen 1975, 159; 2003, 26, 29.
179
2007]
THE SQUARE TEMPLE AT TELL ASMAR
Asmar. Nowhere else are such pronounced architectural changes manifest in an Early Dynastic temple sequence. As understood by the Diyala excavators, the
Abu Temple is truly unique. The three radically different and successive plans in the Abu Temple sequence
formed the basis for the division of the Early Dynastic
period into ED I, II, and III. However, the poorly recorded levels beneath the Square Temple indicate that
the transition from the Archaic Shrine to the Square
Temple was not as radical as the excavators believed. If
the levels between the Archaic Shrine and the Square
Temple had been considered a valid part of the Abu
Temple sequence, instead of necessitating the formation of an ED II period, these building periods might
have characterized ED I as the lengthy, important span
of time in Early Dynastic city-state formation that more
recent excavations have shown it to be.
If sculpture had not been so inextricably tied to establishing the Early Dynastic periodization, a different
assessment altogether of sculpture styles—and of the
Early Dynastic period—also might have emerged. Because of the importance accorded to geometric-style
sculpture, the earliest level in every Diyala temple
from which geometric-style sculpture was retrieved was
dated to the onset of ED II, despite other factors that
might have indicated a different date. Nowhere is the
paradox created by this methodology more clear than
in reference to the Abu Temple sequence, for the ceramic assemblage both below and in Square Temple
I is ED I. The following considerations support the
conclusion that the Asmar hoard was buried from a
level below Square Temple I: the findspot elevation
of the hoard, photographs showing the hole some distance below the Square Temple I floor, the additional
floor levels and associated features below the 32.30 m
Square Temple I floor, the clay packing above the statues in the hoard and directly below a floor that likely
was or would be in use, and Lloyd’s account in the field
notebooks of the discovery of the hoard. The date of
the Asmar hoard should be established by ceramics,
which indicate that the hoard was buried before the
end of ED I. Geometric-style sculpture is, then, present in temple contexts throughout the Early Dynastic
period, and I suggest that the maintenance of temple
traditions encouraged the continuation of relatively
consistent sculpture styles.
There is no longer any evidence, even in the Diyala, that the Fara style is exclusively ED II. The Fara
elegant style first appears in ED I in Square Temple I
629
and levels below it, and in IT IXA at Nippur. The Fara
crossed style appears in Sin VIII at ED II Khafajah but
continues into ED III Diyala contexts and is associated with ED IIIa Fara tablets at the site of Fara. It is
problematic to retain an ED II designation for Farastyle glyptic when the style cannot be correlated with a
precise period of time defined as ED II. According to
the current periodization, Fara-style glyptic is present
from the end of ED I to ED IIIa, with an elegant style
preceding a crossed style. Recognizing ED II as a chronological marker is therefore imprecise, and recognizing
ED II as a stylistic term leaves its dating unresolved.
Due to an inability to identify ED II, some scholars have argued that there is only ED I and ED III
and that ED II is a regional phenomena restricted to
the Diyala. However, I argue that the ED II subdivision should not be understood as Diyala regionalism.
Rather than ED II Diyala, there is only ED II Khafajah
because it is only in Houses 6–4 that ED II diagnostic
pottery was identified. The only Diyala pottery forms
representative solely of ED II are a type of fruitstand
and a type of pilgrim flask that are variations on types
present in ED III. The restriction to a few examples
in Houses 6–4 of two ED II ceramic forms with parallels to ED III ceramics forces the question: are these
two ceramic forms truly diagnostic of a distinct ED II
chronological subdivision or are they merely variations
on forms from other (contemporary?) ED III contexts?
If all geometric-style sculpture had not been de facto
correlated with the onset of ED II, would two such ceramic forms have been designated as ED II diagnostic?
This I understand as the essential dilemma of the ED
II subdivision, which has never been stated so explicitly because the role of geometric-style sculpture in
determining a tripartite Early Dynastic periodization
has never been fully acknowledged.
Even in the current scheme of an ED I, II, and III, a
refinement must be made to the chronology in order
to reflect that the temple levels dated to ED II by the
Diyala excavators are not precisely contemporary with
one another (table 6). Both Square Temple I and the
levels below it are ED I on the basis of ceramics and
therefore earlier than Houses 6–4 at Khafajah, which
were ceramically defined as ED II and have a close relationship to ED III material culture. The Fara crossedstyle seal carved in a local style correlates Square
Temple II with a late Shara Temple level containing
a related Fara crossed-style seal.184 Some references
here have indicated that much of the Shara Temple
184
The Shara Temple yielded one elegant-style seal (Frankfort 1939, no. 883) and one crossed-style seal (Frankfort 1939, no. 875).
The former could be from an earlier context than the latter, but the context is open to interpretation. The findspot elevation of Seal
883 (32.50 m) suggests the Main Level second occupation in the southern section, but the denudation of the walls (Delougaz and
Lloyd 1942, 251, pl. 26) in the M 14:9 room in which it was found suggests the findspot is earlier; the seal would then be from the Intermediate Foundations building period, to which the earlier walls encountered in M 14:9 presumably belong.
630
JEAN M. EVANS
[AJA 111
Table 6. Proposed Relative Chronology of Diyala Levels.
Tell Asmar
“Early Dynastic II”
Khafajah
Tell Agrab
Abu Temple
Sin Temple
Houses
Temple Oval
Shara Temple
Square Temple II
VIII
6
I
Main Level 2
Square Temple I
predecessor
VII
7
Early Dynastic I
8
Archaic Shrine IV
Archaic Shrine III
Main Level 1
Intermediate
Earlier Building
VI
sequence is also earlier than ED II Khafajah and that
part of the Shara Temple sequence is earlier than the
Square Temple.185
The question of whether or not ED II terminology
should be retained is a topic of some contention, and
it may be that scholars prefer to work within the existing tripartite periodization because an adjustment
to Early Dynastic chronology would be too cumbersome. The Fara crossed-style cylinder seals in Square
Temple II and in the Main Level second occupation
of the Shara Temple could correlate these levels with
Sin VIII/Houses 6–4. This correlation could subsequently produce ED II contexts at Tell Asmar and Tell
Agrab. Yet, proceeding in this manner forces a solution
for establishing ED II contexts in the Diyala temples
but does not produce criteria for ED II applicable to
Mesopotamia in general because the Fara crossed style
continues into ED III. As the evidence currently stands,
there is no ED II pottery outside Houses 6–4, and there
is no glyptic or sculpture style that can be considered
representative solely of ED II. Yet, rather than leaving
a gap in the Early Dynastic chronology, the solution
outlined here eliminates ED II terminology by way of
correction: the ED I dating of the relevant Square Temple and Shara Temple levels is based on ceramics, and
the ED II ceramics at Khafajah are better understood
as localized variations on ED III ceramics. The span of
time now accorded to an ED II period could then be
subsumed equally into ED I and ED III.
9
department of ancient near eastern art
the metropolitan museum of art
1000 fifth avenue
new york, new york 10028
[email protected]
Works Cited
Abu al-Soof, B. 1967. “The Relevance of the Diyala Sequence
to Southern Mesopotamian Sites.” Iraq 29:133–42.
Algaze, G. 1983–1984. “Private Houses and Graves at Ingharra: A Reconsideration.” Mesopotamia 18–19:135–93.
Amiet, P. 1980. La glyptique mésopotamienne archaïque. 2nd
rev. ed. Paris: Éditions du Centre National de la Recherche Scientifique.
Behm-Blancke, M. 1979. Das Tierbild in der Altmesopotamischen Rundplastik: Eine Untersuchung zum Stilwandel des
frühsumerischen Rundbildes. Baghdader Forschungen 1.
Mainz: Philipp von Zabern.
Braun-Holzinger, E. 1977. Frühdynastische Beterstatuetten.
Abhandlungen der Deutschen Orient-Gesellschaft 19.
Berlin: Gebr. Mann.
Buccellati, G., and R. Biggs. 1969. Cuneiform Texts from Nippur: The Eighth and Ninth Seasons. AS 17. Chicago: University of Chicago Press.
Buchanan, B. 1956. Review of Stratified Cylinder Seals from the
Diyala Region, by H. Frankfort. AJA 60(1):70–2.
———. 1966. Catalogue of Ancient Near Eastern Seals in the
Ashmolean Museum. Vol. 1, Cylinder Seals. Oxford: Clarendon Press.
Collon, D. 1988. First Impressions: Cylinder Seals in the Ancient
Near East. Chicago: The University of Chicago Press.
185
The date of the Shara Temple is another topic entirely, but Buchanan (1966, 24), Martin (1988, 77), Hansen (2003, 27), Evans
(2005, 119–65), and Marchetti (2006, 35–40) have suggested that certain levels date to ED I.
2007]
THE SQUARE TEMPLE AT TELL ASMAR
Cooper, J. 1986. Presargonic Inscriptions. Sumerian and Akkadian Royal Inscriptions 1. The American Oriental
Society Translations Series 1. New Haven: American
Oriental Society.
Delougaz, P. 1940. The Temple Oval at Khafājah. OIP 53.
Chicago: University of Chicago Press.
———. 1952. Pottery from the Diyala Region. OIP 63. Chicago:
University of Chicago Press.
Delougaz, P., and S. Lloyd. 1942. Pre-Sargonid Temples in
the Diyala Region. OIP 58. Chicago: University of Chicago Press.
Delougaz, P., H. Hill, and S. Lloyd. 1967. Private Houses and
Graves in the Diyala Region. OIP 88. Chicago: University
of Chicago Press.
Dohmann-Pfälzner, H., and P. Pfälzner. 1996. “Untersuchungen zur Urbanisierung Nordmesopotamiens
im 3. Jt. v.Chr.: Wohnquartierplanung und städtische
Zentrumsgestaltung in Tall Chuēra.” DM 9:1–13.
Evans, J.M. 2005. “Redefining the Sculpture of Early Dynastic Mesopotamia.” Ph.D. diss., New York University.
Frankfort, H. 1932. Archeology and the Sumerian Problem. SAOC
4. Chicago: University of Chicago Press.
———. 1934. Iraq Excavations of the Oriental Institute, 1932/33:
Third Preliminary Report of the Iraq Expedition. OIC 17. Chicago: University of Chicago Press.
———. 1935a. Oriental Institute Discoveries in Iraq, 1933/34:
Fourth Preliminary Report of the Iraq Expedition. OIC 19. Chicago: University of Chicago Press.
———. 1935b. “Sumerian Sculpture.” BurlMag 66(384):
110–21.
———. 1936. Progress of the Work of the Oriental Institute in
Iraq, 1934/35: Fifth Preliminary Report of the Iraq Expedition.
OIC 20. Chicago: University of Chicago Press.
———. 1939. Sculpture of the Third Millennium B.C. from
Tell Asmar and Khafājah. OIP 44. Chicago: University of
Chicago Press.
———. 1943. More Sculpture from the Diyala Region. OIP 60.
Chicago: University of Chicago Press.
———. 1954. The Art and Architecture of the Ancient Orient.
Baltimore: Penguin Books.
———. 1955. Stratified Cylinder Seals from the Diyala Region.
OIP 72. Chicago: University of Chicago Press.
Gelb, I., P. Steinkeller, and R. Whiting, Jr. 1991. Earliest Land
Tenure Systems in the Near East: Ancient Kudurrus. 2 vols.
OIP 104. Chicago: University of Chicago Press.
Gibson, M. 1972. “The Archaeological Uses of Cuneiform
Documents.” Iraq 34:113–23.
Gibson, M., D.P. Hansen, and R. Zettler. 2001. “Nippur. B.”
In Reallexikon der Assyriologie und Vorderasiatischen Archäologie. Vol. 9, edited by E. Ebeling and B. Meissner, 546–65.
Berlin and New York: Walter de Gruyter.
Hall, H., and C.L. Woolley. 1927. Al-’Ubaid: A Report on the
Work Carried Out at al-’Ubaid for the British Museum in 1919
and for the Joint Expedition in 1922–3. Publications of the
Joint Expedition of the British Museum and of the Museum of the University of Pennsylvania to Mesopotamia.
Ur Excavations 1. London: Oxford University Press.
Hansen, D.P. 1963. “New Votive Plaques from Nippur.”
JNES 22(3):145–66.
———. 1965. “The Relative Chronology of Mesopotamia.
Part II. The Pottery Sequence at Nippur from the Middle
Uruk to the End of the Old Babylonian Period (3400–
1600 B.C.).” In Chronologies in Old World Archaeology. Vol.
1, edited by R.W. Ehrich, 201–13. Chicago: University of
Chicago Press.
631
———. 1971. “Some Early Dynastic I Sealings from Nippur.” In Studies Presented to George M.A. Hanfmann, edited by D.G. Mitten, J.G. Pedley, and J.A. Scott, 47–54.
Cambridge, Mass.: Fogg Art Museum.
———. 1975. “Frühsumerische und frühdynastische Rundplastik.” In Der alte Orient, edited by W. Orthmann, 158–
70. Propyläen Kunstgeschichte 14. Berlin: Propyläen
Verlag.
———. 1987. “The Fantastic World of Sumerian Art: Seal
Impressions from Ancient Lagash.” In Monsters and Demons in the Ancient and Medieval Worlds: Papers Presented in
Honor of Edith Porada, edited by A.C. Farkas, P.O. Harper,
and E.B. Harrison, 53–63. Mainz: Philipp von Zabern.
———. 2003. “Art of the Early City-States.” In Art of the First
Cities: The Third Millennium B.C. from the Mediterranean to
the Indus, edited by J. Aruz, 21–37. New York: The Metropolitan Museum of Art.
Hansen, D.P., and G. Dales. 1962. “The Temple of Inanna
Queen of Heaven at Nippur.” Archaeology 15:75–84.
Heinrich, E. 1931. Fara: Ergebnisse der Ausgrabungen der
Deutschen Orient-Gesellschaft in Fara und Abu Hatab, 1902/
03. Berlin: Staatliche Museen zu Berlin, Vorderasiatische
Abteilung.
Hrouda, B. 1971. Vorderasien. Vol. 1, Mesopotamien, Babylonien, Iran und Anatolien. Munich: Beck.
Karg, N. 1984. Untersuchungen zur älteren frühdynastischen
Glyptik Babyloniens: Aspekte regionaler Entwicklungen in der
ersten Hälfte des 3. Jahrtausends. Baghdader Forschungen
8. Mainz: Philipp von Zabern.
Kuhrt, A. 1995. The Ancient Near East c. 3000–330 B.C. 2
vols. London: Routledge.
Legrain, L. 1936. Archaic Seal-Impressions. Publications of the
Joint Expedition of the British Museum and of the Museum of the University of Pennsylvania to Mesopotamia.
Ur Excavations 3. Oxford: Oxford University Press.
Lloyd, S. 1933–1934. “The Abu Temple Excavations.” Unpublished field notebook. Chicago: The Oriental Institute of the University of Chicago.
Marchetti, N. 2006. La Statuaria Regale nella Mesopotamia
Protodinastica. Atti della Accademia Nazionale dei Lincei 403. Memorie Classe di Scienze Morali, Storiche e
Filologiche. Ser. 9, vol. 21, fasc. 1. Rome: Bardi.
Martin, H. 1975. “The Tablets of Shuruppak.” In Le temple et le culte: Compte rendu de la vingtième rencontre assyriologique internationale: Organisée à Leiden du 3 au 7 juillet
1972 sous les auspices du Nederlands Instituut voor het Nabije
Oosten, 173–82. Uitgaven ven het Nederlands HistorischArchaeologisch Instituut te Istanbul 37. Leiden: Nederlands Instituut voor het Nabije Oosten.
———. 1982. “The Early Dynastic Cemetery at al-’Ubaid:
A Re-Evaluation.” Iraq 44:145–85.
———. 1988. Fara: A Reconstruction of the Ancient Mesopotamian City of Shuruppak. Birmingham: C. Martin.
Martin, H., J. Moon, and J. Postgate. 1985. Graves 1 to 99.
Abu Salabikh Excavations 2. London: British School of
Archaeology in Iraq.
Matthews, D. 1997. The Early Glyptic of Tell Brak. Orbis Biblicus et Orientalis Series Archaeologica 15. Fribourg:
University Press.
McCown, D., R. Haines, and R. Biggs. 1978. Nippur. Vol. 2,
The North Temple and Sounding E: Excavations of the Joint
Expedition to Nippur of the American Schools of Oriental Research and the Oriental Institute of the University of Chicago.
OIP 97. Chicago: University of Chicago Press.
Moon, J. 1981. “Some New Early Dynastic Pottery from Abu
632
JEAN M. EVANS, THE SQUARE TEMPLE AT TELL ASMAR
Salabikh.” Iraq 43:47–75.
———. 1987. Catalogue of Early Dynastic Pottery. Abu Salabikh Excavations 3. London: British School of Archaeology in Iraq.
Moorey, P.R.S. 1966. “A Reconsideration of the Excavations on Tell Ingharra (East Kish), 1923–1933.” Iraq
28:18–51.
———. 1978. Kish Excavations, 1923–1933: With a Microfiche Catalogue of the Objects in Oxford Excavated by the Oxford–Field Museum, Chicago, Expedition to Kish in Iraq,
1923–1933. Oxford: Clarendon Press.
———. 1979. “Unpublished Early Dynastic Sealings from
Ur in the British Museum.” Iraq 41:105–20.
Moortgat, A. 1965. Tell Chuēra in Nordost-Syrien: Bericht über
die vierte Grabungskampagne, 1963. Wissenschaftliche Abhandlungen der Arbeitsgemeinschaft für Forschung des
Landes Nordrhein-Westfalen. Vol. 31. Cologne: Westdeutscher Verlag.
———. 1967. Tell Chuēra in Nordost-Syrien: Vorläufiger Bericht über die fünfte Grabungskampagne 1964. Schriften der
Max Freiherr von Oppenheim-Stiftung 6. Wiesbaden:
Westdeutscher Verlag.
Moortgat-Correns, U. 1959. “Bemerkungen zur Glyptik
des Diyala–Gebietes.” OLZ 54(7–8):341–54.
Nissen, H. 1988. The Early History of the Ancient Near East,
9000–2000 B.C. Chicago: University of Chicago.
———. 2007. “Archaeological Surveys and Mesopotamian
History.” In Settlement and Society: Essays Dedicated to Robert McCormick Adams, edited by E.C. Stone, 19–28. Los
Angeles: Cotsen Institute of Archaeology, University
of California.
Nöldeke, A., and H. Lenzen. 1940. Elfter vorläufiger Bericht
über die von der Deutschen Forschungsgemeinschaft in UrukWarka unternommenen Ausgrabungen. Abhandlungen der
Preussischen Akademie der Wissenschaften. Philosophisch-Historische Klasse 3. Berlin: Verlag der Akademie der Wissenschaften.
Perkins, A. 1953. Review of Pottery from the Diyala Region, by
P. Delougaz. JAOS 73(1):47–50.
Pittman, H. 1994. The Glazed Steatite Glyptic Style: The Structure and Function of an Image System in the Administration
of Protoliterate Mesopotamia. Berliner Beiträge zum Vorderen Orient 16. Berlin: Dietrich Reimer Verlag.
Porada, E. 1980. “Introduction.” In Ancient Art in Seals, edited by E. Porada, 3–34. Princeton: Princeton University Press.
———. 1991. Review of Fara: A Reconstruction of the Ancient
Mesopotamian City of Shuruppak, by H. Martin, and Untersuchungen zur Älteren Frühdynastischen Glyptik Babyloniens,
by N. Karg. AJA 95(1):170–73.
Porada, E., D.P. Hansen, S. Dunham, and S. Babcock. 1992.
“The Chronology of Mesopotamia, ca. 7000–1600 B.C.”
In Chronologies in Old World Archaeology. Vol. 1, edited by
R.W. Ehrich, 77–121. Chicago: University of Chicago
Press.
Porada, E., D.P. Hansen, and J.R.F. von Beckerath. 1968.
“Summary of Contributions.” AJA 72(2):301–5.
Postgate, J. 1977. “Excavations at Abu Salabikh, 1976.”
Iraq 39:269–99.
———. 1980. “Early Dynastic Burial Customs at Abu Sal-
abikh.” Sumer 36:65–82.
———. 1984. “Excavations at Abu Salabikh, 1983.” Iraq
46:95–113.
Postgate, J., and P.R.S. Moorey. 1976. “Excavations at Abu
Salabikh 1975.” Iraq 38:133–69.
Pruss, A. 2000. “Recent Excavations at Tell Chuera and
the Chronology of the Site.” In Proceedings of the First International Congress on the Archaeology of the Ancient Near
East, Rome, May 18th–23rd 1998. Vol. 2, edited by P. Matthiae, A. Enea, L. Peyronel, and F. Pinnock, 1431–46.
Rome: Dipartimento di scienze storiche, archeologiche
e antropologiche dell’antichità.
———. 2004. Review of Tall Bi’a/Tuttul. Vol. 8, Stadtbefestigungen, Häuser, und Tempel, by P. Miglus and E. Strommenger. ZA 94:146–51.
Roaf, M. 1990. Cultural Atlas of Mesopotamia and the Ancient
Near East. New York: Facts on File.
Schwartz, G. 1990. Review of Tell Chuēra in Nordost-Syrien:
Vorläufiger Bericht über die neunte und zehnte Grabungskampagne, 1982 und 1983, by U. Moortgat-Correns. JAOS
110(4):764–66.
Strommenger, E. 1960. “Das Menschenbild in der altmesopotamischen Rundplastik von Mesilim bis Hammurapi.” BaM 1:1–103.
Tunca, Ö. 1984. L’architecture religieuse protodynastique en
Mesopotamie. 2 vols. Akkadica Supplementum 2. Leuven: Peeters.
Weiss, H. 1975. “Kish, Akkad, and Agade.” JAOS 95(3):
434–53.
Wilson, K. 1985. “A Definition and Relative Chronology of the Jamdat Nasr Period.” Ph.D. diss., New York
University.
———. 1986. “Nippur: The Definition of a Mesopotamian Ğamdat Nasr Assemblage.” In Ğamdat Nasr: Period
or Regional Style?, edited by U. Finkbeiner and W. Röllig, 57–89. Beihefte zum Tübinger Atlas des Vorderen
Orients. Ser. B, Geisteswissenschaften 62. Wiesbaden:
Ludwig Reichert Verlag.
Winter, I. 1984. Review of La statuaire du proche-orient ancien, by A. Spycket. JCS 36(1):102–14.
Woolley, C.L. 1934. The Royal Cemetery: A Report on the Predynastic and Sargonid Graves Excavated Between 1926 and
1931. 2 vols. Publications of the Joint Expedition of the
British Museum and of the Museum of the University of
Pennsylvania to Mesopotamia. Ur Excavations 2. London: Oxford University Press.
Zettler, R. 1978. Review of Die Keramik vom Tell Chuera und
ihre beziehungen zu Funden aus Syrien-Palästina, der Türkei
und dem Iraq, by H. Kühne. JNES 37(4):345–50.
———. 1989. “Pottery Profiles Reconstructed from Jar Sealings in the Lower Seal Impression Strata (SIS 8–4) at
Ur: New Evidence for Dating.” In Essays in Ancient Civilization Presented to Helene J. Kantor, edited by A. Leonard, Jr., and B.B. Williams, 369–87. SAOC 47. Chicago:
University of Chicago Press.
———. 1992. The Ur III Temple of Inanna at Nippur: The Operation and Organization of Urban Religious Institutions in
Mesopotamia in the Late Third Millennium B.C. Berliner
Beiträge zum Vorderen Orient 11. Berlin: Dietrich Reimer Verlag.