Irene J. Winter - On Art in The Ancient Near East - Vol. 2 - From The Third Millennium BCE (Culture and History of The Ancient Near East) (2009)

On Art in the Ancient Near East
Culture and History of the

Ancient Near East
Founding Editor
M. H. E. Weippert
Editor-in-Chief
Thomas Schneider
Editors
Eckart Frahm, W. Randall Garr, B. Halpern,
Theo P. J. van den Hout
VOLUME 34.2
On Art in the Ancient
Near East
Volume II
From the Third Millennium B.C.E.
Irene J. Winter
LEIDEN • BOSTON
2010
This book is printed on acid-free paper.
On the art in the ancient Near East / edited by Irene J. Winter.

p. cm. — (Culture and history of the ancient Near East, ISSN 1566-2055 ; v. 34)
Includes bibliographical references and index.
ISBN 978-90-04-17237-1 (hard cover : v. 1 : alk. paper) — ISBN 978-90-04-17499-3
(hard cover : v. 2 : alk. paper) 1. Middle East—Antiquities. 2. Art, Ancient—Middle
East. 3. Sculpture, Ancient—Middle East. 4. Visual communication—Middle East—
History—To 1500. 5. Social archaeology—Middle East. I. Winter, Irene.
DS56.O5 2009
709.39’4—dc22
2009020832
ISSN 1566-2055
ISBN 978 90 04 17499 3
Copyright 2010 by Koninklijke Brill NV, Leiden, The Netherlands.

Koninklijke Brill NV incorporates the imprints Brill, Hotei Publishing,
IDC Publishers, Martinus Nijhoff Publishers and VSP.
All rights reserved. No part of this publication may be reproduced, translated,

stored in a retrieval system, or transmitted in any form or by any means, electronic,
mechanical, photocopying, recording or otherwise, without prior written permission
from the publisher.
Authorization to photocopy items for internal or personal use is granted by

Koninklijke Brill NV provided that the appropriate fees are paid directly to
The Copyright Clearance Center, 222 Rosewood Drive, Suite 910,
Danvers, MA 01923, USA.
Fees are subject to change.
printed in the netherlands

VOLUME TWO
FROM THE THIRD MILLENNIUM B.C.E.
CONTENTS
Introduction ................................................................................ vii

Acknowledgments ....................................................................... xv
SCULPTURE AND THE EARLY STATE
Chapter Sixteen After the Battle is Over: The Stele of

the Vultures and the Beginning of Historical Narrative in
the Art of the Ancient Near East .......................................... 3
Chapter Seventeen Eannatum and the “King of Kiš”?:
Another Look at the Stele of the Vultures and
“Cartouches” in Early Sumerian Art ..................................... 53
Chapter Eighteen Women in Public: The Disk of
Enheduanna, the Beginning of the Office of En-Priestess,
and the Weight of Visual Evidence ......................................... 65
Chapter Nineteen Sex, Rhetoric, and the Public
Monument: The Alluring Body of Naram-Sîn of Agade ....... 85
Chapter Twenty Tree(s) on the Mountain: Landscape
and Territory on the Victory Stele of Naram-Sîn
of Agade ................................................................................. 109
Chapter Twenty-One How Tall was Naram-Sîn’s Victory
Stele? Speculation on the Broken Bottom ........................... 133
Chapter Twenty-Two The Body of the Able Ruler: Toward
an Understanding of the Statues of Gudea .......................... 151
Chapter Twenty-Three ‘Idols of the King’: Royal Images
as Recipients of Ritual Action in Ancient Mesopotamia ..... 167
EXPERIENCING ‘ART’ AND ARTIFACT
Chapter Twenty-Four Representing Abundance: A Visual

Dimension of the Agrarian State ........................................... 199
vi contents
Chapter Twenty-Five Reading Ritual in the Archaeological

Record: Deposition Pattern and Function of Two Artifact
Types from the Royal Cemetery of Ur ................................. 227
Chapter Twenty-Six “Surpassing Work”: Mastery of
Materials and the Value of Skilled Production in Ancient
Sumer ...................................................................................... 271
Chapter Twenty-Seven The Aesthetic Value of Lapis Lazuli
in Mesopotamia .................................................................... 291
Chapter Twenty-Eight Agency Marked, Agency Ascribed:
The Affective Object in Ancient Mesopotamia ..................... 307
Chapter Twenty-Nine “Seat of Kingship”/“A Wonder to
Behold”: The Palace as Construct in the Ancient Near East .... 333
Chapter Thirty Opening the Eyes and Opening the
Mouth: The Utility of Comparing Images in Worship in
India and the Ancient Near East ........................................... 377
Chapter Thirty-One The Affective Properties of Styles:
An Inquiry into Analytical Process and the Inscription
of Meaning in Art History ..................................................... 405
VIEWING (IN) THE PAST AND THE PRESENT
Chapter Thirty-Two The Eyes Have It: Votive Statuary,

Gilgamesh’s Axe, and Cathected Viewing in the Ancient
Near East ................................................................................ 433
Chapter Thirty-Three Babylonian Archaeologists of The(ir)
Mesopotamian Past ................................................................. 461
Chapter Thirty-Four Exhibit/Inhibit: Archaeology, Value,
History in the Work of Fred Wilson ...................................... 481
Chapter Thirty-Five Change in the American Art Museum:
The (An) Art Historian’s Voice .............................................. 493
Chapter Thirty-Six Packaging the Past: The Benefits
and Costs of Archaeological Tourism .................................... 521
INTRODUCTION
The studies included in the present volume of collected essays range

from specific monuments of the third millennium b.c.e. to issues of
interpretation, understanding and ethics relevant to the third millennium
c.e. The initial preposition of the subtitle has been carefully selected,
as were the prepositions of the subtitle of Volume I and the overall
title to the two volumes combined. On Art in the Ancient Near East in the
main title places emphasis not upon a class of work called ‘Art’, as if
a given and unexamined category, but rather upon discourses about what
we call art as identified within the cultural production of the ancient
Near East. The key preposition of the Vol. I subtitle, then, Of the First
Millennium b.c.e. . . ., implies that the studies included there purport to
speak historically about artistic production and consumption within
that time frame. The comparable preposition for the subtitle of Volume
II, however, signals a less-contained universe. From the Third Millennium
b.c.e. . . . purposely leaves open the sequence and consequences leading
from the historical period of early state formation in Mesopotamia to
both later periods within the tradition and the current era.
This conscious decision was based upon a desire to include in the
assemblage several articles the task of which was to interrogate how
works of ancient ‘art’ have been perceived and dealt with well beyond
the moment of their manufacture and subsequent usage within a single
cultural stream, toward issues of reception, both modern and post-
modern. The word play involved is not merely one of stylistic variation,
therefore; it marks a difference between the contents of Volume I and
Volume II, a part of which is the insertion of the author not only as
historian, but as critic.
As with Volume I, the articles of Volume II reveal a combination
of intellectual and disciplinary approaches—art history, archaeology,
anthropology and philology, inflected by and engaged with current
theoretical and interpretive models. And yet they also mark something
of a shift. One could say that the studies included in Volume I are more
weighted toward archaeological inquiries into various media in their his-
torical setting, along with analysis of political ideology and economy as
manifest in the works and their distribution. Those included in Volume
II on the whole reflect a more integrative bent, one that includes the
viii introduction
role(s) played by belief systems as evident in ritual and religion no less

than politics, as well as an added stress on values attached to experience
and to specific materials, recoverable through representation itself as
well as through language and archaeological deposition.
This shift can be understood not merely as the intellectual trajectory
of a single individual, but historiographically, both within the field of
ancient Near Eastern studies and in intellectual discourse in general
over the past three decades. Initial studies of the 1960s through the
early 1980s into the political, economic and ideological subsystems of
cultures, including this particular ancient old world culture, constituted
a necessary antidote to a prior over-emphasis upon the religious in
scholarship, parallel to the work of the great contemporary theologians
of the 1940s and 50s, but vested, it could be demonstrated, in an
assumed transparency between ‘us’ and ‘them’, hence often consisting
of a series of unexamined projections across time and space. The new
focus on polities and politics yielded important results for its time in the
understanding of historical events and attendant works; however, the
empirical shift, as with any pendulum swing, had its own exclusionary
limitations, and eventually instantiated its own correctives. Furthermore,
as more of the textual and archaeological record was published and
could be put together with the visual/artifactual record, more complex
and cross-cutting data sets could be brought to the service of historical
understanding.
The next phase of inquiry, at least in my opinion, seeks to reveal in a
more nuanced way a cultural system (however blurred the boundaries)
or social/historical moment at work. Re-reading these essays has led me
not only to this reconstruction of the recent history of our subfield, but
also perhaps to a clearer articulation of intellectual processes in general.
Knowledge is cumulative, yes; and scholarship at its best contributes
to the cumulation of knowledge. Yet each new phase of inquiry works
both off of and against previous phases, to the extent that advances in
method and theory as well as in substantive information and heuristics
can only be made in dialogue with the prior phases’ advances and
limitations. It therefore seems too easy to construct a simple binary
model of ever-recurring oscillations between, say, religion and econom-
ics or politics, a counterpoint in which politics always turns back to
religion, which inevitably returns to politics. Rather, I would argue, to
the extent that paradigm shifts are the necessary consequences of either
dead ends or emptied explanatory vehicles, each corrective pendulum
swing itself necessitates a subsequent integrative phase—one in which
introduction ix
the prior oppositions are at least sometimes understood as the results

of competing paradigms, neither wholly right nor wholly wrong. Such
integrations can make clear that what have seemed to be mutually
exclusive interpretive frameworks may in fact be equally embedded in
a more holistic view of culture. Thus, while a given intellectual moment
may stress one governing paradigm over another, or focus on some
aspects of social and artistic production over others, the integrative or
synthetic phase of seeking how the ‘system’ functions must be viewed
as an essential part of the historical enterprise.
* * * * * *
The articles reproduced in the present volume cluster again into three
sub-sets: those dealing with specific monuments of ancient Mesopotamia from
the Early Dynastic Period through the Akkadian to the Neo-Sumerian,
the issues they raise, and the meanings they seem to generate within their
original context (Chapters 16–22); those dealing with ancient experience of
works, charged events, artistic production and materials, hence meaning
at another level (Chapters 23–31); and those with a focus on vision, visual-
ity and display, which clearly builds upon experience but introduces the
additional component, for the modern world no less than the ancient, of
judgment, discrimination and interest related to consumption (Chapters
32–36). Within each section, essays have been ordered more or less in
order of publication, and once again, I have been surprised at the degree
to which individual studies often seem to be in conversation with one
another. At the same time, a number of essays included in Volume II
pick up on themes and issues raised in studies presented in Volume I:
Chapters 1 and 16 cross millennial borders to deal with historical
narrative in the service of hegemonic politics; Chapters 4 and 24 are
joined by concerns with environmental and agricultural abundance as
it related to both ideology and iconography; questions of cultural and
material agency, implied in Chapters 5, 8 and 13, are taken on directly
in Chapter 28. Additionally, several essays, such as Chapters 27, 29,
31 and 32, draw upon materials that span the full sequence from the
third millennium b.c.e. to the first. Although these last might just as
well have been included in Volume I, it was thought to place them
in Volume II because on the one hand, continuities or developments
through the chronological sequence seemed to be most evident if read
from the beginning, and on the other hand, because issues they engage
began to be interrogated later in my own work, and so belonged in
the later volume.
x introduction
A number of cross-cutting threads related to work in the subfield

have emerged for me in the ordering of these essays. First, given the
highly selective range of content in text and image-making in the ancient
Near East, along with excavation strategies at predominantly elite loci,
it comes as no surprise that themes of rule and rulership, kings and
kingship, dominate a number of studies (e.g., Chapters 16, 17, 19, 22,
28, 29, 31). These studies, by their focus on the ruler, always male, also
participate in contemporary inquiries into issues of gender, specifically
the construction of masculinity. They stand in counterpoint to the one
study related to women and two others positing issues of gender in
both funerary and symbolic arenas (Chapters 18, 25, 27). Even work
on landscape and agricultural production (Chapters 20, 24) come back
in the end to tropes of kingship.
Second, it will be noted that over time I have devoted more and
more attention to the relationship between not only text and image
per se (e.g., Chapter 16), but also between verbal and visual strategies of
representation. The indigenous lexicon in both Sumerian and Akkadian
permits us to probe more deeply than formal analysis alone and simple
matching of textual narrative to imagery into Mesopotamian systems of
values and meaning (Chapters 17, 19, 24, 26, 27, 28, 32, 33). A move
in this direction has been due entirely to the continued publication of
the relevant dictionaries and text corpora over the past several decades,
and more recently, to the availability of text editions on line, for all
of which one is immensely grateful. Probing not only how texts and
images represent, but how language itself works on and with imagery
has been very rewarding, and philological colleagues have been most
generous, as well as forgiving of the errors to which non-specialists are
too often prone. [The more seditious argument related to such inquiry
is the suggestion that on occasion, information is preserved/offered
visually when not preserved or offered in the textual record, as when
archaeological deposition is made to demonstrate funerary practice in
the absence of written documents, or imagery is suggested to record
roles that texts do not (Chapters 18, 25).]
Finally, third, in pursuing the construction, consecration and sub-
sequent treatment of public monuments—architecture no less than
sculpture—it has emerged that the relationship between artifacts and
ritual activity is crucial to our understanding, whether by association
with temples, or through direct activation of agentive roles in the public
domain (Chapters 16, 18, 19, 23, 25, 28, 30, 33). Additionally, the use
of, as well as the very nature of, certain materials as recorded in text
introduction xi
and observed empirically, is associated with meaning and argued to

bear charge (Chapters 20, 21, 25–27).
In this respect, it may be observed that many of the same monuments
appear in multiple studies, but with a different focus of attention and
toward differing analytical ends. This redundancy is itself a measure of
the polyvalence of the works with which we are privileged to interact as
scholars—their power as monuments, their complexity of conception,
perhaps even their conscious address of different audiences.
As had been the case with the essays included in Volume I, it will
be observed that the analytic tools of art history—typology, style, ico-
nography, composition, decorative program—are deployed throughout,
and sometimes even become the subject of analysis themselves (e.g.
Chapters 22, 24, 25, 29 and especially 31). Connections to the arts
and practices of other periods and places once again appear not infre-
quently, for purposes of amplification, differentiation, and/or analogy
(e.g. Chapters 19, 25, 23, 31, 32, 34), the validity of which is insisted
upon theoretically in Chapter 30. And theoretical discourses in the field
and beyond—especially those related to the determination of value
through material and symbolic charge, and the affective properties of the
artwork—have been invoked in order to place the various studies within
a broad discursive field (e.g. Chapters 16, 19, 22, 26, 27, 28, 31).
Surprising to me as author was how early in this body of scholarly
work issues of visual affect and the affective appeared in arguments
(e.g., Chapter 16, originally published 1985), and how consistently these
issues have been carried forward in later work. Affect can be stimu-
lated literally by subject matter, of course, but also through materials,
scale, composition, style, placement in charged surrounding space, or
attendant activity, such as ritual (see Chapters 23–24, 26–27, 31). Most
important for my purposes are those instances where the Mesopotamian
evidence (and here, we are greatly privileged by the textual corpus at
our disposal) allows for argument that a work not only has affect in a
sense of eliciting responses, it can also be affective with respect to the
very constitution of its viewing audience and the construction of cul-
tural norms (see, for example, Chapters 19, 29, 32). That works can
be shown to be not merely reflective of their time but affective as well
serves as a generational challenge to unreconstructed Marxian theoriza-
tion of the superstructure (i.e., cultural production) as epiphenomenal,
and engages more recent anthropological and sociological discussions
concerning the agency of the work itself, once it has left the hand of
the patron/artisan (see esp. Chapter 28, but also Chapters 22, 23, 34).
xii introduction
Such arguments at least implicitly address the early modern::postmodern

debate generated largely in literary studies of whether there is ever a
‘text’ as distinct from a ‘reader.’ For the art historian to recuperate his/
her identity as historian in this debate, it is essential to insist that there
are texts, not just readers; there are works with inherent properties, not
just audiences. Nevertheless, by insisting upon the non-transparency
between the scholar in the present and the works/events in the past,
as well as the range of viable responses to any given work, what the
postmodern challenge did for the historian was to force the interroga-
tion of prior assumptions, while ramping up the requirements for the
rigor of arguments and the use of evidence toward interpretation.
Naturally, when dealing with works of ancient art, not only the work
itself (and those responsible for its conceptualization and realization),
but also its audience(s) and its attendant meaning(s) existed in history; its
intended ‘readers’/receivers were not our contemporaries. Part of one’s
historical mandate therefore lies in an attempt to reconstruct reception,
not just conception (see Chapters 16, 19 and 29). This obvious, but, I
would argue, non-trivial observation has occasioned the inclusion of
the third section of the present volume (see especially Chapters 32–33).
Ancient Mespotamians, desirous that their constructions—temples, pal-
aces, and monuments—should be both positively experienced in their
own time and also preserved into an anticipated future, clearly knew
that the works in question would one day represent ‘the past,’ and they
attempted in inscribed curses and constraining orders to control both
treatment and valuation. Furthermore, Assyrians and Babylonians, at
least, were actively engaged in both archaeological recovery and restora-
tion, predicated upon an articulated relationship with their respective
pasts and current identities.
This in turn raises for us the question of audiences with even greater
historical distance from the works under scrutiny, hence the debates con-
cerning the uses of the past in our present! These debates often center
upon rights of ownership and access, the relationship between cultural
products and cultural stewardship, and they introduce an element of
responsibility-cum-ethics as well as understanding into the discussion
(e.g., Chapters 34–36). When seen through the lens of display strategies
in contemporary museums, what has been treated as received wisdom
can be challenged; when seen through the lens of knowledge lost and
knowledge gained in the course of acquisition and/or use, it becomes
clear that some questions can never be asked of unprovenienced works
and/or desecrated sites and monuments. Empirical studies such as that
introduction xiii
of Chapter 25, for example, which yielded insight into Mesopotamian

funerary practice through archaeological deposition, could never have
been undertaken without the excavator’s careful record of archaeo-
logical context. Contemporary practices and governing narratives of
entitlement, therefore, must become the subject matter of the scholar
no less than ancient practices and narratives.
In closing, then, what I trust emerges from the present collection of
essays is the power of sculptural works and architecture in the periods
under discussion, as well as the power they convey in our own time.
In retrospect, some articles seemed to write themselves, while others
had more complex periods of gestation; and interestingly, I find myself
fonder of some than of others, although not always correlated with the
presence/absence of labor pains. As with Volume I, it is hoped that the
whole of Volume II, via interrelated issues and connective tissue, may
stand as more than the sum of its parts. More than anything, I also
hope the point has been made that use of the full range of archaeo-
logical, art historical, and textual data, in conjunction with appropriate
interpretive models, can open windows into an understanding of the
Mesopotamian past—which, as a direct ancestor of our current heritage,
has participated as well in the construction of the present.
The articles included here, and the questions raised in them, in no
way claim to bring closure to understanding. If they have helped or
will help pave the way for the next cycle of questions to be asked and
data to be used, if the works they discuss are deemed to be as compel-
ling and deserving of attention as I believe is warranted, then I shall
be very pleased, indeed.
ACKNOWLEDGMENTS
This collection of essays could not have been put together without the
constant support and patience of Brill USA, particularly Michael J.
Mozina, Jennifer Pavelko and Michiel Klein Swormink. I would also
like to thank Marian Feldman for suggesting to Brill that the project
be undertaken in the first place, as well as Nadav Na’aman and Kapila
Vatsyayan for having offered encouragement throughout. I am espe-
cially grateful to all of the institutions and individuals cited in specific
articles for permission to reproduce once again illustrations used in the
original publications, or for offering substitutions when necessary. The
constancy of Robert Hunt and Piper O’Sullivan has been essential
at many levels in the production of the present incarnation. My grati-
tude to all.
SCULPTURE AND THE EARLY STATE
CHAPTER SIXTEEN
AFTER THE BATTLE IS OVER: THE STELE OF THE VULTURES

AND THE BEGINNING OF HISTORICAL NARRATIVE
IN THE ART OF THE ANCIENT NEAR EAST
To raise the question of the nature of narrative is to

invite reflection on the very nature of culture. . . .
Hayden White, “The Value of Narrativity . . . ,” 1981
Definitions of narrative, generally falling within the purview of liter-

ary criticism, are nonetheless important to art historians. From the
simplest starting point, “for writing to be narrative, no more and no
less than a teller and a tale are required.”1 Narrative is, in other words,
a solution to the problem of “how to translate knowing into telling.”2 In
general, narrative may be said to make use of third-person cases and
of past tenses, such that the teller of the story stands somehow outside
and separate from the action.3 But what is important is that narrative
cannot be equated with the story alone; it is content (story) structured
by the telling, for the organization of the story is what turns it into
narrative.4
Such a definition would seem to provide fertile ground for art-his-
torical inquiry; for what, after all, is a painting or relief, if not content
ordered by the telling (composition)? Yet, not all figural works “tell” a
story. Sometimes they “refer” to a story; and sometimes they embody
an abstract concept without the necessary action and setting of a tale at
* This article originally appeared as “After the Battle is Over: The ‘Stele of the
Vultures’ and the beginning of Historical Narrative in the Ancient Near East,” in
Pictorial Narrative in Antiquity to the Middle Ages [CASVA/Johns Hopkins Symposium in the
History of Art], H. Kessler and M. S. Simpson, eds., Washington, D.C.: National Gal-
lery, l985, pp. 11–32.
1
Robert Scholes and Robert Kellog, The Nature of Narrative (London, 1966), 4.
2
Hayden White, “The Value of Narrativity in the Representation of Reality,” in
On Narrative, ed. W. J. T. Mitchell (Chicago, 1981), 1.
3
Gérard Genette, “Boundaries of Narrative,” New Literary History 8 (1976), 8.
4
Irene J. Winter, “Royal Rhetoric and the Development of Historical Narrative in
Neo-Assyrian Reliefs,” Studies in Visual Communication 7 (1981), 2.
4 chapter sixteen
all. For an investigation of visual representation, it seems important to

distinguish between instances in which the narrative is vested in a verbal
text—the images serving as but illustrations of the text, not necessarily
“narrative” in themselves, but rather references to the narrative—and
instances in which the narrative is located in the representations, the
story readable through the images.
In the specific case of the ancient Near East, instances in which
narrative is carried through the imagery itself are rare, reflecting a situ-
ation fundamentally different from that found subsequently in the West,
and often from that found in the further East as well. We do not have
a cultural situation in which monolithic religious/mythological/heroic
texts stand behind the very fabric of society—as the Bible does for the
Judeo-Christian West, the Odyssey and the Iliad for the classical world,
or the Ramayana and the Bhagavad Gita for the Indian subcontinent.
Nor do we have a tradition of inscriptions directly associated with and
therefore identifying mythological/religious images. Through juxtaposi-
tion with text, as in illustrated books or through labeling, as on Greek
vases, for example, images in later Western art at least are far more
likely to be accurately identified; and these identifiable images then
provide a basic corpus from which to argue for the unknown.5
The problem is further exacerbated by the tendency in the art of
the ancient Near East, as discussed by Ann Perkins in the 1955 Chi-
cago symposium on narrative in ancient art, to be “allusive” rather
than explicit, with the “culminating scene” of a given story standing
for the whole.6 We often find the story behind the image, therefore,
demanding the viewer’s prior knowledge and correct identification
of the scene—a process of “matching” rather than “reading” of the
imagery itself qua narrative.
However, there is one class of work in the corpus of ancient Near
Eastern art—the battle scene, executed primarily in sculptural relief—
that does fall into our definition of pictorial narrative. In content, these
works tend to refer to specific events in time and place; to contain
“action” as it has been distinguished from “description”; and to be
5
How lucky we are, for example, that Euphronios tells us his dying warrior is Sarpe-
don on the Metropolitan Museum krater, so we do not have to waste scholarly energy
arguing whether or why it is not the dead Achilles or the suicidal Ajax instead!
6
Ann Perkins, “Narration in Babylonian Art,” American Journal of Archaeology 61
(1957), 55.
after the battle is over 5
“told,” as if in the third person and in the past tense.7 The subject
matter of these monuments is neither religious nor mythological, but
rather historical. Whether conveyed through the juxtaposition of succes-
sive episodes,8 or in a single image that nonetheless through its action
enables one to “read” the event, these reliefs can by our definition be
called “pictorial narrative,” as we are not required merely to associ-
ate through them to an underlying text or tale. This is not to say that
such works are never linked to texts, since many of the monuments are
either accompanied by inscriptions carved directly upon the stone, or
else can be related to parallel texts, through which our understanding
of the narrative is clarified. Yet they are not dependent upon the texts
for their reading. In the very depiction of historical events, as opposed
to ritual or mythological events, we are presented with a complex inter-
nal sequence and development, through time and across space, which
permits us to “read” the monument itself.
The battle narratives of the Neo-Assyrian period, from the ninth
to the seventh centuries b.c., are perhaps the best known, and have
been the subject of an earlier study, the results of which linked the
exploitation of this particular genre directly to the contemporary
political situation, as Assyrian kings used representations of conquest to
validate the territorial expansion of the empire.9 Comparison with con-
temporary annalistic texts describing actual battles established an exact
correspondence between battles reported and battles depicted. In this
isomorphism between the verbal and the visual, we are presented with
a highly sophisticated and integrated system, in which both means are
expertly manipulated toward the same rhetorical and political ends.
It is therefore of interest to look back to the very beginning of the
sequence, to the late Early Dynastic period (E.D. III) c. 2500 b.c., in
which conditions surrounding the consolidation of the city-state had
given rise to the first monument that could be identified as public,
historical, and narrative: the large, freestanding relief known as the
“Stele of the Vultures,” commemorating a victory of the city-state of
Lagash over the neighboring state of Umma (figs. 1, 2).10
7
White, “The Value of Narrativity,” 3; Genette, “Boundaries of Narrative,” 8.
8
Perkins, “Narration,” 55.
9
Winter, “Royal Rhetoric,” 31.
10
E. de Sarzec, Découvertes en Chaldée II: Epigraphe et planches (Paris, 1884–1912), pls.
3, 3 bis, 4, 4 bis, 4 ter.
6 chapter sixteen
My aim is fourfold: to see whether this monument truly fits the

definition of “narrative” presented above; to bring forth the particu-
lar historical and cultural conditions that produced it and the issues
embodied therein; to determine whether the same degree of integration
between text and image prevailed as in the Neo-Assyrian period and
whether the same degree of political ends were being served; and
finally, to see if it is possible to link this early example to later works
in the history of art by elucidating the uses (and abuses) of public
monuments on which historical narrative appears. For it should be
clear from the outset that the “historicity” of these monuments was a
highly manipulated one. In the Assyrian case, for example, only the
enemies die; Assyrians never lose, never perish. The purported reality
is thus carefully organized to “bring out the central mythos” of the
state;11 and even where events depicted are in fact true, the very fact
of their having been recorded selects the event, its antecedents, and its
consequences as meaningful.
* * *
The Stele of the Vultures, currently in the Louvre, is rather poorly
preserved. Originally it was a large, rectangular slab of white limestone
with a rounded top. The modern restoration of the whole is based upon
only seven fragments. Six were excavated during the 1880s at the site
of modern Tello, ancient Girsu, a satellite town within the city-state of
Lagash in the Early Dynastic period and seat of the deity Ningirsu.12
The seventh fragment, undoubtedly plundered from the site, was
acquired by the British Museum in 1900 and later given to the Louvre
for the restoration.13 As restored, the monument measures 1.80 meters
in height, 1.30 meters in width, and is .11 meters in thickness.14
11
Northrop Frye, “History and Myth in the Bible,” in The Literature of Fact, ed. A.
Fletcher (New York, 1976), 1–19 and especially 7–11.
12
Leon Heuzey, “La Stèle des Vautours,” Gazette archéologique (1884), 164–180 and
193–203; Andre Parrot, Tello, Vingt campagnes de fouilles (Paris, 1948) 95; Louvre, AO 50
+ 2436–2438. That Tello, often identified as Lagash in the early literature, is actually
ancient Girsu has been demonstrated in more recent excavations: cf. Donald P. Hansen,
“Al-Hiba, 1968–9: A preliminary report,” Artibus Asiae 32 (1970), 244.
13
Parrot, Tello, 95; the piece was British Museum 23580, now Louvre AO 16109.
14
See Leon Heuzey and François Thureau-Dangin, Restitution matérielle de la Stèle des
Vautours (Paris, 1909), especially pls. I, II. A recent study by Marie-Therèse Barrelet,
“Peut-on remettre en question la restitution matérielle de la Stèle des Vautours?”
Journal of Near Eastern Studies 29 (1970), 233–258, has argued that the lowest register
on each side might be compressed a bit, but not more than 10–15 cm. in all. On the
basis of our restoration drawing, presented here, this would seem to be highly desirable
The stele is covered on both major faces with carving in relatively high,
well modeled relief; the scenes of the reverse actually wrap around the
two shallow sides as well. The negative space is filled with an inscription
that is incised continuously from obverse to reverse. Much of the visual
contents of the stele can be reconstructed, though there are a few
significant losses. The inscription, however, has major gaps. We know
from the extant portion that the stele was executed by one Eannatum,
ensí (city ruler or governor) of Lagash in c. 2460 b.c. Eannatum pres-
ents the historical background, events, and consequences of his border
conflicts with an unnamed ruler of Umma, the state immediately to
the northwest of Girsu, in a dispute over the Gu’eden, the fertile tract
of irrigated land between the two states.15 Thus, from the text, as well
as from other contemporary references to the same conflict, we have
the historical details to help anchor the visual representations on the
monument. We are therefore in a position to approach the problem of
pictorial narrative both from a reading of the imagery itself and from
a comparison of text and image: that is, we can determine not only
what tale is being told, but also how the tale is told.
The obverse of the stele is divided into two registers, the upper
portion almost twice the height of the lower (figs. 1, 3).16 A large male
figure, centrally placed, holds a mace in his right hand, the mace-head
touching the bald head of a small man who protrudes from a large
net further to the left of the central figure. The victim is nude, and
from the point of view of scale and proportion in relation to the upper registers,
especially for the reverse.
15
The text was originally read and published by Thureau-Dangin, see Heuzey
and Thureau-Dangin, Restitution matérielle, 42–63, and until recently had not been
fully restudied. Thorkild Jacobsen published a new reading, heavily restored for the
missing portions, of the first ten columns in 1976: “Translation of the Stele of the
Vultures, Cols. I–X,” in Samuel Noah Kramer Anniversary Volume (Alter Orient und Altes
Testament 25), ed. Barry N. Eichler (Neukirchen-Vluyn, 1976), 247–259. See now,
however, full edition in Horst Steible and Hermann Behrens, Die altsumerischen Bau-
und Weihinschriften, Teil I: Inschriften aus ‘Lagas’ (Wiesbaden, 1982), 120–145; and, based
upon that, an English translation by Jerrold S. Cooper, in Reconstructing History from
Ancient Inscriptions: The Lagash-Umma Border Conflict (Sources from the Ancient Near East,
vol. 2, fascicle 1) (Malibu, 1983), 45–47.
16
As opposed to Henri Frankfort, The Art and Architecture of the Ancient Orient, 4th ed.
rev. (New York, 1977), 71; Henrietta Groenewegen-Frankfort, Arrest and Movement: An
Essay on Space and Time in the Representational Art of the Ancient Near East (London, 1951),
158; and Perkins, “Narration,” 58—all of whom call it the reverse. But that cannot
be, as the inscription begins on this side, and the text breaks at the bottom, to con-
tinue on the other side. Further arguments in structure will be presented below to
corroborate this visually.
8 chapter sixteen
the uppermost of what would have been at least a dozen other cap-
tives, densely packed and tumbled about within the net.17 With his left
hand posed just above the top of the net, the central figure grasps the
tail feathers of a frontal lion-headed eagle with outstretched wings, its
talons visible below and resting upon the backs of two addorsed lions
(fig. 4).18 This emblem, known from mythological texts, represents the
anzu, or Zu-bird—associated with the thunderstorm—who was identi-
fied with the god Ningirsu in the early periods. Later texts record that
the anzu absconded with the “Tablet of Fate” until Enlil, chief god of
the Mesopotamian pantheon, ordered the hero Ninurta, a later form
of Ningirsu, to retrieve them.19 Like the Gorgon head for Perseus, this
emblem then became associated with, and was adopted as an attribute
of, the vanquisher.
It has been a point of debate whether the standing male on the
obverse of the stele should therefore represent the deity Ningirsu him-
self, to whom the stele is dedicated and in whose name the victory over
Umma was won; or whether it should represent Eannatum, favored
by Ningirsu, who inscribed the stele and literally effected the victory.20
17
The net is a well-known literary topos, clearly reflecting actual battle practices, as
Marduk in a late version of the Creation epic catches his adversary Tiamat in a net
(see James S. Pritchard, Ancient Near Eastern Texts Relating to the Old Testament, 3rd ed.
(Princeton, 1969), 67. Furthermore, in the second half of the inscription on the stele
itself, the vanquished ruler of Umma is made to swear by the “battle-nets” of various
gods that he will not transgress the reestablished border (compare below).
18
In the present state of preservation of the stele and in recent photographs, this
head is broken, and it has been questioned whether or not it was originally a lion’s
head or simply a normal eagle’s head. I have checked the stele, and can clearly see
at both top and bottom the double-curved line that would have been the outline of a
leonine muzzle. In addition, Agnès Spycket of the Département des antiquités orien-
tales, Musée du Louvre, who kindly made the stele accessible to me, pointed out that
the stele had unfortunately suffered damage since its excavation. When first brought
to the Louvre, more of the main figure’s eye and brow, as well as more of the head
of this emblem, were preserved than at present. On a postcard of c. 1910 shown me
by Mile. Spycket, the lion head of the bird held in hand is clearly visible; hence our
reconstruction in the present drawing, fig. 3. I believe that the same outline of a lion’s
head is apparent upon examination of the emblem atop a standard at the far left of
this register; however, here it is possible that reference is to the šir-bur bird (raven?)
sometimes associated with the goddess Ninhursag.
19
Jean Nougayrol, “Ningirsu, vainquer de Zu,” Revue assyriologique 46 (1952), 87–97.
See also, Ilse Führ-Jaeppelt, Materialen zur Ikonographie des Löwenadlers Imdugud-Anzu
(Munich, 1972); B. Husk, Der Mythenadler Anzu in Literatur und Vorstellung des alten Meso-
potamien (Budapest, 1975).
20
See, for example, Perkins, who calls him Eannatum (“Narration,” 58), vs. Frank-
fort, who calls him Ningirsu (Art and Architecture, 158), as does Anton Moortgat, The Art
of Ancient Mesopotamia (London and New York, 1969), 63.
The anzu emblem appears on several “royal” monuments found at

Girsu, however those that are inscribed are all dedicated to Ningirsu
(for example, the silver vase of Enmetena, nephew of Eannatum and
one of his successors, fig. 5, or the votive plaque of Dudu, chief priest
of Ningirsu under Enmetena).21 No royal figure ever actually wields
the emblem.
This would suggest that the main figure on the obverse is in fact the
city god. The question can be settled, I believe, through identification
of the second figure in the scene, smaller in scale and standing as if
behind or to the right of the central male. This smaller figure is female,
as can be surmised from what is visible of her hair. She wears a head-
dress of splayed (cow?) horns, between which are feathers or fronds
flanking a central leonine head similar to that of the anzu emblem held
by the male figure. Behind and to the left of this woman the emblem
is repeated, raised like a standard on a pole. Three maces protrude
from each of her shoulders, the traditional Mesopotamian manner
of showing divine attributes.22 Parallels in headgear and attributes
between this and other figures—as for example on a basalt vase frag-
ment where the figure is identified by inscription23 and on a limestone
votive plaque (fig. 6) where the mountain scales upon which the female
figure sits are a clue to her Sumerian name—allow us to identify this
figure as a divine personage, specifically the goddess Ninhursag, Lady
of the Mountain. She is, further, the mother of Ningirsu, and played
a principal role in the myth of the capture of the anzu, by giving her
son essential strategic advice.24 Because of this role, the anzu emblem
is associated also with the goddess, appearing on a large copper panel
found in her temple at the site of al-’Ubaid (fig. 7).25 This would explain
the presence not only of the emblem in her headdress, but also of the
standard at her side.
It seems clear, therefore, that the figure in the upper register on the
front of the stele represents the city god, Ningirsu, with his emblem,
21
Frankfort, Art and Architecture, fig. 70; Moortgat, Ancient Mesopotamia, figs. 113,
117.
22
See, for example, Henri Frankfort, Cylinder Seals (London, 1939), pls. XVIII, XX,
XXV; and Barrelet, “Peut-on remettre . . . ?” 244.
23
See Moortgat, Ancient Mesopotamia, figs. 114, 115: votive tablet and vessel fragment,
also from Tello, the latter inscribed by Enmetena, nephew of Eannatum and a later
ruler of Lagash (compare Steible and Behrens, Bau- und Weihinschriften, 249 for text).
24
Nougayrol, “Ningirsu,” lines 73ff.
25
Frankfort, Art and Architecture, fig. 63.
10 chapter sixteen
triumphant over the vanquished enemy of Lagash and accompanied

by his mother, a primary figure in his mythical history and a deity
venerated in her own right both in Lagash and in Girsu.26
Although very poorly preserved, the lower register seems to corrobo-
rate this interpretation. It is evident that a chariot was depicted, moving
toward our right; facing the chariot is a repetition of the goddess in the
upper register. Further, it is possible to determine from a bit of visible
skirt that the god himself was standing in the chariot. The wingtip of
the anzu emblem is also visible; it was either held in the god’s hand, or
attached to the protective chest board of the chariot.
Although the draft animals for the chariot are not preserved, we have
accepted Barrelet’s argument that they were more likely to have been
mythological animals than standard equids of the period (as depicted,
for example, on the reverse side of our stele, see below and fig. 11; or
on the “Standard of Ur,” fig. 13). The argument is plausible, first, for
reasons of available space in relation to the yoke pole; and second,
because it would fit better with the mythical character of this side of
the stele, particularly as Ningirsu was said to have harnessed the winds
to his chariot in pursuit of the anzu. In this way, the draft animals
would have contributed to the impression that the power invoked was
supernatural.27
Thus, the whole face of the obverse seems to center around the per-
sonage of Ningirsu, patron deity of the city of Girsu, to whom the stele
was dedicated. The lower register seems to be less devoted to showing
the action described in the inscription on the stele than to setting the
stage for that action by providing attributes and antecedents of the city
26
Adam Falkenstein, Inschriften Gudeas von Lagash I: Einleitung [Analecta Orientalia
30] (Rome, 1966), 104–105.
27
See on this the excellent arguments for restoration on which our drawing is
based, in Barrelet, “Peut-on remettre . . .?” 245–249, 249–254 and figs. 10a, 10b. For
the draft animals, we have followed the subsequent suggestion of Mary A. Littauer
and Jost Crouwel, “The Vulture Stela and an Early Type of Two-wheeled Vehicle,”
Journal of Near Eastern Studies 32 (1973), 324–329, that they must be harnessed in pairs.
While this reconstruction remains problematic, it is not inconsistent with the attested
character of the god. In a later, Neo-Assyrian version of the myth of the Anzu, when
Ningirsu of the Sumerians has been assimilated with Ninurta of the Assyrians, we are
also told that the god’s chariot was decorated with emblems of his victories (compare
Erica Reiner, “Le char de Ninurta et le prologue du mythe de Zu,” Revue assyriologique
51 (1957), 109, and Jerrold S. Cooper, The Return of Ninurta to Nippur: An-gim dím-ma
[Analecta Orientalia 52] (Rome, 1978), 141–154 and especially chart, 143), which
would be quite consistent with having the anzu emblem and the lion represented here
upon the chariot box and yoke pole.
god. It is important to note that no mention of the Zu-bird or of the

chariot of Ningirsu is made in the text on the stele. All these references
come from texts and stories external to the monument, which were
nonetheless likely to have been well known to the residents of Girsu.
Thus, prior knowledge on the part of the viewer would have been
required, to which the representation stood in a referential status.
In the upper register, the god is shown with the vanquished enemy
already in hand. It is not the action itself, but the consequence of action
that is depicted. The forces of Girsu have won. And we have here the
symbolic statement: “after the battle is over.”
In this respect, the obverse of the stele does not conform to our
definition of pictorial narrative. At best, it can be called a “culminat-
ing scene,” because the action is completed. In fact, however, it is
really a summary, a symbolic referent, couched in mythical terms, of
divine intervention, and thereby verging on the “iconic”—intended to
convey a total, abstract concept of divine patronage and favor. Behind
the imagery lie the mythological history and associations of the deity,
which are alluded to but in no way acted out. Hence, this face of the
stele at least is without “action”; without “telling.” It is the icon of the
city and its victory, not its narrative.
* * *
The reverse face of the stele is divided into four registers (figs. 2, 8).
The vultures that give the modern name to the monument are seen
at the upper right, holding in their beaks the severed heads of the
Ummaite enemy of Lagash (see detail, fig. 9). Its fragmentary condi-
tion notwithstanding, the image is a powerful one: ultimate defeat of
the vanquished, absolute success for the victors. The birds of prey fly
above horizontal bands of inscription, so that the upper part of the
stele seems separate from and subsequent to the rest of the action
in the upper register. The main action seems to be that of the battle
itself. From the left edge to virtually the center, a phalanx of twelve
shielded soldiers (nine on the reverse face, three wrapped around the
side) point their spears in attack position as they trample the bodies
of their fallen enemies (fig. 10).28 The soldiers are aligned as if march-
ing behind the figure of the king, who also faces right, wrapped in a
28
On both the first and second registers, the soldiers visible on the reverse are a
continuation in an unbroken line of figures beginning on the shallow left side (11 cm.
thick). Allowing for missing figures, with those on the side we can account for twelve
12 chapter sixteen
flounced garment probably made of sheepskin. The particular action

in which these figures are engaged has been lost in the break; but we
may assume it was a direct military encounter, since at the far right-
hand edge of the register the bodies of some thirteen fallen enemies
are piled in a heap.
In the second register, a group of soldiers again appears at the far left
(fig. 11). This time, however, there are no raised shields. Each soldier
holds a long-shafted spear in his right hand and a battle-ax in his left;
but both weapons are raised in march rather than in attack position.
Here, the soldiers follow behind the royal war chariot. The king holds
the shaft of a spear in his left hand. Unfortunately, both the tip of the
spear and the object of the king’s attention are missing. What is visible
of the chariot itself indicates that it is significantly different from the
one on the obverse.29 This would suggest that a careful distinction was
made between the two, and thereby provides further support for the
argument that it is indeed the god’s vehicle on the front.30
In the third register, the action seems directed from right to left (figs.
8, 12). The central figure faces left; only his feet and a portion of his
flounced skirt are visible; the feet rest on a platform or low socle. In
an unpublished paper, Michelle Marcus has argued that the figure was
seated;31 this reconstruction would conform with seated figures known
from other contemporary reliefs and sculpture.32 In front of the seated
figure, a bull, presumably tethered for sacrifice, lies on its back. In the
field above the bull are a stack of bodies of smaller animals, perhaps
goats, and two large vases containing “sacred” plants with vertical leaves
figures in each register behind the king; thus, an equal number in each register, despite
the apparent inequity of what is visible on the reverse face alone.
29
The equids and the chariot box have been restored on the basis of other Early
Dynastic period representations of chariots and draft animals—for example, on the
Standard of Ur, fig. 13 (see notes on this in Littauer and Crouwel, “The Vulture Stela,”
325–328). The ridge pole is less curved than that on the chariot of the obverse, as if
attached to taller animals and aimed directly at their necks. This would further fit the
restoration of equids here (ass or onager), and possibly also support the argument for
different (mythological?) animals on the obverse.
30
Note also that what seems to be an arm holding a goad is visible behind the king
in the chariot box. This would signify a second passenger in the vehicle, not unlike
the additional figures shown on the running boards of chariots on the Standard of
Ur, fig. 13.
31
Written at the University of Pennsylvania, January 1982.
32
See Moortgat, Ancient Mesopotamia, figs. 114, 116: stele with libation to seated
goddess Ninhursag, from Tello, and plaque with libation poured to seated male god
(Nanna?), from the giparu at Ur.
and drooping date clusters. The bases of these vases are obscured by the
body of the bull, suggesting spatial depth. A naked priest, facing right,
stands upon the stack of animals; he holds the bottom of a spouted
vessel well known from other libation scenes (figs. 6, 16), as the stream
of his poured liquid falls upon the potted plants.33 Finally, at the far
left of the register, naked bodies are stacked in a large mound, outlined
by a double line as if it were a ladder seen in profile or some sort of
covering shown in section. Skirted workers with filled baskets on their
heads hold onto this band and climb up the mound of bodies. Such
climbing figures with earth-filled baskets are the traditional means in
Mesopotamia of representing construction workers, especially in temple-
building scenes.34 Here, they are clearly making burial mounds of earth
poured over the stacked bodies.
The burial-mound, sacrificial animals, and ritual libation all raise
the question of the identity of the seated figure who presides over this
activity. Frankfort has called him the god Ningirsu; Moortgat would
see him rather as Eannatum attending a ritual ceremony.35 I would
agree with the latter interpretation, and have so restored the figure in
the drawing (fig. 8). First, the flounced skirt is the same as that worn
by Eannatum, who is clearly labeled, on the rest of this side of the
stele and different from the skirt of the god on the obverse; second,
other rulers of Lagash wear this same skirt (for example, the stand-
ing statues of Eannatum I and Enmetena, brother and nephew of
Eannatum, respectively, and his successors; and the relief plaque of
Ur-nanshe, grandfather of Eannatum and founder of the dynasty, fig.
15).36 Closest of all is the figure on the Ur-nanshe plaque, where both
the flounced skirt and the seated posture are identical to those of the
primary figure. Moreover, I shall argue below that the designation of
Eannatum in this position fits far better with the suggested reading of
the narrative on the reverse.
33
Compare with this, Moortgat, Ancient Mesopotamia, figs. 114, 116, cited above,
n. 32.
34
For example, on the plaque of Ur-nanshe, our fig. 15, and on the Stele of Ur-
nammu in the Ur III period, c. 2100 b.c. (Moortgat, Ancient Mesopotamia, fig. 194).
35
Frankfort, Art and Architecture, 71; Moortgat, Ancient Mesopotamia, 43.
36
See Moortgat, Ancient Mesopotamia, figs. 85, 87, 88, 109, 111, 112. See also the
seated statue of Dudu, priest of Ningirsu under Enmetena, whose body, flounced skirt,
and seated posture are the model for our reconstruction figure: Eva Strommenger,
500 Years of the Art of Mesopotamia (New York, 1964), fig. 86.
14 chapter sixteen
Only a small portion of the fourth register is preserved, the upper

part, in which one sees a hand at the far left grasping the butt of a
long spear shaft, the tip touching the forehead of a bald enemy near
the center of the band. The enemy faces the oncoming spear; his head
emerges from a group of three additional bald heads before him, all
facing in the opposite direction. The relationship of this cluster of
figures to the menaced figure and the action is not clear. On the basis
of analogies with registers one and two, the figure wielding the spear
is presumably royal. As there is no room for a chariot at the far left,
Marcus has suggested that this ought to be a standing king. Whether
the three bald figures merely flee from the king, or deliberately sur-
round his marked enemy, cannot be determined; whether the interaction
constitutes real battle or just a threat is equally unclear.
Despite the poor state of preservation, we can attempt a reading of
the reverse. Each of the four registers—whether showing battle or ritual
engagement—depicts some contained action and can be individually
read as a narrative, given our requisite of event told, not just referenced.
More intriguing is the question whether, and then how, the registers
may be read successively, in a coherent sequence.
The reverse side of the stele has generally been described preceding
from top to bottom, on the assumption that it was meant to be read
that way: the story proceeding from a battle and heap of enemy bodies
(register 1), to a continuation or second battle (register 2), to a victory
celebration and sacrifice (register 3).37 A major problem is then posed
by the fourth register, in which the battle seems to be resumed after
the celebration/ritual. Discussion has focused upon an early identifi-
cation of the enemy facing the spear thrust as the “King of Kish,”
based upon the proximity of his head to the mention of a king of that
state in the inscription. This problem requires a lengthy discussion not
germane to the present study, but I would argue that the association is
fortuitous, our vision skewed by the unfortunate breakage of the end
of the text.38 Nevertheless, whether or not this is the king of Kish, a
37
Perkins, “Narration,” 58; Frankfort, Art and Architecture, 71; Donald P. Hansen,
“Frühsumerische und frühdynastische Flachbildkunst,” in Propyläen Kunstgeschichte,
Vol. 14, ed. W. Orthmann (Berlin, 1979), 189.
38
In brief, I would argue on typological grounds that this is not a cartouche iden-
tifying the adjacent figure (as is the case with labels for the figure of Eannatum on the
reverse, registers one and two, see fig. 11); rather, it is part of the main body of the
text itself (see fig. 12). Mention of the king of Kish in this place may be understood
in the light of another text, an inscription on a stone mortar dedicated by Eannatum
question remains: why, after sacrifice and celebration, would Eannatum

resume battle with anyone, much less with an enemy not recorded as
such on the stele, when the text describes the conflict with Umma,
whose dead are already heaped up above?
Since this problem cannot be resolved with a vertical reading from top
to bottom, I would propose an alternative reading of the reverse of the
stele, from bottom to top. This proposal is based first on the evidence
from more complete Sumerian monuments where the preferred order
of reading is indeed from bottom to top; and second, on the narrative
sequence in the accompanying text of the stele.
In the first instance, I have argued elsewhere that the major relief
monument of the preceding Uruk period, the cult vessel known as the
“Warka Vase,” is best understood when read from the lower registers
(cultivated plants and domesticated animals) toward the top (cult scene
before the goddess Inanna).39 Certainly the nearly contemporary “Stan-
dard of Ur” from the Royal Cemetery (figs. 13, 14) is to be read this
way. On one side of the standard, three horizontal registers progress
from battle chariots at the bottom to the gathering of prisoners in
the middle to the presentation of the prisoners before a larger central
figure, presumably a ruler, at the top. On the other side, the scenes
proceed from the amassing of pack animals and goods in the lower
register to the procession of food animals and men bearing fish in the
middle to the banquet in the upper register that again is dominated by
a slightly larger figure in a flounced skirt, probably the ruler. In fact,
the dominant, primary position of the ruler at the center of the upper
register on the battle side, the culmination of the narrative sequence,
is comparable to the position of Eannatum in the upper register of
the Stele of the Vultures.
In the second instance, although it will be pursued further below,
we may now consider the text inscribed on the stele, and surrounding
most of the pictorial action. If only the imagery is read, it is difficult
(Steible and Behrens, Bau- und Weihinschriften, 174–175), where the reference is part
of a string of protective curses that traditionally end texts on important monuments.
It is precisely that part of the Stele of the Vultures that has been lost, but we may
hypothesize that the text would most likely have ended with just such curses against
the destruction of the monument, and then a final citation of Eannatum, his titles, and
his special relationship with/dedication to the god Ningirsu.
39
Irene J. Winter, “The Warka Vase: Structure of Art and Structure of Society
in Early Urban Mesopotamia,” paper presented to the American Oriental Society,
Baltimore, 1983. For the image, see Moortgat, Ancient Mesopotamia, figs. 19–21.
16 chapter sixteen
to see the ritual burial and libation as anything other than a postbattle
celebration of victory.40 However, when the imagery is read in con-
junction with the text, we see that before Eannatum actually goes off
to battle with Umma, he enters the temple to receive a dream oracle
from Ningirsu, in keeping with the common practice in Mesopotamia
for those seeking divine instruction.41 In the dream, it is foretold that a
mound of enemy corpses would reach up to the very base of heaven
(text, col. 7:21–22). This prediction is followed by a break of some
seven to eight lines, after which we find Eannatum has already taken
up arms against Umma (col. 8).
It is possible, therefore, to see the scene in the third register as
preceding rather than succeeding the battle. Although Eannatum is
described in the text as “the reclining one,” and in the third register
he is seated, this in itself poses no problem. There is no tradition for
representing reclining figures in Mesopotamian art, while Eannatum
would certainly have had to preside seated over cult performances
in the temple on the same occasion. A parallel depiction of a seated
figure referring to a dream sequence occurs on one of the well-known
statues of Gudea of Lagash in the later, Neo-Sumerian period. The
statue shows Gudea holding the plan of a temple on his lap, having
received instructions to build a new temple by a dream oracle, while
he slept in an older sanctuary.42
Everything represented in register three may thus be seen as anterior
to the battle(s), which itself can be best read in sequence from register
two, with soldiers in battle march, to register one, with soldiers in attack
position behind their ruler, and, rendered at the far right, the resultant
mound of corpses predicted in the temple oracle. The vultures carry-
ing off their prey at the top would then represent the very end of the
narrative. Their action is the narrative counterpart to the symbolic
statement of the obverse: the historical “after the battle is over.”
The problem of the bottom register remains—two apparent battle
sequences are still divided by a register of ritual activity. In my proposed
reading, this apparent anomaly can be explained in either of two ways:
40
It has even been suggested that this represents Lagash burying its own dead
(Frankfort, Art and Architecture, 71), though this seems to me a most unwarranted sup-
position not at all consistent with Mesopotamian traditions, in which only the enemy
are ever shown deceased or defeated.
41
See Leo Oppenheim, The Interpretation of Dreams in the Ancient Near East (Philadelphia,
1956), particularly pages 185–189 dealing with “message dreams.”
42
Oppenheim, Interpretation of Dreams, 211, 245.
first, as corresponding to the preamble of the text (cols. 1–3), which

lays out the previous history of the conflict between Lagash and Umma
under earlier rulers; or second, as corresponding to a portion of the
dream itself, where the promised corpses are after all the results of a
projected battle. The former seems more consistent with the progres-
sion through the action, and parallels the development of the text; but
either way, register four reads best if it is seen as setting the scene for
the subsequent action.43
This development works better in terms of composition as well. If
one reads the stele from bottom to top, the figure of the king seems to
move from the left-hand margin of the lower register to a point almost
exactly one-quarter of the way across the space of register two, to
dead center in register one. If, by contrast, one reads from top to bot-
tom, the king seems to be backing up. It would be far more logical to
see the king progressing steadily toward the culmination in the upper
register, where he occupies the same central position as the “ruler” on
the Ur “Standard.”
Thus, the action on the reverse of the stele can be read both as a
linear progression across each register (synchronically) and as a progres-
sion up the registers (diachronically). If we follow the story parallel to
the inscription, as suggested, the movement of the ruler’s figure is both
spatial and temporal. In contrast to the obverse of the stele, the reverse
conforms to the definition of “narrative” as established at the beginning
of this paper: organized action, progression through time, reference to
a specific event, and readability rather than mere reference.
If, then, we conclude that the reverse of the Stele of the Vultures
is “narrative,” and “historical narrative” at that, while the obverse is
not, we must articulate the ways in which the two faces work with one
another in the single monument.
Virtually everyone who has written on the stele has commented on
the opposition between the mythological and the historical aspects of
the respective sides.44 Perhaps most penetrating is Moortgat’s observation
that the stele reflects the dualism inherent in the “Sumerian age” in
43
Even if I am wrong about the identification of the enemy in the lower register and
it is the king of Kish (see above, n. 38), it should be noted that in the dream/oracle
of our text (Steible and Behrens, Bau- und Weihinschriften, 125: col. 7, lines 1–3) there
is an enigmatic reference: “As is Umma, so is Kish, running about. . . . ”
44
From de Sarzec, Découvertes, 357–358, to Hansen, “Frühsumerische . . . . Flachbild-
kunst,” 181.
18 chapter sixteen
divine acts and events of the mortal world.45 Groenewegen-Frankfort,

in turn, noted the separation of the “actuality” of the depicted battle
on the one side from its “transcendent significance” on the other.46
All commentators have concentrated on the content of the reliefs,
however. When we come to describe how that content is presented, it
becomes evident that different pictorial modes have been selected to
represent the two realms: what we have called the “iconic,” on the one
hand, and the “narrative,” on the other.
In this respect, we may note that the obverse is divided into only two
registers, one larger than the other, while the reverse is divided into four
more or less equal bands. The resulting effect is that the god on the
obverse occupies considerably more space than any other figure. He is
also bulkier in proportion and more nearly frontal in pose. Following
Schapiro’s analysis of the importance of “format” in establishing the
expressive parameters of an image, and his correlation between rank
and size, which assumes an equation between a scale of qualities and
a scale of magnitudes,47 the message is clear. Whatever human agents
have participated in the event(s), the end has been achieved only
through the god’s patronage and guidance. However, the god is not
shown engaged in action, but rather in a symbolic gesture of domi-
nance; the actual events are over. If anything, the relief presents what
Svetlana Alpers has called a deliberate de-emphasis or suspension of
narrative, achieved through a “fixity of pose.”48 Just as the god holds
his emblem, the anzu, he is himself an emblem of the state and of its
victory. The time represented on the obverse, therefore, follows that
of the reverse. But simultaneously it is the summary statement of the
whole. For this reason, the inscription begins on the god’s side, and
ends on the reverse making clear the weighting of the two faces. We
are given important visual clues via composition, pose, and attributes to
serve as aids in deciphering the obverse. But because it requires prior
knowledge to identify the god through his emblems and his association
with the accompanying female deity, and a background in mythology
45
Moortgat, Ancient Mesopotamia, 43.
46
Groenewegen-Frankfort, Arrest and Movement, 158.
47
Meyer Schapiro, “On Some Problems in the Semiotics of Visual Art: Field and
Vehicle in Image-Signs,” Semiotica 1 (1969), 234–236.
48
Svetlana Alpers, “Describe or Narrate? A Problem in Realistic Representation,”
New Literary History 8 (1976), 16.
to fully understand the visual references, the viewer must control more
of the “code” for the obverse to be understood.
The reverse, by contrast, can be read more easily through the depicted
events. Each register seems to have contained a coherent and complete
action, achieving closure of the part before it was to be integrated as
a segment into the whole and read in sequence. The technique of
superimposed registration is not uncommon in later art. As Brilliant
has pointed out, it is one means of transcending the limits of the reg-
ister system, through stacking in a loaded sequence, where the whole
equals more than the sum of its parts.49 On the Stele of the Vultures
as well as on the “Standard of Ur,” we are very close to the beginnings
of this system. In the particular stacking of the four registers on the
reverse of the Stele of the Vultures, we can follow the sequence of the
conflict between Umma and Lagash to the ultimate victory proclaimed
by Lagash. The story unfolds directly as narrative within the pictorial
frame, as a progression of individual actions and their sum total, which
equals success. On the obverse, by contrast, the story lies behind the
pictorial—the “referential” as opposed to the “told.”
Yet this opposition cannot be dismissed as a simple division between
obverse and reverse; for on each side, the dominant mode of representa-
tion also contains within it an echo of its opposite. On the obverse, for
example, though the god is the “icon,” there is an obvious reference to
the “historical narrative” and the political consequences of the battle
won in the net he holds. On the reverse, as one proceeds from bottom
to top through the “historical narrative” of the conflict, the figure of
the ruler as well becomes an effective “icon,” which through repetition
establishes a rhythmic pattern as he ultimately achieves center stage
at the top.
Thus, the ambiguities posed by the opposition are resolved in the
interplay of main and subsidiary themes. The stele is clearly an attempt
to integrate myth with history;50 one in which different modes of
49
Richard Brilliant, Visual Narratives: Storytelling in Etruscan and Roman Art (Ithaca and
London, 1984), 23–26.
50
A not unusual phenomenon: see Scholes and Kellog, Nature of Narrative, 135. In
this, as well as in the role of repetition in creating the “icon” of the king, I am indebted
to Richard Brilliant’s discussion of the repetition of the figure of Trajan on his column
(Visual Narratives, 115–116), and to Wesley Trimpi’s discussion of Aristotle’s comments
on the repetition of personal names in public oratory so that people will think the
individual more significant (“The Meaning of Horace’s Ut Pictura Poesis,” Journal of the
Warburg and Courtauld Institutes 36 [1973], 4).
20 chapter sixteen
representation are used, while neither mode is totally separate from the
other. In this bifacial union of myth and history displayed through the
pictorial modes of icon and narrative, a complex dynamic has been
established between the telling and the already told. The power of the
stele lies in that union, and clearly, both components are required for
the monument to achieve its power.
* * *
A few issues remain which must be considered in order for us to
properly place the Stele of the Vultures into the context of the begin-
nings of pictorial narrative on the historical monuments of the ancient
Near East. First, how did the text inscribed on the stele relate to the
imagery (a necessary question in the evaluation of the overall impact
of the monument)? Second, what is the nature of the stele itself as a
public monument? Third, what was the relationship between the his-
torical event commemorated on the stele and the general history of
the city-state of Lagash during the Early Dynastic period? For these
three issues are crucial to our understanding of the essential questions:
Why narrative? Why historical narrative? Why at that time? It is only
through consideration of these questions that we can reconstruct the
impact of the visual message we are calling pictorial narrative on this
particular stele in this particular period.
The Text
In the incorporation of a lengthy text along with figured relief, the Stele of
the Vultures marks a significant departure from such earlier monuments
as the plaque of Ur-nanshe on which the text consists simply of crudely
incised labels and epithets (fig. 15). Eannatum is identified by label twice
on the stele, in the field adjacent to his own head, much the same as
Ur-nanshe is on his plaque. The rest of the inscription, however, consists
of a continuous text that begins in the upper field of the obverse and
continues in parallel bands down the entire face. It then picks up at the
top of the reverse and continues down that face, even across the raised
bands that separate registers (figs. 1, 2, 9, 10, 11). As on the plaque of
Ur-nanshe, the figures carved in raised relief were executed first. The
incised columns and subdividing “cases” with their enclosed signs were
added later, over virtually all the available background space.
The text was first read by François Thureau-Dangin not long after
its discovery.51 A new edition of the text is included in the collection
of Old Sumerian inscriptions published by Steible and Behrens, and
an English translation and discussion forms part of Cooper’s study of
texts pertaining to the Lagash-Umma border conflict, both containing
full bibliographies of intervening studies.52
In short, the text begins with what must have been, despite major
breaks, a historical account of the background to the dispute, citing
events in reigns prior to that of Eannatum. It then goes on to intro-
duce Eannatum, and describe his generation by the god Ningirsu, his
suckling on the lap of the goddess Ninhursag,53 his heroic qualities, and
the bestowal of his name by the goddess Inanna. This early history
was apparently necessary as justification for Eannatum’s engaging the
ruler of Umma, whose arrogance had led him to encroach into the
arable lands of the Guxeden, the “beloved fields of Ningirsu,” without
recompense to Lagash. But before the actual encounter, there occurs
the dream sequence described above, in which Eannatum receives an
oracle foretelling the encounter. After a short break, the text resumes
in the midst of the action. Eannatum, apparently wounded, carries
on until Umma is defeated. Twenty burial mounds of the enemy are
heaped up, as Eannatum succeeds in restoring to Ningirsu the fields
of the Gu’eden, each one carefully listed in detail.
This section of the text ends along the register division separating the
upper and lower portions of the obverse. Beginning in the first column of
the lower register is an account of the aftermath of the battle, in which
the vanquished ruler of Umma swears on the “great battle-nets” of six
major deities that he will respect the now properly reconstituted border.
By modern dramatic standards, the long description of each oath
and the attendant rituals to secure each one seem to be merely the
working out of terms and hence secondary to the battle itself. Yet it is
actually the longest sequence in the text, continuing from the obverse
through the entire upper register of the reverse and ending along the
51
François Thureau-Dangin, Die sumerischen und akkadischen Königsinschriften (Leipzig,
1907), and again in Heuzey and Thureau-Dangin, Restitution matérielle.
52
Steible and Behrens, Bau- und Weihinschriften, 120–145, and Cooper, Reconstructing
History, 45–48.
53
Note that, if anything, this makes all the more clear the identification of these
two figures on the obverse of the stele, as argued above.
22 chapter sixteen
register line. Toward the end of the same register line we enter into
the final section of the text, in which the titles, epithets, and deeds of
Eannatum are recounted. The litany includes the lands he has con-
quered, how he has restored the Gu’eden to Ningirsu, and finally, how
he has erected this very stele for the god. This last portion falls in the
bottom register of the reverse. It is likely that standard curses aimed at
anyone who would remove or destroy the stele would have comprised
the very end of the text, but are not preserved, due to the breaks in the
lowest register (see n. 38).
The text thus may be divided into six principal sections: background,
introduction and preparation of Eannatum, action, immediate con-
sequences of action, aftermath, and summary plus commemoration.
These six sections conform closely to the requirements of the “Display
Text,” delineated as a separate category of narrative by Pratt:54
1. Abstract = Background
(summary of viewpoint)
2. Orientation (identification) = Preparation and naming of
Eannatum
3. Action = Battle
4. Evaluation/consequence = Restoration of fields
5. Resolution = Oaths toward stability
6. Coda = Recapitulation of titles, dedication
of stele and curses
The text thus constitutes a coherent narrative in itself; it lays out the
position of Lagash with respect to the border conflict with Umma, and
the actions taken by Eannatum. It is immediately apparent (as noted
also by Donald Hansen in a lecture entitled “Early Dynastic Demons
and Monsters” at the Pierpont Morgan Library in April 1983) that
the correspondence between text and imagery is not exact. Significant
elements within the verbal narrative are not depicted on the stele,
while certain of the details so carefully depicted in relief are not fully
described in the text. Read in conjunction with the imagery, the text
does help to identify the specific conflict and interpret certain details
54
Mary Louise Pratt, Toward a Speech Act Theory of Literary Discourse (Bloomington,
1977), 44–45, as cited in Marilyn R. Waldman, Toward a Theory of Historical Narrative: A
Case Study in Perso-Islamicate Historiography (Columbus, 1980), 18–20. Pratt has based her
work upon that of William Labov, Language in the Inner City (Philadelphia, 1976), and
Labov and J. Waletsky, “Narrative Analysis: Oral Versions of Personal Experience,”
in Essays on the Verbal and Visual Arts [ Proceedings of the 1966 Annual Spring Meeting
of the American Ethnological Society] (Seattle, 1967).
of the pictorial narrative. But the visual imagery has its own agenda,
not identical to that of the text.
The major pictorial narrative on the reverse is devoted to a detailed
representation of the preamble and actual conflict with Umma, a
sequence passed over very quickly in the text. By contrast, those por-
tions given the longest description in the text—the delineation of all of
the restored fields and the oaths sworn by Umma—are entirely omitted
from the visual narrative.
In effect, then, the role of the visual portion of the monument is to
convey on the reverse the immediate antecedents of the events and the
action-through-narrative in which Eannatum played a primary role and
on the obverse, a summary of that action through the icon of the state
god as hero and victor. The verbal portion of the monument extends
further in time, both backward and forward—from the historical back-
ground of the conflict through to the consequent binding of Umma
by oath, restoration of the disputed fields, and erection of the stele as
witness to both event and agreement. In other words, the reliefs detail
the immediate action(s), while the text emphasizes the longer-range
antecedents and consequences.55
In fact, the text and the imagery of the Stele of the Vultures come
from two very different traditions. The text derives from a combina-
tion of known celebratory proclamations and legalistic documents that
include background, current intervention, resolution, and proscriptions.
The literary tradition is apparent in the use of certain cuneiform signs,
which are otherwise restricted to poetic usage, while the legalistic tra-
dition is reflected in the wording of oaths and sanctions.56 However
the imagery is relatively new, if we may judge from extant artifacts.
It has developed out of a visual tradition of representing sequences
surrounding, or culminating in, a single event. However, the stacking
of registers that can be read both horizontally, as narrative of a single
55
Only by allusion may the obverse—with its representation of Ningirsu—be said
to evoke visually missing portions of the written text: by reference to the god who was
the progenitor of Eannatum, and by reference to his battle net, which recalls those
of the high gods upon which the vanquished ruler of Umma takes his oaths after los-
ing the battle to Lagash. For a complete discussion of these oaths, and the tradition
surrounding them, see Dietz Otto Edzard, “Die Eide in der ‘Geierstele’ des Eannatum
von Lagash,” Assyriological Studies 20 (1975), 64–68.
56
Personal communication on the use of signs from Hermann Behrens, Babylonian
Section, The University Museum, Philadelphia, and on the legal tradition from Piotr
Steinkeller, Harvard University.
24 chapter sixteen
episode, and vertically, as a sequence of several events, makes the stele

a more complex and sophisticated phenomenon than either iconic
votive plaques such as that of Ur-nanshe (fig. 15) or the single-event
strip narrative of the “Standard of Ur” (figs. 13 and 14) that preceded
it by only some forty years.
We have before us, then, a situation very different from that described
by Kurt Weitzmann, in which imagery may be used to “render liter-
ary content.”57 On the Stele of the Vultures, the literary and visual
trajectories meet, but are not yet either truly parallel or identical in
structure and content. The visual narrative on the Stele of the Vultures
corresponds neither to Weitzmann’s “simultaneous” method, in which
several actions take place within a single scene, nor to his “monoscenic”
method, in which a single action is selected from a part of the story
to stand for the whole. Nor does it correspond to his “cyclic” method,
in which a coherent series of images represents stages within a single
text.58 It is, instead, an “autonomous narrative” method, if we may call
it that, in which the figural representation on the stele has its own logic
of organization and emphasis quite distinct from that of the text.
In fact, text and imagery differ not only in content but also in intent:
the text serving the legal case of the legitimacy of Lagash’s claims over
Umma and the Guxeden at the highest (literate) levels, its audience both
internal and external; the visual portion addressed more to an internal
(not necessarily literate) audience, its message related to the hierarchy
and power of the state itself.
The Stele
The preceding discussion of the text gives rise to the second contextual
point, the nature of the stele as monument; that is, as the vehicle by
which this particular text and imagery were carried.
The Sumerian word translated here as “stele”—na-rú-a—is often
preceded by the determinative, which identifies the noun in the class
of “stone.” rú is the reading for the verb root dú, “to erect.” Thus
na-rú-a was understood in antiquity as a freestanding stone monu-
57
Kurt Weitzmann, Illustration in Roll and Codex: A Study of the Origin and Method of
Text Illustration (Princeton, 1970), 1.
58
Weitzmann, Roll and Codex, 13–14, 14–17, 17–33, respectively.
ment or marker, the size of which could vary considerably. The term
also occurs in Akkadian (narû) as a Sumerian loan word designating a
stone monument inscribed with laws and regulations, a stone monu-
ment used as a boundary marker in situ, or a memorial monument
set up by a king.59
At least one earlier stele is referred to by Eannatum, and again by his
nephew Enmetena, and that is the original stele erected by Mesalim,
king of Kish, some time within the Early Dynastic II period, c. 2700
b.c., when the proper boundary between Umma and Lagash was first
adjudicated.60 This stele would therefore have belonged to the second
category of boundary markers in situ.
The question of where the Stele of the Vultures was originally
intended to have been erected arises from several ambiguities in the
text of the stele itself and from other textual references. Although the
excavated fragments were found at the site of ancient Girsu, they were
dispersed over a rather wide area; it could be argued, therefore, that
this was the secondary, and not the primary, location. At least three
possibilities must be considered: an original location at the border,
alongside the original stele of Mesalim; a location in the É.maª, a
temple at Girsu dedicated to the goddess Nanshe; and a location within
the temple precinct of Ningirsu.
59
See narû, in the Chicago Assyrian Dictionary “N” (Chicago, 1980), 364–367.
60
Steible and Behrens, Bau- und Weihinschriften, 158 (Eannatum text 6, col. 1:6–7);
177 (Eannatum text 63, col. 2:6–7); 230 (Enmetena texts 28, 29, col. 1:8–12). See
also Cooper, Reconstructing History, 49–50. Cooper will publish a more detailed analysis
of these texts in the forthcoming Sumerian and Akkadian Royal Inscriptions, Volume I: Pre-
Sargonic Inscriptions [American Oriental Society Translation Series], the manuscript of
which he has kindly made available to me. There he points out (MS, 48) that there
may have been at least three additional stelae set up at the border, or at a new border
more favorable to Umma, by the ruler of Umma after the destruction of Mesalim’s
stele (based upon Eannatum text 6, cols. 3:9–18 and 4:1, Steible and Behrens, in the
work cited, 159–160). The original stele of Mesalim must have contained a lengthy
inscription recounting the terms of the establishment of the original border, enumerat-
ing the fields and canals on either side (as the Stele of the Vultures includes a long list
of the fields “returned to Ningirsu” by Eannatum—Steible and Behrens, in the work
cited, 128–129 [ Eannatum text 1, cols. 14, 15–16:11]). The smashing of the monu-
ment of Mesalim by the ruler of Umma as he crossed over the border at the time of
Ur-nanshe, grandfather of Eannatum, must have been an act of political defiance of
the terms set forth on the stele. Similarly, when Ur-lumma of Umma moved across the
border against Enannatum I, brother and successor of Eannatum, he set fire to the
stelae marking the boundary—presumably that of Eannatum and the earlier, restored
one of Mesalim (Steible and Behrens, in the work cited, 236–237 [Enmetena text 28,
col. 3:36–38]).
26 chapter sixteen
Arguments for the first possibility are based on mention in the text
of the stele and on two virtually identical clay cones of Enmetena
that Eannatum, after his temporary resolution of the conflict, not only
reestablished the border and restored the original stele of Mesalim,
but also erected his own stele at the site.61 In fact, Perkins goes so far
as to assume that the Stele of the Vultures is that very stele, without
reference at all to the inconsistency of its findspot.62 Arguing against
this, however, are the size of and complex decoration lavished upon
the stele, which are more in keeping with stelae of later periods set up
in urban temple compounds than with the smaller and more schematic
land-grant boundary stelae, as for example, fig. 17.63 An additional
problem exists: why and when the monument would have been moved
to Girsu, where it was discovered. Further evidence may be added from
Cooper’s restoration of an Eannatum text from two previously known
fragments. If Cooper’s collation is correct, the name of the boundary
stele set up by Eannatum has been preserved and is unrelated to the
name given on the Stele of the Vultures.64
Arguments for the second possibility come from a broken reference in
the text of the Stele of the Vultures, in which we are told that Eannatum
set up something in the É.maª, the temple of Nanshe.65 Although the
word “stele” has been restored here by Steible, in fact it is not to be
found in the text, so that the missing noun could be almost anything.66
61
Steible and Behrens, Bau- und Weihinschriften, 126–127 (Eannatum text I, cols.
10:14–11:4) and 234 (Enmetena text 28, col. 2:4–8). The Stele of the Vultures is very
explicit: Eannatum defeated Umma, and set the new border. A certain amount of
the disputed territory he left on the side of Umma, and at that place he set up a stele
(emphasis ours).
62
Perkins, “Narration,” 57.
63
For example, the Stele of Ur-nammu, set up in the temple precinct of the moon
god, Nanna at Ur (see Moortgat, Ancient Mesopotamia, fig. 194).
64
Cooper, Reconstructing History, 27. He has joined two texts, one of Eannatum and
one formerly believed to date to Enmetena, and suggested both belong to the reign
of Eannatum. With that join, the name of the boundary stele can be reconstructed as
“Ningirsu is the lord eternally exalted in the Abzu.” Cooper has noted (27, n. 13) that
this would then have nothing to do with the name of the Stele of the Vultures, which
is preserved on the stele itself as, “Ningirsu, lord, crown of Lumma, is the life of the
Pirig-eden canal” (see complete portion of the Stele of the Vultures text, below).
65
Steible and Behrens, Bau- und Weihinschriften, 128: cols. 12:21–13:2.
66
The word “stele” has been restored here by Steible, but in fact it is not to be
found in the inscription. The sign used for the verb, “to erect” (RÚ), is the same sign
as that used in other contexts for the verb, “to build” (DÙ); so that its translation is
dependant upon the noun, which in this case, we do not know. Thus the missing noun
could be anything that Eannatum either erected or built within the temple compound
of Nanshe, and there is no reason to associate this reference with our stele.
Nor does the temple of Nanshe make any contextual sense for the
Stele of the Vultures, dominated as it is by the figure of Ningirsu and
containing no significant reference to the goddess Nanshe.
In favor of the third possibility, that the stele was originally set up in
the temple precinct of Ningirsu, are the following: first, the distribution
of five of the six excavated fragments on or around Tell “K” at Girsu,
a low mound within the city on which the main temple to Ningirsu was
situated; second, the example of later rulers setting up stelae in temples
(as with the stele of Ur-nammu at Ur in the Neo-Sumerian period,
c. 2150 b.c.); and third, the dedicatory references within the text, not
only to Ningirsu, but also to the erection of the stele.
The most complete reference occurs at the very end, a band of
fourteen separate cases set apart, almost as a label summarizing the
whole object. It is worth quoting in its entirety, as it gives a very real
sense of how the Sumerians conceived of the monument:
The stele,
its name
is not a man’s name; it [its name] is:
“Ningirsu,
Lord, Crown of Lumma,
is the life of the Pirig-eden Canal.”
The stele
of the Gu’eden—
beloved field
of Ningirsu
[which] Eannatum
for Ningirsu
returned to his [the god’s] hand—
he [ Eannatum] erected it.67
Thus, the stele is dedicated to Ningirsu, is named for Ningirsu, makes
reference to the main canal watering the disputed Gu’eden, and is
erected for the god. In addition, I would argue that the obverse is
carved with a monumental figure of Ningirsu as the icon of victory
over Umma. All this, plus the findspots of the major fragments, strongly
67
Steible and Behrens, Bau- und Weihinschriften, 144–145 (Eannatum text 1, reverse
col. 10, unterschcrift: 23–37). This band of inscription is separated from the end of
column 10 proper by a marked space; the enclosing “cases” for signs are slightly nar-
rower than those for the rest of the text. It therefore was clearly meant to stand apart,
almost as a colophon labeling the object, rather than as a part of the narrative text.
28 chapter sixteen
suggests that the stele originally stood as both testimony and votive in
the god’s sanctuary.
The situation of the Stele of the Vultures would seem to correspond
closely to the role of the royal stele as described in a later (Assyrian)
version of the Cuthaean legend of Naram-Sin, king of the Akkadian
dynasty, c. 2290 b.c., in which the king speaks the following words:
I have . . . inscribed a stele for thee [a future king] and in Cutha, in the
É.meslam [a temple], in the shrine of Nergal [a god not unlike Ningirsu
in attributes], I have deposited it for thee. Read this document and listen
to the words thereof.68
The same text later resumes: “Let wise scribes read aloud thy stele”;
while an earlier ruler is described as one who “did not write on a stele
and leave it (for posterity).”69
From these brief references, much can be understood about the
function of the Akkadian narû. I would argue that this is relevant for
the Early Dynastic stele as well. It was not merely intended as a com-
memorative monument; it was rather meant to be a living testimonial
witness to the historicity of the events and the legitimacy of the legal
terms (restoration of fields and reconstitution of the border) that it
recorded. This would fit closely with Pratt’s definition of the “display
text” with which we matched the Stele of the Vultures above: that is, a
text intended not only as a record of events, but also as a (verbal) per-
formance, designed with audience in mind, in which support is elicited
for the particular view held, and the event(s) described is intended to be
morally and/or contractually kept alive.70 The stele functioned, then,
not only as a witness to events but also as a witness to the concluding
agreement and to the consequences that would result were Umma to
break its oaths. It would have been seen as an active agent in ensuring
and enforcing, by virtue of its very existence, the conditions described
on it, at the same time that it celebrated the power of those who were
successful in bringing the present situation about: the ensí, Eannatum,
and the city god, Ningirsu.
68
Oliver R. Gurney, “The Sultantepe Tablets (cont’d),” Anatolian Studies 5 (1955),
106: lines 149–153.
69
Gurney, “Sultantepe Tablets,” 108, line 173, 98, line 1.
70
Pratt, Toward a Speech Act Theory, 136, 138–139.
To borrow Eco’s term, which Richard Brilliant has equally applied

to the Column of Trajan, the stele is itself an “iconogram.”71 A portion
of its message is conveyed simply by belonging to a particular typologi-
cal class: the na-rú-a. We must therefore take into account not only
the literal and rhetorical content of the text and imagery, but also the
rhetorical value of the stele as a public monument—especially if it was
meant to be preserved and read by successive generations.
The Historical Context
As Barbara Herrnstein-Smith has noted, “the fact that something is true

is never a sufficient reason for saying (or depicting) it.”72 Rather, every
narrative is produced and experienced under certain social conditions;
and every narrator or author (in this case, Eannatum) must have some
interest in relating (or displaying) it. It has been suggested above that
the very nature of the stele as a monument would have helped direct
its audience to the intended response. But that response can only be
reconstructed in the context of the larger historical picture of Lagash
and the other city-states of southern Mesopotamia in the late Early
Dynastic period.
The Early Dynastic period marks the full consolidation of the
process of urbanization begun in the preceding Uruk period (c. 3500
b.c.). By Early Dynastic III (c. 2500–2350), southern Mesopotamia
was divided into a series of polities, each consisting of at least one
major urban center that exercised control over surrounding tracts of
agricultural and range lands as well as water sources.73 Of these poli-
ties, Lagash was one of the most important. Thanks to the extensive
archive excavated at its major satellite city of Girsu, it is also the only
state from which sufficient documentation has been preserved to permit
71
Umberto Eco, Theory of Semiotics, as cited by Brilliant, Visual Narratives, 96.
72
“Afterthoughts on Narrative, II: Narrative Versions, Narrative Theories,” in On
Narrative, ed. W. J. T. Mitchell (Chicago, 1981), 229, n. 24.
73
See Robert McC. Adams, Heartland of Cities: Surveys of Ancient Settlement and Land
Use in the Central Floodplain of the Euphrates (Chicago and London, 1981), 137, showing
a chart indicating a large rise in the number and size of sites in the Early Dynastic III
period—78% of which were large urban centers of more than 40 hectares. See also,
Adams and Hans J. Nissen, The Uruk Countryside: The Natural Setting of Urban Societies
(Chicago and London, 1972), 17, which also gives a chart marking a substantial rise of
larger sites and a decrease of smaller ones in the Early Dynastic II/III periods.
30 chapter sixteen
reconstruction of a history of the period and a profile of the workings

of a city-state.74
The state consisted of Lagash (modern al-Hiba) and two satellites,
Girsu (Tello) and Nina (Surgul).75 The total area controlled by Lagash
in the Early Dynasty III period is estimated at 1,160 square miles,76
sustaining a population of some 50,000 free men. At least twenty tem-
ples are attested within the state, of which one of the most prominent
was the temple of Ningirsu at Girsu.77 There is evidence that these
temples were principal landowners with vast holdings, so that when
Eannatum calls the Gu’eden the “beloved field of Ningirsu,” this may
well have been literally true.
Falkenstein has argued that all the major city-states of the period were
more or less equal in size and similarly organized.78 Each was politically
autonomous and was ruled by an individual designated either by the
title lugal (king; literally, great man) or ensí (city ruler, governor).79
It is clear from textual references that at the beginning of the Early
Dynastic period, rulers were considered appointees of the local deity.
By the end of the period, however, the rulers had begun increasingly
to usurp lands and prerogatives formerly held by the temple estates.80
The end of the period was also marked by the expansion of central-
ized urban polities. Buffer zones of potentially rich but undeveloped land
that had separated the rings of irrigated, agricultural fields surround-
ing the cities from those of neighboring states had been reduced to
nothing. This shrinkage accounts for the struggles between Umma and
Lagash, which were largely due to competition over the Gu’eden, the
74
Dietz Otto Edzard, “The Early Dynastic Period,” in The Near East: The Early
Civilizations [Delacorte World History, vol. II], eds. Jean Bottero, Elena Cassin and Jean
Vercoutter (New York, 1967), 52–90.
75
Lagash/al-Hiba has been excavated by Donald P. Hansen (see n. 2); Nina/Surgul
has not been excavated to date.
76
See discussions by Adam Falkenstein, The Sumerian Temple City [Sources and Mono-
graphs in History: Ancient Near East I/1], trans. Maria de J. Ellis (Malibu, 1974), 6;
and Igor M. Diakonoff, Structure of Society and State in Early Dynastic Sumer [Sources and
Monographs in History: Ancient Near East 1/3] (Malibu, 1974), 6–7.
77
Edzard, “The Early Dynastic Period,” 76; Diakonoff, Structure of Society, 6.
78
Falkenstein, Sumerian Temple City, 16.
79
Edzard, “The Early Dynastic Period,” 70–75.
80
William W. Hallo, The Ancient Near East: A History (New York, 1971), 49; Edzard,
“The Early Dynastic Period,” 82, 88; and see especially the primary text, known as the
“Reforms of Uruinimgina,” last king of Lagash in the Early Dynastic Period, in which
direct reference is made to earlier practice: Steible and Behrens, Bau- und Weihinschriften,
313–324 (Uruinimgina text 6), and Cooper, Reconstructing History, 51.
tract of fields and grazing lands along their twenty-eight-mile common

border.81 The conflict persisted for almost two centuries, through the
reigns of the six successive kings of the hereditary dynasty of Lagash
during the Early Dynastic III period (including those of its founder,
Ur-nanshe, his grandson Eannatum, and Eannatum’s successors—his
brother, Eannatum I, and his nephew, Enmetena); it then continued
into the seventh and last reign of the period, that of the probable
usurper, Uruinimgina.82
The basis of this conflict was proprietary: who was entitled to exploit
and derive wealth from the yield of this land? Thus, in the Stele of the
Vultures text (cols. 10:14–11:30), Eannatum does not claim all of the
land that, he declares, actually belonged to Ningirsu; he leaves a por-
tion on the side of Umma. But Umma is expected to pay a percentage
of the yield as rent. It is Umma’s failure to acknowledge the debt and
pay rent that brings about the subsequent conflict under Eannatum I
and Enmetena.83
Underlying the issue of land tenure is the even more fundamental
issue of water rights, without which the land is useless. The state of
Lagash lay to the southeast of Umma, with Girsu approximately halfway
between the two on a direct line; and the contested lands northwest of
Girsu were watered by the same Euphrates canal that passed through
Umma.84 As water is a crucial resource, the upstream user in controlling
the supply of downstream water would clearly have great power—power
that quickly became a political weapon.85 Thus, much of the dispute
recorded on the Stele of the Vultures has to do not only with the fields,
but also with the irrigation canals within the Gu’eden.86
81
Adams, Heartland, 134; and Giovanni Pettinato, “Il conflitto tras Lagash ed Umma
per la ‘frontiera divina’ e la sua soluzione durante la terza dinastia di Ur,” Mesopotamia
5–6 (1970–1971), 281–320.
82
This has been the subject of a number of studies; see Cooper, Reconstructing His-
tory, for the present view and full bibliography; see also Edzard, “The Early Dynastic
Period,” 80–84, for summary.
83
See Steible and Behrens, Bau- und Weihinschriften, 198–202, 230–245.
84
See map, in Adams, Heartland, 157, fig. 28.
85
Robert McC. Adams, “Die Rolle des Bewässerungsbodenbaus bei der Entwick-
lung von Institutionen in der altmesopotamischen Gesellschaft,” in Produktivkräfte und
Gesellschaftsformationen in vorkapitalistischer Zeit, ed. J. Herrmann and I. Sellnow (Berlin,
1982), 119–140; Etienne de Vaumas, “L’ecoulement des eaux en Mesopotamie et la
provenance des eaux de Tello,” Iraq 27 (1965), 81–99.
86
See Steible and Behrens, Bau- und Weihinschriften, 130 (col. 16:25–31), Cooper,
Reconstructing History, 9, and important discussion by Hans J. Nissen, “Geographie,” in
32 chapter sixteen
Although Lagash may have been dominant in the early stages of

this dispute, archaeological survey data would suggest that there was
a rapid growth in the density of sites and in the political importance
of Umma in the late Early Dynastic period.87 At the very end of the
period, Uruinimgina of Lagash was defeated by Lugalzagesi of Umma;
the capital city was plundered, and in fact never fully recovered. Girsu
took over as the principal city of the state, but absolute power had
already shifted elsewhere.88 Meanwhile, Lugalzagesi went on to conquer
the powerful city-state of Uruk, which had formerly been allied with
Lagash, and took upon himself the title, “King of Sumer.”89 By the
end of the period events pointed firmly toward the establishment of the
hegemonic nation-state achieved in the succeeding Akkadian period.
Sumerological Studies in Honor of Thorkild Jacobsen (Assyriological Studies 20), ed. Stephen
J. Lieberman (Chicago, 1975), 9–40. This is a portion of the present study that could
be considerably expanded. As Hunt and Hunt have shown in case studies elsewhere,
conflict between polities over irrigation water has high violence-potential, especially
when both polities are not subservient to a larger, single authority; and when a down-
stream community is superior in political or military might (as was the case with Lagash
before the end of the Early Dynastic Period), unless both reach some mutual agreement,
there is likely to be recurring conflict (Robert C. Hunt and Eva Hunt, “Canal Irriga-
tion and Local Social Organization,” Current Anthropology 17 (1976), 395, and Eva Hunt
and Robert C. Hunt, “Irrigation, Conflict and Politics: A Mexican Case,” in irrigation’s
Impact on Society, eds. T. Downing and McGuire Gibson [Anthropological Papers of
the University of Arizona, 25] Tucson, 1974, 151). Thus, one of the aggressive acts
of Umma that led to resumed hostilities under Eannatum I and Enmetena was the
diverting of water in the boundary channels between the two states. It was apparently
in an attempt to put an end to this dependency upon upstream water that Enmetena
later undertook construction of a very long, new channel from the Tigris, despite the
fact that Euphrates water flowed at a more regular and manageable rate (see Steible
and Behrens, in the work cited, 235–237, 243; and Adams, Heartland, 134).
87
The increase of sites within Umma is recorded in Adams, Heartland, 160. Hallo
(The Ancient Near East, 52) has suggested it was the repeated conflict with Lagash that
provided the stimulus for the development of Umma toward the end of the period.
However this may be, it is indeed unfortunate that the kingdom of Lagash was entirely
outside the intensively surveyed area covered by Adams (in the work cited, 134), and
has been subject only to brief scrutiny in 1953, as reported by Thorkild Jacobsen,
“A Survey of the Girsu (Tello) Region, Sumer 25 (1969), 103–109. There is, then, a
“tantalizing mis-match,” as Adams has called it, between the textual materials and the
archaeological survey data available from the two sites, which is particularly unfortu-
nate with regard to the present issue. A survey of the entire Lagash region is needed,
as is excavation of Umma.
88
Steible and Behrens, Bau- und Weihinschariften, 334–337, and Cooper, Reconstructing
History, 52. One wonders whether this invasion of Lagash by Lugalzagesi would have
not been a likely time for the destruction and dispersal of the Stele of the Vultures, as
it would have been a monument particularly offensive to the Ummaites.
89
Edzard, “The Early Dynastic Period,” 83, and Cooper, Reconstructing History, 10.
It remains to be demonstrated how this historical background ani-

mates our view of Eannatum and the Stele of the Vultures. We now
see Eannatum as the ruler of one of the predominant states of Meso-
potamia in the Early Dynastic III period. The conflict with Umma,
as recorded on the stele, is not merely one event among many but
rather the primary external concern of the state at the time. The issues
involved—control over land and water—were essential to the continued
prosperity and well-being, if not autonomy, of the state.
The Early Dynastic III period was also the time in which the institu-
tion of kingship became fully established; by the end of the period in
particular, the rulers had amassed considerable power and privileges. As
Hallo has pointed out, textual references to the ruler—sired by a chief
god, suckled by a goddess, named and granted attributes by other dei-
ties—constitute an elaborate means of stressing the legitimacy of royal
power.90 These are the very claims made by Eannatum in the text of
the Stele of the Vultures and, together with Eannatum’s central role in
the narrative, they signal the expanded role of the king. I would argue
that the same message is conveyed visually by the prominent position
given to the king in the pictorial narrative of the reverse. As we read
from bottom to top, there would appear to have been a slight but ever-
increasing change in scale, so that the upper figure is larger than the
figure reconstructed at the bottom; at the same time, the ruler progresses
from the margin to the exact center. I would suggest that this movement
is not merely an artifact of compositional and narrative demands, but is
rather a significant statement of the structural development of kingship
as an institution in the late Early Dynastic period.
In the light of this political development, the visual and narrative
tension between the obverse and the reverse of the Stele of the Vul-
tures takes on far greater significance. We are not yet at a point where
the king has assumed power over a national polity, nor has he actually
declared himself a god, as in the Akkadian period.91 But the seeds of
that subsequent development are present, represented on the stele as
they were clearly in the political institutions of the age. Yes, Ningirsu
90
Hallo, The Ancient Near East, 49.
91
This, too, is reflected in the public monuments of the period, as Sargon of Akkad
is shown holding the battle net himself on one stele, as the goddess Ishtar looks on
(Moortgat, Ancient Mesopotamia, figs. 126–127); while on the stele of his grandson,
Naram-Sin, the king occupies the summit of the only face of the relief, wearing the
crown and bearing the determinant of a god (in the same work, fig. 155).
34 chapter sixteen
is given full honors on the obverse, as the “icon” of the city. But at the
same time, a challenge is put forth by the ruler: in the repetition of
his figure and his shifting position, he, too, becomes an “iconic” figure
of heroism and victory.
In short, just as the issues of land and water control raised on the
stele are fundamental for the Early Dynastic period as a whole, so
also is the development of the role of the ruler in the city-states of
the period intertwined with the representation of Eannatum on his
victory monument.
* * *
We are finally in a position to assess the impact of the use of historical
narrative on this work in this period, and to analyze the general impli-
cations of the use of historical narrative within the history of art.
We have tried to demonstrate that the Stele of the Vultures played
an active role in the ideology of the early Mesopotamian state; that it
represented “not merely [a] passive reflection of the political system,”92
but rather a model for that system, by presenting as if given an under-
lying structure that itself had a powerful political agenda. We have
suggested further that while the icon of the god was the traditional
mode of Mesopotamian representation, the historical visual narrative
introduced by Eannatum served new purposes. The latter not only
told the story literally, it also served figuratively to establish the ruler
as the legitimate force in the sequence of events. We have therefore
imputed political ends to the introduction of narrative. These ends
may not have been consciously manipulated (we will never fully recover
intentionality), but the devices employed certainly were not unaffected
by their times.
The principles underlying such a view were best articulated by Her-
rnstein-Smith when she wrote, “no narrative can be independent of
a particular teller and occasion of telling, and . . . therefore every nar-
rative has been constructed in accord with some set of . . . interests”;93
and by Hayden White’s quote from Hegel: “it is the state which first
presents subject-matter that is not only adapted to the prose of history,
but underlies the production of history in the very process of its own
being.”94
92
Robert Layton, The Anthropology of Art (New York, 1981), 85.
93
Herrnstein-Smith, “Afterthoughts on Narrative,” 215.
94
White, “The Value of Narrativity,” 11.
Thus, the primary, official state message presented through the inno-
vative use of historical narrative on the Stele of the Vultures is that of
the legitimate claim of Lagash over the Gu’eden; and here we must
emphasize that there are no comparable monuments from Umma of
this period to give us the other side of the conflict. Despite the asserted
objectivity of the verbal and visual texts, it is important to remember
that the texts constitute instructed rhetoric, intended to provide witness
to a very selected “truth.”
The secondary message presented through historical narrative is
that of the centrality of the ruler in maintaining the state. It is our
assessment that the stele would have been set up in the major temple
to Ningirsu as a votive to the patron deity, who was given titular credit
for the victory at the very time his authority was being eroded and his
territorial prerogatives encroached upon by the secular city ruler. The
god is given pride of place on the obverse; the king is on the reverse.
But the challenge to the older, nominally theocratic base of the city-state
is there. The hierarchy of the religious system is no longer identical
with the administrative hierarchy of the state.
The visual power of the stele lies precisely in the tension between
the two heros (god and king) and their two realms (mythological and
historical). Yet, as we have observed, that tension—reflecting the old,
“theocratic” state and the new, “dynastic” state—avoids open schism
by cross-references. Ningirsu intervenes in battle; Eannatum’s power is
attributed to the divine sphere.
It is particularly striking that distinct pictorial modes have been
selected to represent the different conceptual realms: the iconic for the
mythological world, and the narrative for the historical. It is as if the
mythological were best represented by symbol and allusion, while
the historical should be grounded in action and events. The historical
narrative of the reverse of the stele provides a visual account of real
events; but, as narrative discourse, it also lends weight to that reality,
by providing it with the organizational structure that gives coherence,
fullness, and closure.95 Nevertheless, to get the complete message of the
stele, one must combine the narrative with the icon, so that the ante-
cedents to and the consequences of the specific events are alluded to
and the whole intention made clear. Historical narrative in this period
may thus be said to serve an end, rather than to be an end in itself;
95
36 chapter sixteen
and the end it serves is the establishment of the logical basis for the
resolution and the premises of the actions it records. In this respect,
narrative becomes simply one code among others for a culture to exploit
in the transmission of messages.96
Located in the temple, the stele would transmit in no uncertain
terms the mythic history of the city’s god, as well as the contemporary
validation of that god, of the current ruler, and of the state as a whole.
At the same time, it would have proclaimed the immediate legitimacy
of the state’s position vis-à-vis Umma. And, carried in the historical
narrative would be the covert message of the monument: the increase
in power and authority of the Early Dynastic ruler.
* * *
In the course of this examination of how pictorial narrative—and in
particular, historical narrative—was incorporated into one early monu-
ment, a number of methodological and theoretical points have arisen,
which it seems appropriate to summarize here.
First, with respect to the relationship between narrative and iconog-
raphy: it would seem that the latter, as a central pursuit of the discipline
of art history, constitutes a prior and necessary step in the reading of
narrative, but that the two should be kept separate. The one represents
a process of identification, basically descriptive (Panofsky’s iconographi-
cal analysis proper),97 the other a process of organizational analysis, of
“how” as opposed to “what.”
Second, this analytic process makes clear that there is a very close
relationship between narrative and structure, hence narrative analysis
and structural analysis, to the degree that both attempt to lay out orga-
nizing principles. Yet, if both seek to answer the question “how”—in
our case, how the imagery is organized so that it may be read—I would
suggest that here, narrative analysis is a preliminary step for structural
analysis, which then moves on to seek meaning in the patterns manifest
in orgainizing principles and to answer the question “why.”
Third, it seems evident that narrative is but one representational
code among several. Nothing demands that a historical battle be rep-
resented as narrated action. In ancient Egypt, for example, the motif
of a smiting king was used as an icon for all military victories. Pictorial
96
97
Erwin Panofsky, Studies in Iconology (New York, 1939, 11–12).
narrative is therefore selected as a mode of representation because it

meets the requirements of the particular individual, period, or culture.
Nevertheless, it may be said that narrative requires less prior knowledge
than other sorts of codes; and that it is a particularly effective means
of transmitting an ideological message, simply because it is so seduc-
tive in coercing the viewer to read along with what is given and so be
brought to a desired and seemingly inexorable conclusion.
Fourth, visual narrative has properties very similar to those of verbal
narrative. However, if as Bresson has demonstrated experimentally,
following Piaget, competence in reading visual imagery is a skill that
has its own developmental locus, separate from that of language, and
if one can think through representations, not just through language,98
then one can “make think” via representations—that is, use pictorial
narrative with ideological, rhetorical aims.
Fifth and finally, in the standard communications theory diagram of
sender-message-receiver, one must divide the message into at least two
components: the actual information conveyed, and the extralinguistic or
extravisual referent that is part of the subtext. The ideological message
is often built into the structure of how the message is conveyed, rather
than what the message contains.
* * *
Writing in 1951 about the sculpture of the ancient Near East, Henrietta
Groenewegen-Frankfort stated:
The function of a monument is to penetrate the impermanent, to stress
the transcended signifcance of human action. It requires both symbolic
statement and factual representation—the tension between the ephemeral
and the everlasting.99
I would perhaps substitute the word “temporal” or “immediate” for
“ephemeral”; but what Groenewegen-Frankfort has captured so well
is the separation of the two domains, the need to combine them, and
the tension that remains. All this we have tried to describe for the Stele
98
François Bresson, “Compétence iconique et compétence linguistique,” Communi-
cations 33 (1981), 185–196, especially 194, on the absence of validity for the received
wisdom that the image is more spontaneously and more easily received than text (“a
picture is NOT worth 1000 words”). The two are, Bresson would claim, separate
systems of cognition, equally but differently complex; and he calls for the development
of new methods to analyze perception of figural imagery.
99
Groenewegen-Frankfort, Arrest and Movement, 21–22.
38 chapter sixteen
of the Vultures, in the use of god and king, obverse and reverse, myth
and history, icon and narrative on a major public monument.
But what we have also stressed are the affective, as opposed to the
merely reflective, properties of the monument—the fact that monuments
constitute the objects they represent as much as they mirror them, and are
thus active agents in projecting particular responses in reception.100
Public monuments, such as the Stele of the Vultures, were not neutral
images. The power and awe-inspiring aspects of these monuments can
be partially reconstructed from the enormous energy it took for the
biblical world to shift away from visual foci in divine images and idols,
a move that Northrop Frye has called truly revolutionary.101
Through the innovative use of pictorial narrative, the Stele of the
Vultures engages both history and its audience. The monument stands
at the beginning of a long line of historical narratives in the history of
art. There are many gaps between the Stele of the Vultures (c. 2460 b.c.)
and the fully formed historical narratives of Assyrian reliefs (c. 875 b.c.);
but the linkages exist. And further linkages can be demonstrated to the
classical and Renaissance worlds, as the visual arts strove to achieve first
competence and then perfection in the representation of narrative.102
But at these initial stages, narrative is still balanced against the icon, as
the historical is balanced against the mythological, and even, in many
ways, as the image is balanced against the accompanying text. In the
end, the Stele of the Vultures emerges as a monument with a complex
message, and pictorial narrative but one of several codes activated to
convey that message.
100
On this, see Hayden White, “The Fictions of Factual Representation,” in The
Literature of Fact, ed. A. Fletcher (New York, 1976), 28–29, with regard to the shared
attributes of fact and fiction.
101
Frye, “History and Myth in the Bible,” 4.
102
The other side of this development is discussed by Alpers, “Describe or Narrate?”
15–41. In that article, she describes the development of visual narrative as part of a
process culminating in the Renaissance, then beginning a reverse movement: freezing
or de-emphasizing narrative and turning more frequently to the descriptive, as we
move from the beginning of the seventeenth century into the nineteenth.
Acknowledgments
This study could never have been written without the help of a number of individu-
als. Two graduate students in art history at the University of Pennsylvania—Benjamin
Kessler and particularly Michelle Marcus—laid the groundwork in their initial explora-
tions of this monument in seminars on the art of ancient Mesopotamia. Jerry Cooper’s
recent study of the Lagash-Umma border conflict was essential; and I have profited
from his comments on many of my ideas. Hermann Behrens patiently took me through
the initial steps of the study of Sumerian in 1983–1984 and read with me the difficult
text of the Stele of the Vultures. Without that background, and the timely publication
of his joint effort with Horst Steible, I could not have dealt with many of the ques-
tions of text and image explored here. Renata Holod has been the source of several
important bibliographical references and stimulating discussion on questions of visual
narrative; Robert C. Hunt has provided valuable perspective on questions of irrigation
and the relationship between land and politics. I am grateful also for the thoughtful
comments of Hans Güterbock, Piotr Steinkeller, and Tzvi Abusch upon a reading of
the manuscript—many of which I have incorporated into the final version.
The reconstruction drawings are also the result of many individuals’ efforts. The
figures on the obverse owe most to the work of Marie-Thérèse Barrelet; modifications
in the chariots, both front and back, to the subsequent comments of Mary A. Littauer
and J. Crouwel. On the reverse, the restored king’s figure in the lowest register and the
seated figure in the second register from bottom are the products of an unpublished
seminar paper by Michelle Marcus, used with her generous permission. Some details,
such as the verification of a lion’s head on the several eagle bodies of the obverse,
and the second figure behind the king in his chariot on the reverse, were determined
in direct examination of the stele in the company of Agnès Spycket, who kindly made
the closed Near Eastern galleries in the Louvre accessible to me and was enormously
helpful in the procurement of photographs for publication. Elizabeth Simpson did a
most remarkable job on the reconstruction drawings. Refinements such as the most
logical position of arms, proportions of whole figures, and disposition of draft animals
are frequently the result of her keen eye and mind, and occasionally of our fruitful
collaboration. I have been in close touch with all of the above individuals, and am
most grateful for their contributions.
40 chapter sixteen
Figure 1. Obverse, Stele of the Vultures of Eannatum, ensi of Lagash,

c. 2460 b.c. White limestone; ht. 1.80 m.; found at Tello, ancient Girsu;
Musée du Louvre (AO 50 + 2436–8 + 16109). Photo courtesy Musée du
Louvre.
Figure 2. Reverse, Stele of the Vultures. Photo courtesy Musée du Louvre.

42 chapter sixteen
Figure 3. Reconstruction drawing, obverse, Stele of the Vultures

(drawing by Elizabeth Simpson).
Figure 4. Detail, obverse: The god Ningirsu with battle net and emblem.
Photo courtesy Hirmer, Munich.
Figure 5. Silver vase of Enmetena of Lagash, c. 2420 b.c., Tello; Musée du

Louvre (AO 2674). Photo courtesy Musée du Louvre.
44 chapter sixteen
Figure 6. Votive plaque showing priest before the goddess Ninhursag,

Early Dynastic III period, c. 2500–2350 b.c. Limestone; found at Tello;
Musée du Louvre (AO 276). Photo courtesy Hirmer, Munich.
Figure 7. Copper panel from temple of Ninhursag at al-’Ubaid, Early

Dynastic III period, c. 2500–2350 b.c.; British Museum (WAA 114308).
Photo courtesy Trustees of the British Museum.
Figure 8. Reconstruction drawing, reverse, Stele of the Vultures

(drawing by Elizabeth Simpson).
46 chapter sixteen
Figure 9. Detail, reverse: Vultures and enemy heads, upper register.

Figure 10. Detail, reverse: Phalanx of soldiers, upper register.

Figure 11. Detail, reverse: Eannatum in chariot, soldiers behind, second

register. Photo courtesy Hirmer, Munich.
Figure 12. Detail, reverse: Ritual scene, third register; and spear directed at
enemy head, fourth register. Photo courtesy Hirmer, Munich.
48
chapter sixteen
Figure 13. “Standard of Ur,” battle side, c. 2500 b.c. Shell, carnelian, and lapis inlay;
found in Royal Cemetery at Ur; British Museum (WAA 121201).
Photo courtesy Trustees of the British Museum.
after the battle is over
Figure 14. “Standard of Ur,” banquet side. Photo courtesy Trustees of the British Museum.
49
50 chapter sixteen
Figure 15. Votive plaque of Ur-nanshe, king of Lagash, c. 2500 b.c.

Limestone; found at Tello; Musée du Louvre (AO 2344). Photo courtesy
Musée du Louvre.
Figure 16. Votive plaque from Ur, showing priest and female attendants
before temple facade and deity, Early Dynastic III period, c. 2500–2350
b.c.; British Museum (WAA 118561). Photo courtesy the Trustees of the
British Museum.
Figure 17. Boundary/land-grant stele, Early Dynastic I period, c. 3000 b.c.

Limestone; said to be from the area of Umma; The Metropolitan Museum
of Art (58:29). Photo courtesy The Metropolitan Museum of Art.
CHAPTER SEVENTEEN
EANNATUM AND THE “KING OF KIŠ”?

ANOTHER LOOK AT THE STELE OF THE VULTURES
AND “CARTOUCHES” IN EARLY SUMERIAN ART
The Stele of Eannatum of Lagaš (the so-called ‘Stele of the Vultures’)

is a monument known to most students of the Ancient Near East. In
both text and image, front and back, it documents the victory of the
city-state of Lagaš over neighboring Umma in a battle arising out of
long-standing disputes concerning land- and water-rights.1 The conflict
is attested throughout the entire Early Dynastic III period, both before
and after Eannatum, ruler of Lagaš, who commissioned this particular
stele sometime ca. 2460 b.c.2
In its present state at the Musée du Louvre,3 one is easily led to think
of the stele as a complete monument; yet the restoration actually con-
sists of only 7 fragments—one purchased on the market, and six found
scattered around Tell K, where the temple of Ningirsu was located, at
the site of Tello, ancient Girsu, a city within the state of Lagaš.4 Upon
* This article originally appeared as “Eannatum and the ‘King of Kiš’? Another
look at the Stele of the Vultures and ‘Cartouches’ in Early Sumerian Art,” Zeitschrift
für Assyriologie 76 (1986), pp. 205–212.
1
The text was originally read and published by F. Thureau-Dangin, Restitution
matérielle de la Stèle des Vautours (1909) 42–63 (hereafter RMSV). Th. Jacobsen
published a new reading, heavily restored, for the first ten columns, in “Translation
of the Stele of the Vultures, Cols. I–X,” in: S. N. Kramer Anniversary Vol. (= AOAT
25, 1976) 247–259. See now, however, H. Steible/H. Behrens, Die altsum. Bau- und
Weihinschriften I (= FAOS 5, 1982) 120–145.
2
On the issue of the Early Dynastic conflict, see now J. S. Cooper, Reconstructing
History from Ancient Inscriptions: The Lagash-Umma Border Conflict (= SANE 2/1,
1983). For its continuation beyond ED III, see G. Pettinato, “i7-Idigna-ta i7-nun-šè: Il
conflitto tra Lagaš ed Umma per la ‘Frontera Divina’ e la sua soluzione durante la
terza dinastia di Ur,” in: Mesopotamia 5–6 (1970/71) 281–320.
3
= AO 50+2436–2438+16109.
4
Fragments discovered during the 1880’s in excavations conducted by the French,
cf. L. Heuzey, “La Stèle des Vautours,” in: Gazette archéologique (1884) 164–180
and 193–203; A. Parrot, Tello, Vingt campagnes de fouilles (1948) 95. The purchased
piece was acquired by the British Museum in 1900 (BM 23580), and was given to the
Louvre at the time the restoration was done.
54 chapter seventeen
close examination of the stele, there are significant gaps—both in the

iconography and in the continuous text that covers all available back-
ground, obverse and reverse. The fragmentary nature of the completed
stele notwithstanding, however, we can with some assurance argue that
the obverse represents the god Ningirsu, tutelary deity of Girsu and
chief god of the pantheon of Lagaš as well.5 The god is shown holding
a battle-net filled with defeated enemies—an attribute associated in the
text only with gods; he also holds his emblem, the Anzu—symbolic
reference to the god’s own mythological history.6 The accompanying
figure behind (to the left of ) the god would then represent his mother,
the goddess Nin ursaga, whose characteristic headdress also includes
the head of the Anzu, an allusion to her significant role in her son’s
epic victory. Ningirsu is thus shown after the battle is over, with the
defeated forces of Umma clearly in his power—the “icon” of the state’s
victory.7
Perhaps because the reverse is divided into more registers, the losses
are more significant. We are tantalized with images of battle and
ceremony (cf. fig. 1); but there has been great difficulty in reading the
narrative here represented.8 In the assumed sequence of actual battle
(upper two registers) and celebratory ceremony (third register), one of
the greatest impediements to understanding has been the apparent
resumption of battle in the very fragmentary bottom register—where
5
This identification argued farther in I. J. Winter, “After the Battle is Over: The
‘Stele of the Vultures’ and the beginning of historical narrative in the art of the ancient
Near East,” in: H. L. Kessler/M. S. Simpson (eds.), Pictorial Narrative in Antiquity
and the Middle Ages (= Studies in the History of Art 16, Washington D.C., National
Gallery of Art, 1985) 11–32, esp. 13–16.
6
That the head of this emblem is indeed a lion, appropriate to the Anzu, and not
merely a bird’s head, can be seen clearly on pre-restoration photographs (kindly made
available to me by Agnès Spycket, Départment des antiquités orientales, Musée du
Louvre), before further damage was incurred.
7
As argued in Winter, Pictorial Narrative . . . (see note 5) 16 and 20.
8
Cf. H. Frankfort, The Art and Architecture of the Ancient Orient, 4th revised ed.
(1977) 71;—H. A. Groenewegen-Frankfort, Arrest and Movement: An essay on Space
and Time in the Representational Art of the Ancient Near East (1951) 158;—A. L.
Perkins, “Narration in Babylonian Art,” in: AJA 61 (1957) 58;—A. Moortgat, The Art
of Ancient Mesopotamia (1969) 63;—D. P. Hansen, “Frühsumerisches und frühdynasti-
sches Flachbild,” in: PKG 14 (1975) 189—of which, summary in Winter, Pictorial
Narrative . . . (see note 5) 16–21. See also V. K. Afanasyeva, “K probleme tolkovaniya
sumerskih rel’efov,” in: Kultura Vostoka: Drevnost’ i Rannoe Srednevekov’e (Leningrad
1978) 19–32, with English summary, “Reading Sumerian Reliefs,” 229–230—kindly
made available to me by Piotr Steinkeller in the Spring of 1986.
eannatum and the “king of ki”? 55
we see a hand holding a long spear-shaft, the point of which is directed

toward the forehead of a bald male facing left (fig. 2).
Comprehension of this fragmentary scene has been further compli-
cated by the identification of the enemy as the “King of Kiš”, based
upon the proximity of his head to mention of that ruler in the inscrip-
tion—despite the fact that no king of Kiš is mentioned in that part of
the text which describes the battle or its aftermath.
This association was first made by Thureau-Dangin, in his initial
reading of the text, where we do see two lines: the fragmentary syl-
lable al- presumed to be part of a proper name, followed by the title,
lug[al] Kiš[ki], “king of Kiš”.9 It was subsequently suggested by Th.
Jacobsen that the name should be read [Ka]-al-[bum], a ruler of Kiš
known from the Sumerian King List, with the ka- and the al- signs
written in ligature.10 This latter reading has been taken up by a number
of scholars, particularly archaeologists, although E. Sollberger and
J. Cooper have held out for the uncontorted reading of al-[. . .], an as-yet
unattested king of Kiš.11
It is this problem that I, as an archaeologist and art historian, would
like to address, since underlying the identification is the good archaeo-
logical problem of typology: that is, the question of how personal labels
(what have been called cartouches, borrowed not-entirely accurately from
the Egyptian practise of enclosing and setting apart rulers’ names)12 are
employed on the stele, and elsewhere in early Sumerian art.
Two other occurrences of labelling are preserved on the Stele of the
Vultures—both pertaining to Eannatum. Only one of these is absolutely
certain: clearly visible in the second register, where the ruler’s name
is incised before his forehead, with one of his heroic epithets: kur-gú-
gar-gar dNin-gír-su-[ka], “subjugator of (foreign) lands for Ningirsu”,
added behind (fig. 3). This “label” is quite distinct from the body of
the inscription proper. In the formal text, words are carefully aligned
in columns, and columns are subdivided into cases. Here, however,
9
Thureau-Dangin, RMSV p. 63; followed with qualifications by Hallo, Early Mes.
Royal Titles (= AOS 43, 1957) 27 f.
10
Th. Jacobsen, “Early Political Development in Mesopotamia,” in: ZA 52 (1957)
125, n. 77, and 133, n. 90.
11
Compare Steible/Behrens, ABW II (see note 1) 63, n. 105, with E. Sollberger,
IRSA 58, n. 12, and J. S. Cooper, Sumerian and Akkadian Royal Inscriptions I (1985)
n. 26.
12
Cf. A. H. Gardiner, Egyptian Grammar (1957) 13–14; 33, n. l.
the name and epithet are just free in the field around the ruler’s head,
without case markers.
The ruler in the first register seems to be similarly identified; how-
ever the entire identification is set behind the royal head (fig. 4). While
there is no closure to the cases at the beginning or at the end, divisions
between the three lines are set off by case-markers. What hampers
us in arguing with conviction that this, too, should be considered a
“cartouche” identifying Eannatum is the line of text above the possible
cartouche, and the break before his head, so that we cannot determine
for certain that this is not merely a fortuitous association with the end
of a line of text. In favor of its being a label, however, is the slightly
larger size of the cases and the more ample spacing when compared
with the text directly above, as well as the lack of case-division at the
beginnmg of the phrase—all of which suggest that the identification
may have been set in later.
In addition to open, or partially-open (non)cases, what characterizes
both of these examples, particularly that of register two, is the fact the
actual name of the ruler is always that part of the identification closest
to the depicted individual. Let us then look at several other examples
of such labelling on Early Dynastic monuments.
The closest, and clearest parallel is found on a fragment of a dedica-
tory plaque of Enannatum, nephew of Eannatum and subsequent ruler
of Lagaš (fig. 5). Here, too, the actual name of Enannatum is written in
direct proximity to his forehead, the rest of the label continuing below.
Since this example is dated after that of Eannatum, it is important to
demonstrate that we can bracket his reign with earlier representations
as well. Most important in this regard are the limestone plaques of
Ur-Nanše, grand-father of Eannatum, and founder of the E.D. III
dynasty of rulers at Lagaš. There are three such plaques extant;13 but
I show only the best preserved here, although all exhibit identical
characteristics (fig. 6).
In each of the three cases, we see the ruler standing facing right.
Immediately before his face is his name, and the rest of the inscription
reads out from his face: dNanše:ur [Ur-Nanše], lugal Lagaš, etc. Even
in the one case where Ur-Nanše is depicted seated, facing left, where
the label extends behind the seated figure, the text still moves outward
from his head, his name in closest proximity to his person.
13
Cf. Moortgat, The Art . . . (see note 8) Figs. 109, 111, 112.
Let us now return to the so-called identification of the King of Kiš

on the Stele of the Vultures. Taking our second point first, the broken
lines restored as “al-[. . .], King of Kiš” do not move outward from
the menaced figure’s head; rather, they move in the same direction as
the lines of text directly above. And as regards our first point, unlike the
other cartouches we have noted, there is clearly a case marker at the
beginning of the phrase, suggesting it was actually part of a longer
column of text no longer preserved. What is more, the juxtaposition of
this column (numbered rev. XII), to columns X and XI directly above,
identical in size and proportion, further argues that this mention of Kiš
should be seen rather as part of the longer inscription itself and not
as an identification of the enemy. That is to say, the verbal reference
would not be to the visual action at all, but rather belongs to the clos-
ing of the alas-so-fragmentary text.
This view is further supported by additional evidence. Upon close
examination, columns X, XI and XII are all identically aligned and end
at the same place (cf. fig. 2). [What is generally seen as a continuation of
col. X (ll. 23–37), is actually part of a “postscript,” an added, separate,
and complete statement set off from the rest of the column by a space,
complete in itself and apparently added secondarily.]14 Cols. X–XII,
then, each would have had space for approximately 19 cases, now
broken and missing; the cases preserved are numbered 20–22. Col. X
finishes with the titles of Eannatum; col. XI ends with a statement
of his having erected this stele for Ningirsu. How then to explain the
mention of the king of Kiš to follow?
As noted by Behrens/Steible in the Glossary to ABW,15 there are
relatively few references to Kiš in the texts of the period; only three
others are by Eannatum. The first is indeed within the text of the Stele
of the Vultures, but occurs quite early on, in the dream sequence in
which Eannatum gets inspiration from the god Ningirsu to undertake
his venture against Umma (Obv., col. VII 1–3: Ummaki/kiški-am6/šu
še-dag-ge, “As is Umma, so is Kiš, that it rushes about . . .”, an enigmatic
reference not fully understood.16 The second reference occurs in Ean.
2, a text of dedication to Ningirsu, in which, in a review of the deeds
of Eannatum, we are told that the “kingship of Kiš was given him”
14
Steible/Behrens, ABW I 144–145.
15
Behrens/Steible, Glossar zu den altsum. Bau- und Weihinschriften (= FAOS 6,
1983) 416.
16
Steible/Behrens, ABW I 125 and II 42, n. 41.
(col. VI 4), that “Kiš trembled before him” (VI 9), and, in a summary list
of opponents, that Kiš was among the cities he “struck with weapons”
(VI 21).17 Thus, the historical context is available to support a visual
representation of battle against a ruler of Kiš, and it is this reference
that has generally been cited as context for the bottom register of the
Stele. What is unsatisfactory in this, however, is that nowhere in the text
of the Stele itself is there any discussion of battles or historical events
subsequent to the victory over and re-claiming of disputed lands from
Umma. Furthermore, even in Ean. 2, where conflict with Kiš is men-
tioned, nowhere is a king of Kiš named, not even cited by title, although
other kings are so cited (for example, the “king of Akšak”, VI 10).
Clarification comes, I believe, in the third preserved text, Ean. 62, a
lengthy inscription on a stone mortar dedicated by Eannatum for the
E-ma of the goddess Nanše.18 There, not only Kiš, but the king of Kiš
is mentioned. As on the stele, the reference occurs toward the end of
the inscription—as part of a string of protective curses that traditionally
end texts on important monuments:
Side 4 III 2´ k[a . . .]

3´ na-dib-bé
4´ lugal *Kiški-bi
5´ na-dib-bé
Should the said mortar be removed from its place, or its inscription be
effaced, “X (presumably royal name or title) . . . shall not enter (‘pass’);
the king of Kiš shall not enter . . .” before the goddess Nanše.
With this text as parallel, and having eliminated on typological
grounds the reference to the king of Kiš on the Stele of the Vultures as
a “cartouche”, I would suggest that the lines under discussion must also
have been part of the closing of the narrative text, equally representing
an excerpt from the closing curses should the stele be destroyed.19
Notice now just where the fragment of register four falls in the
organization of the whole stele (fig. 1). With mention of the erection
of the stele in col. XI, directly above, this is exactly the proper place
17
ABW I 150.
18
See Steible/Behrens, ABW I 174–175, and especially Cooper, “Studies in Meso-
potamian Lapidary Inscriptions,” in: Iraq 46 (1984) 87–93.
19
Although Cooper, Iraq 46, 91, n. 10, notes that curses are relatively rare in
pre-Sargonic inscriptions, he is still able to cite Ent. 28/29, Ukg. 16 and Luzag. 2 in
addition to the Eannatum examples. The rarity is most probably due to our lack of
lengthy inscriptions on (stone) objects of value, as opposed to clay cones and tablets.
for some imprecation regarding the consequences if the stele were to

be disturbed—just as was the case for the mortar-stone.
The need to see the line preceeding the title, “King of Kiš” as the
personal name, al [. . .], also disappears.20 In any case, the whole portion
would be not part of a cartouche, but rather part of the inscription that
presumably would have continued into the now-missing portions equal
to cols. XIIIff., on to the end: most likely a final citation of Eannatum
and his special relationship with/dedication to the god Ningirsu.
It will be observed that this has been an essentially visual and typo-
logical, as opposed to philological, argument. With the bottom register
of the Stele thus freed from some reconstructed battle against Kiš, we
are now in a much better position to properly “read” the reverse. I
have suggested elsewhere how I believe this should be done.21 For the
present, I would like to lay aside any necessity to identify the adversary
in register four of the reverse as a king of Kiš.
Acknowledgments
This paper was first presented at the 195th Meeting of the American Oriental Society,
Ann Arbor, Michigan, in March of 1985. I have been much aided in the present effort
by the recent re-publication by Horst Steible and Hermann Behrens of all of the Early
Dynastic period texts from Tello, by the collection and discussion of those texts dealing
with the Umma-Lagaš dispute by Jerrold S. Cooper, and by the comments of both
Behrens and Cooper on the present ms.
20
In fact, further along in the mortar inscription (Ean. 62, side 4, col. IV 2´–3´,
6´–7´, etc.) are lines beginning with the sign AL: géštu-ni/al-zu-zu-a—pertaining
precisely to the curse formulae: if anyone damages/destroys/burns the object, and
“it be brought to the attention of ” the current ruler of Lagaš, “may . . . his personal
god never enter/pass (na-dib-bé) before Nanše.” The wording is close enough to the
mention of the king of Kiš just above that one could argue the lines on the Stele itself
should be restored similarly:
Rs. XII 20: [géštu-ni]
21: al[zu-zu-a]
22: lug[al] Kiš[ki]
XIII 1: [. . .]
. . . either as a reference to the early kings of Kiš having first established the border
between Umma and Lagaš and a subsequent one maintaining it, or as a warning to
a future king of Kiš not to encroach upon it, with dire consequences. In fact, in his
original reading, Thureau-Dangin observed (RMSV p. 3, n. 4) that there was room in
the AL-line for probably two more signs, and the second, of which a small part was
preserved, was “perhaps a ZU”! Since the final A sign would take very little space, his
observations strongly reinforce our reconstruction here.
21
Cf. Winter, Pictorial Narrative . . . (see note 5) 19.
Figure 1. Stele of Eannatum (Stele of the Vultures), reverse. Louvre, AO

50 + 2436–2438 + 16109.
Figure 2. detail, Stele of Eannatum, reverse, registers 3 & 4.
Figure 3. detail, Stele of Eannatum, reverse, register 2.

Figure 4. detail, Stele of Eannatum, reverse, registers 1 & 2.

Figure 5. Plaque fragment of Enannatum of Lagash from Tello. British

Museum WAA 1308228.
Figure 6. Plaque of Ur-Nanše of Lagash from Tello. Louvre AO 2344.

CHAPTER EIGHTEEN
WOMEN IN PUBLIC: THE DISK OF ENHEDUANNA,

THE BEGINNING OF THE OFFICE OF EN-PRIESTESS,
AND THE WEIGHT OF VISUAL EVIDENCE
Perhaps the first question one should ask about women in the art
of ancient Mesopotamia is: “Why so few?” Despite the recent and
comprehensive study of Julia Asher-Greve on old Sumerian women,1
and attention to women in special situations in recent analyses,2 the
number of actual roles in which women appear are relatively few when
compared with many other ancient cultures. As Rivka Harris has
recently demonstrated,3 roles played by women in the epic literature of
Mesopotamia are equally limited. Women are regarded favorably when
they advise, nurture and encourage men in their struggles for success
and fulfillment. Particularly valued is the mother-son relationship, and
those situations in which women act to maintain the norms of society
through ritual or through socially integrative action. And since these
auxilliary roles generally take place on the sidelines, as “catalysts” in the
narrative rather than as main characters,4 they fall outside of the usual
Mesopotamian focus on “culminating scenes” in the visual arts.5
* This article originally appeared as “Women in Public: The Disk of Enheduanna,

The Beginning of the Office of EN-Priestess, and the Weight of Visual Evidence,” in
La Femme dans le proche orient antique (Proceedings of the Rencontre Assyriologique Internationale,
Paris, July, 1986). J. M. Durand, ed. Paris 1987, pp. 189–201.
1
J. M. Asher-Greve, Frauen in altsumerischer Zeit [Bibliotheca Mesopotamica 18]
(Malibu, 1985).
2
For example, articles by K. Grosz, “Bridewealth and Dowry in Nuzi,” and U.
Jeyes, “The Naditu Women of Sippar,” in A. Cameron and A. Kuhrt, eds, Images of
Women in Antiquity, Detroit 1983, pp. 193–206 and 260–272, resp.
3
Cf. R. Harris, “Images of Women in the Gilgamesh Epic,” Abstracts of the 196 th
Meeting of the American Oriental Society, New Haven, 9–12 March 1986, p. 16.
4
Cf. R. Barthes, “Introduction to the Structural Analysis of Narratives,” in Image,
Music, Text, translated by S. Heath (New York, Hill and Wang, 1977), pp. 79–124 and
esp. pp. 93–4.
5
Cf. A. Perkins, “Narratation in Babylonian Art,” American Journal of Archaeology 61
(1957) 54–62.
66 chapter eighteen
Notable exceptions to this occur in the representations of female

deities—not our subject today—and of elite women in socially-sanc-
tioned public roles.
One role that permitted high public visibility was that of the wife
of the incumbent ensí or local chief administrative officer of a given
city-state; the other that of high-ranking priestess in the cult of a major
(male) deity. In the first case, in the Early Dynastic period in Lagash,
for example, it would appear that a significant portion of the state
economy was under the control of the ensí’s wife.6 These are also the
women for whom we have inscribed statues, particularly from Girsu/
Lagash in the Neo-Sumerian period;7 while in the Ur III period, for
which we have an unusual amount of information concerning royal
and highly-placed women,8 we see that the imagery on personal seals
of the wives of ensís have been tailored in subject matter to their
female owners. For example, two seals have been preserved belonging to
Ninªilia, identified by inscription as the wife of A’akalla, ensí of Umma
under Amar-Su’en and Šu-Su’en, along with one seal belonging to
the wife of the previous ensí, Ur-Lisi.9 The husbands of both of these
women had seals that I would like to designate as official seals of the
administrative bureaucracy of the state;10 the womens’ seals, however,
showed quite different imagery. The seals of Ninªilia are particularly
interesting, because the seated figures in the presentation scenes —often
male deities or kings—are here goddesses; and on the one seal where
the presentee is preserved, that figure, too, is a woman—presumably
the seal-owner herself.
These women are elite, but not themselves officials, and their seals
probably represent them in the presence of a patron-deity, not neces-
sarily one in whose cult they function. In the other case of publicly-
sanctioned roles for women, that of highly-placed priestesses in major
6
Asher-Greve, Frauen, p. 169.
7
Cf. A. Spycket, La statuaire du Proche-Orient ancien (Leiden, E. J. Brill, 1981), pp.
188–189, 192, 198–203.
8
Cf. P. A. Parr, “Ninhilia: Wife of Ayakala, Governor of Umma,” JCS 25 (1974)
90–111; P. Michalowski, “Royal Women of the Ur III Period, Part II: Geme Ninlila,”
JCS 31 (1979) 171–176; P. Steinkeller, “More on the Ur III Royal Wives,” Acta Sume-
rologica 3 (1981) 77–92.
9
Parr, op. cit.
10
I. J. Winter, “Legitimation of Authority through Image and Legend: Seals belong-
ing to officials in the administrative bureaucracy of the Ur III state,” in The Organization
of Power: Aspects of Bureaucracy in the Ancient Near East, M. Gibson and R. D. Biggs, eds.,
Chicago, The Oriental Institute, forthcoming.
women in public 67
cults, the women are often represented during cult function, not merely
in presentation scenes.11 We also know, mainly from the later Old
Babylonian period, that priestesses could own property and transact
business, even in cities other than their primary cult residences, and
were active in many levels of economic and social life.12
Just how easily women attained this status is not entirely clear from
the written record, however. Certainly among the most prominent
female holders of priestly office was Enheduanna, who was established
as en-priestess in the cult of the moon-god, Nanna/Su’en, at Ur dur-
ing the reign of her father, Sargon of Agade. She is known to us both
from her literary compositions and from a variety of inscribed objects,
including the carved alabaster disk to be discussed below.13
From these, we get a tantalizing glimpse of a strong and creative
personality; but it is less the individual than the office that I would like
to concentrate on here—perhaps shedding some light in the process
on the institution of en-priestess, as it can be reconstructed from visual
sources.
I certainly wish we had a seal of Enheduanna to compare with the
Ur III examples.14 Nevertheless, we may now turn to the disk of Enhed-
uanna, where the horizontal band of imagery carved on the obverse is
itself not unrelated to the relief impression of a seal rolling.
The disk is of translucent alabaster; it measures ca. 25.6 cm in
diameter and 7.1 cm. thick; and is today in the collection of the Uni-
versity Museum, Philadelphia (fig. 1). It was found in several fragments
during the course of excavations in the area of the giparu at Ur in
11
Cf. Asher-Greve, passim.
12
See on this, Jeyes, in Images of Women, cited above, n. 2., esp. pp. 270–271. N.
Yoffee, “Old Babylonian Texts from Dilbat in the Ashmolean Museum,” Abstacts of the
196th Meeting of the American Oriental Society, New Haven, 9–12 March 1986, No. 142,
has evidence for naditu-priestesses of Shamash, resident at Sippar, retaining rights and
property in Dilbat.
13
A number of studies have been devoted to this intriguing personnage. See par-
ticularly, A. Falkenstein, “Enheduana, die Tochter Sargons von Akkade,” RA 52 (1958)
129ff., and the discussion in W. W. Hallo and J. J. A. van Dijk, The Exaltation of Inanna,
New Haven, 1968, esp. Introduction. Also A. Sjöberg, “in-nin ša3-gur4-ra: Hymn to the
Goddess Inanna by the en-Priestess Enheduanna,” RA 65 (1975) 161–253.
14
Although we do have seals of at least three of her servants: cf. E. Sollberger, “Sur
la chronologie des rois d’Ur et quelques problemes connexes,” AfO 17 (1954–6) 26–9,
regarding the seals of Ada, Dingir-palil and Kitušdug (= UE II, p. 311 and pl. 212, Nos.
307, 308 and 309). The imagery in all three cases is that of animal-contest scenes.
68 chapter eighteen
1927, and has been heavily reconstructed.15 On the reverse, the disk is
inscribed in a column of 11 cases, identifying Enheduanna as the wife
(dam-) of Nanna and daughter of Sargon (ll. 3–4).16 On the obverse,
a central frieze of figures is carved in very high and well-modelled
relief. At the left, a libation is poured by a nude male “priest” from a
spouted vessel into a plant stand set before what has been restored as
a four-staged temple tower.17 Behind (to the right of ) the priest is an
array of three figures, including a well-appointed woman in turban and
flounced garment followed by what appear to be two bald, probably
male, attendants.
Given the placement and delicate attention to detail of the figure
of this central woman, she has been identified as Enheduanna.18 She
faces in the direction of the cultic activity, her right arm bent at the
15
CBS 16665 (Ur 6612). Note that the disk was found in the court of the more
public cult area of the giparu, and is therefore designated as having been found in the
“Temple of Ningal” in early excavation reports (e.g., C. L. Woolley, UE II, p. 334).
Woolley further suggests (UE IV, p. 49) that the stone may have been deliberately
defaced, and if so, one may not argue that the “temple” area of the giparu was the
original locus for the object, despite the temptation to see it as having been publicly
visible. [On the giparu, see P. Weadock, “The Giparu at Ur,” Iraq 37 (1975) 101–128,
and D. Charpin, Le clergé d’Ur au siècle d’Hammurabi, Paris 1986, pp. 192–220, where
he demonstrates the likelihood that what Woolley called the “Temple of Ningal” was
in fact the Ekišnugal, mentioned in texts of the time of Warad-Sin and Rim-Sin of
Larsa (cf. esp. p. 220)]. Amiet (in PPK 14, p. 195) has even suggested that the circular
form of the disk, unique in the archaeological record to date, may be a purposeful
reference to the disk of the full moon of Nanna—although the problem with this is
our lack of knowledge of the accuracy of the present restoration, if indeed the edges
had been defaced in antiquity.
16
Cf. UE I: n. 23; UE IV: Pl. 41d; E. Sollberger and J. Kupper, Inscriptions royales
sumeriennes et akkadiennes (Paris, 1971) IIA i d, restore the text, partly based upon an Old
Babylonian copy (UET I, 28a), as: “E., true lady of Nanna, wife of Nanna, daughter
of Sar[gon, king of all, in the temple of Inan]na-za-za [of Ur, a dais you built (and)
‘Dais, table of heaven [AN]’ you called it.” [See also, Sollberger, RA 63, p. 180, 16.
It should be noted that, since the actual word ‘dais’ is missing on the disk, there is
on guarantee that THIS inscription commemorated such a construction, as opposed
to any other. In any case, we still have no hint as to what the function of the disk
may have been, as dedicatory inscriptions on specific objects almost never name the
object itself, just the fact of its dedication. Nor would it make much sense, to have the
dedicatory inscription on the disk refer to a dais, much less to one set up in the temple
of Inanna, when the iconography of the disk, as will be seen below, seems to refer to
the god Nanna (cf. fn. 24, below). I fear, then, that the OB tablet used to reconstruct
the disk’s text should be used only as a ‘model’ for the form of the inscription, not
as an exact copy.]
17
Note that the upper stage has been entirely reconstructed. In fact, there is evidence
only for the lower three stages.
18
E.g., A. Moortgat, Archaeology of Ancient Mesopotamia (London and New York,
1969), p. 48.
women in public 69
elbow, hand raised before the face, in a well-known gesture of pious

greeting, comparable to those depicted in presentation scenes, from Ur
III seals to the Code of Hammurabi, and finding its literary referent
in the Sumerian verb “to greet”—kiri4 šu-gal2—literally, “to let the
hand be at the nose.”19 The cap she wears has a distinctive rolled-brim,
and has been identified by Renger with the aga—special headgear of
priestesses, particularly Enheduanna.20 But it is important to note that
as en-priestess, she does not pour the libation herself. Rather, she stands
in worship as a male functionary performs the ritual act.21
Compositionally, the nude priest and the highly-ranked en-priestess
share the center of the disk—bound together by his ritual performance
and her overseeing of it (itself an insight into the en-priestess’ role in
ritual, or lack thereof ). Yet, if the reconstruction of associated figures
is accurate, only Enheduanna’s head actually touches the upper margin
of the frieze—the violation of isocephaly serving to emphasize visually
her dominant position. This correlation between rank and relative size,
so well articulated by Meyer Schapiro for European medieval art,22 and
known as well in Egyptian works, first appears in Mesopotamia in the
Early Dynastic III period with the so-called “Standard of Ur” and the
plaques of Ur-Nanše of Lagash—where the primary figures are also
measurably larger than attendants and associates.23
19
Cf. H. Frankfort, Cylinder Seals (London 1939), Pls. XXV–XXVII, passim; Moort-
gat, AAM, fig. 209. In fact, in a hymn of Rim-Sin of Larsa to Nanna and Ningal
(UET VI, 106, republished by Charpin, Le clergé, 295ff., esp. 1. 5), this very gesture is
described in the context of Rim-Sin’s holy (pure) food offerings and libations offered
to the gods, “his hand brought to the nose, his pure hand raised. . .” [kir4-šu-gál-la-zu
šu-kù-il-la-zi]—an almost exact image of what is depicted on the disk.
20
On this, see J. Renger, “Untersuchungen zum Priestertum in der altbabylonischen
Zeit,” ZA 58 (1967) 110–188 and esp. p. 126 and note 100; also Hallo and van Dijk,
Exaltation, Nin-me-šar-ra, 1. 107: aga-zi/nam-en-na, “the true cap/the sign of (appro-
priate to) en-ship”, in Enheduanna’s hymn to Inanna.
21
This male figure may well represent the PA4-E, or “anointing priest”, known to
have played a role in the temple cult of Su’en at Khafaje as early as the Early Dynastic
period (cf. H. Steible, ABW II, pp. 210–211, and discussion in Mark G. Hall, A Study
of the Sumerian Moon-God, Nanna/Su’en, un-published Ph.D. dissertation, University of
Pennsylvania, Philadelphia, 1985, p. 94 and 99). See also Hall, Nanna/Su’en, p. 238,
n. 30, for discussion of various water containers associated with the cult of Nanna.
Silver vessels of the type represented on the disk have been found in tombs of the
Royal Cemetary (= Type 84, known from PG 800, PG 755, etc.; cf. UE II, p. 282
and Pl. 171).
22
M. Schapiro, “On Some Problems in the Semiotics of Visual Art: Field and
Vehicle in Image Signs,” Semiotica 1 (1969) 223–242.
23
H. Frankfort, Art and Architecture of the Ancient Orient, 4th rev. ed. (Harmondsworth
and New York, 1977), fig. 76, and Moortgat, AAM, figs. 109, 111, 112.
70 chapter eighteen
It is tempting to see the stepped architectural element at the far

left as an off-scale representation of the temple tower or ziggurat of
Nanna—known to have been so-staged in the later Ur III period.24 We
do have some precedent for cultic or processional activity outside of
shrines from cylinder seals as early as the Uruk period.25 But closest
perhaps is that seen on the lower register of a votive limestone plaque
found in the Early Dynastic levels of the giparu at Ur (fig. 2).26
There, as on the disk, a nude male pours a libation from a spouted
vessel into a date-palm stand before the exterior of a building, while
three figures stand behind him. The last two figures seem to be atten-
dants—one possibly female, the other male and carrying a kid or lamb,
presumably as sacrificial offering. The lead figure in the group is female,
and is made to dominate the array not by added height, but rather by
her frontal posture. The position she occupies is identical to that taken
by Enheduanna on the disk, and it should be noted also that both wear
the same rolled-brim turban as headgear.
The plaque falls clearly into a stylistic group of ED III votive tablets,
as, for example, from Tello, where another nude male priest pours a
libation before a seated figure identifiable by attributes as the goddess
Ninhursag (fig. 3).27 These plaques are surely pre-Akkadian; yet their
contents clearly mirror that of the disk.
The upper register of the Ur plaque is particularly interesting. Here,
a libation is poured before a seated male deity. Three women—all
cloaked and wearing the same wide-brimmed headgear—stand behind
the libating figure, still identifiable as male, but now represented with
24
Cf. comment by Woolley, UE II, p. 334. This would follow the tradition in Akka-
dian and Early Dynastic seals of showing an architectural form very small in relation to
human or divine figures in scenes of “building a ziggurat”—e.g., P. Amiet, La glyptique
mésopotamienne archaique, Paris 1980, 1441–8, and discussed recently by L. al-Gailani
Werr, “Catalogue of Cylinder Seals from Tell Suliemeh, Himrin,” Sumer 38 (1982)
70–72, nos. 7–10. I should observe here that the only way to reconcile Sollberger’s
reconstruction of the word “dais” in the dedicatory inscription with the imagery on the
disk would be if the word, bara2, were to refer not only to single platforms, but also to
superimposed ones, where the scene would then show ritual activity before the VERY
platform mentioned in the dedication (but cf. PSD ‘B’, Philadelphia 1984, 134ff., esp.
1. 3, where the term can be used as an epithet of a temple or a city as a whole, but
does seem to always imply the “base” or “platform” of a dais.
25
Frankfort, Art and Architecture of the Ancient Orient, fig. 27.
26
= BM 118561; cf. UE IV, Pl. 39c, Moortgat, AAM, fig. 116, and Asher-Grève,
Frauen, n. 557 and pp. 88–89, whose discussion, available only after the initial draft of
the present paper was completed, closely parallels my own.
27
Cf. Moortgat, AAM, fig. 114. The Ur plaque has been dated by J. Boese, Altmesopo-
tamische Weihplatten [ UAVA 6] (1971), p. 80, to ca. the time of Enmetena of Lagash.
women in public 71
long hair. It has been argued elsewhere that narrative order in Sume-
rian art is such that the temporal sequence proceeds from bottom to
top (for example, on the battle side of the Standard of Ur, as one
proceeds from chariots in combat to prisoners to presentation before
the central figure, and also on the reverse of the Stele of Eannatum,
known as the “Stele of the Vultures”).28 I would suggest that the lower
register of the Ur plaque equally represents temporal anteriority, in
the procession, offering and libation outside the temple, prior to ritual
performed before the resident deity within the shrine—just as on the
Stele of Ur-Nammu of the Third Dynasty of Ur, the scene of temple
building at the bottom preceeds the ritual activity within the temple
at the top.29 On this basis we might even suggest that certain aspects
of the ritual activity, such as animal sacrifice, took place out of doors
and not inside; and this seems to be confirmed by a text from Emar
dealing with the installation of the en-priestess of the Storm God, in
which the sacrifice of animals on the second day preceeds the entrance
of the priestess into the temple.30 In any case, because Enheduanna
represents the first written attestation of en-priestess (Akk. entu), it has
occasionally been argued that she marks the first occurrence of the
office. Yet, considering the findspot of this plaque in the giparu and
the presence of female participants in the ritual, as well as the striking
parallels with the Enheduanna disk, a case could be made for suggest-
ing that this plaque, too, shows an en-priestess, or en-priestesses, both
before the shrine and in the presence of Nanna.31
28
Cf. Winter, “After the Battle is over: The Stele of the Vultures and the beginning
of Historical Narrative in the Ancient Near East,” in H. Kessler and M. S. Simpson,
eds., Pictorial Narrative in Antiquity and the Middle Ages, Studies in the History of Art 16
(Washington, National Gallery, 1985), pp. 11–32.
29
Frankfort, Art and Architecture of the Ancient Orient, fig. 110.
30
Cf. D. Arnaud, Recherches au pays d’Astarté: Emar VI: 3, Textes sumeriennes et akkadiens,
Paris 1986, pp. 329–335, esp. ll. 7–8. This is further supported by a late text describing
the “Festival of Ishtar” (publ. by S. Lackenbacher, “Un nouveau fragment de la ‘Fête
d’Ištar’,” RA 71 (1977) 39–50, esp. ll. 26’–27’, in which the priest pours a libation from
a gold vase outside the temple before the king and goddess enter the sanctuary.
31
The point argued here is that the visual evidence suggests rather continuity in
this office, from at least Early Dynastic III. It is possible that the Akkadian period is
the first in which one individual emerges as paramount from what had hitherto been
a group of priestesses, or has such high rank (here, and in many subsequent cases,
the daughter of a king), that the office takes on new significance. This could partly
explain the use of the same Sumerogram, nin-dingir, for two Akkadian words: entu and
ugbabtu—the latter a title of lower rank priestesses, and translated as such when the
context requires (cf. CAD “E”, 1958: entu, pp. 172–3 and Renger, ZA 58, pp. 134–149).
72 chapter eighteen
The case is strengthened by a cylinder seal of the Early Dynastic

period (fig. 4),32 which shows the composite scene of the plaque collapsed
into a single register: woman in cape and widebrimmed hat behind
long-haired male holding a spouted vessel before both a seated deity
and a temple facade.33 On stylistic grounds, the seal is of E.D. III date.
Although there is nothing in the god’s attributes or headgear to identify
him, at the god’s feet are a pair of addorsed bulls or calves—animals
known to have been associated with the moon-god, Nanna, in this and
other periods.34 In addition, the careful attention to rendering the rolled-
brim cap of the female figure can only be understood in light of the
aga, the special cap associated with the en-priestess, referred to above
Moortgat further suggests (AAM, p. 42) that the scene on the plaque actually represents
the induction of the nin-dingir to office.
32
Currently in Berlin, cf. Moortgat, Vorderasiatische Rollsiegel (Berlin, 1940), No.
144.
33
That the figure is male is clear from the rolled waist of his skirt and bare chest.
This garment is typically worn by men, in contrast to the standard women’s garment
that covers the breast and wraps over the left shoulder. In an early study of Mesopo-
tamian libations ( J. Danmanville, “La libation en Mésopotamie,” RA 49 (1955) 57–68),
it was determined from a sample of some 150 monuments that ritual nudity seems
to disappear after the Ur III period, and may have been a Sumerian practice (p. 67).
This early example of a clothed individual pouring a libation suggest a mixed practice
even at this early date.
Note also the rampant goat-construction in the upper field. Moortgat has called this
an offering table (VAR, p. 96), on which two dishes and bread have been placed, and
this does have good visual parallels in seals where food offerings are clearly placed on
tables (cf. UE II, Pl. 193: 21). In addition, it has been suggested that the two rampant
goats from the Royal Tombs at Ur once formed the supports of a table (cf. Woolley,
ibid., p. 265, re Pls. 87–90). However, I should also like to point out that the triangular
placement of the vertical struts are equally congruent with the frame of the musical
instruments also found in the Royal Tombs, one of which had as a protome a rampant
goat (Ibid., Pls. 75, 11 and esp. 112). In that case, the vertical strokes at the top would
be tuning pins, not pieces of bread set on edge. In favor of this is the frequent mention
of the balag—“lyre”—in the offering texts of the Ur III period (cf. PSD ‘B’ (1984),
p. 75), suggesting that the lyre was a regular feature of the cult. Also in favor of the
goat-as-cult-lyre are the two horizontal lines below, which are standard in Sumerian art
for the representation of a dais. While it might not be inappropriate to set an offering
table on a platform, we have no precedents for such practice. By contrast, the dais is
frequently mentioned as allocating sacred space, on which kings or gods are seated
(cf. J. Klein, Three Šulgi Hymns: Sumerian royal hymns golorifying King Šulgi of Ur [Ramat
Gan, Bar Ilan University Press, 1981], Šulgi X, 1. 57). Such a dais might be more
appropriate for display of the balag in the cult of the moon-god—the care and feeding
of which as a ritual object is well-attested (cf. PSD ‘B’, 75f.).
34
Cf. Hall, Nanna/Su’en, p. 94 et passim, where the epithets amar-banda-dEnlil,
“beloved calf of Enlil”, as applied to Nanna/Su’en, and gu4-dEN.ZU, “bull of Su’en”,
are discussed. See also on this, A. Sjöberg, Der Mondgott Nanna-Su’en in der sumerischen
Uberlieferung (Stockholm, 1960), p. 24.
women in public 73
(note 20) and indicated here to assure identification. I would therefore

suggest that the representation on this seal is precisely a conflation of
what we have seen on the E.D. III votive plaque, with en-priestess and
male functionary before the god Nanna, whose temple facade we see
at the far right.
In all three cases, then—seal, plaque and disk—we are presented
with the same visual theme: woman identifiable through attributes or
titles as en-priestess in the cult of the moon-god, Nanna/Su’en at Ur,
and shown present at rituals pertinant to the cult, either before the
god’s shrine or in his presence.
There are significant stylistic and compositional shifts from the Early
Dynastic to the Akkadian periods—particularly in the greater modelling
and higher relief of the Enheduanna plaque, and the more sophisticated
visual device of substituting slightly greater height for frontality in the
chief female figure, such that her prominence is discernable without
interrupting the narrative focus on cult activity. But, those shifts notwith-
standing, essential components in each of the scenes remain constant,
and argue strongly for continuity of activity and function as well.
Now, it must be remembered that the earliest references to the moon-
god occur in texts from Ur, Khafaje, Abu-Salibiª and Fara dated to the
Early Dynastic IIIA period, ca. 2600 b.c., while the title of en-priestess
employed by Enheduanna is not attested before her usage.35 Neverthe-
less, Renger, and after him, Hall, have argued that, since the appellation
MUNUS-NUNUZ-ZI-NANNA (“true lady of Nanna”) employed by
Enheduanna on the disk inscription (1. 2) may go back to the earlier
MUNUS.ZI.(ZI).dNANNA of the Fara and archaic Ur texts, the very
tradition of the en-priestess may also date to this time.36 I would suggest
that the close visual continuity in representations of officiating women
in the cult of the moon-god from the E.D. III to the Akkadian period
speaks eloquently in favor of the existence of the priestly office at that
earlier date as well.
That there should be such continuity is not at all surprising. Even
without textual evidence for the early office of en-priestess, it has been
assumed that Enheduanna was placed in the service of the moon-god
35
Cf. Hall, Nanna/Su’en, pp. 98, 107ff.; Sollberger, AfO 17, p. 26.
36
Renger, ZA 58, p. 115; Hall, Nanna/Su’en, pp. 108 & 238, n. 26. The use of the
title for a male as early as (En) Me-barag-si, as demonstrated by Edzard, would also
support this (cf. ZA 53, 1959, 9–26).
74 chapter eighteen
precisely to consolidate the Akkadian dynasty’s links with the traditional

Sumerian past in the important cult and political center of Ur.37
Again, the visual evidence speaks eloquently in this regard. In present-
ing characteristic works of Sumerian and Akkadian art, the differences
have more often been emphasized—citing, for example, the Stele of
Eannatum as contrasted with the Stele of Naram-Sin, and indicating
the latter’s lack of registration, landscape elements and dominance of
the royal, as opposed to the divine, figure. Yet, if one looks to works
earlier in the Akkadian dynasty, one sees the intermediary steps in what
was clearly a continuous process of development.
On an early Akkadian stele now convincingly demonstrated to have
belonged to Rimuš, son and immediate successor of Sargon,38 we see
a series of battle-encounters divided into horizontally-stacked registers,
not unlike the upper registers of the reverse of the Stele of the Vultures.
The same registration can be observed on a stele-fragment of Sargon,
found at Susa, where the king is identified by a cartouche before his
face, just as Eannatum is identified on the Stele of the Vultures, and
as a number of Early Dynastic kings are similarly identified (cf. figs. 5
and 6).39 And finally, on another fragmentary Sargon stele from Susa,
we see a topos similar to that of the Stele of the Vultures in enemy
captured and tumbled about within a net (figs. 7 and 8).40
One could say that this last Sargon stele was in some ways intermedi-
ary between the Stele of the Vultures and the Stele of Naram-Sin, for it
is the king, Sargon, who wields the net here, not the god Ningirsu, while
his patron deity, Ištar, sits at the right. Nevertheless, the maintenance of
Sumerian dynastic topoi of victory and traditions of labelling attest to
a general attitude of conservatism early in the Akkadian period. This
would in fact have been an extremely important way of establishing the
legitimacy of the new Akkadian dynasty, by taking over and deploying
elements of the old symbolic system in its own behalf.
37
See on this, Hallo and van Dijk, Exaltation, pp. 7–9; also P. R. S. Moorey, “What
do we know about the people buried in the Royal Cemetary,” Expedition 20 (1977) 38.
Moorey notes that the archaeological evidence for Early Dynastic levels of the giparu
are sketchy; however, it is his thesis that the graves at Ur are best explained as burials
of cult personnel, not royal family-members per se, and thus, on independent grounds,
argues that the institution of en-priestess was well-established by the E.D. III period.
38
Cf. B. Foster, “The Sargonic Victory Stele from Telloh,” Iraq XLVII (1985)
15–29.
39
For example, Ur-Nanše and Enannatum of Lagash, cf. Moortgat, AAM, fig. 109
and E. Strommenger, 5000 Years of the Art of Mesopotamia (New York, 1964), fig. 71.
40
Moortgat, Ibid., figs. 126–127.
women in public 75
It is in this context that we must see the strategic placement of the

king’s daughter, Enheduanna, in a traditional office at Ur.41 The one
difference we may note between the Early Dynastic (Sumerian) and
Akkadian representations of the en-priestess attending ritual functions
is the careful inscription identifying the person of Enheduanna on the
reverse of the disk. One likely explanation for this phenomenon—in
addition to the strength of ego and presence of the individual—is that
part of the significance of her occupying that office lay precisely in her
Akkadian identity: as daughter of Sargon, and sister of the succeeding
rulers, Rimuš and Maništusu, she serves through her office to solidify
the traditional base of the whole Akkadian regime. This would only
have been possible had the office of en-priestess already existed in
Sumerian cultic practice.
In closing, I would emphasize four points. First, it is important to
underscore the limited number of ways in which women were rep-
resented in Mesopotamian art. In order to demonstrate that this was
indeed a reflection of the limited number of roles available to women
in public life, one would have to probe further into those cases in which
women do emerge—situations in which they own property, exercise
economic and cultic offices; with whom they have contact, and even,
who their witnesses are on tablets (family members or functionaries?).42
That task I turn over to my text-oriented colleagues, noting however
that there certainly was in Mesopotamia nothing like the public visibility
of women of the royal household in Republican and early Imperial
Roman art, who were openly extolled and displayed for their civic and
familial virtues (as on the Ara Pacis of Augustus).43 And one must won-
der whether this general invisibility of women in Mesopotamia does
41
A tradition continued by subsequent kings of the dynasty; for example, a daughter
of Naram-Sin also served as en-priestess of Nanna in the later Akkadian period (cf.
Hall, Nanna/Su’en, p. 111), while a daughter of Išme-Dagan of Isin, and a sister of
both Warad-Sin and Rim-Sin of Larsa held the office in the early second millennium
(cf. Charpin, Le clergé, p. 195).
42
I quote from a comment by Martha Roth (personal communication, 5 May 1986)
on reading this: “My own work (on the Neo-Babylonian period) has found that in a
great many cases, when a woman acts seemingly independently, her husband, father
or brother will be the first recorded witness—almost certainly thereby signifying his
consent to the transaction. Whether such consent was a legal prerequisite or not
remains to be seen.”
43
On this, see D. Kleiner, “The Great Friezes of the Ara Pacis Augustae,” in Mélanges
de l’Ecole française à Rome 90.2 (1978) 753–785, and comments in N. B. Kampen, “The
Muted Other,” paper given as part of sympsium on Classicism” at the College Art
Association annual meetings, New York, February 1986.
76 chapter eighteen
not reflect deeper cultural mores regarding distinctions of public and

private along gender lines—a practise entirely consonant with parts of
the Middle East today. That is to say, if the subjects of public art are
public themes, and the protagonists of public acts are seen to be men,
then the absence of women in public art asserts their lack of public-
ness, to the extent that “woman” is to be equated with “private”—our
priestesses being a notable exception.
Second, it has been argued that the ritual office of en-priestess of
Nanna had to have pre-dated Enheduanna, occurring at least as early
as the Early Dynastic III period,44 and in addition, some of the office’s
associated activities have been delineated: that the en-priestess did not
herself perform ritual libations; that the ritual sequence began with
activities outside of the shrine and then continued inside the sanctu-
ary; that animal sacrifice is likely to have been performed out of doors,
not inside.
Third, the situation of Enheduanna, daughter of Sargon, as en-
priestess of Nanna/Su’en at Ur should be seen in the larger context
of a display of ritual continuity with the past that served some of the
political ends of the early Akkadian dynasty—particularly divine sanc-
tion for its legitimacy.
And finally, fourth, it will be noted that throughout, I have con-
structed a plea for the weight and significance of visual evidence, even
in the absence of textual confirmation, by suggesting that we may legitimately
use the votive plaque from Ur and seal of the Early Dynastic period
to demonstrate the existence of the office of en-priestess prior to writ-
ten attestation. In such a way, we see the extraordinary visual record
preserved in Mesopotamian art not merely as illustration of (and hence
secondary to) texts, but rather as a highly-developed system of com-
munication in its own right.
44
There is another line of argument, not pursued here, that could support our
case, and even suggest pushing the likely date of the office back beyond the Early
Dynastic III period. As Edzard has shown (D. O. Edzard, “Enmebaragesi von Kiš,”
ZA 53 (1959) 9–26, and esp. l0ff. ), the alternative writing Me-bara(g)-si for Enme-
baragesi argues strongly in favor of the initial EN- already signifying the priestly office
by E.D. II. As we know, since the office of en-priest is gender coded to the opposite
sex of the deity being served, if there is a male en in the period, and if there is also
evidence for male gods, then there is likely to have been an office for female en’s as
well. In fact, since the really formative period for urban Mesopotamian cult activity
seems to have been the Uruk/Jemdet Nasr period, it would not be entirely untoward
to assume that the office—both male and female—would have been in existence as
long as the religious system.
women in public 77
Acknowledgments
My sincere thanks to H. Behrens, M. Hall, N. Kampen and M. Roth for comments

on an early draft of this paper.
78 chapter eighteen
Figure 1. Disk of Enheduanna, from the giparu at Ur. Akkadian period;

alabaster, diam. 25.6 cm. [ The University Museum, Philadelphia,
CBS 16665].
women in public 79
Figure 2. Votive Plaque, from the giparu at Ur. Early Dynastic III period;
limestone, ht. 22 cm. [ The British Museum, London, WAA 118561].
Figure 3. Votive Plaque, from Tello. Early Dynastic III period; limestone,
ht. 17 cm. [ Musée du Louvre, Paris, AO 276].
80 chapter eighteen
Figure 4. Seal (impression), from Umma (?). Early Dynastic III period;
translucent greenish marble, ht. 4.3 cm. [Staatliche Museen, Berlin,
VA 3878].
women in public 81
Figure 5. Sargon stele, found on the acropolis, Susa. Akkadian period;

diorite, ht. 91 cm. [ Louvre, Paris, Sb1].
Figure 6. Stele of Eannatum, from Tello, detail: reverse. Early Dynastic III
period. [ Louvre; cf. caption, fig. 8].
82 chapter eighteen
Figure 7. Sargon stele in the round, found at Susa. Akkadian period;

diorite, ht. 54 cm. [ Louvre, Paris, Sb2/6053].
women in public 83
Figure 8. Stele of Eannatum, from Tello, obverse. Early Dynastic III

period; limestone, ht. 1.8 m. [ Louvre, Paris, AO50+2346-2348+16109].
CHAPTER NINETEEN
SEX, RHETORIC, AND THE PUBLIC MONUMENT:

THE ALLURING BODY OF NARAM-SÎN OF AGADE
Sex
What, the modern viewer may ask, is Naram-Sîn of Agade doing on

his Victory Stela (cf. figs. 1–3), displaying for us not only his victory in
battle but his well-rounded buttocks, his muscled calves, his elegantly
arched back, his luxuriant beard? More Baryshnikov than Stallone, he
is nonetheless, within our cultural lexicon of value, well proportioned,
lithe, fit, and simply “divine”!
That the Stela of Naram-Sîn has found a responsive audience in the
West since its discovery in 1898 is clear from its inclusion in virtually
every survey of art, from Gardner to Janson.1 However, if we are to
pursue what this particular mode of representing the body of Naram-
Sîn might have meant in his own time (c. 2254–2218 b.c.e.), it must be
viewed, insofar as we are able to reconstruct it, within a Mesopotamian
lexicon of value. At issue is whether those values include sexuality, along
with other positive qualities of form and substance. By virtue of the
title of this chapter, my position is clear. The problem posed for the
modern viewer by the eroticized body of the political leader was not a
problem for the ancient Mesopotamians; rather, I shall argue, sexuality
was inextricably linked to potency, potency to male vigor, and male vigor
to authority and dominance, hence rule. Indeed, for the royal body,
a rather unusual set of ideal attributes emerges in the Mesopotamian
* This article originally appeared as “Sex, Rhetoric and the Public Monument:
The Alluring Body of the Male Ruler in Mesopotamia,” in Sexuality in Ancient Art,
N. B. Kampen et al., eds. Cambridge and New York: Cambridge University Press,
1996, pp. 11–26.
1
See the most recent publication of the stela, in P. O. Harper et al., The Royal City
of Susa: Ancient Near Eastern Treasures in the Louvre (New York: Metropolitan Museum of
Art, 1992), #109, for bibliography from discovery to date. General surveys include
H. W. Janson, History of Art (Englewood Cliffs, N.J.: Prentice-Hall, 1962), Fig. 76, and
H. Gardner, Art Through the Ages, 6th ed. (New York: Harcourt Brace, 1975), Figs.
2–17.
86 chapter nineteen
lexicon: an accumulation of good form or breeding, auspiciousness,

vigor/vitality, and, specifically, sexual allure or charm—all of which
are not only ascribed in text, but equally to be read in imagery.
The first attribute, from the Akkadian verb, banû, meaning “to build,”
also “to generate”—hence “form plus breeding”—yields as an adjective
one who is well built, well-formed.2 The term is applied to women,
especially goddesses, and also to men. Of the epic hero, Gilgamesh,
for example, his companion-to-be, Enkidu, is exhorted: “Look at
him . . . well-formed in young manhood.”3 Good conformation, therefore,
a term used in modern times with reference to show animals, is conveyed
by the physically positive to indicate both form and breeding.
The second quality, from the Akkadian term damqu, is used in
descriptions of persons and things and may be rendered as “auspi-
cious/good.”4 That Naram-Sîn is himself both auspicious and protected
by the auspicious is marked in two ways on his body. First, there is
archaeological evidence to confirm that his neck bead (see detail, Figure
3) was not only ornamental but also protective. Examples have actu-
ally been found, inscribed with the name of the ruler and his favored
status with respect to a particular deity, thereby becoming wish-fulfilling
markers.5 The beads are auspicious insofar as they offer the protection
of the deity named and thereby keep away the equivalent of the “evil
eye.”6 Further, Naram-Sîn’s inherent auspiciousness is manifest in the
king’s perfect body itself. Note that he stands in such a way that his
right side is entirely visible to us. From Mesopotamian omen texts, we
know the right is the side that, if deformed in any way, portends ill for
the country and the ruler. If no right ear, “the days of the prince will
be at an end;” if no right forearm or hand, “the outlying districts will
become wasteland.”7 In other words, the king NEEDED to show, the
2
See Chicago Assyrian Dictionary (henceforth, CAD) (Chicago, University of Chicago
Press). B:81–83, banû.
3
Cited CAD, B:142, baštu.
4
CAD, D:68–74, damqu, gives “good, gracious, propitious.” I prefer “auspicious.”
5
For example, an inscribed turquoise pendant, literally called a “neck stone” (writ-
ten in Sumerian logograms as NA4.GU2), belonging to a Middle Assyrian ruler, which
identifies the king as the “Favorite of (the gods) Enlil and Ninurta”: see B. K. Ismail
and M. Tosi, “A Turquoise Neck-Stone of King Ninurta-apal-Ekur,” Sumer, 32 (1976),
105–12.
6
M. L. Thomsen, “The Evil Eye in Mesopotamia,” Journal of Near Eastern Studies,
51 (1992), 19–32.
7
See, for example, E. Leichty, The Omen Series Summa Izbu, Texts from Cuneiform
Sources IV (Locust Valley, N.Y.: J. J. Augustin, 1970), Tabl. II, lines 12, 16, 19; Tabl.
sex, rhetoric, and the public monument 87
Mesopotamian viewer NEEDED to see, that the right ear was there,
that the right limbs were perfectly formed; indeed, as Mary Kelly has
described it with respect to contemporary maleness: that there was a
lack of a lack.8 With those conditions fulfilled, the auspiciousness of the
person is apparent and can be read from his body. As such, the goddess
Ishtar looks with desire upon the damqu of the hero Gilgamesh, and
Enkidu is asked to reflect upon his companion, damqu dGilgameš.9 I would
suggest that this auspiciousness was manifest physically as wholeness
and perfection, the underlying referent actually being a more ineffable
quality of being.
The third attribute, baštu, is translated as life force, vigor, vitality.10
Thus, Gilgamesh, once identified as “well-formed in young manhood,”
is further described as “having (lit., bearing) vitality.”11 That this is also
visually perceptible is clear, since in the poem we are first exhorted to
“look upon Gilgamesh” in order to see the quality of vitality, or life
force, he possesses. Evidence for this vitality is likewise inscribed in/on
the body. An inverse illustration may be found in the account of the
Assyrian king Sennacherib, who, in cutting off his enemies’ beards, is
said to rob them of their baštu.12 Here, I would suggest that what is mani-
fest in men’s beards—an important secondary sexual characteristic—is
precisely their fully developed manhood; and note Naram-Sîn’s beard,
most abundant by far on the stela (see fig. 3). Vitality is therefore con-
flated with manliness (for men), and is articulated visually by facial hair,
along with breadth of chest and virile stance.
Finally, a fourth quality, kuzbu, connotes attractiveness of a different
sort. It is an attribute applied to temple, palace, garden, sculpture, and
cultic paraphernalia as well as to gods, kings, and lovers; when applied
to persons, it refers to sexual allure or charm. From the range of uses
it is clear that both the Akkadian term and its Sumerian counterpart,
ªi-li, carry multiple significations—sometimes quite concrete, sometimes
III, lines 2, 48–56, 85; etc. For a discussion of the opposite, the god-given positive
features of the ruler’s body, like a strong right arm, see I. Winter, “The Body of the
Able Ruler: Toward an Understanding of the Statues of Gudea,” in H. Behrens et al.,
eds., Dumu É-dub-ba-a: Studies in Honor of Åke W. Sjöberg, Occasional Publications of the
Samuel Noah Kramer Fund, 11 (Philadelphia: University Museum, 1989), 573–84.
8
Mary Kelly, “Masculinity on Display,” lecture given at Harvard University, March
16, 1994.
9
Cited, CAD, D, damqu (supra n. 4).
10
CAD, B:142–4, baštu.
11
Ibid., 142, baštu 1: e lūta bani balta iši (note that baštu can also be written baltu).
12
CAD, B:143, baštu 3a: sapsapāte unakkisma baltašun ābut.
88 chapter nineteen
more abstract; for that reason, the Chicago Assyrian Dictionary includes
“luxuriance,” “abundance,” “delight,” “attractiveness,” “charm,” “sexual
vigor,” and “voluptuousness” among its meanings.13
Wilfred Lambert seems to have come closest to the underlying
meaning of the term in associating the range of metaphoric usage and
generally “fruity language” with the “luscious natural attractiveness of
(ripe) fruit” and other conveyors of sweetness, such as honey.14 When
applied to gardens, especially in the Neo-Assyrian period, the obvious
association is with abundant water and vegetation, as in mušarê kuzbi,
luxuriant gardens.15 But the gateway sculptures installed by Sennacherib
in his palace are also endowed with kuzbu.16 Indeed, they are possessed
of both kuzbu and baltu, a not uncommon pairing of attributes, whereas
a processional boat built by Nebuchadnezzar of Babylon is said to be
“adorned with kuzbu, (and) full of delight” [= lalû, another frequent
pairing].17
Deities are also described as possessing kuzbu, as when the goddess
Ninlil is referred to as the balti Ešarra kuzbu Ekur, “the vital one of the
(temple) Esarra, the k. of the (temple) Ekur.”18 Although female deities
are more frequently described as possessing this quality, male deities
such as the moon god Nanna, Enlil, and Ninurta are also credited with
being adorned with, or charging the heavens or a particular sanctuary
with kuzbu.19
In a study of the nature of divinity in Mesopotamia, Elena Cassin
has suggested that implied in these references is some balance between
voluptuousness and generative power.20 There are also clear occasions
13
CAD, K:614–15, kuzbu.
14
W. Lambert, “Devotion: The Languages of Religion and Love,” in M. Mindlin
et al., eds., Figurative Language in the Ancient Near East (London: SOAS, 1987), 30–1.
Indeed, the signs used in the writing of the Sumerian ªi.li further suggest that what
is conveyed is some play on “sweetness” + “joy.”
15
Cited CAD, M2:234, mušaru, and CAD, K:614, kuzbu, re Sennacherib and Assur-
banipal.
16
Cited CAD, L:51, lalû.
17
ša kuzba zānatu lalâ malātu, cited CAD, K:615 kuzbu, and CAD, L:51, lalû.
18
Text cited in E. Cassin, La splendeur divine: Introduction à l’étude de la mentalité méso-
potamienne, Civilisations et Sociétés 8 (Paris and La Haye: Mouton, 1968), 88, n. 27,
and discussed by Å. Sjöberg, in Sjöberg and E. Bergmann, The Collection of the Sumerian
Temple Hymns. Texts from Cuneiform Sources, 3 (Locust Valley, N.Y.: J. J. Augustin,
1969), 119.
19
Riekele Borger, Die Inschriften Asarhaddons Königs von Assyrien, Archiv für Orientfor-
schung, Beiheft 9 (Osnabrück: Biblio-Verlag, 1967), 77, text 49, p. 1; see also Cassin,
La splendeur divine, 88.
20
Cassin, La splendeur divine.
when kuzbu is used to refer directly to sexuality, as in the first tablet of

the Gilgamesh epic, when the harlot is instructed to expose herself to
Enkidu: ūrki pitêma kuzubki lilqi, “expose your nakedness, so that he can
take (in) your k.”21 In such a context, one would have no trouble to
translate this either explicitly as sex or voluptuousness, or euphemistically
as charms or lusciousness. Similarly, in a reference to Gilgamesh himself,
we are told that he, too, is adorned with k.: zu’una kuzba kalu zumrišu.
This has been variously translated: “His whole body is covered with
attractiveness;22 “adorned with charm is his whole body;”23 “his entire
body exudes voluptuousness.”24 As such, it is a direct variant of ªi-li-su3
as sexual virility or attractiveness in Sumerian love poetry, where there
are frequent references to the beloved’s ªi-li.25
At the same time, when ªi-li, or kuzbu, is ascribed to the king in the
guise of a lion, as in a text of Lipit-Ishtar, then we get a sense not of
the sex appeal of the lion, but of his open mouth and powerful, roaring
charge.26 Here, the term may come closer to what Cassin would see as
a powerful force issuing from its possessor. That sense of “emanation”
is particularly useful when the term is ascribed to a temple, or to an
object, where an alternative to sex appeal is obviously called for.
For purposes of scholarly rather than literary translation, I find I
am most comfortable with attempting to render the literal underlying
concept with one consistent term whenever possible, and then activating
the various connotations from context, not as variation in translation.
Lusciousness, or luxuriance, deriving from the sweet, ripe, or fruity
range of associated metaphors and associations, seems to serve well as
the ascribed quality being signaled. What it does not do is suggest the
quality as an active force radiating from the subject. The most suit-
able term in English I have found for this aspect, in all of its usages, is
“allure.” Like attractiveness, with its base in physical pull or magnetism,
it implies not just a passive attribute, but one in which energy passes
21
M. Kovacs, The Epic of Gilgamesh (Stanford: Stanford University Press, 1989), 8
(= Tabl. I, 163). For urû as nakedness, see W. Von Soden, Akademisches Handwörterbuch
III (Wiesbaden: Harrassowitz, 1981), 1435.
22
CAD, Z:49, zânu.
23
Benjamin Foster, “Gilgamesh: Sex, Love, and the Ascent of Knowledge,” in
Love and Death in the Ancient Near East: Essays in Honor of Marvin H. Pope, ed. John H.
Marks and Robert M. Good (Guilford, Conn.: Four Quarters Publishing, 1987), 29 =
Nineveh I, v, 16–18.
24
Kovacs, The Epic of Gilgamesh, 10 (= Tabl. I, 218).
25
See discussion of Lambert, in Figurative Language, 30–1.
26
Text cited and discussed in Cassin, La splendeur divine, 88–9.
90 chapter nineteen
between the possessor and the receiver. But whereas for attractiveness,
magnetism even, the direction of movement is a pulling IN toward the
possessor, for ªi-li/kuzbu the force seems to emanate OUT from the
possessor, like some sort of pheromone.
As for Gilgamesh, so I would suggest for Naram-Sîn. Not only is his
body beautifully proportioned and lithe, but his girdle is tied seductively
over his right hip, the masterfully executed pleats and knot pulled to
reveal the outline of the perfectly rounded buttocks—a Glutäenerotik par
excellence (see Figure 3).27 Indeed, all of the qualities—inner states and
outer manifestations—discussed above are readable at a glance from the
sculpted body of the king: good conformation, auspiciousness, (male)
vigor, and (sexual) allure.
Rhetoric
What virtue would there be, one might then ask, in representing the
victorious ruler as one not only well-formed, auspicious, and virile,
but also sexually alluring? If one distinguishes “sex” as a necessary
component of biological reproduction from the distinctively cultural
determination of “sexuality,” one seeks the referent within a particular
“set of norms and practices developed in response to the fact of human
biology” (emphasis mine).28 The sexuality manifest by the individual is
then to be situated within a social body. Moreover, Naram-Sîn as ruler
is not just any male, but the dominant male within the state hierarchy;
so one must further inquire how these values fit within, or work for,
the state and the ruler’s position in it, and why these attributes and not
others have been selected for the royal image? In short, how does the
body of the king relate to the body politic?29
The revealed and developed body of Naram-Sîn seems to be a new
development within the Akkadian period, the well-rounded and muscled
27
See Paul Brandt, Beiträge zur antiken Erotik (Dresden: Paul Aretz, 1924), 31–4:
“Viewed from an aesthetic standpoint, nothing more beautiful or more perfect can be
imagined than the form of the human buttocks.”
28
See on this, Jeffrey Henderson, “Greek Attitudes toward Sex,” in Civilizations of
the Ancient Mediterranean: Greece and Rome, ed. M. Grant and R. Kitzinger (New York:
Scribners, 1988), 1250. Also, Peter Tatham, The Makings of Maleness: Men, Women and
the Flight of Daedalus (New York: New York University Press, 1991), esp. xvi.
29
For a different set of desirable physical characteristics, as exemplifying a different
social body, see John J. Winkler, “Phallos Politikos: Representing the Body Politic in
Athens,” differences, 2 (1990), 29–45.
upper torso of what may have been a sculpture of a ruler in the Uruk
period (fourth millennium b.c.e.) notwithstanding.30 The one stela suffi-
ciently preserved from the previous, Early Dynastic period—the Stela of
Eannatum, ruler of the city state of Lagash c. 2500 b.c.e.—is organized
very differently, and the ruler represented very differently: Eannatum is
identified by name and wrapped in a distinctive garment, but otherwise
barely distinguished physically from the phalanx he leads into battle
(fig. 4).31 Certainly, there is power inherent in the solid proportions of
his body, but missing is the full development of the revealed body of
Naram-Sîn, with its subtle articulation of parts. That this “new” body
happens to coincide with an equally new elevation of the king to divine
status will be argued below to be anything but accidental.
Naram-Sîn’s grandfather, Sargon of Agade—founder of his dynasty,
and first unifier of the previously autonomous city-states into a nation—
appears to have begun to disengage his body from others’, as is seen
on his fragmentary victory stela from Susa.32 However, I would suggest
that the immediate model for the fuller unwrapping and display of
the body of Naram-Sîn is not to be found in representations of prior
rulers. Rather, it is to be seen in representations of a type of generic,
semi-divine “heroic genius,” often engaged in combat with wild preda-
tors on earlier cylinder seals (for example, fig. 5).33
As suggested above, this fusion of hero-plus-ruler into a new royal body
occurs at exactly the historical moment when for the first time the king
is elevated to divine status. This shift is manifest in inscriptions, in which
both Naram-Sîn and his son after him write their names preceded by
the determinative for god, not man, and take on particular epithets in
their titularies, such as “god [not king] of Agade.”34 The helmet worn
30
See A. Moortgat, Art of Ancient Mesospotamia (London and New York: Phaidon,
1969), Pl. 13.
31
See discussion in I. Winter, “After the Battle is Over: The Stele of the Vultures and
the Beginning of Historical Narrative in the Art of the Ancient Near East,” in H. L.
Kessler and M. S. Simpson, eds., Pictorial Narrative in Antiquity and the Middle Ages, Studies
in the History of Art, 16 (Washington, D.C.: National Gallery of Art, 1985), 11–32.
32
See Harper et al., The Royal City of Susa, #106.
33
These seals are best seen as a group in R. M. Boehmer, Die Entwicklung der Glyp-
tik während der Akkad-Zeit (Berlin: De Gruyter, 1965), Figs. 2–8 of the Early Dynastic
period, and Figs. 13–158 of the Akkadian period prior to Naram-Sîn. The figure has
been identified by F. M. Wiggerman as the semi-divine la mu, a benevolent protec-
tive genius—see “Exit Talim! Studies in Babylonian Demonology, I,” Jaarbericht van het
Vooraziatisch-Egyptisch Genootschap Ex Oriente Lux, 27 (1981–2), 90ff.
34
See most recently on this, W. Farber, “Die Vergöttlichung Naram-Sîns,” Orientalia,
52 (1983), 67–72, and J.-J. Glassner, La chute d’Akkadé (Berlin: Reimer, 1986). Indeed the
inscription on the seal impression, Figure 6, uses this title in the second line.
92 chapter nineteen
by Naram-Sîn on the stela has long been seen as the visual counterpart
of this phenomenon, bull horns having heretofore been reserved for
the headgear of gods alone (see, for example, fig. 6).35 In addition, I
would argue that the innovation of a perfect, alluring, “heroic” body
for the ruler, coming as a visual quote from a pictorial tradition hitherto
reserved for a semidivine hero, also represents a conscious strategy of
representation. Along with the divine determinative and the divine
headgear, the “divine” body equally marked the elevation of the ruler
to the status of a god.
Two strands are woven together in this argument that the divine body
is correlated with divine kingship: a long-standing Mesopotamian tradi-
tion of rhetorical statements that the gods are responsible for shaping the
royal body, and an explicit association of the ruler with the legendary
hero Gilgamesh, who was himself a semidivine king. For the first strand,
many Mesopotamian rulers make use of epithets that emphasize the
king’s maleness and valiance—for example, zikaru qardu, heroic male,
or e lu qardu, ferocious (male) youth, terms that carry with them both
a gender marker and also a sense of associative potency.36 Naram-Sîn
himself uses the term dannum, “strong one,” as his most frequent epithet
in his titulary.37 This focus on the (male) potestas of the ruler as part of the
formal title, I would argue, is rendered visually not only by the weapons
he carried in hand, but also by the life force/vitality of his perfect and
alluring (hence sexually desirable, for which, read “potent”) body.38 That
35
Note, however, that the divine crown worn by Naram-Sîn is not one of multiple
tiers of horns, as worn by the high gods of the pantheon, but rather consists of a single
tier of horns, often used to distinguish lesser divinities on seals—as, for example, in
Figure 6 (= Boehmer, Die Entwicklung der Glyptik, Fig. 725—belonging to an official of
Naram-Sîn himself ). This detail would have permitted the contemporary viewer to
read simultaneously the king’s divine standing and yet his relatively lower status with
respect to the high gods.
36
See, for example, the texts of Adad-nirari II of Assyria (911–891 b.c.e.), in A. K.
Grayson, Royal Inscriptions of Mesopotamia 2: Assyrian Rulers of the Early First Millennium
(Toronto: University of Toronto Press, 1991), 157 = AN II.4.4’ and 13’; also, for
Assurnasirpal II (885–53 b.c.e.), in Grayson, Royal Inscriptions of Mesopotamia 2: Assyrian
Rulers of the Early First Millennium 2, 196 = ANP II.1.33. In one couplet, Adad-nirari
all but beats his chest, as he announces: labbāku u zikaraku, “I am a lion and I am a
(potent) male.”
37
H. Hirsch, “Die Inschriften des Könige von Agade,” Archiv für Orientforschung, 20
(1963), 1–82; D. Frayne, Royal Inscriptions of Mesopotamia: Sargonic and Gutian Periods
(Toronto: University of Toronto Press, 1993), 88–144.
38
This potency on the part of the ruler is also clearly articulated as desirable with
respect to the engendering of male issue. Its contrary, the negative attached to lack of
male issue and an heir, is preserved in a text of Naram-Sîn himself, in which a closing
this body is not merely the product of good breeding and/or physical
training may be inferred from a text of the Assyrian ruler, Adad-nirari
II, which, in the voice of the king, relates that the gods “altered my
stature to lordly stature, . . . made perfect my features and filled my
lordly body with wisdom . . . [and] surrounded my head with the aura
of rulership.”39 Comparable assertions are also recorded by Sumerian
rulers of the third millennium, both before and after Naram-Sîn, with
the specific bodily parts and qualities each delineated as the separate
bequest of a particular deity.40 Royal rhetoric thus invests the royal body
with the authority of the high gods themselves. Despite the absence of
such a statement exactly contemporary with Naram-Sîn, I would sug-
gest that this tradition helps to illuminate the Akkadian period as well.
The notion that the gods are said not only to select but also to shape
the ruler underscores the importance of the viewer’s ability to discern
lordly stature and perfect features from the royal image, and all the more
so in a period in which the ruler is himself accorded divinity.
The second strand in this argument, which explicitly adds the ele-
ment of sexual allure to that of male power and perfection in the king’s
body, is founded on a relationship between the youthful and alluring
Gilgamesh, king of Uruk, and Naram-Sîn, king of Agade. Not only is
there a parallelism in attributes, but both rulers also shared a rhetorical
ascription of divine status. The two kings thus participate—verbally for
the one, visually for the other—in virtually identical representational
practices.
Given the fact that the Akkadian version of the Gilgamesh Epic exists
in its developed form as a relatively late, Neo-Assyrian composition
of the first millennium b.c.e., and the only Sumerian fragments that
curse invokes the wrath of the goddesses Ninhursag and Nintu not to grant someone
male offspring—see Frayne, Royal Inscriptions of Mesopotamia: Sargonic and Gutian Periods
99 = text E2.1.4.3, vi, 29–35.
39
Grayson, Royal Inscriptions of Mesopotamia 2: Assyrian Rulers of the Early First Mil-
lennium, 147 = AN II.2.6–7; see also a text of Assurbanipal, in which he asserts that
“[the great gods] gave me a splendid figure, and made my strength great” (cited CAD,
Š:289, šamā u).
40
See, for example, the text of Eannatum of Lagash, written on his victory stela
( J. S. Cooper, Reconstructing History from Ancient Inscriptions: The Lagash-Umma Border Conflict
[Malibu: Undena Publications, 1983]) of the previous, Early Dynastic period; also, the
texts of Gudea of Lagash (discussed by Winter, “The Body of the Able Ruler . . .”), of
the subsequent Neo-Sumerian period. The separation of parts in such a context is then
not unlike the familiar Charles Atlas comic-book ads, in which the various parts of
the body to be developed are illustrated separately—chest, arms, legs, shoulders—their
sum total adding up to “magnetic personality” (see fig. 7).
94 chapter nineteen
survive are dated to the second millennium, it would be quite accept-

able simply to point to this parallelism as a sign of continuing values
and representational strategies within a long range of Mesopotamian
history. I would like to offer a more radical stress on the relationship
between Naram-Sîn and Gilgamesh, however, which I believe has
important implications for understanding both the erotics and the
poetics of the stela.
The second-millennium copies of the Sumerian version of the epic
were executed by scribes of the Old Babylonian period but are gener-
ally presumed to have been composed in the preceding Ur III period
(c. 2110–2050 b.c.e.).41 Certainly, by Ur III times, the person of Gil-
gamesh as legendary hero and king had already entered into popular
culture, which is clear from the fact that rulers at the very beginning
of the period already claimed filiation with him: Ur-Namma of Ur,
founder of the dynasty, for example, is referred to as the “brother of
Gilgamesh,” and a similar claim was made by his son, Šulgi.42 With no
evidence in the form of early text fragments, it is impossible to demon-
strate with certainty that the epic existed earlier than the Ur III period;
however, circumstantial evidence suggests that the poem, or at least
significant elements of the legend that would eventually be redacted as
epic, must have existed considerably earlier.43 This would fit well with
the recent argument put forward by Piotr Steinkeller that there was
a considerable early literature, now lost to us, for which the only
remaining traces are the apparently mythological narratives on cylinder
seals of the Early Dynastic and Akkadian periods.44 Indeed, I believe
one would have to prove that the legends of Gilgamesh were not well-
formed and disseminated by the time of Naram-Sîn. In such a universe,
the kuzbu-bearing body of Gilgamesh as ideal prince, would actually
41
J. H. Tigay, The Evolution of the Gilgamesh Epic (Philadelphia: University of Penn-
sylvania Press, 1982), 242–4.
42
See Thorkild Jacobsen, Treasures of Darkness (New Haven: Yale University Press,
1976), 158, and J. Klein, “Šulgi and Gilgameš: Two Brother-Peers (Šulgi O),” Alte Orient
und Altes Testament, 25 (Neukirchen-Vluyn: Harrasowitz, 1976), 271–92.
43
See, for example, K. Afanasieva, “Gilgamesh and Enkidu in Glyptic and in the
Epic,” Klio, 53 (1971), 59–75, esp. 72; also implied in already existing Early Dynastic
terminology related to Gilgamesh presiding over the deceased rulers in the underworld
(see P. Steinkeller, review of J. Marzahn, Altsumerische Verwaltungstexte aus Girsu/Lagaš for
the Journal of the American Oriental Society, forthcoming; I am grateful to the author for
making the ms. available to me prior to publication).
44
P. Steinkeller, “Early Semitic Literature and Third Millennium Seals with Mytho-
logical Motifs,” Quaderni di Semitistica, 18 (1992), 243–75.
lie quite explicitly behind the visually alluring body of Naram-Sîn.45

In the epic, the narrative specifically calls for Gilgamesh to be alluring,
since the goddess Ishtar becomes enamored of the young ruler; however,
the visual signs of his desirable qualities are equally discernible to his
companion Enkidu. For Naram-Sîn to identify himself with the epic
hero on his own victory stela would have served the same rhetorical ends
as the divine Augustus identifying himself with Alexander or the Greek
gods46—in short, legitimation through association, and participation in
an honored heroic tradition that itself had rhetorical power.
The Public Monument
The ultimate question to be posed here concerns the consequences of

using this particular rhetorical strategy, this particular image of Naram-
Sîn, on a freestanding stone stela (Akkadian narû). How does sex(uality)
mix with politics in public? And, indeed, can we be certain that the
monument was in the public domain? Since the stela was carried off
as booty to Elam in the twelfth century b.c.e.,47 we cannot know for
certain where it was placed originally; however, circumstantial evidence
is helpful in positing a likely venue in the temple precinct of the sun
god Shamash at Sippar. First, other stelae, of periods both earlier and
later, are known to have been set up in such locations; second, spe-
cific Akkadian period monuments were apparently installed in temple
courtyards; and third, the Elamite ruler who carried off the actual
45
Herbert Sauren has implied the same thing, noting that Gilgamesh is described
as two-thirds divine in the epic and Naram-Sîn is two-thirds the height of the moun-
tain opposite which he stands on the stela: see “Die Königstheologie in der Kunst des
3. Jahrhundert,” Orientalia Lovaniensia Periodica, 13 (1982), 50. A more subtle inference
may be drawn from the use of the title mut dINANNA, “beloved of Ishtar,” by Naram-
Sîn, comparable to the status of Gilgamesh in the epic: see B. Kienast, “Naramsîn mut
d
INANNA,” Orientalia, 95 (1990), 196–203.
46
Cf. John Pollini, “Man or God: Divine Assimilation and Imitation in the Late
Republic and Early Principate,” in Between Republic and Empire: Interpretations of Augustus
and His Principate, ed. K. A. Raaflaub and M. Toher (Berkeley: University of Cali-
fornia Press, 1990), 334–63. For Greek athletic heroes identified with Herakles, see
also Leslie Kurke, “The Economy of Kudos,” in Cultural Poetics in Archaic Greece, ed.
C. Dougherty and L. Kurke (Cambridge: Cambridge University Press, 1993), n. 13,
and Wendy J. Raschke, “Images of Victory: Some New Considerations of Athletic
Monuments,” in The Archaeology of the Olympics (Madison: University of Wisconsin
Press, 1988), 43, 45.
47
See P. Amiet, L’art d’Agadé au Musée du Louvre (Paris: Editions des Musées Nation-
aux, 1976), 29.
96 chapter nineteen
Naram-Sîn stela and subsequently inscribed it himself, tells us that he

took it from Sippar and that he set it up as a trophy in the precinct of
his god at Susa, where it was actually found some 4,000 years later.48
In addition, the Mesopotamian tradition of copying inscriptions on
earlier monuments indicates that scribes at least had access to them
well beyond the period in which they were made.49 Indeed, the fact
that an Elamite ruler carried the Naram-Sîn stela off in the twelfth
century demonstrates that it was still around a millennium after its
manufacture. For want of evidence to the contrary, therefore, even if
we cannot determine for certain the range of individuals who would
have had access to the temple precinct or to the stela, I feel it is not
inappropriate to speak of it as set up in “public” space.
Still at issue, however, is an assessment of the stela’s particular
rhetorical strategy. The public monument constitutes a special subset
of material culture per se, in that it not only has physical substance,
it also has as part of its ontology a key aspect of display. I shall nei-
ther contest nor attempt to document the often noted likelihood that
in some measure, the god to whom the stela was dedicated, or the
Mesopotamian gods in general, as well as the king himself, constituted
a portion of the stela’s intended audience.50 At the same time, the
mix of erotics/allure with military victory that constitutes the subject
matter of the stela is of great interest when projected at an audience
neither royal nor divine.
The paintings of J.-L. David have been used recently to probe the
association of beautiful male bodies and idealizing political messages
in the period following upon the French revolution, suggesting that
48
See the discussion of freestanding stelae in general in Winter, “After the Battle
is Over,” 23–5; for Akkadian-period monuments in the courtyard of the Enlil temple
at Nippur, see P. Michalowski, “New Sources Concerning the Reign of Naram-Sîn,”
Journal of Cuneiform Studies, 32 (1980), 236 and 239; for the later inscription on the
Naram-Sîn stela, see Amiet, L’art d’Agadé, 27; for its findspot, see Suzanne Heim,
“Royal and Religious Structures and Their Decoration,” in Harper et al., The Royal
City of Susa, Fig. 41.
49
See reference above to Michalowski, “New Sources Concerning the Reign of
Naram-Sîn,” 239, where Akkadian inscriptions in Nippur were subsequently copied by
Old Babylonian period scribes. In addition, a whole genre exists, known as narû literature,
which purports to be copied from monuments, although this was apparently a fictional
attribution designed to enhance the importance of the contents: See J. Westenholz,
“Heroes of Akkad,” Journal of the American Oriental Society, 103 (1983), 327–36.
50
The original inscription of Naram-Sîn, badly effaced, still preserves a portion of
the closing, dedicatory verb; see now, Frayne, Royal Inscriptions of Mesopotamia: Sargonic
and Gutian Periods, 143–4 = E2.1.4.31, col. iii, 3)–5).
aesthetic pleasure can serve as the vehicle by which a highly seductive

ideal vision is articulated for the citizen-viewer. Thus, for Alex Potts,
“beautiful and graceful nudity . . . is . . . a formal idealizing device,” trans-
forming the subject of the painting into a “more effective signifier of
heroic virtue than a naturalistic image.”51 Potts argues that the rhetori-
cal power of images that combine an erotic charge with violence, as
represented in military battles, derives precisely from an engagement
with the sensual body. At the same time, Ewa Lajer-Burcharth suggests
that this engagement occurred precisely at a moment in which the work
was inserted into a new and very public discourse.52
One is, of course, in a far better position to assess the viewer’s engage-
ment with the attractive male body in late eighteenth-century France,
when contemporary critical reactions have been preserved. Chaussard,
for example, exhorts the viewer to look at David’s Romulus in the Sabine
Women (fig. 8), and exclaims, “How these extremities, how these legs
shine with beauty!”53 Yet, surprisingly, there are similar hints in Meso-
potamian literature to suggest that a comparable association of viewing
pleasure and inspiring response existed. An exhortation to view the
physical perfection of Gilgamesh has been noted above. In addition, a
first-millennium bilingual text about a ruler uses an intensive conjuga-
tion of the verb “to see” in order to rhapsodize that “the people of
the land stared at his tall, perfect [lit., fitting, appropriately endowed],
princely body.”54 And far closer in date to Naram-Sîn is a line in the
Sumerian epic poem “Enmerkar and the Lord of Aratta,” in which
the king declares: “May the people look at me admiringly!”55
This articulation of a cathected viewing of the royal person illustrates
for us the value vested in the stela, set up to (re)present the ruler at
his moment of triumph. If, in the absence of additional evidence, we
use Ishtar’s response to Gilgamesh as an illustration of the normative
heterosexual Mesopotamian female response to male attractiveness, the
similarly idealized manly body of Naram-Sîn would equally constitute
an obvious locus of pleasurable engagement—situating the woman
51
A. Potts, “Beautiful Bodies and Dying Heroes: Images of Ideal Manhood in the
French Revolution,” History Workshop Journal, 30 (1990), 1–21, esp. 4.
52
E. Lajer-Burcharth, “David’s Sabine Women: Body, Gender and Republican Culture
under the Directory,” Art History, 14 (1991), 397–430.
53
Cited by Lajer-Burcharth, “David’s Sabine Women,” 400.
54
ibtarrâ nišī māti lānšu elâ šūšumu etella, cited CAD, E:383, etellu.
55
[nam]-lu2-lu6 u6 du10-ge-eš ªu-mu-un-e, cited in T. Jacobsen, “Lugalbanda and
Ninsuna,” Journal of Cuneiform Studies, 41 (1989), 78.
98 chapter nineteen
libidinally in relation to the ruler as today before a cinema idol.56 Of

course, Ishtar’s response to Gilgamesh may not be normative and may
have been possible precisely because she was a goddess and so freer
to respond than an ordinary female. Social constraints placed in an
attempt to control female sexuality are certainly not rare;57 however,
even if that were to have been the case in Mesopotamia, the very
fact that the goddess is permitted what the ordinary woman was not
argues for the importance of this area of suppression, in need of liter-
ary transference.
More difficult to account for is the projected response of a normative
heterosexual Mesopotamian male. The construction of an ideal mas-
culinity in works of art that is to serve as a model for social formation
has been at the forefront of much recent work, as has the question of
the gaze of men upon men.58 How to assess it for an ancient culture
in which sources are limited is a serious methodological problem, how-
ever—even when, as in the present case, we have literary evidence that
Enkidu looked as appreciatively (if nonsexually) upon the charms of
Gilgamesh as did Ishtar. One can project that for the not-fully-formed
(adolescent) male, the king may serve as inspiration, an ideal to be
emulated; for the mature adult male, however, things get more compli-
cated. Does the image of the king represent the projected self, if not
of the individual, then of the self-in-the-nation?59 Or, as emphasized
in the king’s greater size and divinely engineered perfection, does the
ruler represent an impossible standard—competitor/father—never to
be achieved by any but the king himself ? Or, in fact, could both reac-
tions be operative simultaneously, such that the normative viewing male
56
As noted by Mary Ann Doane, theories of female spectatorship are rare; but the
concept of a necessary distance between self and the image (as in the modern cinema)
permits one to account for the emergence of a meta-desire bonding the female viewer
to the icon before her. See “Film and the Masquerade: Theorizing the Female Spec-
tator,” in Femmes Fatales: Feminism, Film Theory, Psychoanalysis (New York and London:
Routledge, 1991), 20–1.
57
See Tamar Garb, “The forbidden gaze: women artists and the male nude in late
nineteenth-century France,” in The Body Imaged: The Human Form and Visual Culture Since
the Renaissance, K. Adler and M. Pointon, eds. (Cambridge: Cambridge University Press,
1993), 33–42. For Mesopotamia, the topic has yet to be pursued.
58
See, for example, Michael Hart, “Muscles, Morals, Mind: The Male Body in
Thomas Eakins’ Salutat,” in The Body Imaged, 57–69; see also the editors’ comments
on “Bodies of Masculinity,” 53–5.
59
As discussed by Raschke, “Images of Victory,” 48; and by Jonathan Rutherford,
“Who’s That Man?” in Male Order: Unwrapping Masculinity, ed. R. Chapman and
J. Rutherford (London: Lawrence and Wishart, 1988), 24.
is at once dwarfed/engendered by his king and united with him in his

embodiment of/as nation—thereby merging himself in the sum larger
than the total of its individuals?
Here it is important to keep in mind that the stela displays not only
the physical build, grace, vitality, and military prowess of the king but
also his “sexuality.” Within the Mesopotamian lexicon of value, sexual
allure has been associated with potency and desirability, linked to ideal
maleness. This element thus constitutes a variation within traditional
genres of Machtkunst—images of power that serve to reinforce hier-
archies of dominance. Just how sexuality adds to that discourse as an
independent variable is at issue. I can only point the way toward the
tantalizing problem, knowing that the present case is not one that is
likely to provide a solution.
As Judith Butler has noted, “sexuality” is a primary means of estab-
lishing gender, its display closely associated with identity-formation.60 To
display sexuality, therefore, is to perform gender; and to display sexuality
in a work of art is to give form to a particular cultural construction of
gendered identity. The admiration with which the individual is to look
upon the allure of the ruler in Mesopotamia is thus more than aesthetic;
it calls forth an act of social approbation on the one hand, and on the
other, it induces in the viewer a state of what Michael Fried has called
“specularity,” in which he or she, oblivous to the membrane of the
medium, engages in a quasi-corporeal relationship with the image.61
Viewed in this light, Naram-Sîn’s display of male attributes on a public
monument does more than just narrate his role as victorious potentate.
By setting up active currents of positive value through seductive allure,
the display also facilitates identificatory processes that elicit a series of
vicarious associations and projections that have a socializing function:
for women, their subordination to desire and by men; for men, their
fusion with authority at the same time as they are subject to it.62
60
J. Butler, Bodies That Matter: On the Discursive Limits of “Sex” (New York and London:
Routledge, 1993), 2–3.
61
M. Fried, “Between Realisms: From Derrida to Manet,” lecture given at Harvard
University, December 6, 1993.
62
Note that there are no female images on the stela. The victory and the display
take place in a landscape peopled only by male figures. An important aspect of this,
apart from its “narrative” value, may well be, as noted by Rutherford, “Who’s That
Man?” in Male Order, 47, that “the appeal of the hero” (in this case, Naram-Sîn or
Gilgamesh) “is his freedom from women”; “a man’s man.” At the same time, the enemy
have long hair, wear animal skins, and cringe—in other words, they are gendered
female, uncivilized, and submissive.
100 chapter nineteen
Underlying this issue may be a clue to one of the most vexing

questions for political analysts: namely, by what mechanisms of social
control are subordinate populations subjected to, or do they accede
to, subordination?63 Recent work on the psychoanalytic foundations of
hegemony sees the modern West grounded in a dominant discourse of
masculinity.64 There is certainly ample evidence for a parallel discourse
of masculinity in ancient Mesopotamia, and I would see in sexual allure
one of the mechanisms by which dominance can be brought into play.
In other words, for Naram-Sîn, as for Gilgamesh, sexuality, couched as
part of an ideal poetics, is put in the service of the political, toward
the social, not merely the libidinal, seduction of the viewer.
I have used the term “poetics” advisedly. If cultural poetics may
be defined as that process “whereby a society and its sub-groups
construct widely-shared meanings—behavioral conventions, social
distinctions . . . aesthetic values . . . [and] gender roles,”65 then what I
am emphasizing here may better be called “cultural rhetorics”—that
process whereby meaning and representation are turned into ideology.
Potts argues with respect to David’s heroes that “the rhetorical power
of [the] forms derives from an erotic engagement with the body.”66
On the stela, the erotic allure of Naram-Sîn would be at least one of
the triggering mechanisms that turns the viewer of a military victory
into an identified, engaged, but subordinated enthusiast; and seduction
may be seen as a goal, a product, of the work. The generative power
of the body of Naram-Sîn then produces the viewer as his seed, his
consort, his surrogate self.
The metaphor of the ruler’s potency, and hence reproductive poten-
tial, functions as a very literal operator within Althusser’s construct that
every social formation must be constantly reproduced in order for the
system to be maintained.67 If it is the job of ideology to turn the living
subject into one subjected to the system, then the polyvalence reflected in
63
L. Althusser, “Ideology and Ideological State Apparatuses (Notes Towards an
Investigation),” Lenin and Philosophy and Other Essays, tr. B. Brewster (New York: Monthly
Review Press, 1971), 127–86.
64
See Kaja Silverman, Male Subjectivity at the Margins (New York and London: Rout-
ledge, 1992), esp. 23–35.
65
David M. Halperin, John J. Winkler, and Froma I. Zeitlin, eds., Before Sexuality:
The Construction of Erotic Experience in the Ancient Greek World (Princeton, N.J.: Princeton
University Press, 1990), 4.
66
Potts, “Beautiful Bodies and Dying Heroes,” 6.
67
Althusser, “Ideology and Ideological State Apparatuses,” 128, 180.
the English word “subject”—subject of the work, subject (ego) viewer,

and subject (political) dominated—is beautifully played out on the Stela
of Naram-Sîn, where all three are brought together through the seduc-
tive allure of the body (politic). And as a corollary to the now widely
accepted premise that “representation constitutes its subject” (by which
Louis Marin meant subject matter, the king whose portraiture he was
discussing),68 I would argue further that representation also constitutes
its subjects (plural), meaning its subordinates!
The attributes identified as positive signs of value inscribed in the
ruler’s body—good conformation, grace, vitality, and allure—had to
exist within a lexicon of cultural value before they could be deployed as
part of a politicized aesthetic. They could be deployed precisely because
they had a prior value. Naram-Sîn’s strategic deployment of these
attributes—and especially the innovative correlation of sexuality and
divinity—coincided with a number of strategic political moves during
his reign, as the king continued the consolidation of the nation-state set
up by his grandfather.69 The eroticized body of the ruler can therefore
best be seen in conjunction with a moment in time when new mecha-
nisms of social control were required within a new system of political
organization. Since the stela is unique in the Mesopotamian repertoire
as recovered to date, we cannot say for certain that this eroticized body
posed no problem for the Mesopotamian viewer, as heroic nudes of the
Emperor Augustus apparently did not pose a problem for the Romans.
We can project the normative viewer from the king’s rhetoric, but s/he
cannot speak with her or his own voice. Nevertheless, since virility and
sexuality were intimately linked to concepts of power and leadership
for nearly two millennia in Mesopotamia, it is quite possible that this
coupling was seen as not unnatural.
The context of military victory and territorial expansion in which
the alluring ruler was depicted on the stela brings into one unified field
68
Louis Marin, Portrait of the King, tr. Martha M. Houle (Basingstoke: Macmillan,
1988), 5.
69
Cf. P. Michalowski, “Thoughts about Ibrium,” in Wirtschaft und Gesellschaft von
Ebla, Heidelberger Studien zur Alten Orient, Bd. 2, H. Hauptmann and H. Waetzoldt,
eds. (Heidelberg: Heidelberger Orientverlag, 1988), 267–77. Note that I am in no way
imputing any universality to this particular strategy. For a study of contrasting strate-
gies, corresponding to different but neighboring constructs of kingship, see Maurice
Bloch, “Tombs and States,” in Mortality and Immortality: The Anthropology and Archaeology
of Death, ed. S. Humphreys and H. King (London and New York: Academic Press,
1981), 137–48.
of vision the three dominant “P” ’s of maleness: “(im)pregnate, protect,

provision.”70 In so doing, the stela not only engages the subject-viewer’s
moral commitment to the values of his or her own society, as noted
by David Gilmore for art in general;71 it also, I would argue, truly
subjects the viewer to the hegemony of the ruler. In other words, the
stela provides a powerful visual form of rhetoric, in which physical and
social characteristics can be shown to have political as well as cultural
coordinates.72
The sexual allure of the ruler’s body serves as a positive attribute;
it also functions as a triggering mechanism within a powerful rhetoric
of rulership. At issue is not merely the “self-fashioning” of king and
viewer,73 but the fashioning of the social body. Viewed in this light,
the monument on which Naram-Sîn is depicted may be said to have
played an affective, not merely a reflective, role—shaping its constituency
according to the political strategy of its leader, shaping its constituency
in the image of its leader.
Acknowledgments
My sincere thanks to the many students who have worked on aspects of this stela over
the years; without their frequent insights and serious endeavors, my own thoughts
would be significantly limited. In particular, I would like to acknowledge the stimulus
of Cynthia Hall, Lisa Pon, and Kishwar Rizvi. In addition, I would like to thank a
number of colleagues with whom I have discussed the Stela of Naram-Sîn and related
issues, and who will no doubt find pieces of themselves herein: Bettina Bergmann, Nor-
man Bryson, Natalie Kampen, Leslie Kurke, Ewa Lajer-Burcharth, Peter Machinist,
Michelle Marcus, and Piotr Steinkeller.
70
David Gilmore. Manhood in the Making: Cultural Concepts of Masculinity (New Haven
and London: Yale University Press, 1990), 223.
71
Gilmore, Manbood in the Making, 224.
72
See M. Pointon and K. Adler, “Introduction to Part IV: The Body as Language,”
in The Body Imaged, 126, in which the authors argue that the importance of visual
forms of rhetoric has not been adequately appreciated. And yet, we seem to have no
trouble at all accepting claims about the degree to which gender is shaped, not merely
mirrored in contemporary advertising—see Erving Goffman, Gender Advertisements (New
York: Harper and Row, 1979). Perhaps because it is so problematic to establish agency
in periods other than our own, the visual arts have been less subject to analysis as agents
in social formation, rather than as mirrors of form.
73
M. Gleason, “The Semiotics of Gender: Physiognomy and Self-Fashioning in the
Second Century c.e.,” in Before Sexuality, 389–413.
Figure 1. Stela of Naram-Sîn (c. 2250 b.c.e.), found at Susa; height 2 m.

Photo by the author, reproduced courtesy Musée du Louvre,
Département des antiquités orientales.
Figure 2. Detail, Stela of Naram-Sîn.

Figure 3. Detail, Stela of Naram-Sîn.
Figure 4. Stela of Eannatum of Lagash (c. 2500 b.c.e.), detail, reverse.

Found at Tello. Photo courtesy Musée du Louvre,
Réunion des Musées Nationaux.
Figure 5. Drawing of cylinder seal impression, Early Dynastic/Early

Akkadian period (c. 2340 b.c.e.), from Ur. After R. Boehmer,
Die Entwicklung der Glyptik, Pl. IV, 31.
Figure 6. Drawing of cylinder seal impression, Akkadian period (reign of

Naram-Sîn, c. 2250 b.c.e.) from Ur. After R. Boehmer,
Die Entwicklung der Glyptik, Pl. LXIII, 725.
Figure 7. Detail of advertisement for Charles Atlas body-building methods,

from inside cover of Dare Devil, Marvel Comics, March 1974.
Printed by permission of Charles Atlas, Ltd., New York.
Figure 8. Jacques-Louis David, The Sabine Women. Photo Courtesy Musée du

Louvre, Réunion des Musées Nationaux.
CHAPTER TWENTY
TREE(S) ON THE MOUNTAIN:

LANDSCAPE AND TERRITORY ON THE VICTORY STELE
OF NARAM-SÎN OF AGADE
If any work of art can be said to have the status of an “icon” in our
field, comparable to the Last Supper of Leonardo da Vinci, it is the
Stele of Naram-Sîn of Agade—commemorating that ruler’s victory
over the confederation of mountain peoples identified as Lullubi in
what remains of the stele’s own inscription (fig. 1).1 The special status
accorded to this monument is not entirely inherent in the work itself;
rather, we come predisposed to appreciate it by our own “habits of
viewing” in the West—habits that privilege a correlation between rank
and size, symmetry or balance disposed on either side of a vertical
axis, and culmination of action in the upper field: not unlike what we
find in religious paintings such as Titian’s Assumption of the Virgin, where
Mary seems to float in the air amid a heavenly host, larger than and
observed by the faithful below.
But then, scholars have gone a step further, in finding in the Stele—as
did Henrietta Groenewegen Frankfort in what is the best art histori-
cal description of the Stele to date—evidence for the Akkadian ruler’s
personal striving against the mountain of his fate: “Man alone before
* This article originally appeared as “Tree(s) on the Mountain: Landscape and

Territory on the Victory Stele of Naram-Sîn of Agade,” in Landscapes: Territories, Fron-
tiers and Horizons in the Ancient Near East, Part I: Invited Lectures [History of the Ancient Near
East/Monographs—III/1], L. Milano et al., eds. Padova: Sargon srl, 1999, pp. 63–72.
1
Discovered in 1898; originally published by J. de Morgan et al., Recherches
Archéologiques, première serie: Fouilles de Suse en 1897–1898 et 1898–1899 (Mémoires de la
Délégation en Perse I), Paris 1900. For the most recent publication of the monument, see
P. O. Harper et al. (eds.), The Royal City of Susa: Ancient Near Eastern Treasures in the Louvre,
New York 1992, No. 109, which includes bibliography; also the recent dissertation
by Dana Bänder, Die Siegestele des Naramsîn und ihre Stellung in Kunst- und Kulturgeschichte
(Beiträge zur Kunstgeschichte, 103), Idstein 1995. The inscription was in the field directly
above the head of the king; it is badly eroded, with only a few cases preserved, which
has led to the frequent misapprehension that the monument was uninscribed in its
original form (published now in I. J. Gelb – B. Kienast, Die altakkadischen Königsinschriften
des dritten Jahrtausends v. Chr. [FAOS 7], Stuttgart 1990, 90f., NSin 4).
110 chapter twenty
his destiny.”2 Here, I fear, the paradigm has been determined by a

particularly 19th century Romantic vision—whether that of Thomas
Cole, whose Voyage of Life: Manhood of 1840 shows the mature individual
riding the rapids of life in his rudderless boat, or that of Caspar David
Friedrich, whose Wanderer of 1818 depicts the individual, male, having
scaled a mountain peak to stand alone before the sublime power of
nature.
Neither of these constructs maps particularly well over Naram-Sîn,
however. He is certainly the victorious ruler and the focus of the work,
occupying a privileged place in the upper field. But in the end, he is not
alone. Naram-Sîn’s own “deification” notwithstanding, he has arrived at
his destination accompanied by an Akkadian cohort; while the protective
amulet around his neck, the symbols that accompany him as standards
into battle, and the celestial elements at the top all attest to the ruler’s
embeddedness within a system of divine power and patronage.3 Also,
his victory is political and military, over a physical enemy, as distinct
from personal triumph or religious ascendance.
Nor can the natural elements of mountain, upward-sloping ground-
lines and trees that provide the setting for the scene be said to represent
the genre of “landscape” as we know it—this despite the apparent
resemblance to Western and even Chinese images of trees in mountain-
ous landscapes. Such images were especially popular in Europe in the
19th century, conforming to then-current Principles of Landscape Painting,
and serve, like the familiarity of the formal composition, to connect
modern audiences to the ancient monument. In many European land-
scape paintings—e.g., John Constable’s Haywain of 1821—the land itself
becomes the “subject” of the work.4 In such cases, our own cultural
2
H. A. Groenewegen-Frankfort, Arrest and Movement: Space and Time in the Art of the
Ancient Near East, Cambridge MA 1987 (orig. 1951).
3
For the battle standards, see now R. Mayer-Opificius, Feldzeichen, in H. Gasche –
B. Hrouda (eds.), Collectanea Orientalia: Histoire, arts de l’espace et industrie de la terre. Études
offerts en hommage à Agnès Spycket (Civilisations du Proche-Orient, Serie I, Archéologie et environment
3), Neuchâtel-Paris 1996, 213–226; for the symbols and deification of Naram-Sîn, see
J. Asher-Greve, Observations on the Historical Relevance of Visual Imagery in Mesopotamia, in
Histoire et Conscience historique dans les Civilisations du Proche-Orient Ancien, Leuven 1990, 180;
for the person of Naram-Sîn and his relationship to divine patronage, see I. Winter, Sex,
Rhetoric and the Public Monument: The Alluring Body of Naram-Sîn of Agade, in
N. Kampen (ed.), Sexuality in Ancient Art, Cambridge 1996, 11–26.
4
The degree to which “trees” were important in the genre is reflected in the vol-
ume devoted to them by Alexander Cozens, drawing master at Eton and the author
of “Principles of Landskip” (sic), when he wrote a book in 1771 entitled The Shape,
Skeleton and Foliage of 32 Species of Tree. His own paintings generally include at least one
tree(s) on the mountain 111
investment in notions of “land” more often becomes an impediment,

rather than an avenue of access to the ancient image.
In this paper, I wish to pursue the “Tree on the Mountain” of the
Stele of Naram-Sîn and its “meaning” within a Mesopotamian system
of referencing the landscape. I may anticipate my conclusions by stat-
ing at the outset that what it is NOT is a tree-for-its-own-sake, of the
sort painted by the 17th century Dutch artist Jan Wijnants in his Dune
Landscape—the typical “landscape-element-cum-portrait” that generally
ended up on the walls of Dutch houses, or even appears in paintings
of walls of Dutch houses, as in Vermeer’s interior, The Guitar Player.
It is certainly true that the Victory stele of Naram-Sîn includes its
landscape “elements” in ways that the previous Early Dynastic Period
stele of Eannatum of Lagash, known as the “Stele of the Vultures,”
did not.5 Given the latter’s fragmentary condition, it is possible that in
the missing areas of the upper two registers, some plant could have
suggested the battle terrain between Lagash and Umma; however, in
view of the absence of such elements on the “battle” side of the simi-
larly registrated “Standard of Ur,”6 it is not likely.
Helene Kantor, in her pioneering article on “Landscape in Akkadian
Art,” argued that the inclusion of, and interest in, landscape was an
innovation of the period, manifest not only on the Stele of Naram-
Sîn, but on a number of cylinder seals as well—from the attention to
the watercourse on a seal of the scribe of Šarkališarri, Naram-Sîn’s
son and successor, to the hunt in a terrain on a seal in the Boston
Museum of Fine Arts (fig. 2).7 Kantor suggested that these elements
represent the “realism” of a unified topographic setting, and considered
them among the “greatest achievements of Akkadian art.”8 Indeed, it
has further been suggested that this approach to terrain was so unlike
previous Mesopotamian tradition that it must be attributed to contact
with Egypt, not just to new “Akkadian” interests—especially as the
tree in the foreground to balance an architectural or natural element in the distance

(for example, High Tor, Matlock, 1756).
5
This division between the Early Dynastic and Akkadian periods, both in terms
of material culture and mentality, has recently been the subject of a number of the
inquiries in M. Liverani (ed.), Akkad: The First World Empire, Padova 1993.
6
E. Stommenger, 5000 Years of the Art of Mesopotamia, New York 1964, Fig. 72 bot-
tom and Pl. XI.
7
Helene J. Kantor, Landscape in Akkadian Art, JNES 25 (1966), 145–152. The seals
are also published in R. M. Boehmer, Die Entwicklung der Glyptik während der Akkad-Zeit,
Berlin 1965, Nos. 232, 721.
8
Kantor, JNES 25, 148, 151.
112 chapter twenty
undulating ground line and occasional floral element incorporated into

a broader narrative field finds close parallels to contemporary Egyp-
tian imagery—as on a Vth Dynasty relief sequence from the tomb of
Sahurē at Abusir, showing the deceased engaged in a hunt of animals
at the desert fringe.9
Before we look at our tree, however, it is useful to review briefly the
human players on the stele: from the left, the soldiers of Naram-Sîn
follow him up the slopes, two behind him in the upper-most line car-
rying the battle standards referred to above (fig. 1). The leading figure,
immediately behind the ruler, has a beard somewhat more developed
than the others. He also carries an axe and wears a bracelet, which
the others do not. By analogy with later Neo-Assyrian reliefs and mili-
tary organization—as seen on a relief of Assurbanipal in the British
Museum, where an officer awards a bracelet to a soldier after battle—it is
most likely that this lead figure himself represents an officer, his bracelet
a mark of field valour and/or status, and not merely ornament.10
The soldiers all direct their attention toward the ascendant ruler,
who stands in triumph with his complete battle arsenal—bow, arrow,
axe, javelin.11 His feet rest upon the twisted bodies of nude, defeated
enemy, while the rest of his enemy are distributed along the right side
of the stele in a variety of humiliated, subservient and/or supplicatory
postures—tropes of defeat well-known from later, Neo-Assyrian art.12
The clothing and hairstyles of these enemy are particularly noteworthy
(fig. 1 and detail, fig. 3). They are consistently shown with a long pig-
tail, and those that are dressed are shown with a garment apparently
9
For the Egyptian parallels, see W. Stevenson Smith, Interconnections in the Ancient Near
East: A Study of the Relationships between the Arts of Egypt, the Aegean, and Western Asia, New
Haven–London 1965, 152f. Additional examples from the Old Kingdom include his
Fig. 178, from Abu Gurob, and Fig. 179, from Saqqara. Bänder has gathered all of the
Egyptian parallels (above, fn. 1, Ch. IV, 247–268), but does not press for dependence,
in my opinion correctly. Kantor’s hypothesis must be tempered, in any case, by more
recent publications of Uruk Period seals on which hunt scenes in an open landscape
not unlike that of the seal in the Museum of Fine Arts indicate that the Akkadian
phenomenon may well demonstrate a special predilection, but not an innovation.
10
On this, see the discussion in F. M. Fales – J. N. Postgate, Imperial Administrative
Records, Part I: Palace and Temple Administration (SAA 7), Helsinki 1992, xxiii, and illustra-
tion on a relief of Assurbanipal (BM 124955; ibid., Fig. 2).
11
My thanks to Marc Van De Mieroop for having drawn my attention to the array
of weapons, comparable to those at tributed later to deities—a topic beyond the range
of the present study (personal communication, 9 February 1993).
12
An excellent discussion of this issue is presented by Megan Cifarelli, Enmity,
Alienation and Assimilation: The Role of Cultural Difference in the Expansion of Assyrian Ideology,
Ph.D. Dissertation, Columbia University, New York, 1994, esp. 166–170.
made from an animal skin. It is striking that when later Assyrian rulers
campaigned in the same region of Iraqi Kurdistan that we believe the
Lullubi to have inhabited,13 they referred explicitly to tribal adversaries
who “wore their hair like women;” and while we cannot be absolutely
certain that it is this coiffure they meant, circumstantial evidence points
in that direction.14 In addition, both Sargon II and Sennacherib repre-
sent enemy from the area in garments that are clearly meant to be the
equivalent of those shown on the stele some 1500 years earlier—pat-
terned texture resembling animal skins, and two-pronged hems that
indicate the uncut skins with limbs still in place.15
Now, in our own culture, since the 17th century and the exploitation
of the New World, we have tended to romanticize the “noble savage,”
and even today, make reference to the “animal” in the virile male as
an attractive quality, played upon by modern advertising. However, for
Mesopotamia, despite the occasional reference to the king as a “lion,”
and with the possible exception of Enkidu, there was no such romantic
counterpart to this noble savage. To gender the enemy “female” by
13
On the region of the Lullubi, see the entry by H. Klengel, Lullu(bum), R1A 7,
3/4 (1988), 164–168, which follows upon his earlier article in MIO 11 (1966), 349–371.
More recent evidence from the second millennium also places Lullubi-country within
travelling distance from Nuzi by donkey—J. W. Carnahan et al., Texts and Fragments:
Nuzi Texts, JCS 46 (1994), 115f.
14
Assurnasirpal II speaks of combat against the Lullu in the area around Mount
Lara, describing the latter, appropriately for our purposes, as “a rugged mountain:” A.
K. Grayson, Royal Inscriptions of Mespotamia II, Toronto 1991, 207 and 249 (= ANP II
A.0.101.1.ii.76b–77 and A.0.101.17.iii.98b–102). In the same area, he describes tribute
from in habitants “who do their hair like women,” ša gim munus. meš-te ap-ru-ni (ibid.,
1.ii.75 and 17.iii.17.97). Although the coiffure is not described further, I believe it is
important to note the long pigtail worn by the enemy on the stele as a hairstyle never
worn by men in Assyria. In earlier Mesopotamia, it seems to be worn occasionally by
men whose gendered identity is complex, and possibly feminized: as, for example, on
the Early Dynastic “Standard of Ur,” where the singer directly behind the lyre-player
on the “banquet” side, as well as the priests standing behind the goddess on the Warka
Vase, all have long hair down the back (see H. Frankfort, The Art and Architecture of the
Ancient Orient, 4th revised edition, Harmondsworth 1970, Figs. 77 and 10–11). These
figures, possibly related to the gala-priest and the assinnu, may fall into a trans-sexual
category; the problem is beyond the scope of the present paper, however.
15
For the animal-skin garments worn by enemy on the reliefs of Sargon II at
Khorsabad and of Sennacherib at Nineveh, I am indebted to John Russell for first
having drawn my attention to this parallel, and would point to the discussion in M.
Wäfler, Nicht-Assyrer assyrischen Darstellungen (AOAT 26), Neukirchen-Vluyn 1975. Indeed,
one of the Sargonid sieges is shown taking place in a mountainous terrain marked
not only by the citadel upon a large rock outcrop, but by an independent, tall, conical
“mountain” nearby, from which just trees project (see drawing in E. Bleibtreu, Die Flora
der neuassyrischen Reliefs [WZKM, Sonderband I], Vienna 1980, Fig. 35).
114 chapter twenty
hairstyle or by action, to unclothe him, and/or to render him “ani-

mal”-like by an emphasis on his “natural” garment, was to deprive him
of civilization, dignity and status.16 Naram-Sîn, therefore, represents
anything but the noble savage: he was clothed, his garment reveal-
ing, but carefully rendered; he was vigorously male; “super-human,”
especially through his deification, rather than sub-human; and in his
dignity, he exuded “civilization”—the victory of culture over nature.
If one must seek a contemporary counterpart, it is images where the
capable individual masters the natural universe through his physical
and mental gifts.
As noted above, Naram-Sîn is positioned at the center of the upper
field of the composition—a complex layout recently studied by Lorenzo
Nigro.17 What I would stress, however, is the bi-partite distribution of
figures in the field below him—another attribute that accords well with
our Western “habits of viewing,” as seen, for example, in paintings of
the Last Judgment, from Giotto’s in the Scrovegni Chapel to Michel-
angelo’s in the Sistine. Naram-Sîn, like Christ, occupies the center, in the
upper field. And just as in the Christian images of the day of judgment,
the “blessed,” or the loyal soldiers—are on our left, the central figure’s
right hand; the “damned,” or the enemy—are on our right, the central
figure’s left. As the blessed rise, so do the Akkadian’s soldiers; while
the enemy and the damned fall. The line of lateral bifurcation on the
stele descends from the heel of the advanced left foot of the king, and
in general the distribution of figures adheres to that divide. Beneath
the left foot of the ruler are the bodies of two contorted, naked, and
clearly dead enemy, their limbs poignantly pointing downward toward
an equally nude figure who plunges head-first down the mountain
(fig. 3), just as later Assyrian enemy plummet from citadel walls and
ramparts in siege scenes.18 The plunging figure’s outstretched arms
16
For the “animalization” of people from the mountains, see A. W. Sjöberg, The Ape
from the Mountain who Became the King of Isin, in M. E. Cohen et al. (eds.), The Tablet
and the Scroll: Near Eastern Studies in Honor of W. W. Hallo, Bethesda 1993, 211–218.
17
L. Nigro, Per una analisi formale dello schema compositivo della stele di Naram-
Sîn, Contributi e Materiali di Archeologia Orientale 4 (1992), 61–100.
18
I am grateful to Zainab Bahrani for having insisted upon the importance of these
figures some years ago. Note that here, too, later Neo-Assyrian (literary) images parallel
closely the Akkadian (visual) imagery. For example, Tiglath-Pileser III tells us, concerning
his Western campaign against Gurgum and Melid: “I filled the mountain gorges (with
their corpses)” (H. Tadmor, The Inscriptions of Tiglath-Pileser III King of Assyria, Jerusalem
1994, 50f., Annals 17). For enemy who fall from walls of besieged settlements on Assyr-
ian reliefs, see the dissertation of Cifarelli, cited above (fn. 12), 267, 296.
lead the eye directly to yet another fallen enemy, who seems to have
landed in a flexed (broken?) position on the mountain slope just above
the preserved lower edge of the stele (fig. 1).
While the headlong figure remains to the absolute right of the median
line, on the enemy side, the buttocks, legs and feet of the lower, landed,
enemy extends part-way into the Akkadian side. The two Akkadian
figures that break this division in reverse, do so to indicate the army’s
successful penetration of enemy territory. The uppermost of these is
positioned with his back to the right of the vertically falling enemy, his
left leg raised in ascent toward our tree. He is clothed in standard-issue
Akkadian headgear, wearing a short battle kilt like Naram-Sîn and his
officers, and is the only figure beside the king armed with bow and
quiver. Based upon the seal of Kalki, a scribe apparently serving the
brother of the Akkadian king (fig. 4)—where a similar figure is shown
in rather rustic dress, carrying a bow and quiver, at the head of a pro-
cession of five men—I would suggest that this individual represents a
“scout,” one who literally leads the way into enemy territory.19
The scout and the fallen enemy, rather than obscuring the median
division, actually strengthen it. Indeed, once recognized, this device of
dividing fields into two distinct halves can be found in other Akkadian
representations as well: on a seal in the British Museum, for example
(fig. 5), the action in a mythological scene of a divine battle on the
left is separated from an offering-cum-presentation scene before the
god Ea/Enki on the right.20 A plant or tree appears to be the divider
on the seal; yet, if one looks at the total space allocated to the two
scenes, mythological and ritual, then the tree actually belongs to the
mythic narrative, which takes place, like the action on the stele, in a
mountainous domain.
The seal imagery thus provides us with an independent example
of flora as landscape marker, which brings us, finally, to Naram-Sîn’s
19
BM, WA 89137, published in J. E. Reade, Mesopotamia, London 1991, 64; also
in Boehmer, Glyptik, No. 717. Inscriptions seem to have been used in the same way
in some instances (examples in Boehmer, ibid., Nos. 31, 148). See also, CAD D, 27f.,
s.v. dajālu, “scout,” inspector. In warfare in unknown territory, and particularly in the
mountains where expert knowledge was essential, this was not a trivial occupation. An
illuminating description of the role is found in the autobiography of the Dalai Lama
(Freedom in Exile, London 1990, 14), who describes the scouts responsible for leading
parties across difficult routes: “These men spent their lives working the . . . routes of
Tibet, and were indispensable to any long journey. They knew exactly where to cross
each river and how much time it took to climb the mountain passes.”
20
WA 116586 (= Boehmer, Glyptik, No. 305).
116 chapter twenty
“Tree on the Mountain.” The tree stands clearly on the “enemy” side,
on rising ground by virtue of a line moving from lower left to upper
right. If we isolate that tree and see it as a discreet element, it is easy
to compare it to other Mesopotamian images, from Akkadian seals (e.g.,
fig. 6) to one of Middle Assyrian date from Assur.21 But that would be
a mistake, I feel, since the stele tree represents a “setting” element in
a broader, historical narrative, while the seals’ trees have emblematic
status as part of a formulaic and generic motif: the mountain landscape.
In short, all plant elements are not equal, and to situate our tree it is
necessary to consider it within a typology of images and their contexts,
such as Jeremy Black is currently pursuing for categories of landscape
in general.22 In a very preliminary sorting, I have isolated several sub-
sets, for each of which I would distinguish “referential,” “symbolic”
and/or “narrative” uses.
First, in the frequent juxtapositions of domesticated plants and
domesticated animals, from the Uruk period onward—as on the lower
register of the Warka Vase23—the reference is clearly to that fertilty
achieved through the yield of the internal landscape, properly managed
by the socio-political system. What suggests that a domestic, domesti-
cated “landscape” is being referenced, in addition to the specific plants
and animals, is the wavy lines representing a watercourse that circle
the bottom of the vase, below the plants—a terrain distinct from the
despoiled and unyielding land of the desert and the desertified. The
divine beneficence to which this abundance is due finds its expression in
the upper registers of the Vase, and its counterpart in European Medi-
eval imagery, where the realia of produce and “labors of the months,”
illustrated in manuscript and often in sculpture on cathedral façades,
suggest production within a cosmic, not just a social universe.
A second, but equally referential, case is that where a “garden” of
managed plants and trees stands for the ideal productivity and abun-
dance desired throughout the land—accounting for the many garden
21
Berlin, VA 1229, found in Tb. 45, most recently published in P. O. Harper (ed.),
Assyrian Origins: Discoveries at Ashur on the Tigris, New York 1995, No. 95.
22
Black is pursuing the classificatory categories for the Sumerian landscape as a
whole: marsh, field, etc. (“Sumerian Mythology and Landscape,” lectures given to
the British Academy and to the Altorientalisches Seminar, Freie Universität Berlin,
in 1996–7).
23
Frankfort, Art and Architecture, Figs. 10–11; and see also a similar usage on a stone
bowl in the Iraq Museum, of similar date (ibid., Fig. 15).
and fruity metaphors in Sumerian and Akkadian poetry.24 This is prob-

ably intended in Assurbanipal’s symposium (the heads of his enemy
suspended from the fir tree notwithstanding), as well as in idealized
Islamic “paradise” gardens (for example, a 16th century ms. of the
Baburnama).25 The domain can be represented by fruit trees or by
husbanded groves of fir and palm, and for it we have ample evidence
in the textual record.26
I would distinguish that “referential” realm of the first two “types”—
field and garden—from the more “symbolic” realm, as exemplified by
the “Assyrian tree.” There, a composite is based upon natural reality,
the palm, but at the same time is constructed and abstracted to sug-
gest simultaneously a temporal sequence and a complex of ideas.27 I
would stress this difference from the more naturalistic and referential
images. The tree of the Assurnasirpal reliefs from Nimrud, for example,
incorporates the trunk and branches of the living palm tree at its core
with a lattice of immature trees that in nature bud off by underground
shoots surrounding the parent. This new growth appears along with the
parent tree as the schematic surround of palmettes, and thus signals
the full temporal growth cycle. The “symbolic” nature of the image is
precisely that it combines aspects that would not be observed combined
in nature into a single schematic motif.28
All of the “types” noted thus far represent plants or trees within a
cultivated landscape, “inside” the state. To this may be added an action
zone “outside” the state, perhaps, but exploited by the state through
hunting and gathering: that is, the forest, steppe, and foothills, in which
24
W. Lambert, Devotion: The Languages of Religion and Love, in M. Mindlin et
al. (eds.), Figurative Language in the Ancient Near East, London 1987, 30f.
25
P. Albenda, Landscape Bas-Reliefs, BASOR 224 (1976), 49–72; BASOR 225
(1977), 29–48. On the garden of Babur, see Wheeler M. Thackston, The Baburnama:
Memoirs of Babur, Prince and Emperor, Washington 1996, 359f.
26
See, for example, fruit trees for Dūr-Šarrukin (S. Parpola, The Correspondence of
Sargon II, Part I: Letters from Assyria and the West [SAA 1], Helsinki 1987, 176f.), that include
apple, medlar, almond, quince and plum; and also the Sennacherib relief of a procession
of men carrying bundles of trees with roots in sacks (ibid., fig. 37). More generally, see
A. L. Oppenheim, On Royal Gardens in Mesopotamia, JNES 24 (1965), 328–333.
27
On this, most recently, B. N. Porter, Sacred Trees, Date Palms, and the Royal
Persona of Ashurnasirpal II, JNES 52 (1993), 129–139; and Iconographic Evidence for
Change in the Significance of Assyrian Tree Images, Paper to the American Oriental Society,
1996. Also, Bleibtreu, Die Flora . . . , passim.
28
Lest it be thought that the Assyrians could not distinguish between the naturalis-
tic representation of the tree and the symbolic, I would draw attention to the glazed
pendants found recently in one of the Queens’ tombs at Nimrud, where the two
representational modes are carefully and clearly differentiated.
118 chapter twenty
game and a variety of trees are sought. This domain is often seen in
Medieval and Renaissance painting in the hinterground, beyond the
walled zone of the “town.”29 This region is amply represented in Akka-
dian seals (e.g., figs. 2, 6), where a conical shape corresponds to the
“mountain” we see on the stele, but is generally covered with scales—the
standard way of rendering terrain.30 The Boston seal incorporates a
complex visual field, with plants and animals disposed throughout the
zone along an irregular line that must indicate a watercourse. Note,
however, that, as with most contemporary Egyptian tomb relief rep-
resentations of the desert fringe, these scenes do not represent “pure
landscape;” rather the natural environment serves as a ground for
human activity in the form of hunting.
While the figures doing the hunting may seem quite undifferentiated
and generic to us, it is important to call to mind the later topos of the
royal hunt, and to note that Naram-Sîn himself records having hunted
aurochs in the mountains following upon a military campaign.31 There-
fore, I would suggest that these scenes do not merely signify generic
passtimes, but rather become a statement of royal activity, and most
specifically of the exercise of royal power within the extended territory.
By the Neo-Assyrian period, the rulers are even importing trees of this
forest zone into the capital, which then serve as anthropogenic game
parks adjacent to the palace.32 I have suggested elsewhere that the reliefs
of Sargon II from Room 7 at Khorsabad, which show a hunt amid fir
trees not unlike that on Akkadian seals, is likely to replicate Sargon’s
own park off the citadel—particularly as the entrance to Room 7 is on
a direct axis off the edge of the citadel toward the Northwest and the
Khosr river, where that park is most likely to have been situated.33
This last is a case of bringing the “out there” to the “in here”, and
can be done with plants, trees and animals. But the whole package of
inhabitants-plus-mountain terrain must still signify the truly “out there,”
where the hills are—that is, to the northwest, north, northeast and east
29
E.g., British Library, Royal MS 18 D II, f. 148, showing Pilgrims leaving Can-
terbury.
30
Kantor, JNES 25, 148; also, Boehmer, Glyptik, Nos. 721, 722.
31
Gelb–Kienast, Die altakkadischen Königsinschriften (FAOS 7), Stuttgart 1990, C6, ll.
5–14.
32
See fn. 23, above.
33
I. J. Winter, Art as Evidence for Interaction: Relations between the Neo-Assyrian
Empire and North Syria, in H.-J. Nissen and J. Renger (eds.), Mesopotamien und seine
Nachbarn. XXV. RAI (Berlin, 2–7 Juli 1978), Berlin 1982, 355–382.
of Babylonia. In just this way, a pair of mountain-peaks are often used

to designate the terrain from which the sun(god), Šamaš emerges; and
on a seal belonging to one Lugal-ušumgal, a tall tree with high trian-
gular canopy not unlike the firs of the Metropolitan Museum hunting
seal is also used to mark the terrain.34
Finally, a fifth and more explicitly “narrative” category—into which
the Stele of Naram-Sîn would fall—is one in which the terrain and its
plants are indicated in the context of a particular military event. Here,
the sequence is not well represented between the stele and the historical
reliefs of the Neo-Assyrian period; however it has been shown that a
military narrative is preserved (albeit poorly) on the base of one of the
altars of the Middle Assyrian period from the Ishtar Temple in Assur,
where soldiers, prisoners and horses proceed across mountainous ter-
rain indicated by an irregular scale-pattern, to be presented before the
ruler, presumably Tukulti-Ninurta I.35 There do not seem to be plant
or tree elements marking the terrain, however.
Certainly, by the early Neo-Assyrian period such flora represent
quite accurate observations of the specific territory, and correspond
directly to verbal narratives of campaigns. For example, in a relief of
Assurnasirpal II from the throneroom of the Northwest Palace, we see
the Assyrian soldiers cutting down fruit trees as they also decapitate
enemy; and this sequence is paralleled in the king’s annalistic text from
the Ninurta temple, in which we are told that in a campaign against
Damdammusa on the northern Tigris orchards were leveled and piles
made of enemy heads (fig. 7).36 By the reliefs of the 7th century rulers,
from Sennacherib to Assurbanipal, botanical elements in profusion help
to create a unified visual field in which the action takes place. Michelle
34
Boehmer, Glyptik, Nos. 431, 432, dedicated to Naram-Sîn and Šarkališarri, respec-
tively, by the ensi of Lagaš. The juxtaposition of a “cedar-like” conifer to the mountains
of the sungod calls to mind the fragment published by G. Steiner, Huwawa und sein
‘Bergland’ in der sumerischen Tradition, AcSum, 18 (1996), 187–215, esp. lines 17–18, in
which Gilgameš calls upon “Utu in the Bergland” to be his helper against uwawa
in bringing down the cedars. This literary association would resolve the ambiguity of
so frequently showing a tree associated with the west on the seals when the sun(god) is
apparently rising in the east; it is equally possible, however, that in ancient times, variet-
ies of conifer were clearly distinguished, the cedar being associated with the northwest,
the pine and/or fir being known to have a wider distribution both west and east.
35
This monument has been discussed recently by U. Moortgat-Correns, Zur ältesten
historischen Darstellung der Assyrer, AfO, 35 (1988), 111–116, esp. drawing, Fig. 2.
36
I. Winter, Royal Rhetoric and the Development of Historical Narrative in Neo-
Assyrian Reliefs, Studies in Visual Communication, 7 (1981), 2–38, esp. 15.
120 chapter twenty
Marcus has recently discussed the issue of Neo-Assyrian representations

of landscape with respect to the vocabulary utilized, finding that the
Assyrians penetrate and dominate the conquered landscape in gendered
ways,37 just as Naram-Sîn represented his enemy female, emasculated,
and animal. In short not only does topography serve an indexical role,
indicating the literal ground, the specific place where the campaign
has taken place and over which victory is claimed; it also serves as the
figurative ground against which the ruler positions himself.
What then is the landscape against which and in which Naram-Sîn
positions himself on the stele? To answer such a question requires a
closer look at the particular landscape elements on the monument—the
mountain and the tree.
Since the mid-16th century of our era, the Western landscape tradi-
tion has rendered its trees in such a way as to both give a sense of place
and also identify that place—as in Herman van Swanevelt’s Landscape
with Tall Rocks of 1643.38 The closest bond between tree and land with
which we in the field of ancient Near Eastern studies are acquainted
is that of the cedar (Cedrus libani ) with the Amanus range and Mount
Lebanon—seen in many European engravings of the mid-19th century,
but also attested in literary and historical texts from earliest times.39
What is clear when one compares the tree on the stele with images
of those cedars, or indeed any coniferous tree—whether in modern
botanical prints or in ancient Near Eastern representations40—is that
our tree is not of that family.
In fact, I think I can with some assurance now identify the tree of
the Naram-Sîn stele as an OAK—as depicted in a painting by Jacob
van Ruisdael called Hilly landscape with large oak of about 1652.41 In
general, the tree is characterized by a full but contained leaf-canopy,
and occasionally has a twisted trunk, or boll, especially when ground
37
M. I. Marcus, Geography as Visual Ideology: Landscape, Knowledge, and Power
in Neo-Assyrian Art, in M. Liverani (ed.), Neo-Assyrian Geography (Quaderni di Geografia
Storica, 5), Roma 1995, 193–202.
38
Rijksmuseum, Amsterdam, A. 2497.
39
As, for example, in the Sumerian “Gilgameš and the Cedar Forest”—A. Schaffer,
Gilgamesh, The Cedar Forest and Mesopotamian History, JAOS, 103 (1983), 307–313.
On the cedar, see N. Liphschitz, Levant Trees and Tree Products, in J. N. Postgate – M.
Powell (eds.), Trees and Timber in Mesopotamia (BSA 6), Cambridge 1992, 33–46.
40
For example, Boehmer, Glyptik, Nos. 181, 189, 251, 264, 657, with upward-curving
branches and often three pinecone-like projections at the top; see also discussion in
Bleibtreu, Die Flora . . ., 103–114, for Sargon II at Khorsabad, and passim.
41
Herzog Anton Ulrich Museum, Braunschweig, No. 376.
soil is poor and meagre. Now, “oak” in botanical terms is a genus

(Quercus), with many different species the result of adaptation to differ-
ent climatic conditions and terrains.42 But fortunately, the British Naval
Intelligence Handbook on Iraq of 1944 and the volumes on the Flora of
Iraq published by the Republic of Iraq since 1966 provide the species
native to Mesopotamia.43 Indeed, the hilly to mountainous environment
of the Kurdish Mts. to the northeast and east were once predominantly
open oak forest, although largely denuded today.44 The species Q. aegilops
(also known as Q. persica) is the most widespread variety.45 It is char-
acterized, according to the British Naval Handbook, by “twisted gnarled
boles,”46 which corresponds quite well with the twisted trunk of the
tree on the stele. What is more, the leaves of this species tend to be
more lanceolate, with barely serrated edges, unlike the great European
and American oaks (fig. 8).47 This fits nicely with the pinnate leaves
that make up the canopy of the tree on the stele, which are just barely
discernable despite surface erosion (detail, fig. 3).
I do not have an opinion on the Akkadian name for this oak species:
whether aluppu, allānu, or something other.48 However, that aluppu
(“oak?”) makes only an occasional appearance in Old Babylonian inven-
tory texts dealing with wood and wood products, and that neither is
among those timber trees cultivated or husbanded for the production of
tools and furniture in Pre-Sargonic Lagash, is consistent with the fact
42
Liphschitz, Trees and Timber (BSA 6), 41.
43
Iraq and the Persian Gulf (Geographical Handbook Series B.R. 524), Oxford 1944, esp. Fig.
81; Flora of Iraq, Vol. I, edited by E. Guest, and Vol. IV, Pt. I, edited by C. L. Townsend –
E. Guest, Baghdad 1966 and 1980, esp. 81–85 and Figs. 24–25, and 44–51, resp.
44
J. N. Postgate, Trees and Timber in the Assyrian Texts, in Postgate – Powell (eds.),
Trees and Timber in Mesopotamia, 180.
45
Flora of Iraq, Vol. IV, Pt. I, 45.
46
Seen in a photograph of 1948: Flora of Iraq, Vol. IV, Pt. I, Fig. 25B.
47
Flora of Iraq, Vol. IV, Pt. I, 50–51 and Pl. 9, 4; also, Liphschitz, Trees and Timber,
illustration on p. 32 (= our Fig. 7). Lorenzo Nigro reminds me that the regularized
shape of the canopy, with its pointed top, also played a role in the overall composition,
by directing lines of vision toward the ruler at the top (personnal communication). I
would take no issue with this; but the shape is still consistent with the generally full and
rounded canopy of the species Q. aegilops; and the carefully depicted leaves still visible
within the canopy make clear that this is not a member of the conifer family.
48
See, on this, CAD A, 354f., s.v. allānu, “oak;” CAD , 55f., s.v. aluppu, “oak?.”
See also Trees and Timber in Mesopotamia (BSA 6), Cambridge 1992, particularly J. N.
Postgate, Trees and Timber in the Assyrian Texts, 179, 182, who suggests that aluppu
may have been the oak native to the east, while allānu represents the western species;
also M. Van De Mieroop, Wood in the Old Babylonian Texts from Southern Babylonia,
159, whose references are exclusively to aluppu, “oak?.”
122 chapter twenty
that, unlike the Western oaks that are so good for building, from ships’
timbers to furniture, the scrub oak does not provide good, evenly-grained
wood, and so should not appear frequently in such inventories.49
I would argue that the identification of Q. aegilops on the Stele of
Naram-Sîn is more than a footnote in the iconography of the ancient
Near East. In the region of Iraqi Kurdistan, from the Diyala up toward
Rowanduz, this variety of oak is associated as well with another oak
species, Q. infectoria, the “gall oak,” as well as the wild pistacio and
juniper pine; but Q. aegilops is often the sole species in the lower slopes,
and still dominates in the middle ranges.50 The distribution map of the
Iraqi forests makes this clear; and when one translates that distribution
onto the topographical map of Iraq, it corresponds to the area in which
we would put not only the Lullubi of Naram-Sîn, but also the people
designated Lulumi in Middle and Neo-Assyrian inscriptions, from the
region around the Diyala to above Sulaimaniye, toward Penjwin, and
up into what was ancient Zamua.51 In short, I would suggest that what
we have on the stele is not just a generic “tree,” but a reference to the
dominant vegetation in the very area of the military campaign being
recorded, of the sort well known for later Neo-Assyrian reliefs.
But I would also like to draw out this single observation into a more
complex picture of all that Naram-Sîn was attempting to pack into
his monument. As is well-known, the actual monuments of Naram-
Sîn are few; but some texts and monuments do exist, and later copies
of this king’s texts have been preserved as well, the full corpus most
49
Van De Mieroop, loc. cit., 159; M. A. Powell, Timber Production in Pre-Sargonic
Lagaš, loc. cit., 121. As an aside, the pairing of the allānu with giš.butnu, “pistacio,”
in a late text (cited CAD A, 355) fits well with the frequently combined habitats of
oak and pistacio, esp. in the Zagros. By contrast, mention of aluppu as one of the
“eastern” woods imported into Mesopotamia from Makan and Melu a, disappearing
with the interruption of eastern trade routes (CAD , 56), is less likely to be oak from
the nearby eastern mountains, and rather suggests a more exotic wood from further
east, such as aloeswood or sandalwood, the fragrance of which would suit for the rare
chairs cited by Van De Mieroop above.
50
Flora of Iraq, Vol. IV, Pt. I, 89. See also the 1952 aerial photograph of the mountain
area in north-eastern Iraq near the Iranian frontier, covered by oak forest and grass,
in P. Buringh, Soils and Soil Conditions in Iraq, Baghdad 1960, Fig. 11. What is more, Q.
aegilops can stand isolated, literally like the representation on the stele, as a “relict:”
a lone or occasional growth on de- or un-forested slopes that need not be very high
in the ranges—as can also be observed for the Turkey oak, Q. cerris, on the Anatolian
coast today, which similarly has elliptical leaves and tends to grow in sandy or limey
soils produced on sandstone and limestone bedrock (G. Aas – A. Riedmiller, Trees of
Britain and Europe, trans. M. Walters, New York 1994 [orig. Munich 1987], 178).
51
See above, fn. 13 (= R1A 7, 167).
recently published by Gelb and Kienast.52 On one stone tablet from

Nineveh, we are told that the king sent an expedition to the source of
the Tigris and Euphrates, during the course of which he cut cedars
in the Amanus for a temple of Ištar.53 He was therefore quite aware
of the primary and most desirable species in that part of the world,
the acquisition of which similarly became a trope in the Neo-Assyrian
period. What is more, a text from Ur tells us that in the course of his
campaign to Armanu, Ebla, the Amanus and the cedar mountains, and
only in that campaign, the regional deity Dagan was among those who
provided the king with weapons and supported his victory.54 In other
words, whether for topographic features or for gods, references were
carefully tailored to be specific to the region in which the campaign
was taking place.
This pattern of regional reference is further manifest on the base
of a diorite statue found at Susa, where Naram-Sîn tells us that dur-
ing a campaign against Magan and Mani um, he cut diorite blocks
from their mountains, which he then brought to Agade to make his
triumphal image, just as Gudea later commemorated his own bring-
ing of diorite from Magan for his statues.55 In the Akkadian case, it
is significant, I believe, that the ruler used the occasion of a military
campaign to acquire the highly-valued hard black stone, a major local
resource, which he then rhetorically celebrated as having used for the
image commemorating that very campaign.
Scholars have long wondered at the unusual pinkish sandstone of
the Naram-Sîn stele, unique among the monuments we know from
Mesopotamia. I would suggest that the use of that stone must be
understood together with the fact that the habitat of our oak, Q.
aegilops, is found on mountain slopes of limestone and sandstone; also
that sandstone outcrops in Iraq are relatively rare, occurring only in
Jurassic and Cretaceous deposits in the west, near Rutba, and in Mio-
cene/Pliocene deposits in a belt above Kirkuk toward Sulaimaniye in
the northeast—in other words, like the tree, in precisely the area of
the Lullubi campaign.56 Therefore, I would further suggest that, as in
52
Gelb – Kienast, Altakkadischen Königsinschriften, cited above, fn. 24.
53
Ibid., NSin 2, 90, ll. 56–60.
54
Ibid., NSin C5, ll. 21–24.
55
Ibid., NSin 3, 90, ll. 37–45.
56
See the discussion and map in Flora of Iraq, Vol. IV, Pt. I, 50–51; Vol. I, Fig. 3;
also, Iraq and the Persian Gulf, 82–83 and Fig. 23, describing the broad sandstone ridge
124 chapter twenty
the Magan campaign, Naram-Sîn is likely to have come home with the
stone as a trophy of his campaign; in fact, it would have constituted a
not-insignificant part of his victory statement, that he had conquered
and possessed himself of a piece of the territory itself.
Unfortunately, in the very damaged portion of the original inscrip-
tion, it is not possible to reconstruct a reference to the stone of which
the stele was made; but between the mention of the Lullubi and the
final verb indicating that the stele had been dedicated to a deity, there
is at least enough space for there to have been some mention of the
monument and its material. Either way, that the stone is indigenous
to the region of the campaign is important. Put together with the
references to the god Dagan for his northwestern campaign and the
diorite from Magan, what emerges from the texts of Naram-Sîn is a
very strong sense of “place;” and a multi-vocalic way of evoking or
marking place.
In pursuing this argument, it is important that Naram-Sîn is the one
king of the Akkadian dynasty to employ the title “King of the Four
Quarters.”57 His direct referencing of the places of his campaigns in
his texts and monuments back home is an equal, if oppposite, phe-
nomenon to the Neo-Assyrian practice of marking the place itself with
rock reliefs or monuments.58
With respect to places he penetrates, Naram-Sîn actually uses a phrase
that becomes a familiar trope in the later periods, of having had to
forge a path through difficult terrain never before explored. In a text
recording a campaign against Elam and Subartu, it is stated: “this way
no one, no king amongst kings, had ever taken; Naram-Sîn, king of
Agade, he took . . .”59 If the annalistic account of Naram-Sîn’s campaign
against the Lullubi had been preserved, one could perhaps anticipate
very similar rhetoric with respect to the northeastern Zagros.
This rhetoric finds direct parallels in later, Neo-Assyrian usage. Assur-
nasirpal II’s account of his accession year campaign in the early 9th
century b.c. speaks similarly of “difficult routes over rugged mountains,”
and is particularly evocative when read in conjunction with the stele:
of reddish clays and sandstones between the Diyala and the Little Zab, which give it
the name of Jebal Hamrin, “the red range.”
57
Gelb – Kienast, Altakkadische Inschriften, NSin 3, C1, C4, C5, for example; and
review by D. Frayne, JAOS 112, 4 (1992), 631.
58
This topic is pursued in the doctoral dissertation of Ann Shafer, Carving of an
Empire: Neo-Assyrian Monuments on the Periphery, Harvard University 1998.
59
Gelb – Kienast, Altakkadische Inschriften, NSin C3, ll. 22–29, emphasis mine.
The mountain was as jagged as the point of a dagger, and therein

no winged bird of the sky flew . . . For three days the hero (the king)
explored the mountain . . . He ascended on foot (and) overwhelmed the
mountain.
Pursuing the enemy, the king in the first person then states further,
I felled 200 of their fighting men . . . with their blood I dyed the mountain
red . . . ; the rest of them the ravines . . . of the mountain swallowed.60
Here, then, is the context in which I believe we must see the “Tree
on the Mountain” of the stele of Naram-Sîn. In an excerpted detail,
our tree may be juxtaposed to other trees used as markers of place in
the history of art: the palm used by Exekias to indicate the setting for
the Suicide of Ajax on a black-figure Greek vase of the 6th century
b.c.—even if there were no palm trees at Troy, it signified “over there”
to the Greeks; the plane tree found frequently in Indian paintings of
the Gita Govinda as a marker of the Gangetic plains on which Lord
Krishna danced with the gopis; and even Michelangelo’s tree in the
Garden of Eden that divides the temptation from the expulsion of
Adam and Eve from paradise on the Sistine Chapel ceiling.
But it will be remembered that, in fact, there is not just one tree on
the stele. There are at least two, set vertically one above the other; and
other reconstructions of the monument are possible, given the breakage
of the stone, that could permit of a third. If in fact these trees, now
plural, do represent Quercus aegilops, the oak of the Zagros, then what
they accomplish is to contribute in no uncertain way to the specificity of
place indicated in the inscription by the naming of the specific peoples
against whom the Akkadian king campaigned. And indeed, if correct,
the stele then manifests a veritable “cluster of elements” to indicate
that specificity of place visually: the stone itself, the carefully selected
landscape elements, and the locals’ hairstyles and garments.
In a recent article, Steven Grosby has discussed the articulation of
frontiers and territories as constitutive characteristics of nationality, and
this fits in well with current historical studies of ideology and land-
scape.61 What the Akkadians, as well as the Assyrians whom Grosby cites,
60
Grayson, RIMA 2 = ANP II A.0.101.1.i.42–43 and 49–54.
61
S. Grosby, Borders, Territory and Nationality in the Ancient Near East and Armenia,
JESHO 40 (1997), 1–29; also, A. R. H. Barber – G. Biger (eds.), Ideology and Landscape
in Historical Perspective: Essays on the Meaning of Some Places in the Past (Cambridge Studies in
Historical Geography), Cambridge 1997; E. Hirsch – M. O’Hanlon (eds.), The Anthropology
126 chapter twenty
were marking was their awareness of the distinction us/not us, projected
onto the land. Thus, unlike Dutch or English landscape painting, the
trees on the mountain do not signal an identification with a place,
but rather an identification of a place: an event that occurred in that
landscape and no other; in that territory, on that frontier.
In such perspective, the “Tree(s) on the Mountain” of the Stele of
Naram-Sîn take on significant meaning: not merely as indicators of an
early interest in rendering “true landscape,” so appealing to historians
and viewers of Western art, but as markers in the developing tradition
of historical narrative. The trees, no less than the mountain itself, serve
as topoi—special topographical features. Through their verisimilitude,
they mark a shift from cognitive mapping—that mental image projected
onto an unknown territory—to a kind of physical mapping achieved
by specific indicators associated with a known, if foreign, geography.62
As such, they work to sustain the reality of the ruler’s claim to victory
in his campaign against the Lullubi.
There is no question but that the primary focus of the stele, com-
positional and ideological, is upon the victorious ruler, Naram-Sîn.
Nevertheless, within the context of the whole scene, the very specificity
of the ground upon which the king stands, and the accuracy of the
landscape in which his victory was achieved, would have served essen-
tial ends: as noted, these markers of the actual time and place visually
quicken the event in the viewers’ eyes; but, perhaps even more impor-
tant, they further serve to naturalize, to present as equally “natural,”
the transcendant stance and status assumed by the ruler.
This is not to deny that the unified field of action and its accompa-
nying landscape—no less than the physically perfect body of Naram-
Sîn—would not have given visual pleasure to Mesopotamians, as they
do to the modern viewer. However, while a part of our pleasure depends
upon the fact that trees, mountain, and sandstone combine in familiar
ways, it is important not to lose sight of the fact that within the Akka-
dian context they serve not merely as value for its own sake, landscape
of Landscape: Perspectives on Place and Space (Oxford Studies in Social and Cultural Anthropol-
ogy), Oxford 1995; C. Nicolet, Space, Geography in the Early Roman Empire, Ann Arbor
1991; D. Cosgrove – S. Daniels (eds.), The Iconography of Landscape: Essay of the Symbolic
Representation, Design and Use of Past Environments, Cambridge 1988.
62
On this distinction, see P. Michalowski, Mental Maps and Ideology: Reflections on
Subartu, in H. Weiss (ed.), The Origins of Cities in Dry Farming Syria and Mesopotamia in the
Third Millennium b.c., Guilford 1986, 129–156; also, P. Gould – R. White, Mental Maps,
Harmondsworth 1974.
for the sake of landscape; nor can those elements be reduced to their
value merely for purposes of identification. Like the vanquished enemy,
they reference a topography—a territory—dominated by the state;
and as such, they also provide a (literal) ground for the representation
of royal ideology.63 In short, the landscape elements on the Stele of
Naram-Sîn should be considered a stratagem as much as an attribute
of the Akkadian period.64
Acknowledgments
I should like to thank Zainab Bahrani, Jeremy Black, Annie Caubet, John Gage, Robert
C. Hunt, David McKetterick, Marc Van De Mieroop, Lorenzo Nigro, Hans J. Nissen,
Nicholas Postgate, Michael Roaf, Duncan Robinson and John M. Russell for providing
perspective, comments and/or sources.
63
Additional cases of political agendas expressed through topographical setting are
discussed by W. J. T. Mitchell, Imperial Landscape, in W. J. T. Mitchell (ed.), Landscape
and Power, Chicago-London 1994, 5–34.
64
It is perhaps no accident that one can point so often to Assyrian narrative and
descriptive traditions as parallels to what is seen on the Akkadian stele some 1500 years
earlier. From the topos of the individual tree or mountain peak to landscaped settings
for battles in general—all of which heighten verisimilitude and therefore historical
claims to reality—these parallels may well signal not only a continuous engagement
with narrative, but a far broader cultural continuum within the north of Mesopotamia
from the Akkadians of the third millennium to the Assyrians of the first . . .
128 chapter twenty
Figure 1. Victory Stele of Naram-Sîn, found Susa [ Musèe du Louvre, Sb4 ].
Figure 2. Akkadian Seal: hunting scene [ Museum of Fine Arts, Boston, 34.199 ].
Figure 3. Detail, Stele of Naram-Sîn: tree, scout and enemy.
Figure 4. Akkadian Seal of Scribe Kalki: scout [ British Museum, WAA 89137].
130 chapter twenty
Figure 5. Akkadian Seal: presentation scene and divine combat [ British

Museum, WAA 116586 ].
Figure 6. Akkadian Seal impression: paired combats with mountain+tree

[ Pierpont Morgan Library, Corpus 159 ].
Figure 7. Relief of Assurnasirpal II, Nimrud: Northwest Palace B.3a, siege

against Damdammusa (?) [ British Museum, WAA 124555].
Figure 8. Leaves of oak species, Q. aegilops [ Flora of Iraq, Vol. 1, p. 32].

CHAPTER TWENTY-ONE
HOW TALL WAS NARAM-SÎN’S VICTORY STELE?

SPECULATION ON THE BROKEN BOTTOM
The Victory Stele of Naram-Sîn of Agade (ca. 2254–2218 b.c.e.) is surely

one of the major monuments preserved from the ancient Near East (fig. 1).
Unfortunately‚ it is partially broken at the top‚ making the curve of
the upper portion of the Stele difficult to reconstruct with surety.1 Even
more difficult is to assess just what and how much is missing from the
bottom (see detail‚ fig. 3).
From what is visible‚ it is clear that at least three lines of soldiers
were represented following Naram-Sîn diagonally from left to right.
They move visually upward through space‚ into the mountainous terri-
tory indicated by a wavy groundline and landscape elements of at least
two trees; and they move through time‚ into battle against the Lullubi‚
mentioned in the partially eroded inscription in the upper field‚ above
the head of the victorious ruler.2
* This article originally appeared as “How Tall was Naram-Sîn’s Victory Stele?
Speculation on the Broken Bottom,” in Leaving No Stones Unturned: Essays on the Ancient
Near East and Egypt in Honor of Donald P. Hansen, E. Ehrenberg, ed. Winona Lake, IN:
Eisenbrauns, 2002, pp. 301–311.
1
See the reconstruction in J. Börker-Klähn‚ Altvorderasiatische Bildstelen und vergleichbare
Felsreliefs (BaF 4; Mainz: Philipp von Zabern‚ 1982)‚ fig. 26k (our fig. 4)‚ where the
curve is gradual‚ more like that of the better-preserved Stele of Eannatum of Lagash
of the preceding Early Dynastic Period (ca. 2460 b.c.e.; = fig. 2). Börker-Klähn then
expands the three extant astral elements preserved in the upper field to seven‚ which
is not convincing. See also die reconstruction in L. Nigro‚ “Per una analisi formale
dello schema compositivo della stele di Naram-Sîn‚” Contributi e materiali di Archeologia
Orientale 4 (1992)‚ fig. 15 (our fig. 5)‚ where the sides tapers more acutely‚ and the
top is more pointed (what Holly Pittman‚ personal communication‚ suggests is most
appropriate‚ as it echoes the shape of the “mountain” before which the ruler stands
in the relief below).
2
I. J. Gelb and B. Kienast‚ Die altakkadischen Königsinschriften des dritten Jahrtausends v.
Chr. FAOS 7 (Stuttgart: F. Steiner‚ 1990)‚ 90f. (= NSin4).
134 chapter twenty-one
The figures of the first two lines of march are well-preserved‚ heads
tilted upward toward their leader‚ weapons and/or standards in hand.
One leg of each soldier is firmly planted on the ground‚ the other bent
at the knee and raised to illustrate most graphically the rising terrain.
For the third row of figures‚ however‚ only the head of one and the
upper half of a second Akkadian soldier are visible‚ along with most of a
tree and the upper torso of a defeated enemy at the right hand margin‚
his arms bent‚ fists clasped at the brow of his upturned head.
As it is presently preserved‚ the Stele measures some 2 m in height.
To complete just these bottom figures; at least 30 cm more must be
added to the overall height of the monument‚ to make room for the
lower bodies and legs of the humans and the base of the tree (see‚
for example‚ fig. 4). Some reconstructions of the Stele have simply
completed the obviously missing portions‚ indicating the outline of
the lower edges of the monolith with dotted lines not closed at the
bottom (e.g.‚ fig. 5). By modern conventions‚ this is sufficient to imply
uncertainty whether there might have been (a) immediate closure at
the horizontal below the feet of the figures‚ (b) a blank or inscribed
dado of uncertain height below‚ or even (c) some further continuation
of the relief imagery.
Before we engage in any speculation as to the missing portions‚ it is
useful to consider the particular stone from which the Stele is carved.
It is a pinkish sandstone that has been argued to come from Miocene/
Pliocene deposits above Kirkuk‚ in the very same area as the Lullubi
campaign.3 The sandstone is grainy‚ and like all sedimentary rocks‚
especially sandstone‚ it is water-absorbent and friable‚ eroding or
flaking easily. It is possible‚ therefore‚ that if the monument had been
embedded in the ground at some point‚ rather than standing on a
more protective base‚ it could have taken in moisture‚ such that either
at its original site in Lower Mesopotamia or when moved to the Susa
acropolis by Šutruk Nahunte to be displayed in the Inšušinak temple
complex where the Stele was found‚4 it would simply have disintegrated
up to the point of present preservation.
3
See I. Winter‚ “Tree(s) on the Mountain: Landscape and Territory on the Victory
Stele of Naram-Sîn of Agade‚” in L. Milano et al. (eds.)‚ Landscapes: Territories‚ Frontiers
and Horizons in the Ancient Near East. History of the Ancient Near East/Monographs III/1
(Padova: Sargon‚ 1999)‚ p. 71.
4
The history of discovery is outlined in Börker-Klähn‚ Altvorderasiatische Bildstelen‚
pp. 134–35‚ and in Harper et al.‚ The Royal City of Susa‚ pp. 159–62.
how tall was naram-sîn’s victory stele? 135
This explanation is less likely‚ however‚ due to the jagged and uneven
lower edge. Had water seepage and disintegration been the cause of
destruction‚ the lines of exfoliation of the surface and the breakage of
the edge would show a more even horizontal or slightly wavy‚ continu-
ous line‚ consistent with the point up to which the stone would have
taken in water‚ rotted and ultimately disintegrated‚ to fall once the seep-
age and disintegration reached above ground level and/or sufficiently
weakened the fabric of the stone.
A second possibility must be explored‚ therefore: that the stone was
instead actually broken in antiquity. This would better account for the
jagged lower edge we see today‚ where the fracture would follow die
crystalline stress lines of the sandstone‚ producing an uneven break line.
Yet‚ if indeed the bottom portion has been broken off‚ we have another
problem: that the fulcrum for a hard snapping of the stone is not likely
to be so close to the bottom‚ were the Stele to be missing only the 30
cm of the readily restored human figures and their putative ground
line. Far more likely is that a sharp blow or impact from falling over
would create a break between two-thirds to mid-way up the stone. The
implications of this are surprising: if we restore the missing portion as
at least 50% again of the present height of the monument (with the
2 m preserved equaling approximately two-thirds of the original and the
approximately one-third now gone adding an additional meter)‚ then
the overall monument would have stood some 3 m tall. This would
leave room for not one‚ but two additional lines of soldiers moving up
the terrain‚ were the decorative portion to have occupied the entire
obverse!
Of course‚ it is certainly possible that‚ even if the monument were
to have stood at least one-third again taller‚ an undecorated dado
or an inscribed band could have completed the bottom‚ rather than
a continuation of the visual narrative. However‚ with a dedicatory
inscription by Naram-Sîn already extant at the top‚ plus the lack of
precedent in the Akkadian period for lengthy texts as opposed to brief
dedications and/or labels‚ and the more usual placement of inscriptions
in the upper field‚ generally in proximity to the ruler‚5 there is really
no compelling reason to hypothesize a lower text. There is at present
no way to prove that there would not have been a blank portion at the
5
See‚ for example‚ Naram-Sîn’s own Pir-Hüseyin stele‚ in A. Moortgat‚ The Art of
Ancient Mesopotamia (New York: Phaidon‚ 1969)‚ fig. 153.
base‚ a relatively small dado band that could have occupied the missing
expanse‚ but even that hypothesis is not fully satisfying.
Whether there are actually two missing register-lines or only one in
the overall scene‚ the vexing problem is how to resolve the bottom of
the visual field. The upper-most line of soldiers‚ two carrying standards
and the third a spear‚ move left to right on an undulating and ascend-
ing ground line that seems to end at the median of the crushed enemy
who lie contorted at the feet of the Akkadian king. The second line of
Akkadian soldiers also moves upward on a wavy and ascending ground
line‚ which in this case does have a visual continuation past the median.
The “scout”6 at the head of the line and the uppermost tree in the
landscape are seen higher up along the same contour‚ which seems to
terminate in the high plateau on which stands a defeated enemy with
broken spear (see figs. 1‚ 4‚ and 5).7 If‚ following this same organizing
principle‚ one sees the evidently missing third line of soldiers moving
along an equally wavy and ascending ground‚ then that would be the
line on which the lower of the two trees on the Stele also stands; and
this line again seems to end as a plateau on which a third vertically-
stacked enemy is positioned‚ supplicating arm raised to the victorious
king. But this leaves the fourth‚ partially preserved and brow-beating
enemy of the bottom right corner on yet another ground line. Now‚
if the resolution of the bottom edge of the representational field were
a horizontal line‚ then that last enemy could be the only figure on his
own‚ independent ground line‚ serving as a filler of the corner‚ with his
feet more or less on the same level as those of the ascending soldiers of
“line 3” (as in the reconstruction published by Börker-Klähn‚ fig. 4).
If‚ however‚ each enemy is to be seen as the endpoint of a rising
ground line‚ then there had to have been a fourth row of ascending
Akkadian soldiers as well! This solution would be consistent with the
recent study by Lorenzo Nigro that argues for narrative coherence
plus rhythmic regularity in the Stele’s opposition of loyal soldiers and
defeated enemy.8 Such an added register however‚ significantly changes
6
Winter‚ “Tree(s) on the Mountain‚” p. 66.
7
Megan Cifarelli has undertaken a study of the defeated enemy of Assurnasirpal II
from the Northwest Palace reliefs; the physical characteristics of posture and gesture
that she identifies are equally applicable here (“Gesture and Alterity in the Art of
Ashurnasiripal II of Assyria‚” Art Bulletin LXXX (1998)‚ pp. 210–28.
8
L. Nigro‚ “Visual Role and Ideological Meaning of the Enemies in the Royal
Akkadian Relief‚” in J. Prosecký (ed.)‚ Intellectual Life of the Ancient Near East: Papers of the
the height and proportion of the monument as we presently envision

it‚ as well as its relief imagery.
The Victory Stele of Naram-Sîn would then have been closest in
height to the original restoration of the Stele of Ur-Namma of Ur in
the University Museum (ca. 2110 b.c.e.; estimated ht. 3 m);9 taller than
the Law Stele of Hammurabi of Babylon (ca. 1760 b.c.e.; ht. 2.25 m;
= fig. 6)‚ and certainly taller than the Stele of Eannatum (ht. 1.80 m
as currently restored; = fig. 2). Most comparable to the Law Stele of
Hammurabi would be the figure of the ruler‚ who stands approximately
54 cm in a visual field of 65 cm on the Old Babylonian monument‚
while the actual body of Naram-Sîn as measured from crown of head
to heel of hind foot would be ca. 58 cm on the Akkadian work (see
comparison‚ figs. 7 and 8). The added height of the Naram-Sîn Stele
would furthermore raise considerably the position of the king vis-à-vis
the viewing audience‚ setting him well above eye level.
In anticipation of such a possibility‚ with the kind permission of Annie
Caubet‚ Conservateur en Chef du Département des Antiquités Ori-
entales‚ Musée du Louvre‚ and the generous cooperation of Prudence
O. Harper‚ then Curator of the Department of Ancient Near Eastern
Art at the Metropolitan Museum of Art‚ I was able to commission two
drawings of the Stele of Naram-Sîn in 1992 (figs. 9 and 10). The first‚
more conventional version‚ simply completes the visible elements as
discussed and does not differ significantly from the previous drawings
of Börker-Klähn and Nigro cited above. Even with such a conservative
reconstruction‚ the monument is taller than the generally triangular form
preserved to us. The second version adds a fourth line of soldiers to
meet the fourth register of gesturing enemy. In that case‚ the problem
of resolution of the bottom right-hand corner remains: it is not clear
how that corner would have been filled in the original. Yet this version
has at least one advantage that should be considered.
In the taller reconstruction‚ with added pictorial register‚ the number
of followers of Naram-Sîn increases: from the nine visible figures plus
presumably a tenth or even an eleventh at the left-hand side of the
bottom line‚ to a probable thirteen or fourteen. If we understand
the two standard-bearers on the upper-most ground line to represent the
43e Rencontre assyriologique internationale‚ Prague 1996 (Prague: Oriental Institute‚ 1998)‚
p. 291.
9
The restored version is reproduced in Moortgat‚ The Art of Ancient Mesopotamia‚ fig.
194. The Stele has since been dismantled and has not yet been reconstructed.
elite guard accompanied by a probable officer‚ just behind the king‚10

then the more conventional reconstruction would allow for only six to
eight foot soldiers‚ depending upon whether one counted the “scout” in
the second line and whether there might be two missing soldiers along
the third line. A possible fourth line would add an additional three or
four figures‚ bringing the total of foot soldiers closer to twelve (again‚
depending upon whether one counted the “scout” in the second line).
In this case‚ the added line conforms more closely with the represen-
tation on the upper register of the reverse of the Stele of Eannatum
(fig. 2)‚ where a phalanx of twelve soldiers in battle formation follows
the ruler—three wrapping around the left hand side of the monument‚
nine continuing along the same line onto the reverse.11 This num-
ber was then replicated in the second register from the top‚ where a
similar set of twelve soldiers (now four on the side and eight on the
reverse) are represented in two tiers marching toward battle behind
Eannatum in his chariot. The Early Dynastic precedent could suggest
that the number twelve‚ or a multiple thereof‚ was used to indicate a
representative sample of a military cohort‚ thereby privileging the taller
reconstruction.12
In order to argue further for such a reconstruction‚ with its attendant
shift in relationship of height to width and incorporating an addition to
the representational field‚ the Stele would have to be understood as the
product of the availability and/or transport of the stone plus a desire
to heighten visual impact. The result would not only foreground the
victorious and perfect ruler‚ as would be the case in either solution;13
it would also raise him further above the viewing public at ground
10
Winter‚ “Tree(s) on the Mountain‚” p. 64.
11
I. J. Winter‚ “After the Battle is Over: The Stele of the Vultures and the Beginning
of Historical Narrative in the Art of the Ancient Near East‚” in H. Kessler and
M. Simpson (eds.)‚ Pictorial Narrative in Antiquity and the Middle Ages (Studies in the History
of Art 16; Washington‚ D.C.: National Gallery of Art‚ 1985)‚ p. 16 [note‚ however‚
that while there are twelve heads‚ there are twenty-four pairs of hands grasping spears
in battle formation before the protective shields].
12
Note also that the reverse of the Stele of Eannatum is divided into four registers‚
albeit along straight horizontal groundlines. Nigro‚ “Visual Role and Ideological Mean-
ing‚” p. 287‚ further observes that a stele from Tello attributed to Rimush (2278–2270
b.c.e.)‚ an Akkadian predecessor of Naram-Sîn‚ was also subdivided into at least four
horizontal registers. Although the Stele of Naram-Sîn is innovative in its representa-
tion of the terrain as a unified field‚ it is possible that the sub-division into four (here‚
diagonal) bands is not entirely unrelated.
13
See on this‚ I. Winter‚ “Sex‚ Rhetoric and the Public Monument: The Alluring
Body of Naram-Sîn of Agade‚” in N. Kampen (ed.)‚ Sexuality in Ancient Art (New York:
Cambridge University Press‚ 1996)‚ pp. 11–26.
level. Indeed‚ one wonders whether such a strategy would not have
been consciously deployed‚ such that in order to focus on the king‚
the viewers’ eyes/heads would have to be angled upward just like those
of the defeated enemy along the right-hand margin of the Stele—in short‚
turning the viewers into subservient subjects as well! This would not
be at all inconsistent with the assumption by Naram-Sîn (and his son‚
Šar-kalli-šarri) to (semi-)divine status in the period.14 We would then
have to see the height of the Ur-Namma Stele and the Law Stele of
Hammurabi not as unique‚ later developments‚ but as part of a gen-
eral trend toward greater monumentality in the later part of the third
millennium and the first half of the second millennium b.c.e. Indeed‚
the 18th-century Old Babylonian ruler confronting/in special and
proximate relation to the sun-god suddenly becomes a powerful visual
counterpart‚ in composition as well as scale‚ to Naram-Sîn confronting
his mountain.
In the end‚ I believe there are not yet sufficient grounds on which to
insist that the taller reconstruction would better represent the original
Victory Stele of the Akkadian ruler. Nevertheless‚ enough problems
remain with the more conservative‚ shorter reconstruction‚ to permit
us to entertain this possible alternative. In that spirit of inquiry and
constant examination of data which Donald Hansen always espoused
in his classroom‚ and of which I was a frequent beneficiary‚ I leave the
question open‚ and invite continued speculation.
Acknowledgments
The pleasure and intellectual stimulus of having had the Stele of Naram-Sîn in New
York as part of the Susa loan exhibition from the Louvre to the Metropolitan Museum
of Art in 1992 (on which‚ see P. O. Harper et al. [eds.]‚ The Royal City of Susa: Ancient
Near Eastern Treasures in the Louvre [New York: Metropolitan Museum of Art‚ 1992]‚
no. 109) was heightened for me by the experience of crawling about on the floor with
Donald Hansen the night of the grand opening‚ as we examined the ragged edge of
the lower portion of the monument. It seems therefore most appropriate to include
this speculation in his honor.
14
The visual implications of this “deification” are discussed by J. Asher-Greve‚
“Observations on the Historical Relevance of Visual Imagery in Mesopotamia‚” in
Histoire et Conscience historique dans les Civilisations du Proche-Orient Ancien (Leuven: Peeters‚
1990)‚ p. 180.
Figure 1. Victory Stele of Naram-Sîn‚ found at Susa. Akkadian Period‚

ca. 2250 b.c.e. Pinkish sandstone; ht. as preserved‚ 2 m. Louvre‚ Sb 4
(Photo: author‚ reproduced Courtesy of the Musée du Louvre).
Figure 2. Reconstruction drawing of the Victory Stele of Eannatum of

Lagash‚ found at Tello. Early Dynastic III Period‚ ca. 2460 b.c.e. White
limestone; ht. as preserved 1.80 m. Louvre‚ AO 50 + 2436–8 + 16109
(Drawing: Elizabeth Simpson).
Figure 3. Detail‚ bottom of Victory Stele of Naram-Sîn (Courtesy of the Metropolitan Museum
of Art).
Figure 4. Reconstruction drawing of Victory Stele of Naram-Sîn.

Börker-Klähn‚ Altvorderasiatische Bildstelen‚ fig. 26k.
Figure 5. Reconstruction drawing of Victory Stele of Naram-Sîn. Nigro‚

Contributi e Materiali‚ fig. 15.
Figure 6. Law Stele of Hammurabi of Babylon‚ found at Susa. Old Babylonian

Period‚ ca. 1760 b.c.e. Diorite; ht. 2.25 m. Louvre‚ Sb 1 (Courtesy of the
Musée du Louvre).
Figure 7. Detail‚ visual field of the Law Stele of Hammurabi; ht. approx.
65 cm (Courtesy of the Musée du Louvre).
Figure 8. Detail‚ ruler and mountain from Victory Stele of Naram-Sîn;

ht. of figure‚ approx. 58 cm (Courtesy of the Musée du Louvre).
Figure 9. Reconstruction drawing of Victory Stele of Naram-Sîn‚ with one

register added; ht. approx. 2.30–2.50 m (Drawing: Denise Hoffman‚ 1992).
Figure 10. Reconstruction drawing of Victory Stele of Naram-Sîn‚ with two

registers added; ht. approx. 2.80–3.0 m (Drawing: Denise Hoffman‚ 1992).
CHAPTER TWENTY-TWO
THE BODY OF THE ABLE RULER: TOWARD AN

UNDERSTANDING OF THE STATUES OF GUDEA
For the twelve years that I taught at the University of Pennsylvania,

Åke Sjöberg was a source of constant information, affection, sup-
port, sometimes exasperation, always laughter. My debt to him is
great—beginning with the first semester, when he would appear in
my survey of Mesopotamian art each week, graduate students from
the Tablet Room in tow, to demonstrate by example his commitment
to the teaching of that material; and continuing throughout, as he
generously made sure certain books came my way, his own files and
library always open. What is most important, Åke has been a significant
factor in the special environment at Penn, where students of philology
have been exposed to material culture from several perspectives, and
students of art history and archaeology have gotten serious training
in language—thereby redressing the limited vision that has too often
characterized our field, in which texts and images have been seen as
belonging to separate domains, rather than as aspects of a single and
mutually-referential cultural system.
It is in such a vein that I offer the present essay—as an initial inquiry
into what will eventually be a more comprehensive study, in which
the combined verbal and visual message of selected works of early
Sumerian “art” will be explored, with a view toward an understanding
of the affective intent of the whole within the specific Mesopotamian
context.
I have purposely chosen to begin with the Gudea statues—those old
staples of the Mesopotamian repertoire—where in fact text and image
are brought together on the body of the statue itself. Gudea, ensi of
Lagash, is probably the second most well-known ruler of ancient Meso-
potamia, next only to Assurbanipal and his lions. He appears in virtually
* This article originally appeared as “The Body of the Able Ruler: Toward an under-
standing of the statues of Gudea,” in Dumu-É-dub-ba-a: Studies in Honor of Å. W. Sjöberg,
H. Behrens et al., eds., Philadelphia: The University Museum, 1989, pp. 573–583.
152 chapter twenty-two
every secondary survey of the history of art, Mesopotamian or global, as

the arch-icon of the oriental ruler: relatively narcissistic, else how could
there be at least 20 remaining statues of him (“a great patron of the
arts,” explained Woolley);1 not very imaginative (his statues “resemble
one another so closely” that a small sample “adequately represent(s)
the whole group,” said Frankfort);2 imbued with traditional values of
permanence and piety (his “immutability” manifest in the cylindrical
block, witness to a “tendency which permeated all the Eastern world,”
according to Moortgat).3
Most of these characterizations constitute projections from the
perspective of 19th and 20th century Europe and the West. Gudea
is seen as a patron of the arts-for-arts’-sake, rather than an individual
with specific, culturally-determined functions and goals for his several
representations. Although in general his images closely resemble each
other, minor variations in representation and significant variations in
accompanying text serve as operative factors in distinguishing one work
from another, while providing continuity in the basic form. And finally,
Gudea’s goals of piety and permanence are not to be seen merely as
expressions of his personal ambitions, but as acting on behalf of his
people and the state, within a Mesopotamian framework of the just
and effective ruler.
Who, then, was this ruler, Gudea of Lagash, and how did his images
and their texts interact as communicators on their own terms? I hope
ultimately to pursue these questions in seven stages: I. The historical
individual; II. The visual properties of his sculptural corpus; III. The
verbal properties of the accompanying texts; IV. The combined verbal
and visual message; V. The ritual and metaphorical activation of the
message; VI. The placement of the images; and finally, VII. The overall
purpose of the statues.
In this essay, I shall present a very brief introduction to the historical
individual, and concentrate on the intersection between certain visual
aspects of the free-standing statues of Gudea and some of the verbal
representations of the ruler evident in his own texts, in order to dem-
1
C. Leonard Woolley, The Development of Sumerian Art (New York, 1935) 107–8.
2
H. Frankfort, Art and Architecture of the Ancient Orient (Pelican History of Art, NY
and Harmondsworth, 1954) 47.
3
A. Moortgat, The Art of Ancient Mesopotamia (Phaidon, London and NY, 1969)
62–63.
the body of the able ruler 153
onstrate that what have been considered to be merely formal properties

of the works are in fact highly-coded signifiers of meaning.
I. The Historical Individual
By his own account, Gudea ruled as ensí (governor) of the autono-

mous city-state of Lagash in southern Mesopotamia. Successor of his
father-in-law, Ur-Baba, and predecessor of his son, Ur-Ningirsu, he was
the most prominent figure in what has been identified as the Second
Dynasty of Lagash, dating toward the end of the third millennium
b.c. A large corpus of texts and images preserved from his satellite city
of Girsu, modern Tello, and various finds from other sites such as Ur
and Nippur, provide us with a picture of public works and activities
during his reign.4
The estimated 20 years of Gudea’s reign were long thought to have
preceded the emergence of the complex and powerful Third Dynasty
of Ur, which ultimately came to dominate the remaining city-states
and surrounding territories of southern Mesopotamia. Recently, how-
ever, evidence has emerged that seems to demonstrate that the reign
of Gudea overlaped with that of Ur-Nammu, first ruler of the Ur III
dynasty, with a complexly interwoven set of relations existing between
Ur and Lagash into and beyond the reign of Ur-Nammu’s son and
successor, Šulgi.5
This is important for our purposes, because it helps to account for
striking similarities in both verbal and visual expression in works associ-
ated with the two polities. It allows us, also, to use the arts and docu-
ments of each to illuminate with greater security the other. In addition,
it permits us to explain the prominent position alotted to Gudea by
subsequent Ur III rulers in the calendar of ritual activities.6
4
Cf. historical summary by A. Falkenstein, “Gudea (A. Nach Texten),” RlA 3,
676–679; related monuments presented by E. Strommenger in the same volume,
“Gudea (B. Archäologisch),” pp. 680–688, and most recently studied by A. Spycket,
La statuaire du proche-orient ancien, Handbuch der Orientalistik 7 (E. J. Brill, Leiden-Köln,
1981) 184ff.
5
Cf. summary of literature and evidence in P. Steinkeller, “The Date of the Gudea
Dynasty,” JCS, 40 (1988), 47–53.
6
See mention in Falkenstein, RIA 3,679 re offerings by Amar-Suxen and Ibbi-Suxen,
and discussion in B. Perlov, “The Families of the Ensís Urbau and Gudea and Their
Funerary Cult,” CRRAI 26, B. Alster, ed. (Copenhagen, 1980) 77–81, as well as the
II. The Visual Properties of Gudea’s Sculptural Corpus
During the course of intermittent excavation at Tello between 1877 and

1933, some twenty statues of Gudea were brought to light—a solid core
from excavated contexts, with which we shall mainly concern ourselves
here, others attributed to clandestine digs carried out (see examples, our
figs. 1–3) at the site in the 1920’s, and yet others via the art market,
without known provenience.7 Formal descriptions and chronological
analyses of some or all of this corpus abound,8 and I shall not attempt
to recapitulate, or even review, that literature here. Virtually universal
are references to the basically cylindrical form maintained in the block
(although opinions differ in the degrees to which the underlying body is
[Frankfort] or is not [ Woolley] apparent beneath the wrapped garment
that is the skin of the cylinder); the sense of overall mass; the muscular
arms; the clasped and therefore immobilized hands, associated with an
attitude of pious concentration; the enlarged eyes, recognized as part
of a continuing Mesopotamian tradition; and the characteristic facial
features of broad, flat cheekbones and prominent chin (that have led
some to speak of “portraiture” and others of ideal “essences”).9
These factors constitute ready criteria for recognizing and classifying
“Gudea” figures and their contemporaries, allowing us to place them
in general between the Akkadian and Old Babylonian periods, and
recent dissertation by R. E. Averbeck, “A Preliminary Study of Ritual and Structure in

the Cylinders of Gudea,” (unpubl. diss., Dropsie College, Philadelphia 1987) 17.
7
See the account in A. Parrot, Tello, Vingt campagnes de fouilles (1877–1933) (Éditions
Michel, Paris, 1948) based upon the original publication of E. de Sarzec, Découvertes
en Chaldée (Paris, 1884–1912).
8
Cf. Woolley, Frankfort, Moortgat, cited above; E. Strommenger, “Das Menschen-
bild in der altmesopotamischen Rundplastik von Mesilim bis Hammurapi,” BaghMitt
1 (1960), esp. 63–69 on Gudea; F. Johannsen, The Statues of Gudea, Ancient and Modern,
Mesopotamia 6 (Copenhagen, 1978) which challenges the authenticity of a number of
the attributed works; Spycket, 1981, cited above, n. 4; the dissertation by G. Colbow,
Zur Rundplastik des Gudea von Lagaš (Munich, 1987) which I have not yet been able to
see; Donald P. Hansen on the Stoclet Gudea, recently acquired by the Detroit Institute
of Arts: “A Sculpture of Gudea, Governor of Lagash,” Bulletin of the Detroit Institute of
Arts, 64:1 (1988) 5–20; and a lecture by Horst Steible, “An attempt at establishing a
chronological order for the statues of Gudea of Lagash from the point of view of a
Sumerologist,” given at the Universities of California at Berkeley and Pennsylvania in
the Spring of 1989, and kindly made available to me.
9
Cf. on this, comments in Winter, review of Spycket, La statuaire du proche-orient ancien,
in JCS, 40 (1984) 107–8, where an intermediate stage, between purely “idealized” forms
and “true portraiture” is discussed, one that includes “signature elements” that permit
identification of a particular individual, and yet idealizes some features.
more specifically in sequence with regard to each other.10 However, it

is my point here that while formal properties of style and rendering
are valid criteria for attribution, they also have another value: that of
expression.
Just what those particular properties are designed to express can
only be understood if taken in conjunction with textual references to
the intent of the statues and the nature of the subject represented.
Particularly relevant in this regard is the text inscribed on Statue B
(fig. 2),11 which opens with a list of attributes ascribed to Gudea. These
attributes are couched as gifts bestowed by a series of gods, and serve as
indicators of those qualities deemed important for and by the ruler. A
number of them echo closely aspects of the gods Ningirsu and Nindub
as they are described in Gudea’s dream, preserved in the lengthy text
of Gudea’s Cylinder A.12 By comparing the verbal representations of
the ruler with his physical representation in the statues, what emerges
is a consistent and coherent picture in both domains—constituting
what Louis Marin, in his study of the images of Louis XIV of France,
has referred to as “representations of power,” brought into being and
reinforced by the “power of representation.”13
In brief, it is argued that qualities of expression associated with visual
imagery are retrievable, but must be seen as culturally and historically
specific, for the understanding of which internal evidence must be
adduced, before value can be ascribed. It is further argued that the
common art-historical equation of “style” with the manner of rendering,
hence mere form—thereby viewing “iconography” or subject matter as
the sole conveyor of meaning—must be challenged; for I shall proceed
now to several instances in which it may, I believe, be demonstrated that
with the statues of Gudea many of those properties we would place in
10
Spycket, for example, would see some of the statues as more “youthful” than
others, thus representing earlier phases within Gudea’s reign (La statuaire 190–191). It
is interesting that the relative chronology thus arrived at would correspond with the
sequence proposed by Steible on the basis of the dedicatory inscriptions (cf. lecture,
cited above, n. 9).
11
Louvre AO 2, de Sarzec Découvertes, pls. 14–19. For the text, I am grateful to have
had access to the ms. of H. Steible, Die neusumerischen Bau- und Weihinschriften [Freiburg
altorientalische Studien 9 (forthcoming)].
12
Steible, FAOS 9, cited above; and see also the recent translations included in
T. Jacobsen, The Harps that Once . . . (Yale University Press, New Haven and London,
1987) 388–425, and in Averbeck, Ritual and Structure 589–678.
13
L. Marin, Portrait of the King, M. H. Houle (transl.), (University of Minnesota
Press, Dexter MN, 1988) 4.
the domain of form are actually charged in the Mesopotamian context

to carry meaning in quite significant ways.14
1. Height
A number of the standing Gudea figures are quite tall (e.g., 1.42 m
[our fig. 1] and 1.25 m, both without heads);15 one of the seated stat-
ues is downright colossal (1.57 m);16 and even those that are smaller
than lifesize give a general appearance of being “larger-than-life,” by
virtue of their overall proportions and massiveness. The importance
of size as a measure of distinctiveness is implied in the couplet that
records Gudea’s “rightful head made to stand out in the assembly by
(his personal god) Ningišzida”:
B iii 3–4: sag-zi ukkin-na pa-e3-a
d
nin-giš-zi-da
While this is the weakest of the parallels that can be cited between text
and statue, I feel it is still worth pointing out. On the same basis as
“outstanding” in English implies height, “standing out in a crowd” as
a concrete allusion to having exceptional and noteworthy qualities in
general, I would suggest that the “largeness” conveyed by the Gudea
statues equally gives expression to his “outstandingness,” which in
turn led to recognition of his capacity for leadership—much as the
designation of a “big” man in the Middle East today, one who “casts
a big shadow,” uses a metaphor of physical size to imply charisma and
social power.
2. Breadth of Body, Chest

The use of the adjective dagal, indicating width or breadth, as a posi-
tive attribute implying something “lavishly” or “abundantly” present, is
14
Gudea was clearly not the first ruler to portray himself in stone—even diorite:
witness the partially-preserved figures of the Akkadian ruler, Manishtusu (Moortgat, Art,
Fig. 141) and of Gudea’s father-in-law and predecessor, Ur-Baba (Fig. 164). However,
while it may well be due to the accidents of archaeological discovery, it is only with
this ruler and his large assemblage of associated images and texts that we are in a
position to evaluate the particular confluence of style and expression toward intended
meaning that I wish to pursue here.
15
E.g., Louvre AO 6, and AO 5, resp.
16
Louvre AO 1, known as Statue “D.”
generally important, as will be seen in section 4, as well. Here, Gudea

is described as, “his life within him abundantly (lit. widely) supplied
by (the God) Šulšaga”:
B iii 1–2: zi-ša3-gal2-la šu-dagal-du11-ga
d
šul-ša3-ga-ka-ke4.
I would suggest that the ascription of width, or breadth, or fullness of
life is expressed precisely where in the body the fullness of breath is
manifest, that is, in the chest (cf. fig. 1). The implication is reinforced
by the description of Ningirsu himself later on in the Statue B inscrip-
tion (ix 27) as gaba-gal2-dingir-re-ne-ke, “great-chest-having (one) of
the gods,” with the attributes of largeness, fullness of life and power
contained therein.
3. Full-Muscled Arm
The fringed garment worn by Gudea is draped around his body,
tucked with a fold under the right armpit, with the free end falling
over and covering the left arm. The right arm and shoulder is thus left
fully exposed (e.g., figs. 1 and 2). The well-developed, almost bulbous
musculature of that right arm and shoulder is often remarked upon
as a notable feature of the Gudea “style.” Yet, only when this stylistic
feature is read in conjunction with one of the most frequently-employed
epithets of Gudea, can we account for that massive musculature in a
meaningful way. Among the divinely-sanctioned attributes of the ensí
occurs the epithet, “strength-given one of (the God) Nindara”:
B iii 12–13: a2-sum-ma
d
nin-dar-a-ke4
Essential here is an understanding of the play between the cuneiform
sign for arm with the meaning “strength,” a2 [ DA+šeššig]. In this
way, one literally reads the epithet as “(strong) arm given by the god
Nindara.” The reference closely parallels Gudea’s description of the
god identified as Nindub, who appears in his dream sequence recorded
in Cylinder A:
A v 2–3: mina-kam ur-sag-ga2-am3
a2 mu-gur8
Translated most recently by Thorkild Jacobsen as, “the second one
(in the dream) was a hero; he was mighty of strength,” it literally means,
“he was of massive arm”; and indeed, in a repetition later on, as the
dream is being interpreted, Jacobsen himself originally translated the
same phrase poetically as, “the arms bulged on him.”17 Thus, one who
knew the physiognomic code of Mesopotamian somatotypes could not
fail to recognize in Gudea’s exposed arm the heroic power vested in
the ruler, as manifest in his image.
4. Broad-Faced; Wide-Eared
I believe the revealing passage here is that in Cylinder A, where Gudea
describes himself as responding to the divine instructions given him
by Ningirsu:
A i 12: ensi2 lu2-geštu3-dagal-kam
geštu3 i3-ga2-ga2
Loosely translated, we are told that “the ensi, a man of wisdom, was
giving ear”—that is, paying attention, listening, giving heed; but literally,
the phrase used to indicate a wise man is one “of wide ear,” geštu3-
dagal. Hence, the ensi, a man (who is) wide of ear, was listening hard.
Once again, we see width as a positive attribute: one who is wide-eared
is one who is able to pay attention, be attentive and wise, listen well.
This calls to mind the title, “attentive prince,” applied to Neo-Assyrian
rulers as one of the primary attributes of the just and appropriate ruler.18
While it is perhaps less compelling than the parallel between power-
fully muscled arm and strength, cited above, it is nonetheless tempting
to see a correlation between the broad-faced and wide-eared heads of
Gudea and the qualities of attentiveness and wisdom attributed to the
ensi (cf., e.g., Fig. 3).19
17
Jacobsen Harps 393 for lines v 2–3; his translation, “the arms bulged on him,” for
line vi 3 was used in an early, privately-circulated ms, later amended for publication
to conform with the first reference, as “mighty of strength.”
18
I. J. Winter, “Royal Rhetoric and the Development of Historical Narrative in
Neo-Assyrian Reliefs,” Studies in Visual Communication, 7 (1981) 21.
19
Cf. both heads with cap, e.g., AO 13 (our Fig. 3), and without cap. AO 12: de
Sarzec Découvertes pl. 12:1 and 2. One would expect the same feature of divine rep-
resentations in the period as well, in conformation with the reference by Nammaªni
of Lagash, another son-in-law of Ur-Baba and successor to Gudea and his direct line,
who requests of the goddess Baba that she “turn her ear” to the dedicator (cf. Steible,
FAOS 9, N1: col. 2:4–6)—thus emphasizing the ability of the god, no less than that
of the ensi in governing, to hear and respond to the supplicant, as well as the ability
of the obedient servant to respond to the commands of god or ruler.
5. Large-Eyed
As a final case, we are told toward the beginning of the Statue B
inscription that Gudea is “looked upon with favor by Nanše”:
B ii 10–11: igi-zi-bar-ra
d
nanše-ke4
And we are told toward the end of the same inscription that the focus
of the gaze of the statue, (Sum. alan) is upon the god Ningirsu, to whom
the image is dedicated, and in whose sanctuary it was to be placed:
B vii 58–59: alan igi-zu
d
nin-gir2-su-ka-kam
literally, “statue, your eye is of (belongs to/is upon) Ningirsu.” The
phrase has been translated by some to mean that the god Ningirsu
actually sees through the eye of the statue, making the reference into
an important theological point concerning the god acting through the
king, and the special importance of maintaining the offerings to and
identity of the statue by later rulers, since not only Gudea, but the god
is involved.20 Yet, the reference is placed immediately after a descrip-
tion of the setting up of the statue, and I would rather suggest that
it serves as part of the instructions to the statue: Gudea had clearly
instructed the statue earlier (B vii 21–24) that it was to speak directly to
the god; here (B vii 58–59), the statue is reminded that its objective is to
constantly keep its attention upon the god (its eye belongs to the god).
Seen in this way, the two references cited make clear the importance
of the gaze: on the one hand, the ruler is gazed upon with favor by
the gods, and presumably returns that favor with reciprocal attention,
the eyes being an important vehicle for the expression of attentiveness,
just as the ears are for giving heed to verbal commands; and on the
other hand, a prime purpose of the statue of the ruler is to maintain
focussed attention upon the deity, expressed as keeping its eye perma-
nently on the object of its gaze.21 Small wonder, then, that the eyes of
20
P. Steinkeller, personal communication.
21
The association of the gaze with “piety,” as distinct from mere “attentiveness”
needs to be pursued. It is possible that the eyes must be seen in conjunction with the
clasped hands; although the meaning of that gesture as salutation, as indicating prayer,
or merely of respect, is not at all clear (cf. on this, M.-Th. Barrelet, “La ‘figure du
roi’ dans l’iconographie et dans les textes . . .,” in Le palais et la royauté, CRRAI 19, P.
Garelli, ed., (Paris 1974) 49–50, citing Landsberger, “Das gute Wort,” in MAOG 4/1
[1928–9] 295).
the Gudea statues should be enlarged, just as they had been throughout
the history of Mesopotamian votive statuary for over a millennium, as
images of high-ranking individuals were placed within sanctuaries, to
serve as permanent representatives of their patrons.22
With these five cases in mind, we may conclude that the stylistic
features described above are not merely formal properties of the works,
but rather have been deployed as signs, carrying definite and identifiable
value, to accord with the rhetorical ends of the statues. We, as scholars
at a historical remove from the later third millennium b.c., must have
recourse to the textual glosses in order to identify the meanings car-
ried by physical attributes. But for the Sumerian, as distinct from the
Sumerologist, we may speculate that, quite independent of the texts,
simply by seeing the ruler represented thusly, the knowledgeable viewer
would have been informed at once that the ruler was a leader, full of
life and strength, wise and attentive to the gods—all attributes claimed
in the inscription as well, and conforming to Mesopotamian ideals of
rulership.23 Indeed, these qualities seen collectively represent requisites
for the status claimed as Gudea’s very first attribute, “shepherd found
in the heart of Ningirsu:”
22
As found, in situ, in the Early Dynastic Ishtar temple at Assur (cf. reconstruction
in W. Andrae, Das wiedererstandene Assur [ Leipzig 1938] Fig. 24). The likely placement of
the seated, as opposed to the standing statues of Gudea will be pursued in a subsequent
amplification of the present study. The text of the seated Statue ‘B’ suggests that it was
installed in the ki-a-nag, located within the Eninnu; and whether that disputed place
represents a chapel specifically dedicated to Gudea, where the statue itself was the
recipient of worship, or whether it would have been placed within sight of an image or
symbol of the god to whom it was dedicated, does not diminish the importance of the
enlarged eyes, since the first reference cited, that of Nanše regarding Gudea, indicates
the importance of the cult-object’s gaze, while the Assur temple context emphasizes
the importance of the votaries’ gaze.
23
These very attributes are precisely those stressed by Kramer as well, in his study
of the ideal ruler, using literary texts of the Ur III period: S. N. Kramer, “Kingship in
Sumer and Akkad: The Ideal King,” in CRRAI 19, 176. Kramer describes the king as
necessarily “brave, wise, pious and benevolent.” I was further struck by the continuity
therefore in overall attributes of rulership from the third to the first millennium, as
witnessed in the epithets apparent from the titles of Assurnasirpal II in his Standard
Inscription at Nimrud: i.e., “attentive prince, keeper of the gods, fierce predator and
hero in battle” (cf. discussion in Winter, Studies in Visual Communications 7 [1981] 21–2);
and would even go so far as to suggest that there is great comparability between the
attributes of Gudea (attentiveness in the head/ears; piety in the eyes, and perhaps also
in the hands; powerful life/virility being comparable to the ferocity of the predator; and
heroic strength in the arms) and those of Assurnasirpal, although the latter king chose
to illustrate those attributes through narrative, while Gudea portrayed them as personal,
physical attributes. Still, there are many similar features in the powerful proportions of
the two rulers’ figures, despite stylistic differences appropriate to their dates.
B ii 8–9: sipa ša3-ge-pa3-da

d
nin-gir2-su-ka-ke4
—where, for shepherd, we are to understand “good ruler of his people,”
favored by the chief god of his land.24
Visual attributes, no less than verbal epithets, thus function as part
of a signaling code, with “style” very much a carrier of meaning. The
particular physical traits represented would be seen in conjunction
with the major iconographic signifier of “rule” seen on many of the
Gudea statues: the round-brimmed cap associated with kings from Ur-
Nammu to Hammurabi.25 The overall sense conveyed lies in the very
nature of the representation of power and the power of representation
noted above; the statues at once articulate and constitute the right of
Gudea to rule.26
The Gudea statues, then, take their place alongside those of rulers
from East to West, ancient to modern, in which the image of the ruler
was at once the likeness of an animate body (the representation of an
actual person) and also an archetype (a fictionalized and idealized symbolic
type possessed of all the appropriate attributes of rule). In addition, as
the object of offering and sacrifice, the image itself takes on a presence,
not only within the cult, but within society. These three functions cor-
respond to the three bodies of the king, made manifest in his statues,
as discussed by Marin for Louis XIV: the historical, the political, and
the sacramental.27
The particular expressive qualities of Gudea are specific to time
and place; the nature and function of representation are not. In the
litany of attributes proferred by the panoply of gods in the beginning
of the inscription on Statue B, we find the verbal equivalents of the
visual attributes of Gudea—attributes that proclaim him strong, wise,
attentive, pious—marked for leadership and shepherd of his people.
24
The epithet frequently employed by Gudea as the summary statement of his
leadership qualities (e.g., Cylinder “A” ix 5; Cyl. “B” ii 7, etc.), “able,” or “righteous
shepherd,” is one which bears further study—particularly in view of the statue of Šulgi
found at Tello, where the king is shown holding a young kid in his arms (cf. Parrot Tello
222, and good photograph plus text published by J.-M. Durand in Barrelet, CRRAI
19, Pl. I and pp. 96, 130).
25
On the cap, see Barrelet in CRRAI 19, 38–39.
26
And, as with representations of late antique rulers, Julian and Theodosius, “those
who looked upon them (their images), looked upon the only constitution the state
possessed.”—Peter Brown, “The Philosopher and the Monk,” unpubl. ms., gratefully
acknowledged.
27
Marin Portrait 13.
Certainly, Gudea also has aspirations for the continuation of his seed
and his name:28 however, the maintenance of his ritual offerings and
cult by later Ur III kings suggests that by revering him, they were also
revering what he stood for.
It has been my thesis here that what Gudea stood for was not only
explicitly articulated in text, it was also articulated through expressive
physical attributes manifest in the ruler’s image. What we see in the
Gudea statues is thus a remarkable confluence of form, expression,
and meaning appropriate to the particular socio-political context of
the Mesopotamian city-state in the late third millennium b.c. While
some details of facial physiognomy may have served as signature ele-
ments of the individual, the statue as a whole stood (or sat) as the literal
embodiment of the ideal able and righteous ruler.
Acknowledgments
This article could not have been written without the knowledge and insight of Hermann
Behrens, with whom I initially read the Statue B inscription, and who first drew my
attention to the “strong arm” of Gudea, and of Piotr Steinkeller, with whom I first read
the text of Cylinder A. Both are absolved, however, from responsibility for any lapses
in my Sumerian. I am grateful to each of them, as I am to Donald Hansen, Robert
Hunt, and Susan Taylor for expanding my vision of the scope of this problem.
28
Hence the curses at the end of this and other texts on statues and stelae, that
attempt to insure the continued maintenance and servicing of the images.
Figure 1. Standing Gudea, known as Statue E, Louvre AO 6, Ht. 1.42 m

[ Photo courtesy of Département des antiquités orientales, Musée du Louvre].
Figure 2. Seated Gudea, known as Statue B, Louvre AO 2, Ht. 0.93 m [ Photo

courtesy of Département des antiquités orientales, Musée du Louvre].
Figure 3. Head of Gudea, Louvre AO 13, Ht. 0.23 m [ Photo courtesy of

Département des antiquités orientales, Musée du Louvre].
CHAPTER TWENTY-THREE
‘IDOLS OF THE KING’: ROYAL IMAGES AS RECIPIENTS

OF RITUAL ACTION IN ANCIENT MESOPOTAMIA
Abstract
Far from being mere reminders or representations, sculptural images of rulers in ancient
Mesopotamia were subject to complex rituals of consecration, installation, and maintenance,
as a result of which they were deemed to be living manifestations, empowered to act and speak
on a ruler’s behalf. These rituals may be seen as “rituals of constitution” insofar as through
their performance the objects were brought into and maintained in being. Both sculptures
and related texts reflect continuous practice in Sumerian tradition from the middle of the
third millennium B.C.E. to well into the second millennium. Later Assyrian and Babylonian
documents suggest that many aspects continued to the middle of the first millennium B.C.E.
Concentrating here on the period, ca. 2100–2000 B.C.E., it can be demonstrated that con-
secrated images were intended to be placed in temple sanctuaries or in funerary chapels. It
can also be shown that in the course of this period, as rulers were elevated to divine status,
royal images were introduced into worship in temples dedicated to the living ruler. This fact
suggests that ritual and cultic practice, however traditional in format and however apparently
unchanging, in fact co-varies with socio-political change.
Nowhere is the power of sacred images more evident than in the

energies required to argue against them in biblical texts: from the
prohibition against the making and worship of images in the Ten
Commandments (Exodus 20:4–5) to the passionate denunciations of
Isaiah (30:22; 44:9–20; 46:6) and Exekiel (23:30), reiterated in Psalm
115, I Corinthians 8:4, and throughout the exegetical literature (see
Jacobsen 1987b; Camille 1989). Yet, for cultural traditions in which
image-oriented cults represented the primary focus of religious wor-
ship, the images were not merely the “work of human hands,” having
“mouths, but speaking not, having eyes, but seeing not” (Psalm 115:4–5);
nor were they reducible to the matter of which they were made (as in
Isaiah 44:15). Rather, through a process of ritual transformation the
* This article originally appeared as “ ‘Idols of the King’: Royal Images as Recipi-
ents of Ritual Action in Ancient Mesopotamia,” Journal of Ritual Studies, Vol. 6. No. 1
(1992), pp. 13–42.
168 chapter twenty-three
material form was animated, the representation not standing for but
actually manifesting the presence of the subject represented. The image
was then indeed empowered to speak, or to see, or to act, through
various culturally-subscribed channels.
This is certainly the case with both divine and royal images in
ancient Mesopotamia, insofar as we can reconstruct cultic practice
from archaeological and textual remains. For this reason, having used
the word “idol” in the play of the title of this study, I shall not refer
to it again. It is important to keep our distance from the essentially
pejorative term, which suggests mis-/displacement of investment in the
re-presentation in lieu of an absent referent. Instead, I propose to focus
upon images as manifestations, pursuing, on the one hand, the series of
ritual actions addressed to the empowerment and maintenance of such
images, and, on the other, the consequences of their empowerment.
While one naturally thinks of this process with respect to images of
divine figures, I shall concentrate rather on the class of royal images
for which we have considerably more evidence in Mesopotamia, and
which in fact provide significant information about the sacred domain
through inscriptions that attest to their role therein.1
Our concern with the meaning and role of royal images in ritual
practice begins with a reading of the Sumerian texts inscribed directly
on statues of the Mesopotamian ruler, Gudea of Lagash, ca. 2100
b.c.e., (fig. 1) and his son and successor, Ur-Ningirsu (figs. 3 and 4).
These texts tell us that each statue was dedicated to a particular deity
in the Mesopotamian pantheon. They also state that the statue was
intended to be placed in a temple or shrine, provided with messages
to be communicated to the god through direct discourse. Each was to
be the recipient of regular offerings.2
1
While we do have some information about divine images from religious and eco-
nomic texts, especially in the later periods, and will draw on some of that data below,
it is the royal images that are both attested in texts and found in the archaeological
record. The reasons for this are many. Foremost, it would appear, is the fact that divine
images were often made of wood, which would not be preserved, or of precious met-
als, which would have been subject to capture and re-use. It is certainly the case, as
noted by Oppenheim almost 30 years ago (1964:183), that “the role and function of
the divine image in [Mesopotamian] civilization have never been considered important
enough to merit a systematic scholarly investigation.” It is my hope eventually to address
this problem. The present inquiry into better-attested royal images should provide a
foundation for such an undertaking.
2
For the most complete text, that on Statue B (fig. 1), see Steible (1991:156–179).
idols of the king 169
Documents pertaining to both earlier and later royal images suggest

similar purposes. With recent evidence that the reigns of Gudea and
Ur-Ningirsu overlapped the beginning of the important Third Dynasty
and the neighboring city-state of Ur (Steinkeller 1988), it becomes all
the more interesting to incorporate data related to the Ur III rulers
into our picture.
By including this paper in a collection of studies of “Art in Ritual
Context,” I assume membership of these royal statues in a category
called “art,” although there is no word for art in either Sumerian or
Akkadian, the languages of ancient Mesopotamia, nor is there a term
specifically designating sculpture. Rather, the Sumerian term alam3
(Akkadian almu)4 seems to mean simply “image,” and while used
directly to refer to these works in the inscriptions themselves (as in
Statue B, Col. I:3, called alam Gudea, “image of Gudea”), it is also
used to refer to any representation, whether in relief, in the round, or
painted.5 Nevertheless, that there is a category for “image”, and that
we can attest to targeted goals in representation (e.g., “a statue whose
appearance represented my own features”), suggests that at least some
of the criteria for “art”—a class of affective objects to which are applied
standards of correctness—were operative in Mesopotamia and justify
our use of the term as an analytic category.
These affective images of rulers, then, were animated by a series of
ritual acts and incantations and served by priestly functionaries. Thus
they existed throughout their viable lifetime in what I would argue to
be a ritual context. Recent persuasive conceptions of ritual as opera-
tive in secular as well as sacred domains notwithstanding (e.g., Moore
and Myerhoff 1977:3–4), our understanding of the nature and role of
ritual in culture has proceeded essentially from studies of the nature
and role of religion. Debates of the past twenty-five years in anthropol-
ogy over a definition of religion as a system of beliefs (Geertz 1966:4)
vs. a system of institutions and behaviors (Spiro 1966:96) are largely
unproductive for our purposes. Once one accepts the necessary role in
religion of some non-or supra-human agency to which or whom sanctity
3
It is conventional to set Sumerian words in normal type.
4
Akkadian terms are conventionally underlined or italicized.
5
Here I differ from the Chicago Assyrian Dictionary, which “translates” almu (= alam)
as “statue in the round, relief, drawing, figurine, likeness, constellation”. Rather, I would
say the term is used in reference to all of these media without distinguishing one from
the other, because in fact it means “representation” or “image.”
is imputed (see Goody 1961; Spiro 1966; Rappoport 1971), then what
emerges is a conjunction of beliefs and behaviors in which ritual is the
major mode of expression, permitting communication between such
agencies and the human domain.
The nature of ritual has itself been the subject of exhaustive debate
(recently summarized in Kelly and Kaplan 1990). Neither space nor
the goals of the present paper permit a lengthy exposition of conflict-
ing (functionalist, Marxist, cognitive) positions here. Most useful for
our purposes are the views of Rappaport (1971:62–3), who sees ritual
as a set of conventional actions performed at regular intervals fixed
by clock, calendar, or circumstance, having affective, emotional value,
and through which is communicated information concerning physical,
psychological, or social states. Another useful definition is that of Kelly
(1988:40,41,52), who argues against dividing components of ritual into
the real vs. the symbolic, and who, then, sees ritual as “conventional
form set within systems of meaning.” In his view ritual includes both
practice and discourse, both behavior and belief.
To the extent that “religious belief involves a prior acceptance of
authority” (Geertz 1966:25), and that ritual practice engages power
dynamics in the exercise of, or acquiescence to, that authority or its
representative (cf. Holdredge 1990; Bourdieu 1991:126), religious ritual
becomes an arena in which political leaders not only may but must
function, if they are to establish and maintain their own authority. In
the play of Rappaport’s conundrum (“authority being contingent upon
sanctity . . . [and ] . . . sanctity being contingent upon authority” [1971:72])
is summarized a pattern of political engagement characteristic of ritual
action that has been observed by many historians and anthropologists.
Much of the literature has concentrated upon explicitly royal rituals,
such as coronations, court ceremonials, and funerary practices (cf. Gie-
sey 1960; Fortes 1968; Kuper 1972; Cannadine & Price 1987; Kertzer
1988; Bloch 1989). In the Mesopotamian case we are dealing not with
rituals engaged in by the physical person of the ruler—although these
are certainly attested—but with rituals addressed to images of rulers,
images that were introduced as actors into the sacred space of the god
and received worship themselves.
The sacred character of the royal image in Mesopotamia is not with-
out its parallels elsewhere. Marin’s important study (1988) of Louis XIV,
for example, emphasizes the role of the royal portrait as a sacerdotal
object, at the same time as it represents a historical personage (Louis)
and an exemplar of a class or institution, namely, the king. However,
the particular combination of ritual action, sacrality and imagery in

Mesopotamia is distinctive. The historical circumstances of Mesopo-
tamia in the late third millennium b.c.e. introduces constraints and
conditions that cannot be excluded from an account of ritual practices
related to royal images.
This brings us to the issue of context and the argument of Kelly
(1988:40) and Kelly and Kaplan (1990:139) that ritual systems must
be seen as historical phenomena. Insofar as ritual has been understood
as the repetitive performance of highly conventionalized form, studies
of ritual have tended to assume conservatism and hence continuity in
both form and meaning over time. However, as Kelly notes, rituals are
also performed at concrete moments, in specific places, by agents with
historical identities. He not only contends that ritual form can change
(often in the face of public assertions of unchanging tradition and
denials of historical agency); he also argues that ritual as performed
in real time and real space is itself history-making to the extent that
it serves to fix social relations. Indeed, to have any understanding of
ritual it all, it is necessary to consider “the historical construction and
reconstruction of ritual systems, and also the crucial role of ritual in
praxis” (1988:54).
I believe the Mesopotamian case has something to contribute to this
discussion. First, in a long diachronic sequence it provides evidence for
a tradition of consecration of Sumerian royal statues over a period of
at least 500 years. Second, it documents some major shifts that empha-
size the possibilities of creative change in ritual practice. And third, it
provides evidence for concomitant shifts in political organization, which
permits us to argue for both a reflective and an affective role for ritual.
In addition, the specific series of ritual acts that served to animate and
maintain royal (and also divine) images in Mesopotamia permit us to
distinguish a subset of ritual activities that I would call “rites of con-
stitution,” that is, rituals without the performance of which the object
would not be what it has been designed to be.
The statues of Gudea of Lagash remain the most complete docu-
mentation we have of the ritual processes associated with royal imag-
ery; however, they stand neither at the beginning nor the end of the
sequence. Votive statues with brief labelling texts identifying rulers,
as well as texts describing the making and care of statues, are known
from at least the Early Dynastic Period, ca. 2500 b.c.e. These include
examples from Gudea’s own state of Lagash (cf. Spycket 1981: figure
54; Steible 1982: Urnanše 24). A few inscribed statues and textual
references to others now lost are also preserved from later dynasties
of Ur, Isin, and Larsa that bracket the turn of the third to the second
millennium (see for example, Civil and Zettler 1989; Steible 1991: Šulgi
25,26,27; Ibbisu’en 3). Nevertheless, it is the Gudea material that rep-
resents the point of departure, because the corpus of inscriptional as
well as sculptural finds together enable us to reconstruct the meaning
and function of these images.
The city-state over which Gudea ruled was one of the most important
of early Mesopotamia. Its territory was considerable, including several
satellite cities and towns, of which the site of Tello (ancient Girsu), where
the statues were found, was one of the largest (Falkenstein 1971:676).
Girsu was the center of the cult of the god Ningirsu, tutelary deity of the
state, and temples to that god and other related deities were prominent
in the town, many of them built or reconstructed by Gudea himself.
Throughout his reign, which can be documented for at least 11 years
(though it may have been longer),6 Gudea consistently took the title
ensi (“governor” or “steward” in Sumerian) rather than lugal (“king”).
This practice does not suggest any diminished status, however. It can
be shown that different city-states in ancient Mesopotamia had differ-
ent traditions in nomenclature, with a consistent preference throughout
the historical sequence in Lagash for the term that emphasized the
ruler’s stewardship vis-à-vis the tutelary deity, who in title at least was
the primary ruler of the state. Thus, Gudea as ensi was at the top of
his state’s administrative hierarchy, entirely comparable to rulers called
lugal in other states, such as Ur. Indeed, in at least one inscription
of Gudea’s reign (Falkenstein 1971:677), he is actually referred to as
“king.” I mention this practice because only those individuals (or their
wives) who did stand at the top of their respective hierarchies seem to
have inscribed their statues. It is important for our purposes to establish
equality between rulers despite any differences in their public titles. For
this reason, I prefer to call all such individuals “rulers” without specify-
ing the titles they themselves assume.
6
The dates of Gudea have recently been opened to discussion. Falkenstein (1971),
Steible (personal communication), and others think it likely he reigned as long as 20
years. It now appears that Gudea’s rule in Lagash had to overlap the Third Dynasty of
Ur by at least 10 years, perhaps more (cf. Steinkeller 1988). This conclusion will have
important consequences for our understanding of the special relationship of the kings
of Ur to Lagash and the funerary cult of Gudea. If indeed Gudea did rule some 20
years, then tentatively ca. 2120–2100 now seem the most likely years of his reign.
Unfortunately, virtually all of the statues of Gudea uncovered in

legitimate excavation at Tello were found in a secondary context. They
were gathered in antiquity to stand in the courtyard of the palace of
a ruler of the second century b.c.e. (cf. fig. 2; see also the discussion
in Parrot 1948:152–6).7 Nevertheless, each inscribed figure names the
deity to whom the image was dedicated, and in several cases we are
told that, and where, the statue was intended to be placed in the deity’s
temple.
The inscribed statues also include in their dedications a description
of the bringing of the highly prized diorite stone (na4 esi) from a land
to the southeast, the creation of the image, the name given to the
image (in which is contained some positive aspect of Gudea’ to be kept
in mind before the god), and a record of the statue being brought in
to the temple. In the most complete text, that of the seated Statue B
(see fig. 1), we are told specifically that the image is to be set up in a
shrine associated with libations (the ki-a-nag) within the temple to the
god Ningirsu. We are given a list of foodstuffs, regular offerings (sa2-
du11) from the temple stores with which the image is to be provided in
perpetuity (cf. Steible 1991: SB.7.10–20, 55; 1.8–12). In the text, as the
image is brought into the temple, Gudea the ensi gives the command
(literally, “puts the word”) to Gudea the alam: “Image, to my lord [the
god Ningirsu], speak!” (7.21–25). He then proceeds to enumerate his
pious deeds so they may be communicated to the god. At the end of
the recitation we are told that he (Gudea) installed the image (in order)
that it recite his words/perform his command. (7.4 48).
Implicit in this passage is the understanding that in the course of
its making the statue has undergone a process of animation in order
to be able to engage in direct discourse with the deity. Here we may
begin to investigate the ritual actions associated with these and related
statues. Circumstantial and direct evidence from the statuary inscrip-
tions, plus evidence available from other texts, suggests three points at
which the images are the recipients of ritual attention: (1) rituals of
making or consecration; (2) rituals associated with passage of the image
7
Some nine or ten statues or parts thereof were recovered from the palace courtyard.
An additional ten figures are known. Several, along with the figure of Ur-Ningirsu (figs.
3 and 4), seem to have been the fruits of clandestine digging at Tello in 1924. The
authenticity of some others has been questioned ( Johansen 1978:29 et passim). Since
these latter do not substantially change or add to the points I shall make below, I will
confine my discussion to works whose origins have not been questioned.
into the temple and installation in the shrine; and (3) placement and
ritual maintenance of the installed image.
Rituals of Making or Consecration
A primary clue to the animated nature of these royal images is con-

tained in the texts of the Gudea and Ur-Ningirsu statues themselves
(e.g., figs. 1, 3, 4 & 7). Whereas cult paraphernalia, stone stelae, and
even protective stone gateway lions are described with the standard
Sumerian verb for making [dim2], a different verb [tud ] is used to
refer to the making of the statues.8 This distinction is not unique to
the Gudea dynasty; it can be demonstrated both earlier (Steible 1982
Urnanše 24) and later (Steible 1991: Šulgi 26.13; Lugirizal 1.rs.1); in
Sumerian texts, continuing into the first millennium b.c.e. with the
Akkadian equivalent walādu (cf. B.N. Porter n.d.:5), and is used only
when stone images of human or divine figures are involved. The verbs
used designate birth. Indeed, on one of the Gudea statues dedicated
to the goddess Ninhursag, who is associated with birth and fertility, the
goddess is referred to by epithet as “Divine lady of birth [dingir nin.
tud ], mother of the gods” (Steible 1991: SA 3.5–6). And when Gudea
himself is provided with the epithet: “Legitimate son [dumu tu-da] of
the god Gatumdu”, it is expressed through the verbal form that indicates
physical birthing (SB 2.16–17). When Gudea says of Statue B that he
brought precious diorite from the land of Magan and created of it his
image [alam-ni-še3 mu-tud], the engendered and animate nature of
the image is evident.
Far more explicit information about the ritual process of animation
undergone by images upon their physical completion is to be found
in bilingual texts in Sumerian and Akkadian of later Babylonian and
Assyrian origin (see Smith 1925; Ebeling 1931:esp. 100ff.).9 Although
the Sumerian version is concerned with the making of divine rather
than royal images, the statue [alam] is described not as having been
8
See, for example, Steible (1991: Gudea SB.6.9–10 and 17–18); compare with
7.10–13; Ur-Ningirsu II.6.
9
These texts were restudied in the Ph.D. dissertation of Christopher Walker and are
being prepared by him for publication. I am grateful to have seen copies of the texts
in the files of the Philadelphia Sumerian Dictionary, and to have had an occasion to
read through the texts with Markham Geller of the University of London in 1988.
“made” on earth, but as having been “born” in heaven, using a form

of the verb tud (Tablet 4, line 23a in the Assyrian version; Smith 1925:
obverse lines 3–4, for the Babylonian). At the same time, it is made
clear that the mere physical form of the image is not sufficient. One
of the incantations to be recited in the course of the ritual is called
“Sacred/pure image [alam.kù] perfected by a great ritual” (Smith 1925:
Rev.48; emphasis mine), which emphasizes the role of the ritual process
for completing the image.
Starting out in the workshop where the image was made and then
proceeding outdoors to the riverbank, the image was subject to a series
of ritual actions over the course of a night. In a penetrating analysis
Peg Boden has suggested that the action recapitulates a birth the next
morning.10 The ritual actions are accompanied by incantations of which
only the first line is given. Presumably, the officiating priests would have
known the complete texts of each. In the Assyrian bilingual version the
rituals serve to “open the mouth” of the image (Sum. ka.du - a; Akk.
pīt pî; cf. Tablet 4, ll.14–15), thereby rendering it vital.
While these texts are known only in late, first millennium b.c.e.
copies, they clearly preserve kernels of much older material (Civil
1967:211). A text of the last ruler of the Ur III dynasty, Ibbi-Su’en, from
Ur, attests to this ritual in the third millennium in the king’s account
of his dedication of a gold vessel or basin to be used in the bathing
festival of the moon-god Nanna in the “place of the mouth-opening
ritual” (Steible 1991:Ibbi-Su’en A9–10.17–27).11 Furthermore, buried
in economic texts of the Ur III period (ca. 2050 b.c.e.) from Tello are
references to provisions for performances of the “mouth-opening” ritual
to statues of the deceased and now-deified Gudea. These references
both attest to the ritual and also suggest that it was repeated after the
initial consecration, presumably as renewal.12
10
I am most grateful to Ms. Boden, a graduate student at Johns Hopkins University,
for having shown me a copy of her paper, which draws upon Arnold van Gennep’s
Rites of Passage to suggest the metaphorical birthing of the image through ritual.
11
Here I am grateful to Piotr Steinkeller, Harvard University, and Åke Sjöberg, Uni-
versity of Pennsylvania, for their reading of line 25 as ki dub-sen-e ka!-du - u- a.
12
Noted by Civil (1967). Further references (generously provided by Piotr Steinkeller)
in Ur III texts from Tello, record that the offerings provided for the ka-du - a of the
Gudea images occur in each case in the 3rd month, suggesting that it was performed
regularly and annually (see, for example, de Genouillac 1910:6927, dated to the second
year of Šu-Su’en).
In the Babylonian version of the mouth-opening ritual (Smith 1925:

line 53) the statue’s eye—always prominent in Sumerian sculpture (see
fig. 3)—is also ritually opened. Once the mouth has been opened, the
image may be offered food to eat (Tablet 4,1.19). Once the eye is opened,
the incantations that follow, “In your going forth . . .” and “Image born
in heaven . . .” (Smith 1925:ll.53–54), suggest that the statue is now fully
animate. There follows a plea directed to the image itself not to stay
in heaven, but to “Enter the form. . . .” (Smith 1925:ll.56–7).13 This I
understand to be the final invocation of the essence of the deity into
the image. The late bilingual text actually states that “the image without
the mouth-opening ritual [ka-nu-du .u3.da] cannot smell the [offered]
incense” (CAD , 1962:78). This last reference makes absolutely clear
the process of enlivening effected by the mouth-opening ritual and the
animate presence henceforth in the image. Considering the verb used
for their creation and their subsequent animation, a similar presence
must have been called into the images of rulers as well, and this mani-
fest essence was most likely reflected in inscriptions that make a point
of describing such images as endowed with the ruler’s own features
(cf. Frayne 1990:67, vi.6’–10’).14 Finally, one quite explicit reference
comes from a text of about 100 years after Gudea, in which a ruler of
Larsa states concerning a votive image, “Having entered the temple,
may it [the alam] be a living thing in the temple!” (Frayne 1990:195.
v.3’–4’), thus making clear the animate nature of the non-divine figure
as well.
Rituals of Installation
Upon animation, the image can be both fed and adorned. At this point
it is ready to be brought to the temple, where special offerings and ritual
acts accompany its entrance into the sanctuary. The late Assyrian and
Babylonian bilingual texts discussed above deal with the entrance of
13
For the last portion I am indebted to the reading of Peg Boden.
14
While we cannot suggest anything like true portraiture for Mesopotamian statues,
it has long been pointed out that the chin and broad face of Gudea (cf., for example,
Winter 1989:figure 2) and the resemblance of his son, Ur-Ningursu (fig. 3) seem too
consistent to be merely idealized form and must contain some reference to actual
features. Perhaps the best way to understand this is as “signature elements” that sig-
nal salient aspects of the individual and thus are tied to his or her person, while still
retaining an overall idealized general form of face or body.
images of divinities into the temples; however, since we know that the
royal statues were also placed in the temple (cf. discussion below), we
may infer that they, too, underwent some special ceremony to mark
this transition into sacred space.
In the Babylonian version of the mouth-opening ritual, once the
deity has been invoked into the image, it is brought in procession to the
temple, accompanied by incantations related to the process of transit
(Smith 1925:1.59). At the gate of the temple offerings are made, and
the god is led into the shrine (Smith 1925:ll.60–61). The installation
is complete only when the deity (the image) is placed upon a throne
covered by a canopy, a repetition of the ritual of the washing of the
mouth has been performed, and the god has been fed in the new abode
(Smith 1925:61–65). From this point on, the god is treated as a living
person, fed, adorned, and cared for with rites and actions appropriate
to daily and festival worship (cf. Oppenheim 1964:184). This ritual of
installation, including the transition from outside to inside the shrine
seems to illustrate Turner’s (1969:94ff.) concept of liminality in the cross-
ing of the threshold. Unfortunately, however, of the rituals associated
with images, both royal and divine, it is the one least well-documented
in the literature.
Nevertheless, it is useful to signal this important juncture in the hope
that further evidence may become available. The phrase used in our
late Babylonian text, “taking the hand of the god” (Smith 1925:60),
is one known from other texts to indicate procession of the divine
image (cf. Black 1981:56). It seems appropriate to suggest as a working
hypothesis that royal statues introduced into the temple would have
been subject to rituals of installation, just as the divine images were.
There is evidence that some kings of the period actually underwent
highly ritualized coronations and consecration in the temple when they
assumed office (cf. Klein [1991] for Šulgi of the Third dynasty of Ur).
We also know that on occasions such as the installation of a new chief
priest or priestess (en) in the cult of a deity, sacrifices were made at the
temple gate to mark the initial entrance of the new functionary (Arn-
aud 1986). If we put this fact together with the suggestion of Renger
(1967:120) that the occasion of the installation of a new en-priestess
may also have been the occasion of the setting up of her statue in the
temple, we may leave open a similar possibility for the installation of at
least some royal images as well. In any event, that the entrance of the
statue into the temple was part of the formulaic account on all of the
inscribed Gudea images and is emphatically referred to in many texts
describing the making of royal statues15 does suggest that the installation
was both distinct from and the culmination of the process of making
such images. And since neither the deity as image nor the living ruler
was fully installed until enthroned and attended by ritual activities, we
may also argue for a similar ritual process upon the occasion of the
installation of the royal image.
Placement and Ritual Maintenance of the

Installed Image
The most eloquent reference to a non-divine image in the temple is

the one cited above in a text of Sin-iqišam of Larsa, who ruled in the
early 2nd millennium. He declares that the votive image, once brought
in, should be a living thing in the temple (Frayne 1990:195.v.3’–4’).
Other sources suggest that there were at least three places where royal
statues would have been installed in early Mesopotamia: (1) the god’s
shrine, (2) the ki-a-nag, literally, the “place of libation”, or funerary
chapel, to the deceased ruler, and (3) during the Ur III period, in a
shrine to the living ruler.
Placement of worshippers’ statues in shrines is attested in Mesopota-
mia from the Early Dynastic period (see Winter 1991:64). The multiple
images of Gudea with their various dedications suggests that the ruler
would have been represented not just in the shrine of the dominant
deity but in each important shrine. Later, when Ur had hegemony over
Gudea’s state of Lagash, a text found at Tello records offerings to three
images of the king Šu-Su’en: one each in the temples of the principle
Lagash deities, Ningirsu and Baba, and one for a temple dedicated to
Šu-Su’en himself (de Genouillac 1910:793). If this pattern were repeated
in every major city-state over which Ur had hegemony, as well as within
the capital city of Ur, then there must have been a huge number of
such images for each reign.16 That we have relatively few preserved may
15
The lines stating that the image was “created” are always followed by the giving
of a name to the image and then the final statement: “[and] into the temple [e2-a]
he [Gudea as agent] brought [literally, entered] it [mu-na-ni-ku4].” Compare Steible
(1991: Gudea SA 4.4). Frayne cites similar formulae for statues of later kings of Larsa
(e.g., 1990: Sin-iddinam 12.13–15).
16
For example, for one of the early Ur III rulers, Šulgi, at least three inscribed stone
statues are known, two from Ur, the third of uncertain provenance (cf. Steible 1991:
Šulgi 25, 26; Civil and Zettler 1989:27).
well be because many were made of precious metals and would have
been subject to later appropriation and melting for re-use.17
Steible (1988) has pointed out that the ten diorite statues of Gudea
found gathered in the later palace in Tello seem to go in pairs: one
seated and one standing image per deity. Only the chief god Ningirsu
has four: two standing and two seated. The standing images, with only
one exception (Statue A), have their inscriptions contained on the back
(Statues C, E, G and K), as does the statue of Gudea’s son, Ur-Ningirsu
(fig. 4). This pattern is quite remarkable when contrasted with known
seated images, whose inscriptions tend to be either on the front of the
skirt or circle the statue front and back (see Statues B [fig. 1], D [fig.
7], F, H, and I). When this observation is put together with evidence
that the seated position indicates dominant status in Mesopotamian
representations (cf. Winter 1987:71), and that rulers in the presence
of the deity are shown standing before the seated god,18 then we may
suggest that it is most likely the standing Gudea images that were placed
before the divine image in the god’s shrine.
Textual evidence supports such a reconstruction. In a 1st millen-
nium text the Assyrian ruler Assurnasirpal II tells us that he created
his “royal image with a likeness of his own countenance” and placed
it before the god Ninurta (Grayson 1976:679). Even more apt are the
lines in a hymn composed for Šulgi of Ur, in which we are told that
the king’s image, “made brilliant like the heavenly stars,” was set up
“before the good eyes” of the god Enlil at Nippur (Klein 1991:308).
While not all standing images may have been placed in the cella, this
fact would certainly account for those images of Gudea and his son on
which inscriptions are confined to the back, as presumably the royal
image would be situated in a privileged place facing the deity (as in
fig. 3) and thus be seen and identified from behind by anyone else in
the space (see fig. 4).19
17
Texts of the reign of Šu-Su’en of Ur tell us that the ruler had his statue made
of the gold he brought home as booty from a victorious military campaign (Kutscher
1989:73). A slightly later ruler from Larsa recorded 14 bronze statues brought into
Nippur in a given year, along with three gold thrones and gold statues of two deities
(Frayne 1990:192f.). Many year-names of rulers similarly record events including the
presentation of gold and/or bronze images, none of which has ever been found.
18
As on Gudea’s own cylinder seal and on a detail of the stele of Ur III ruler,
Ur-Nammu (fig. 5).
19
Texts describing some royal images of the Isin-Larsa period (early 2nd millennium)
clearly state that they were intended for the temple courtyard (cf. Frayne 1990:67,86).
At least one of these explicitly describes the king holding a votive goat for sacrifice
The function of royal images placed before a god is apparent from

the messages embedded in the name given to each image, for instance,
“. . . Ningirsu has decreed a long life for Gudea . . .” (Statue D, fig. 7; see
Steible 1991:SD.5.2–7), and in the command to speak directly to the
deity explicit on Statue B. From a text dated to the reign of a late ruler
in Gudea’s dynasty, we learn that the deity is entreated to turn an ear
to the ruler, that is, respond to him with favor (Steible 1991:Namma ni
1; cf. Winter 1991:66). If a statue of Šulgi holding a libation vessel
(Civil and Zettler 1989) is included in the repertoire, then there may
also have been instances in which the ruler’s image was intended to be
performing in perpetuity some of the cultic functions carried out by
the living king for the deity (Klein 1991:296; cf. our fig. 5).
By contrast the text of the seated Gudea Statue B tells us explicitly
that the image is to be set in a place designated the ki-a-nag (Steible
1991:SB.7.55). The Sumerian refers literally to a place (ki) of the pour-
ing (nag) of water (a), and when mentioned in conjunction with royal
names, seems to be associated with libations in memory of deceased
individuals. Although not fully understood, the ki-a-nag seems to be
the locus of funerary activities, in which liquid offerings were made to
the deceased (cf. Michalowski 1977:221). At the same time, at the end
of Statue B’s inscription it is stated that the image was in the Eninnu,
which is known to be the principal shrine of the god Ningirsu in Girsu
(Steible 1991:SB.7.60–1). Therefore, at least in this case, the place of
the ruler’s ki-a-nag was in the temple.
Such funerary “chapels” are attested from the Early Dynastic period
(see Behrens 1987; Bauer 1969) and continue through the Ur III period,
during which we have mention of the ki-a-nags of deceased Ur III
kings by their successors (Michalowski 1977). It makes good sense that
(67, vi.8’, comparable to a known statue of Šulgi of Ur holding a kid; Barrelet 1974:
figure 3a). Placement in the temple courtyard for such images makes perfect sense,
since we know animal sacrifices were conducted outside the cella (Arnaud 1986). Within
the cella, or indeed in the courtyard as well, one may imagine that it could get rather
crowded with images of successive rulers, and there must have been some protocol
for ceding pride of place to the figure of the reigning ruler. A glimpse of intention to
the contrary is seen in some of the curse-formulae related to statues that claim a place
in perpetuity. An example is recorded at the end of the 3rd millennium by a ruler
of Larsa for the cult center of Nippur. He instructs any future king not to displace a
statue from its dedicatory location, nor to place it in a storehouse, and certainly not
to stand his own statue in that place (Frayne 1990: Sin-iqišam 1). Similarly, a text of
Amar-Su’en of the Third Dynasty of Ur curses any man who changes the place of his
(Amar-Su’en’s) image, or tears out its socle (Steible 1991: Amarsu’en 3).
such a chapel would house an image of the deceased as a focus for

commemorative cultic activity (cf. Spycket 1968:59ff ). In this regard,
it is important to keep in mind that the Gudea Statue B destined for
the ki-a-nag is a seated figure (fig. 1). It is therefore possible to suggest
that, just as the standing Gudeas were those most likely to have been
placed in attendance upon divine images, the seated figures, while
equally empowered to speak directly to the gods, would themselves
have been the objects of cultic attention.
The purpose of such a cult is clear when we read the aspiration
expressed in the inscriptions of Gudea Statues I and P, namely, that his
name not be forgotten (literally, “that his name be called”) in conjunc-
tion with references to the libations poured in the ki-a-nag, which are
said to open up channels to the underworld, as “Water of the Name-
calling” (cf. Steible 1991:64, n. 15).20 In other words, these are rituals
of commemoration.
What is not clear is for how long such libations were poured—in
perpetuity, for the duration of a particular dynasty, or (in the case of
Gudea’s dynasty commemorated by the kings of Ur III) for a recent
but valued past. In all or any of these cases, it is evident that continuing
ceremonies to deceased rulers in the ki-a-nag would have functioned
on at least two levels. While obviously serving to perpetuate the name
of particular individual, they also, by evoking previous rulers, celebrate
a dynasty and/or the very institution of rule. Put this way, it is clear
that succeeding rulers would have been heavily invested in these com-
memorative practices.
From the middle of the reign of Šulgi on, rulers of the Third Dynasty
of Ur were deified. The epithet, “god of the land,” was added to their
royal titulary (Wilcke 1974:179), and they received worship during their
lifetime, not just after death. Textual evidence from Tello (Steible 1991:
Šu-Su’en 13) and an inscribed building excavated at Ešnunna (Frankfort
et al. 1940; fig. 6) make clear that temples to the ruler Šu-Su’en existed
20
The desire or need to be remembered is even more explicitly stated in a text of
Hammurabi, ruler of Babylon in the early 2nd millennium, which describes a part of
his building program. The text is preserved in two copies and ends with a statement
of the consequences of his building: “My good name, [which] like [that of ] a god [is]
to be said daily [and] which for all times should never be forgotten in the mouth of
the people, truly I established [it]” (cf. Frayne 1990:336, lines 75–81.)
in cities outside of Ur, and it is most likely that shrines to the rulers
existed inside the capital as well.21
The similarity of the main shrine plan of Šu-Su’en’s temple at Ešnunna
(fig. 6) to the incomplete remains of the facade of a Gudea temple
uncovered at Tello within the foundations of the late palace in which
so many of the Gudea statues were uncovered (fig. 2) makes it rather
tempting to wonder whether Gudea, too, might not have been enshrined
during his lifetime. We know that the relatively small seated Gudea B
(.93 cm. without head) was destined for the funerary chapel. The colossal
seated Gudea D (1M.58 without head; fig. 7), given its scale, is the most
likely candidate for occupying a dedicated shrine of its own. It would
then have been placed in a niche in the cella not unlike that preserved
in the Šu-Su’en temple. Against such a reconstruction is the argument
(as in Hallo 1990:63) that royal images would only have been in wor-
ship during their lifetime in a period when rulers were deified. It must
be remembered, however, that at least two Akkadian rulers preceding
Gudea had been deified (ibid.: 58), and it is not impossible that practices
related to royal statues during that time were maintained by Gudea,
whose own practices would then have prefigured those of the kings of
Ur to come, as we know was true in other domain (Klein 1989).
In any case, during Ur III and in the subsequent Isin and Larsa
periods we have extensive records not only of ritual libations but
food offerings to royal images in both temples and funerary chapels
(Bauer 1969; Kutscher 1974; Wilcke 1974:n.57; Lafont and Yildiz
1989:791,875,973; Civil and Zettler 1989; Frayne 1990:192.iv.1–8,
196.v.12’–16’;16’). Indeed, the Ur III rulers continue to allocate offer-
ings to images or the ki-a-nag of Gudea and his dynasty in Girsu
(Tello) after the Ur dynasty takes over (Perlov 1980; Lafont and Yildiz
1989:791,886,985,1021–2; Nies 1920:39).
These offerings, consisting variously of breads, beer, meat, fish, milk,
cheese, clarified butter, honey, and dates, replicated offerings made to
the gods (Steible 1982: Urnanše 24, 26). Some rations were provided
daily, others at the new and full moon and at particular festivals
(Oppenheim 1964:187; Lafont & Yildiz 1989:875,973). These were
often accompanied by gifts of clothing, jewels, and other objects (Bauer
1969; Civil and Zettler 1989). Again, the text of Gudea Statue B is
21
Wilcke (1974:179–80 and notes 51 and 57) points out the possibility of temples
to Šulgi in Girsu, Umma and Ur, and cites offerings to four statues of the founder of
the Ur III dynasty, Ur-Nammu.
most explicit. The very first column, separated from the rest of the text,
describes the beer, bread, and grains that are to be provided the image
(alam-Gudea) in the ki-a-nag as regular offerings (sa2-du11) in perpetuity
from the stores of the Ningirsu temple (Steible 1991:SB.1.1–20). This
practice, as Steible has noted (1988), echoes Early Dynastic practice
as preserved on a statue inscription of an earlier ruler of Lagash (cf.
Steible 1982: Entemena 1).
Statues of rulers set in gods’ shrines also received food offerings,
according to Early Dynastic period texts (Bauer 1969:111) and later,
Neo-Babylonian texts of the 6th century b.c.e. (Beaulieu 1989:133ff.).
Since royal images within the gods’ shrines were called upon to be no
less animate than those of the gods whom they faced, obviously they
too required provisions.
That provisioning was not merely a secular enterprise is suggested by
the fact that included in the temple personnel was a specific office, that
of the gudu4-priest, charged with the care and feeding of the image
( J. Westenholz, personal communication; see also Oppenheim 1964:187).
In addition “kitchen” facilities for temples are referred to in texts as
providing provisions for the gods (Frayne 1990:141–2), and actually
have been found adjacent to at least one major shrine (Trumpelmann
1990), along with offering tables within the temple complex (Jacobsen
1989:85). The sacred nature of the offerings is further suggested by
the timing of offerings at important junctures of the lunar calendar
and festivals, including the festival days of the kings represented (cf.
Lafont and Yildiz 1989:875; Genouillac 1910:5271, and cf. endnote 8).
As with the food provided to the image at consecration, then, the daily
and festival provisioning of the image would have been a fundamental
part of ritual activity.
Thus, textual references to ritual activities marking the moments of
consecration, installation, and the subsequent maintenance of the royal
image in Mesopotamia demonstrate that from the moment of creation,
the image was indeed enmeshed in a ritual context that continued
throughout its existence. While details of this have been known for some
time within ancient Near Eastern studies, the entire sequence has not,
to my knowledge, been put together as a framework for understand-
ing Mesopotamian sculpture. This focus on a ritually consecrated and
maintained image may be seen as one of the Mesopotamian contribu-
tions to the broader study of royal ritual in general.
The presence of the royal image as votary in the god’s sanctuary
both speaks to a need for the god’s munificence and also testifies to a
special relationship between the ruler and the divine. It makes manifest
a ruler’s privileged mediating role between the gods and the people, a
status sanctioned in the Mesopotamian cosmology. At the same time, I
would also argue that the royal image is affective, and serves to reinforce
the hierarchical order that privileges the ruler through its very presence
in the shrine. The ruler’s image in the god’s shrine is not only the result
of his special status, it also works to establish the special status. Simi-
larly, the maintenance of a funerary cult of deceased rulers not only
serves to perpetuate the memory of a particular individual but it also
emphasizes the continuity in, and the essential role of, the institution
of rulership by virtue of the focusing of resources and ritual attention
carried out by subsequent rulers.
From ca. 2100 to ca. 2010 b.c.e., within the Ur III period, royal
images were introduced into worship during the lifetime of the ruler.
This shift in ritual practice occurs parallel to the deification of the
ruler in that period. I have argued elsewhere (1987) that elevation to
divine status is a function of the development of the nation-state and
the need of the ruler to rise in status above previously equal rulers
of once autonomous city-states now under the hegemony of a single
center. If so, then it may be proposed that the concomitant increase
of ritual contexts for royal images co-varies with, and indeed serves,
the increase in royal status. To the extent that the ruler, as embodied
in his image, is in worship, he is also inscribed in the world order at a
level above that of any other mortal, giving him and the institution of
rulership authority sanctioned by and answerable to the gods.
With Kuper (1972), Kelly (1988:54), Bell (1990:310) and others,
then, I would argue that evidence for change in ritual processes over
time (here, for example, the introduction of a divine ruler’s image into
direct worship) must be seen in relation to other historical phenomena
(in this case, the capture of power by the ruler elsewhere in the society).
In fact, the Mesopotamian situation during the Third Dynasty of Ur
seems to exemplify the discursive relationship between ritual-in-politics
and politics-in-ritual (Kelly and Kaplan 1990:141).
The linkages existing among politics, power, ritual practice, and reli-
gion have been underscored often (Holdredge 1990; Bell 1990; Abélès
1988; Kertzer 1988; Price 1984). Indeed, for Mesopotamia, Barbara N.
Porter has argued that there are few ways to talk about politics other
than in religious terms.
To the extent that the Ur III king is a god, he has been introduced
into the religious (and political) domain with a new status. As to statues
of Šulgi of Ur holding a libation vessel or sacrificial animal demonstrate,

in the god’s shrine he may still be a performer, but he is also now a
recipient of worship. His divination confers sanctity. As Rappaport
argues (1971:69), sanctity confers “the quality and unquestionable
truthfulness . . . to unverifiable propositions,” and “people are more likely
to accept sanctified messages than unsanctified ones [as] true.” While
there would be little dispute that the expanded context for the ritual
treatment of royal images was reflective of sociopolitical change in the
Ur III period,22 I believe it is essential to underscore that such practice
is also affective. Since the king is now divine, he has his own temples
and images in worship; but equally, since the king has his own temples
and images in worship, he is now divine. The practice itself functions as
a sort of self-fulfilling prophecy. The worship of rulers (hence of royal
images) in Ur III becomes one way not only of demonstrating, but of
establishing (cf. Bloch 1989:198) royal authority as beyond challenge.
In all of the contexts in which royal statues functioned (particularly
when in worship in their own shrines), I see their presence as part
of a larger strategy, namely, the introduction of the ruler into, and
the appropriation of, ritual space hitherto belonging to the god. Pres-
ence in the god’s shrine constitutes power; it permits direct access to
supra-human authority. Presence and service in the funerary chapel
concentrates power through an investment of resources and attention
that underscores not only the person but the institution of which the
person was a representative. Presence in a shrine dedicated to the ruler
himself confers power commensurate with the gods, thereby significantly
increasing the position of the ruler in the social hierarchy.
What may be particular to the Mesopotamian situation is the power
vested in the image of the ruler, the alam-lugal, as it takes its place
in the various ritual contexts outlined above. Identified by likeness,
inscription, and name as the ruler,23 empowered and instructed by
the ruler, ritually consecrated to be the ruler, the image plays upon
representation and manifestation, man-made and heaven-born. At
issue, therefore, is not only the relationship of ritual and politics to
22
Kertzer (1990:350) has reviewed the issue of the relationship between political
change and ritual change in the legitimation of authority.
23
Many of the statues are identified by a brief label on the shoulder, for instance,
“Gudea, ensi of Lagash” (Steible 1990:Statues A and E), then given a ritual name in
the longer text inscribed over the body, for example, “ ‘My lady [the goddess Baba], you
have raised me up! Grant me life! On the appointed day have I built [your temple],’
is its (the statue’s) name” (Steible 1990:Statue E.9.1–4).
power, but that of representation to power, or what Marin (1988:6) has

referred to as the power of presence both of and in the image. With
the Gudeas in particular, to see the image merely as a substitute for
the ruler, as has often been suggested, denies the levels of careful con-
structedness according to appropriate physical form (cf. Winter 1989)
and consecration according to appropriate ritual action that together
transform inert material into the animate subject, Gudea. The nature
of the ritually empowered royal image is such that it brings signifier
(the statue) and signified (the ruler) together in the same vehicle/sign,
either to stand before a similarly manifest deity or sit itself in receipt
of devotional service.
As Marin has shown for the portraits of Louis XIV (1988:13), the
Mesopotamian royal image is (1) the particular historical personage
whose features it resembles and on whose body the name may be
inscribed (Gudea); (2) the representative of a class, “ruler” (ensi of
Lagash); and (3) a sacred, animate entity identical with its referent.
As such, the royal image stands as the absolute embodiment of the
ruler. Indeed, as Marin argues (1988:8), the ruler is absolute only in
his images.
The rituals of consecration, installation, and maintenance that dif-
ferentiate Mesopotamian (and other) “manifestations” from European
(and other) “representations” further intensify the three simultaneous
representational identities cited above, and underscore the absolute aspect
of the image. I would distinguish the particular rituals I have described
here from both rites of passage and rites of institution (Bourdieu
1991:117). Rather, they may be seen as rituals of constitution that first
bring the image into being through its animation, then maintain that
being. These rituals of constitution provide the context in which the
Mesopotamian royal image exists and functions. They are constitutive
insofar as by their performance (cf. Tambiah 1985) they constitute the
object of the ritual’s focus: in this case, the manifestation within the
material form. They give the image its life—and its power.
The unusually long historical sequence of ritual practice associated
with images in Mesopotamia permits us to document both continu-
ity and change. Continuity in Sumerian practice is attested from the
Early Dynastic period (ca. 2500 b.c.e.) through the Ur III period (ca.
2000 b.c.e.) and into the succeeding dynasties of Isin and Larsa in the
early second millennium b.c.e. Akkadian texts of the later Assyrian
and Babylonian periods show that many of these practices were still
in place well into the mid-first millennium (see Porter, n.d.; Beaulieu
1989), extending the over 500-year Sumerian sequence into a pan-

Mesopotamian sequence covering a period of some 2000 years.
I have concentrated in the present article on Sumerian material in
the period at the end of the third millennium b.c.e., around the time
of Gudea of Lagash, for that is when the best confluence of texts and
archaeological finds happens to occur. The partly contemporary and
immediately subsequent period of the Third Dynasty of Ur marks
an important change, albeit short-lived. There was an increase in the
places for images of rulers—adding independent royal shrines to gods’
shrines and funerary chapels. This increase in venues can be correlated
with social and political developments that culminated in the equally
short-lived elevation of the ruler to divine status. The agency of the
ruler in consciously and singularly manoeuvering strategic changes
should not be over-valued, however. As Bourdieu has noted (1991:126),
“The belief of everyone, which pre-exists ritual, is the condition for the
effectiveness of ritual.” Clearly, Mesopotamian belief included notions
of ritual efficacy and of matter as the physical vehicle for manifesting
essence.24 In this belief system the royal (as well as the divine) image
was born, lived, and was maintained in a constant ritual context of
service and attendance. It has been the task of the present article to
gather evidence that would demonstrate this claim to have been the
case and at the same time to engage the recent literature on royal
ritual through the introduction of the royal image, ritually constituted
to be the ruler. By arguing further that these images were affective as
well as reflective, it is suggested that they served as the loci of a set of
values and relationships, political no less than religious, around which
functioned the entire Mesopotamian state.
24
A number of texts state explicitly that when the ruler is well-received by the god,
the reward is not his alone. Abundance of harvest, good health and prosperity are
assured for his people (e.g., Gudea cylinder A: Jacobsen 1987:A.9.1–18). That these are
all official and public state texts does not necessarily imply that the sentiments expressed
were not also popularly held to be true. Only in cases where there is evidence to show
the division between popular sentiment and political assertion (as in Porter, n.d.) can
we speak of propaganda or rhetoric as distinct from belief.
Acknowledgments
I am grateful to a number of individuals for their part in providing stimuli, sources,

help and criticism, particularly H. Behrens, P. Boden, M. Civil, M. Geller, R. C. Hunt,
Barbara A. Porter, Barbara N. Porter, A. Sjöberg, P. Steinkeller. In addition, although
the present article does not pursue comparative materials, I should like to acknowledge
the extraordinarily enabling experiences I have had in discussing related issues and in
observing ritual treatment of images in India, particularly in the Rādhāramana Temple
of the Sri Caitanya Prema Sansthana, Vrindaban, and to thank specifically Kapila
Vatsyayan, Srivatsa Goswami, and Asim Krishna Das for their generosity of time and
resources. I would not have sought much of the Mesopotamian evidence had I not
been prepared by current Hindu practice.
References
Abélès, Marc. 1988. “Modern Political Ritual.” Current Anthropology 29:391–404.

Arnaud, Daniel. 1986. Recherches au pays d’Astarté: Emar VI:3, Textes sumeriennes et akka-
diennes. Paris: CNRS.
Barrelet, Marie-Thérèse. 1974. “La figure du roi.” Le palais et la royauté, 27–138. Ed.
P. Garelli. Paris: Geuthner.
Bauer, J. 1969. “Zum Totenkult im altsumerischen Lagasch.” Zeitschrift des deutsches
Morgenlandes Gesellschaft, Supplementum, 107–114. Wurzburg.
Beaulieu, Paul-Alain. 1989. The Reign of Nabonidus, King of Babylon 556–539 B.C. New
Haven: Yale University Press.
Behrens, Hermann. 1987. “Statues, Tablets and the Cult of Ancestors in Ancient
Mesopotamia.” Unpublished lecture given at the University of Pennsylvania.
Bell, Catherine. 1990. “The Ritual Body and the Dynamics of Ritual Power.” Journal
of Ritual Studies 4(2):299–313.
Bourdieu, Pierre. 1991. Language and Symbolic Power. Trans. G. Raymond and M. Adam-
son. Cambridge: Harvard University Press.
Black, Jeremy A. 1981. “The New Year Ceremonies in Ancient Babylon: ‘Taking Bel
by the Hand’ and a Cultic Picnic.” Religion 11:39–59.
Bloch, Maurice. 1989. “The Ritual of the Royal Bath in Madagascar: The Dissolution
of Death, Birth and Fertility into Authority.” Ritual, History and Power: Selected Papers
in Anthropology, 187–211. London School of Economics, Monographs on Social
Anthropology No. 58. London: Athlone.
Camille, Michael. 1989. The Gothic Idol: Ideology and Image-Making in Medieval Art. Cam-
bridge: University Press.
Cannadine, David and Simon Price, eds. 1987. Rituals of Royalty: Power and Ceremonial
in Traditional Societies. Cambridge: University Press.
Chicago Assyrian Dictionary. 1962. Volume S:78–85. “ almu.” Chicago: University of
Chicago Press.
Civil, Miguel. 1967. “Another Volume of Sultantepe Tablets,” Journal of Near Eastern
Studies 26:200–211.
Civil, Miguel and Richard L. Zettler. 1989. “The Statue of Šulgi-ki-ur5-sag9-kalam-ma,
Part One: The Inscription,” and “Part Two: The Statue.” Dumu-E2- dub-ba-a: Studies
in Honor of Åke W. Sjöberg, 49–64, 65–78. Ed. H. Behrens and others. Occasional
Publications of the Samuel Noah Kramer Fund, 11. Philadelphia: The University
Museum.
Ebeling, Erich. 1931. Tod und Leben nach den Vorstellungen der Babylonier. Berlin: de
Gruyter.
Falkenstein, Adam. 1971. “Gudea (A. Nach Texten).” Reallexikon der Assyriologie 3:676–
679. Berlin: de Gruyter.
Fortes, Meyer. 1968. “Of Installation Ceremonies,” Proceedings of the Royal Anthropological
Institute for 1967, 5–20.
Frankfort, Henri, Seton Lloyd and Thorkild Jacobsen. 1940. The Gimil-Sin Temple and
the Palace of the Rulers at Tell Asmar. Oriental Institute Publications, 43. Chicago:
Oriental Institute.
Frayne, Douglas. 1990. Old Babylonian Period (2003–1595): Royal Inscriptions of Meso-
potamia, 4. Toronto: University of Toronto Press.
Geertz, Clifford. 1966. “Religion as a Cultural System.” Anthropological Approaches to the
Study of Religion, 1–46. Ed. M. Banton. ASA Monographs, 3. London: Tavistock.
Genouillac, Henri de. 1910. Inventaire des Tablettes de Tello. Paris: Leroux.
Giesey, Ralph E. 1960. The Royal Funeral Ceremony in Renaissance France. Geneva: Droz.
Goody, Jack. 1961. “Religion and Ritual, The Definitional Problem.” British Journal
of Sociology 12:142–164.
Grayson, A. Kirk. 1976. Assyrian Royal Inscriptions, 2. Wiesbaden: Harrassowitz.

Hallo, William W. 1988. “Texts, Statues and the Cult of the Divine King” Supplement,
Vetus Testamentum, 40:54–66.
Holdredge, Barbara A. 1990. “Introduction: Towards a Phenomenology of Power.”
Journal of Ritual Studies 4(2):5–37.
Jacobsen Thorkild. 1987a. The Harps that Once. . . : Sumerian Poetry in Translation. New
Haven: Yale University Press.
——. 1987b. “The Graven Image.” Ancient Israelite Religion: Essays in Honor of Frank Moore
Cross, 15–32. Ed. P. D. Miller, Jr. and others. Philadelphia: Fortress Press.
——. 1989. “The Mesopotamian Temple Plan and the Kititum Temple.” Eretz Israel
20:79–81.
Johansen, Fleming. 1978. Statues of Gudea—Ancient and Modern. Copenhagen: Akademisk
Verlag.
Kelly, John. D. 1988. “From Holi to Diwali in Fiji: An Essay on Ritual and History,”
Man (N. S.) 23:40–55.
Kelly, John. D., and Maureen Kaplan. 1990. “History, Structure and Ritual.” Annual
Review of Anthropology 19:119–150.
Kertzer, David I. 1988. Ritual, Politics, and Power. New Haven: Yale University Press.
——. 1990. “Review Essay: Ritual and Politics.” Journal of Ritual Studies 4(2):349–
354.
Klein, Jacob. 1989. “From Gudea to Šulgi: Continuity and Change in Sumerian Liter-
ary Tradition.” Dumu-E2-dub-ba-a: Studies in Honor of Åke W. Sjöberg, 289–302. Ed. H.
Behrens and others. Occasional Publications of the Samuel Noah Kramer Fund,
11. Philadelphia: The University Museum.
——. 1991. “The Coronation and Consecration of Šulgi in the Ekur (Šulgi G).” Ah,
Assyria . . ., 292–313. Ed. M. Cogan and I. Eph’al. Jerusalem: The Magnes Press.
Kuper, Hilda. 1972. “Royal Ritual in a Changing Political Context,” Cahiers d’études
Africaines 12(48):593–615.
Kutscher, Rafael. 1974. “An Offering to the Statue of Šulgi.” Tel Aviv 1:55–59.
——. 1989. Royal Inscriptions: The Brockmon Tablets at the University of Haifa. Haifa: Uni-
versity Press.
Lafont, Baernard and F. Yildiz. 1989. Tablettes cuneiformes de Tello au Musée d’Istanbul, I.
Istanbul: Nederlands Historisch-Archäologisch Institut.
Marin, Louis. 1988. Portrait of the King. Trans. M. H. Houle. Dexter, MN: University
of Minnesota Press.
Michalowski, Piotr. 1977. “The Death of Šulgi,” Orientalia (N. S.) 46:220–225.
Moore, Sally F., and Barbara G. Myerhoff, eds. 1977. Secular Ritual. Assen, The Nether-
lands: Van Gorcum.
Nies, J. B. 1920. Ur Dynasty Tablets. Leipzig: J. C. Hinrichs’sche.
Oppenheim, A. Leo. 1964. Ancient Mesopotamia: Portrait of a Dead Civilization. Chicago:
University of Chicago Press.
Parrot, André. 1948. Tello, Vingt campagnes de fouilles (1877–1933). Paris: Editions
Michel.
Perlov, B. 1980. “The Families of the Ensis Urbau and Gudea and Their Funerary
Cult,” 77–81. Death in Mesopotamia. Ed. B. Alster. Copenhagen: Akademisk.
Pollock, Sheldon. 1990. “From Discourse of Ritual to Discourse of Power in Sanskrit
Culture.” Journal of Ritual Studies 4(2):315–337.
Porter, Barbara N. n.d. (forthcoming). “Gods’ Statues and Political Policy: Esarhaddon’s
Return of Marduk to Babylon,” in L. Martin, ed. Religious Transformation and Socio-
Political Change.
Price, S. R. F. 1984. Rituals and Power: The Roman Imperial Cult in Asia Minor. Cambridge:
Cambridge University Press.
Rappaport, Roy. 1971. “Ritual, Sanctity, and Cybernetics.” American Anthropologist
73(1):59–76.
Renger, Johannes. 1967. “Untersuchungen zum Priestertum in den altbabylonischen

Zeit, 1. Teil.” Zeitschrift für Assyriologie 58:110–188.
Smith, Sidney. 1925. “The Babylonian Ritual for the Consecration and Induction of
a Divine Statue.” Journal of the Royal Asiatic Society: 37–60.
Spiro, Melford. E. 1966. “Religion: Problems of Definition and Explanation.” Anthro-
pological Approaches to the Study of Religion, 85–126. Ed. M. Banton ASA Monographs,
3. London: Tavistock.
Spycket, Agnès. 1968. Les statues de culte dans les textes mesopotamiens des origines a la I re.
dynastie de Babylone. Cahiers de la Revue Biblique, 9. Paris: Gabalda.
——. 1981. La statuaire du proche-orient ancien. Handbuch der Orientalistik, 7. Leiden-
Köln: E. J. Brill.
Steible, Horst. 1982. Die altsumerischen Bau- und Weihinschriften. Freiburger altorientalische
Studien, Bd. 5. Wiesbaden: Steiner.
——. 1988. “An Attempt at Establishing a Chronological Order for the Statues of
Gudea of Lagash from the Point of View of a Sumerologist.” Unpublished lecture,
University of Pennsylvania.
——. 1991. Die neusumerischen Bau- and Weihinschriften. Freiburger altorientalische Studien,
Bd. 9. Stuttgart: Steiner.
Steinkeller, Piotr. 1988. “The Date of the Gudea Dynasty.” Journal of Cuneiform Studies
40:47–53.
Tambiah, Stanley J. 1985. “A Performative Approach to Ritual,” in Culture, Thought and
Social Action: An Anthropological Perspective, 123–166. Cambridge: Harvard University
Press.
Trumpelmann, Leo. 1990. “Bemerkungen zum ‘Kitchen-Temple’ in Nippur.” Baghdader
Mitteilungen 21:125–129.
Turner, Victor. 1969. The Ritual Process: Structure and Anti-Structure. Chicago: Aldine.
Wilcke, Claus. 1974. “Zum Königtum in der Ur III Zeit.” Le palais et la royauté, 177–232.
Ed. P. Garelli. XIXe Rencontre Assyriologique Internationale, Bruxelles 1971. Paris:
P. Geuthner.
Winter, Irene J. 1987. “Legitimation of Authority through Image and Legend: Seals
Belonging to Officials in the Administrative Bureaucracy of the Ur III State.”
The Organization of Power: Aspects of Bureaucracy in the Ancient Near East, 69–116. Eds.
M. Gibson and R. D. Biggs. Studies in Ancient Oriental Civilization, 46. Chicago:
The Oriental Institute.
——. 1989. “The Body of the Able Ruler: Toward an Understanding of the Statues of
Gudea.” Dumu-E2-dub-ba-a: Studies in Honor of Åke W. Sjöberg, 573–784. Ed. H. Behrens
and others. Occasional Publications of the Samuel Noah Kramer Fund, 11. Phila-
delphia: The University Museum.
——. 1991. “Reading Concepts of Space from Ancient Mesopotamian Monu-
ments.” Concepts of Space, Ancient and Modern, 57–73. Ed. K. Vatsyayan. New Delhi:
Abhinav.
Figure 1. Gudea B, seated. Dedicated to

the god Ningirsu. Found at Tello (ancient
Girsu), in palace courtyard. Diorite; ht. 0.93
m. (Louvre, AO 2).
Figure 2. Plan, pal-

ace of Adad-Nadin-
A e (2nd c. b.c.e.),
with part of the
foundations of the
Gudea temple (ca.
2100 b.c.e.) in dark,
solid lines visible
in courtyard. Tello,
Mound A. (= Parrot
1948: Figure 33b).
Figure 3. Ur-Ningirsu, son of

Gudea, standing, detail.
Dedicated to the god Ningišzida.
Found at Tello during clandestine
excavations of 1924. Gypseous
alabaster; ht. 0.46 m. (Louvre,
AO 9504).
Figure 4. Ur-Ningirsu (cf. Figure 3),

rear view. Inscription picked out in
white chalk.
Figure 5. Stele of Ur-Nammu, first king of Third Dynasty of Ur. Detail,

second register. Found at Ur in precinct of the moon-god, Nanna.
Limestone; ht. of whole, ca. 3 m. (University Museum, Philadelphia,
CBS 16676).
Figure 6. Plan of temple to Šu-Su’en,

fourth king of Third Dynasty of
Ur (ca. 2030 b.c.e.). Tell Asmar
(ancient Ešnunna). (=Frankfort et al.
1940 Pl. III).
Figure 7. Gudea D, seated.

Dedicated to the god Ningirsu.
Found at Tello in palace
courtyard. Diorite; ht. 1.58 m.
(Louvre, AO 1).
EXPERIENCING ‘ART’ AND ARTIFACT
CHAPTER TWENTY-FOUR
REPRESENTING ABUNDANCE:
A VISUAL DIMENSION OF THE AGRARIAN STATE
Abstract
The symbolic systems and representational strategies of the early agrarian state in ancient
Mesopotamia remain relatively unexamined. One important rhetorical device in text is
the assertion of, or hope for, abundance as the result of divine beneficence brokered by the
state apparatus. It is argued here that the use of visual motifs such as repetitive friezes of
domesticated plants and animals—often dismissed as merely ornamental—not only paral-
leled textual references to abundance, but actually offered an independent and highly charged
articulation of the same governing concept.
Analysts of the early agrarian state have stressed the agricultural and
pastoral surplus necessary to sustain the specialized labor force and social
hierarchy characteristic of an urban population, while field archaeolo-
gists have sought remnants of large-scale storage facilities, increased scale
of irrigation systems, and other features to document the production
and accumulation of such surplus (Childe 1951; Adams 1960; Sherratt
1980; Hunt 1987; 2000; Dolce 1989; Pollock 1992; H. T. Wright 1994;
Breckwoldt 1995–96; Smyth 1996; Algaze 2004; Rothman 2004; M. E.
Smith 2005; Steinkeller, this volume). Remaining underdocumented,
however, are the symbolic systems and representational strategies that
mirrored such necessities, or even—dare one say!—actively served to
create a receptive social environment in which the authority structures
needed for a requisite surplus could be reinforced.
In the present paper, I wish to pursue ways in which the accumula-
tion of such a surplus, through the production of domesticated plants
and animals and the yield of the Tigris-Euphrates Rivers, was rein-
forced through visual representation in the early stages of the agrar-
ian state. I shall argue that surplus in our terms was in Mesopotamia
* This article originally appeared as “Representing Abundance: The Visual Dimen-

sion of the Agrarian State,” in Settlement and Society: Essays Dedicated to Robert McCormick
Adams, E. Stone, ed., Los Angeles: Cotsen Institute, 2006, pp. 117–138.
200 chapter twenty-four
equated with a concept of “abundance” in natural production, and

that the representation thereof constituted a virtual “iconography of
abundance.” This iconography was then consciously deployed to signal
the surplus that on a practical level permitted the maintenance of an
urban, specialist population. At the same time, on the rhetorical level
the iconography also served to lay the foundations for social cohesion
through a shared visual vocabulary for success issuing from a projected
cosmic/natural order.1
Such a study could not have been undertaken by an art historian
until relatively recently, and on many levels, it owes its conceptual base
to the work of anthropologically trained archaeologists, notably Robert
McCormick Adams, whose focus on early state formation significantly
changed the sorts of questions one asks of artifactual and archaeologi-
cal materials. From Land Behind Baghdad (1965) to The Uruk Countryside
(Adams and Nissen 1972) and Heartland of Cities (1981), Adams helped
to sensitize the field to think beyond the immediate locus of excavation
to broader issues of region and system. His work—and subsequently
that of his students and the field at large—demanded that one consider
site distribution with respect to the sustaining area of agricultural and
animal production, themselves essential components in the dynam-
ics of state development and maintenance. More recently, with the
development of the subfield of “landscape archaeology,” the study of
topography, region, and space have focused attention on these broad
issues (such as Wilkinson 2003; A. Smith 2003; Gleason 1994). Their
work further opens the door to seeing artistic production not as separate
from, but in conjunction with, the particulars of a given land/environ-
ment, as a locus for both biological and social reproduction.2
Indeed, this conjunction between agrarian and artistic production
from the Uruk to the Neo-Sumerian periods, fourth to third millen-
nium b.c.e., argues for continuity of a central theme of “abundance,”
despite changes in representational strategies over time. I shall argue
1
The present paper has grown out of a presentation in a symposium honoring
Bob Adams at the meetings of the American Oriental Society in Toronto in 2001. A
portion of that presentation, pursuing the rhetorical component of this claim for the
Neo-Assyrian period, has recently been published as Winter 2003.
2
Indeed, as noted by Adam T. Smith (2003: 10, 54), the land provides the condi-
tions for aesthetic as well as pragmatic pleasure. In the visual domain, a very specific
case was noted in a recent study by Trudy Kawami (2001), which demonstrates a
correlation between the introduction of a new variety of milk cow in the Uruk period
and the appearance of the same breed of bovine figurines on stamp seals.
representing abundance 201
it is no accident that an iconography of abundance appeared precisely

at a moment in history when, although domestication had been under
control for millennia, the need for sufficient surplus to both provision
the city/state and afford a medium of exchange in long-distance trade
was of primary importance in both the rhetoric and the governance
structures of the stratified state and its agrarian-based economy.3
The rhetorical aspect was instantiated in both text and image. To date,
the textual components have been better integrated into the broader
picture. Virtually every literary text attributes “abundance”—Sumerian
e2.gal2, e2.nun; Akkadian egallu, nu šu, tu du—directly to one deity
or another. And virtually every Mesopotamian ruler who left royal
inscriptions has at some point declared himself to be the “provider” of
abundance for the land, thanks to the privileged relationship he enjoys
with a god, or the gods. Clear from the context is that this abundance
issues from the natural world—water, plants, animals, fish—rather than
from material wealth or manufactured goods.
The agency of gods and rulers with respect to provisioning the
land was articulated most explicitly by Gudea, ruler of the city-state
of Lagash in the Neo-Sumerian period (ca. 2110 b.c.). Preliminary to
building a temple for the state god, Ningirsu, the ruler is visited in a
dream by the deity, who makes a promise [Cyl. A, xi 1–17; Edzard
1997: 75–76]:
When you, true shepherd Gudea, will start (to build) my House . . .
then I will call up to heaven for a humid wind, so that surely abundance
3
Congruent with more recent archaeological studies of domestication, the Sumerians
themselves seem to have preserved a memory of bringing domesticated species into the
Tigris-Euphrates Valley. In the literary composition, “How Grain Came to Sumer”
(ETCSL t.1.7.6), we are told of a time when men did not know barley or flax, and
then barley was brought down from the mountain and introduced into Sumer. The text
breaks off at this point, but as a story of origin, it is clear that the introduction of these
two species was taken as generic for the bringing of “civilization,” since, before that
time, “Men used to eat grass with their mouths like sheep. In those times, they did not
know grain: barley or flax” (lines 1–2). Modern scholars have argued for one variable
or another as the prime mover in the process toward urbanization—most recently,
Algaze (2004) sees trade as the principal engine in what he calls the “Sumerian takeoff.”
The issue of what produced the urban phenomenon in southern Mesopotamia in the
fourth millennium b.c. is beyond the scope of the present study. What I would insist
upon is that, whatever the accountable factors (and I emphasize the plural; similarly,
Wilkinson 2003: 71), the rhetorical strategies employed to maintain the system had to
have been polyvalent, and the assertion of abundance (read surplus) with respect to
agricultural/pastoral production was one of them.
[ e2.gal2] will come to you from above and the land will increase in
abundance.
When the foundations of my House will be laid, abundance will surely
come at the same time:
the great fields will “raise their hands” to you . . . ; water will . . . rise . . . where
it never reaches;
. . . more fat will be poured, more wool will be weighed in Sumer (than
ever before). . . .
The trope of promised abundance is woven through the entire account,
ending with the point when the temple has not only been physically
completed, but has begun to function as a cultic center, the deity
installed within. As part of the temple equipment, (images of ) the seven
daughters of Ningirsu and his consort, the goddess Bau, are placed
near the deity to offer permanent prayers on behalf of the ruler, and
these daughters are explicitly identified as the ones who “create/provide
abundance for the people” [ e-gal2-lu2-šar2] (Cyl. B xi 15–23; Edzard
1997: 95).4
. . . that the vast fields might grow rich, that the ditches and canals of
Lagash be full to the brim, that in the plain . . . the grain goddess . . . might
proudly look up . . . , that after the good fields have brought barley, emmer
and all kinds of pulses, enormous grain heaps, the whole yield of the
land of Lagash might be heaped up. . . .
Finally, the account closes with a series of summary refrains on the
same theme (Cyl. B xix 12–15; xxii 19; Edzard 1997: 98–99, 100):
The earth makes mottled barley grow for him (Gudea), and the abundance
of the land of Lagash increases under (its) ruler.
May the people lie down in safe pastures under your reign, (enjoying)
abundance. . . .
No other ruler of the second half of the third millennium offers as
complete an account. However, Shu-Su en of the later Third Dynasty
of Ur designates the god Enlil as the “lord of abundance” [en e2-gal2-
la], and then asks that he himself be granted “years of abundance”
[mu e2-gal2-la] (ETCSL t.2.4.4.a, lines 21, 31), while the text known
as “Erra and Naram-Sîn” associates the earlier Akkadian ruler with
a “reign of abundance” [ pa-la-i nu-úh-ši ] (Westenholz 1983: 199),
4
This episode is followed immediately by the introduction of the inspector of fisheries
of the Gu-edenna to the deity [Cyl. B xii 7–15], and his intervention that the birds and
animals and fish of the steppe might also produce abundantly (Edzard 1997: 95).
suggesting that this was not limited to a Sumerian topos.5 In addi-

tion, as evidenced in various Sumerian literary texts known from Old
Babylonian Period copies, almost all of the great gods, including An,
Enlil, Enki, Utu, Nanna-Su en, Inanna, Nisaba, Dumuzi, Ninurta, and
Nanše, are credited with providing this same abundance ( e2-gal2) in
the land.6
At least a millennium separates the Neo-Sumerian royal inscriptions
and the largely Old Babylonian Period copies of literary texts from the
early phases of urbanization at Uruk/Warka. The Uruk Period has
received the lion’s share of attention in political analyses of the early
stare (for example, Algaze 1993a; G. Stein 1999a; Rothman, ed. 2001;
see on this also, the historiographical perspective provided by Bernbeck
and Pollock 2002), due to interest in the formative phase of urban
processes. Not surprisingly, this same theme of desired abundance, par-
ticularly with respect to domesticated animals and plants, is to be found
referenced extensively from the beginning. It is not yet explicit verbally
in the archaic texts from Uruk (Nissen, Damerow, and Englund 1993),
but is attested visually, in the cultic objects associated with the temples
of the Eanna precinct of Uruk, dedicated to the goddess Inanna.
The theme may be demonstrated in the low-relief decoration on
a rectangular alabaster trough found in the Eanna, for example
(Moortgat 1969: pls. 17–18; our fig. 1). On each long side, a biological
couple—male ram and female ewe—flank a reed hut that may well
correspond to the “sheep-fold” or “birth hut,” since from it there
emerges a pair of lambs, one in each direction, toward each couple.
That this is not merely a single occurrence, but rather represents a
known motif in the period, is demonstrated by a pair of stone bowls,
5
Jacobsen’s discussion (1993: 72–73) of the term lú argues that “provider” (“sug-
gesting the responsibility of providing food for others”) is part of the meaning of the
term, usually translated as “man.” He cites Shulgi Hymn G, which indicates that the
Ur III ruler himself was considered such a provider by the god Enlil, thereby implying
an equivalence between the notion of providing abundance and the provider thereof,
which constitutes a major rhetorical trope of kingship.
6
ETCSL t.1.1.4, line 2; t.1.2.2, lines 159–70; t.1.1.3, lines 52–60, 71b, 90; t.4.32.
e, line 34; t.1.5.1 lines 186–97; t.4.16.1, line 33; t.4.08.01, lines 53–56; t.1.6.2, lines
362–64; Sjöberg 1975; Jacobsen 1993: 72. The theme also recurs in the Old Baby-
lonian debate texts. In the text known as “The Hoe and the Plough,” the hoe claims
that “the abundance I create spreads over all the land” [ e2-gal2-gu10 kur-kur-ra šu-bi]
(ETCSL t.5.3.1, line 79); and in the debate between Sheep and Grain, before grain is
accorded victory in the debate, both are credited with bringing abundance [ e2-gal2] to
the assembly of the gods (ETCSL t.5.3.2, lines 52–54). (All ETCSL texts in this note
were accessed on 04/12/2004.)
one from Khafaje (Hansen 1975: pl. 71b), the other in the Louvre
(fig. 2). In these cases, the flanking animals are cows or bulls, and the
emerging animals immature calves, but the parallel is clear. On the
Louvre vessel, two strap-handled pitchers are shown in the upper field,
on either side of the hut, suggesting a reference to the important new
breed of milk-producing cattle appearing in the Uruk period (Kawami
2001). On the trough, pairs of reed bundles at the far left and right,
associated with the cult of Inanna (Szarzynska 1987–88), serve as
frames to the whole scene, and also, as on the stone bowls, rise from
the roof of the reed hut, towering above the emerging young. These
reed bundles are repeated on the two short sides of the trough as
well, along with free-floating rosettes, another element associated with
the goddess, adjacent to two ewes or lambs. The juxtaposition of reed
bundle, rosette, and domesticated quadruped echoes similar juxtaposi-
tions on cylinder seals of the same period, also from the Eanna precinct
at Uruk (see, for example, Moortgat 1969: pl. B1). For the trough and
the bowls, at least, it is not unlikely that depiction of the nuclear units
of reproduction (male, female, and offspring), along with references to
the divine power generating reproduction, actually have significance
with respect to their functions as containers: the trough possibly used
to water animals intended for the temple, the bowls possibly holding
milk as divine offering.7
The cereal component is as well documented as the animal. On a
fragmentary limestone stele from Warka, dated stylistically to the Uruk
period, for example (fig. 3), we see in the upper register several em-
paneled boxes that may be a convention for indicating agricultural
fields, containing what look like young stalks of grain (followed in the
second and third registers by work scenes employing nude (?) and bald
male figures; Becker 1993: pl. 39 no. 785). Unfortunately, the stele
was found in a secondary context, reused in antiquity in a later water
7
Joy McCorriston (1997) has shown that one of the important by-products of the
particular breed of wool-bearing sheep introduced into southern Mesopotamia was a
fourth-millennium textile industry of woven woolen fabrics, complementary to earlier
flax-made textiles. Algaze discusses this industry in terms of the dynamic between
imports and the stimulation of indigenous production (2004). It is important to note
that plants and animals may carry meaning on more than one level. However, it
may also be the case that one level of meaning tends to be privileged in particular
instances of representation—in the present case, for example, the wealth/abundance
represented by the herd encompassing the wealth/abundance available from the pro-
duce of the herd.
channel (Becker 1993: 58). However, the imagery seems especially

appropriate to Warka and the goddess whose powers govern fertility
in the plant as well as the animal domain.8 Thus, from the beginning
of the early Mesopotamian state, there is a well-structured cosmology
of reproduction in which gods, their temples, rulers, and rituals are set
in an active relationship to living things.
On several occasions, domesticated animals are actually paired with
domesticated plants in repeating motifs. For example, on a stone vessel
of the Uruk period found at Ur (Strommenger 1964: fig. 28; Hansen
1975: fig. 71a; see fig. 4), one sees a repeated pairing of a sheaf of
grain behind a standing bull, which reflects visually the dual domains
of agricultural and animal productivity referred to in the Gudea text
cited above. Both plant and animal represent not just any species, but
domesticates. Their duplication—echoing the grammatical reduplication
of nouns, adjectives, and verbal stems in the Sumerian language used
to indicate the plural/scale/quantity/intensity—seems to announce in
visual terms the necessary multiplicity (read abundance) of reproduction
and its resultant surplus, upon which the agrarian state was founded.
The emphasis on production and reproduction of both plants and
animals is also found in the conjunction of grain and domesticated
quadruped on a number of cylinder seals of the period (for example,
Amiet 1980: nos. 396–399; see fig. 5) and is most evident on the lower
portion of the alabaster cult vessel known as the Warka Vase, also of
the Uruk period (Heinrich 1936: pl. 36; Moortgat 1969: pls. 19–21;
discussed by Winter 1983; Bahrani 2002; Bernbeck and Pollock 2002:
187–190; see figs. 6 & 7). Above a pair of wavy lines indicating
water,9 two bands are joined by a narrow dividing strip: the bottom
8
Representatives of the Iraq Board of Antiquities found another stele fragment in
the Spring of 2004, which seems to belong to the same stylistic period, if not monu-
ment; but until it is published, this cannot be determined for certain.
9
Identification of these paired wavy lines as water conforms with the ideographic
sign for water in early Uruk tablets and is replicated visually on a Uruk-period cyl-
inder seal that shows such a band running through a terrain marked by drilled blobs
to indicate “land” (see, for example, Moortgat 1969: pl. A2). The presence of a water
channel immediately below the grain and animals suggests a hierarchy, with water as
the source for the growth in the registers above. This would represent the symbolic, and
even cosmic, rendering of the importance of water management known to have been
of primary concern in the early state, and with which a good deal of recent scholarship
has been concerned. See, in particular, the various papers published in the proceedings
of the Sumerian Agriculture Group (for example, Postgate and Powell, eds. 1988).
shows the alternation of stalks of grain, most likely barley,10 with

what can now be identified as flax;11 the upper shows five alternating
pairs of identical rams and ewes.12 The two bands are bound into a
single register with a minor subdivision. Their repetition around the
circumference of the vase calls to mind H. Groenewegen-Frankfort’s
eloquent description of a prehistoric beaker from Susa, where the form
of the vessel is iterated and reiterated through the repetitive decorative
program (Groenewegen-Frankfort 1987: 146–147 and pl. I). Yet, for
the Warka Vase, the repetitive nature of the lower register has led most
Western viewers to pass over the bands quickly, in a rush toward the
more dynamic and linear narrative of the upper register.13 I would argue,
however, that this repetition was carefully planned and highly charged
with meaning. Predating by millennia the Neo-Platonic elevation of
the circle as the perfect geometric form because it has no beginning
and no end, the procession of plants and animals is acephalous and
inexorably conveys the eternal nature of the productivity depicted. Just
as in Mesopotamian literary texts, drama is actually achieved through
a tension built up by repetition, it is suggested that pictorial repetition,
10
The extensive scholarly discussion of the increase in ratio of barley to emmer/
wheat from the Uruk period through the third millennium may be found summarized
in Powell 1985; and more recently, in A. Cohen 2003.
11
Groenewegen-Frankfort (1987: 151) had earlier associated the plant with the
date palm. Flax has now been identified independently by H. Crawford (1985) and
the present author (Winter 1983)—the former on morphological grounds, where she
then discusses the likelihood of flax fiber production by this period; the latter both on
morphological grounds and also by virtue of the frequent pairing in literary texts of
Sum. še and gu, “barley and flax” as the two paradigmatic archetypes of agricultural
production (for example, Enlil and Sud, version B, line 20 = ETCSL t.1.2.2; Enlil
and Ninlil, line 147 = ETCSL t.1.2.1; A Hymn to Utu, line 34 = ETCSL t.4.32.e;
The Debate between Winter and Summer, line 6 = ETCSL t.5.3.3—all accessed
04/12/2004). See also the images and discussion in Ertuğ 2000: esp. figs. 3 and 4, for
visual comparanda with the actual plant.
12
The rams and ewes are obviously gendered, and this parallels the reference on the
Warka through to the biologically reproducing nuclear couple. I argue elsewhere that the
barley and flax also have gendered associations, the shaft of barley reading “male” for
Dumuzid, consort of Inanna; the flax connected to fabric and the weaving of the sheets
for the bridal bed, as in the text of Dumuzid-Inanna A (ETCSL t.4.08.01, accessed
04/19/2004), hence female. For now, I am persuaded that the base of the vase echoes
the cultic event signaled in the upper register, on which see Hansen 1975: 180.
13
Even the perspicacious Groenewegen-Frankfort, while lauding the Susa vessel,
dismissed such repetition in Mesopotamian art in terms of a loss of “vigor,” stating
that in the reduction to pattern, the “temporal aspects of action and . . . space . . . (are)
robbed of their truly dramatic character” (Groenewegen-Frankfort 1987: 162).
too, serves to raise a scene from the level of the actual, or temporal,
to the (hoped for) perpetual, or recurrent.14
In order to understand the nature of repetitive imagery in the ancient
Near East, it is essential to resist the Western relegation of pattern
and repetition to the merely “decorative” (see, on this, Winter 2003).
The union of not just any plants and animals, but domesticates, in the
lower registers of the vessel makes clear that several millennia following
upon Neolithic breakthroughs, those responsible for the program of
the Warka Vase were not unaware that the economic and subsistence
base for urban society lay in the surplus possible through that very
domestication.15 I would even go so far as to suggest that the repeating
figures of plants and breeding animals circling the lower portion of the
vessel with no apparent beginning or end can be read symbolically as
fundamental (basic) to the urban enterprise—representing that perpetual
chain of productivity, that very abundance, in which the state (as well as
its rhetoric) was grounded, and upon which it depended.16
The middle zone of the Warka Vase also often gets short shrift.
There, a fairly homogeneous row of nude, bald men moving from
right to left carry vessels, again in an acephalous procession. The roll-
out drawing of that register makes clear that at least three different
vessel types are represented: a shallow bowl of uncertain material, the
contents of which are not visible; a spouted pitcher, of a type known
from pottery of the period, presumably containing a liquid of some
sort; and a basket with at least two sorts of contents visible mounded
above the rim. A palm frond in one of the baskets, as well as in the
14
Noted in Winter 1976, a review of Groenewegen-Frankfort 1951 [1987] (upon
its initial reissue in 1974).
15
Adams himself emphasized early on that the agricultural “revolution” of the Neo-
lithic was what ultimately made it possible for both specialization and the aggregation
of communities into urban centers (Adams 1960). What is interesting for our purposes
is that this development and dependence was not unremarked by the Sumerians of
the fourth/third millennium b.c.
16
Although in the literary text, “The Debate between Sheep and Grain,” grain is
ultimately declared by Enki to be predominant, the whole debate tradition—“hoe and
plough,” “palm and tamarisk,” “bird and fish,” “farmer and shepherd” “winter and
summer”—tends to pair sets that are in fact essential dyads, one hardly more important
than the other. As noted above, in the Sheep and Grain text, before the ultimate judg-
ment is made, the two protagonists are brought before the gods, where it is made clear
that both of them “brought abundance ( e2.gal2) to the assembly” (ETCSL t.5.3.2, line
54). In “The Debate between Winter and Summer,” as well, both protagonists are
credited with providing abundance: winter in terms of water and planting, summer in
terms of harvest (ETCSL t.5.3.3, line 23; both texts accessed 04/12/2004).
baskets of the upper, narrative register representing a scene oriented to

the goddess Inanna, suggests that the contents of at least those baskets
would be dates—entirely fitting, considering the association of Inanna
with the fruit of the date palm.17 Given the presence of a nude, bald
male in the upper register identical to those in the middle register, who
offers a basket directly to the goddess, it would seem that the unend-
ing procession of the middle zone is meant to conceptually culminate
in the offering of all of the contents of the various vessels, that is, the
productivity of the land, to the goddess above in a cultic setting.18
This fits well with the cult offerings recorded in the reign of Ibbi-
Su en of the Ur III period to the moon-god Nanna/Su en at Ur, in
which festival offerings of fruit, especially dates (Sum. su11-lum), are
presented in baskets of 3-sila capacity, sometimes accompanied by
beer and animals.19 Although here the recipient is the late third millen-
nium moon-god at Ur, whereas it is the earlier Inanna of Uruk who is
depicted on the vase, one may assume that the broad pattern of offerings
17
For the association between Inanna and the date palm, see Abusch 1995, with
reference to Jacobsen 1976: 34–36. For the presentation of dates to the deity, see the
reference in the poem “Inanna and Šu-kale-tuda” (ETCSL t.1.3.3, lines 83–84, accessed
04/17/2004) to “ . . . dates . . . fit for the temples of the great gods.” In yet another Hymn
to Inanna (Sjöberg 1988: 169, col. ii, line 7), the goddess is called “the one who makes
the dates be alluring in their panicles/baskets.”
18
For the cultic setting of offerings to Inanna in archaic Uruk, including liquids
in jars and various comestibles in bowls, see Szarzynska 1997. The paired elements
that are gathered behind the goddess on the upper register of the vase presumably
also represent received offerings, including animals, vessels, and baskets. Indeed, one
wonders whether the assemblage represents not just numerical pairs, but actually, as
with reduplication in Sumerian verbs and adjectives to indicate myriads/plurality,
might in fact stand for a general visual category of “multiples,” hence to be read as
“abundance.” That abundance was signaled in royal texts as having been provided
by individual rulers to the temple of a particular god is apparent in later hymns of the
Third Dynasty of Ur: for example, in the text known as Shulgi A (ETCSL t.2.4.2.01,
line 51), where the king is made to state in the first person: “I filled with abundance
the temple of Nanna/Suen” [e2 dsuen-na . . . e2.gal2-la e2-bi2-du8]. In addition, in the
text known as “Enki’s journey to Nibru” (ETCSL t.1.1.4, line 50), Enki’s temple in
Eridu is described as one “whose inside is full of abundance” [šag4-bi e2-gal2 sug4-
ga]. If one understands the reed bundles behind the goddess on the Warka Vase as
representing her shrine (compare Szarzynska 1987–88), then the presentation of these
offerings to the goddess at her shrine could well link the Uruk-period cultic practices
in urban temples with those reflected in later texts.
19
See the discussion in M. G. Hall 1985: esp. 260–95, where different types of offer-
ing are enumerated; also Sallaberger 1993. The other fruit most frequently mentioned
in offerings is Sum. aš ur, “pomegranate” or “apple” (M. G. Hall 1985: 291), and
this could account for the other type of basket content depicted on the vase. Beer is
also mentioned in conjunction with offerings to the sun-god in the Hymn to Utu cited
above (ETCSL t.4.32.e, line 42, accessed 04/12/2004).
within a Sumerian cultic context would be similar (see Szarzynska

1997). This picture is consistent with a point made by Nicholas Postgate
(1987) that, in general, third-millennium references to fruits are largely
contained in offering lists of materials supplied to temples for use in
cultic contexts. In addition to mention of “Dilmun dates in baskets,”
such texts also refer to nuts, apples/pomegranates, figs, and grapes, and
to dates, grapes, and pomegranates in jars—suggesting that they have
been rendered as syrup (see, for example, the poem “Enlil and Sud” =
ETCSL t.1.2.2, lines 118–123, where the produce is dispatched by
Enlil to Ereš as marriage gifts). While most of these texts are again
Neo-Sumerian in date or Old Babylonian in copy, there are also Early
Dynastic references to both figs and apples/pomegranates, so it may
well be that the differential sizes of round objects in the baskets of the
Warka Vase also represent some of these other fruits, along with a pre-
dominance of dates. And while it is impossible to determine, without an
extensive study of liquid offerings in cultic settings plus archaeological
sampling and analysis of residues found in appropriate contexts in the
field, what liquids (from beer to fruit syrups to water) might have been
contained in the spouted vessel on the Warka Vase, the point remains
that the offerings mirror primary agrarian production.
What we see on the Warka Vase represents the inverse of the
abundance desired from the goddess, she who is described as “keeping
the abundance of the land in her hands” [ e2-gal2-kur-kur-ra šu-še3
la2/-a-e (Sjöberg 1988: 167, col. i, line 16)]; it is the offering back
to the goddess of the fruits of her proffered abundance, manifest in
the previous harvest. There is a retrospective aspect to such offerings,
comparable to the offering of harvest fruits in the Jewish festival of
Sukkoth (Schauss 1973: 170). A similar ceremony is also attested for
ancient Mesopotamia—for example, in the “Debate between Winter
and Summer” (ETCSL t.5.3.3, line 601, accessed 04/12/2004), which
makes reference to “the harvest, the great festival of Enlil.” But there
is also a prospective aspect: the offerings, like Gudea’s construction of
the temple itself, constitute service to the deity which will then stimu-
late the next round of positive response(s), in an unbroken chain of
reciprocities. In just this way, in the temple bathing festivals (Sanskrit
abhishēka) observed in Hindu India, ritually offered materials are prof-
fered not just symbolically, but in great abundance. Furthermore, the
bathing festival to Krishna Rādhāramana-ji in his Vrindaban temple,
as but one example, is celebrated both at the beginning and at the
end of the monsoon season that will determine the next harvest, and
so may be seen in relation to the fertility of the agricultural year to

come (Winter 2000).
Robert Hunt (1987) has discussed the “provisioning of cities” from
the perspective of what is necessary at the physical and practical level
for the city/state to be maintained; and also (2000) the technology
plus labor force required to successfully manage natural resources. His
analysis complements the textual evidence for cultic offerings in the
context of ritual activity, which works at the symbolic and cosmological
level to maintain the same system (see Sallaberger 1993). And I am
arguing here that similar principles may be observed in the visual
sphere, marked on cultic objects like the Warka Vase and also on public
monuments like the fragmentary stele mentioned above.
Similar themes are referenced throughout the third millennium b.c.,
from the Early Dynastic period through the Third Dynasty of Ur. On
the silver vase inscribed by Enmetena of Lagash, for example (Moort-
gat 1969: pl. 113; see fig. 8), attention has been largely focused on the
major imagery around the body of the vessel: the lion-headed eagle
viewed frontally, his claws grasping the hindquarters of pairs of heraldic
lions and ibexes. This composite bird-creature is the familiar of the
god Ningirsu, to whom the vase is dedicated, and it is clear that the
principal iconography is tailored to the deity in whose cult the vase
would presumably have been used. Still, incised above the main panel,
around the shoulder of the vessel, is a continuous frieze of couchant
calves facing left, ordered like the rams and ewes of the Warka Vase,
with no clear beginning or end. One wonders, therefore, whether there
is not more to be read from the iconography of the vase as a whole
when the two registers are combined: the lion-headed eagle of Ningirsu
controlling the wild animals of the steppe, in order that the domesticated
breed depicted on the shoulder should be able to reproduce abundantly
and enduringly. That this is not a unique juxtaposition is suggested by
an Early Dynastic seal in the Louvre (fig. 9), where in the lower register,
along with a combat scene, the lion-headed eagle has its talons into the
backs of couchant ibexes, while in the upper register bovids emerging
from a hut appear with a milking scene. This motif, on both vase and
seal, would then provide the visual equivalent of the “safe pasture”
cited in Gudea’s Cylinder B above. In the text, the safe pasture is used
as a metaphor for the security of the people, but underlying it is the
importance of security from human and/or animal predators that will
permit the herd to flourish, multiply, and produce.
Related plant and animal motifs appear regularly on works dated

to the Early Dynastic period: for example, on a number of objects
from the Royal Cemetery of Ur, including vessels, gaming boards, and
personal jewelry (Zettler and Home, eds. 1998: esp. pls. 29–30; Miller
2000; Gansell 2004; see fig. 10); in the representation of the goddess
Ninhursag on a basalt vessel inscribed by Enmetena of Lagash from
Tello, on which the goddess holds a date spathe in her right hand
(Moortgat 1969: pl. 115; see fig. 11); and in the several examples of
priests or rulers pouring libations into a potted plant, usually some
form of date frond with clusters, before a seated deity (for example,
on a limestone plaque from Tello: Moortgat 1969: pl. 114). This motif
continues well into the Third Dynasty of Ur, as seen on the Stele of
Ur-Namma from Ur (Canby 2001: pl. 10; see fig. 12). Semi-divine
figures holding a flowing vase on Akkadian seals (Boehmer 1965: 232,
280, 502, 523, 525) or the fish depicted in the streams that flow as
attributes from the shoulders of the god Ea/Enki (Boehmer 1965: 377,
488, 493–499, 508–513, 520) may also fall into this same category
of referencing natural abundance, especially as produce in the rivers,
from fish to reeds, is frequently mentioned within the literary topos of
Sumerian e2-gal2, Akkadian egallu.20
Clearly, the visual motifs change over the two millennia from early
Sumerian urbanization and the city-state to Neo-Sumerian territorial
expansion of the fully developed and bureaucratized nation-state, and
for purposes of classification, the iconographer would be likely to put
the themes into different categories (animal files, repeating plant chains,
libations before deities, semi-divine figures with flowing vases). What I
am suggesting, however, is that once one identifies the textual continu-
ity of claims to and desire for abundance, then the general theme of
an “iconography of abundance” also becomes apparent, with multiple
ways to signal the concept.
This variation in the visual repertoire is parallel to the literary
component, which also employs variant imagery in how abundance
20
The flowing streams with leaping fish are not only a visual attribute of Enki;
they also appear as a literary trope, as in the text “Enki’s journey to Nibru” (ETCSL
t.1.1.4, line 80, accessed 04/17/2004), where it is stated: “When Enki rises (from the
Abzu), fish leap before him like waves.” The Akkadian seals cited in our text, plus those
explicitly representing some sort of grain/vegetation deity (Boehmer 1965: 536–52),
make as clear as the citation above relating the Akkadian ruler Naram-Sîn to Akk.
nu šu, “abundance/prosperity,” that these concepts were not solely Sumerian.
is ascribed and described. In Ur-Namma C, for example, a hymn

celebrating Nippur and its god, Enlil, the king is made to proclaim:
“Anu opened his pure mouth and rain was produced for me, He made
it fall right into the deep earth, and abundance came forth to/for me”
(Flückiger-Hawker 1999: C 20–21). Ur-Namma goes on to state that
he is “the faithful shepherd who has increased his flock” (Flückiger-
Hawker 1999: C 76), declaring his “enormous harvest which raises itself
high” (Flückiger-Hawker 1999: C 80). The hymn ends with libations to
Enlil, and the paean: “thanks to him (Ur-Namma) abundance entered
the Royal Canal and the temple of Enlil” (Flückiger-Hawker 1999: C
103). In Ur-Namma D, the hymn describes the king as having been
provided with broad wisdom by Enki, digging a “canal of abundance”
in his home city of Ur, which then produces plentiful fish, birds, and
plants, as well as watering large arable tracts where barley sprouts like
reed thickets (Flückiger-Hawker 1999: D 21–22; 33'–36').
While royal hymns celebrate abundance, lamentations mourn what
has been lost, thereby providing us with the same value system through
inversion. In the lament known as the “Death of Ur-Namma,” the king
is made to complain, rather like Job: “Despite having made abundance
manifest to the Anuna-gods (that is, having made prolific offerings?) . . . no
god assisted me, nor was my heart relieved!” (Flückiger-Hawker 1999:
A158–160). And yet, the lament ends with the reassurance that the
canebrakes which the king has drained, and the vast barley fields which
he has produced, will ultimately be productive and looked upon with
admiration (Flückiger-Hawker 1999: A227–230). These references in
Ur-Namma A and D may well provide the key to the elusive reverse of
the Ur-Namma stele in Philadelphia, where we know a small fragment
of text makes reference to the canals dug in the reign of that ruler,
while the second register shows elements of a canebrake (Canby 2001:
pl. 11). A variety of other laments, including The Lamentation over the
Destruction of Sumer and Ur (Michalowski 1989) and several of the texts
known as the Canonical Lamentations (M. Cohen 1988: 511, line 1; 589;
637, lines a+11 and 12) equally provide the inverse of abundance in
the negative tropes of floods that drown the harvest, sheep and goats
that cry out, and forests and orchards that have been destroyed. The
catalog of disasters in the Lamentation over the Destruction of Sumer and Ur,
for example, includes the loss of cattle, palm trees, and fertile reeds
(Michalowski 1989: lines 411–418); while the more encouraging res-
toration promised by Enlil at the end envisions full rivers, good rains,
abundant grain, fish and fowl in the marshes, plentiful reeds, and
fruit-bearing trees in the orchards (Michalowski 1989: 498–505)—in

short, along with plentiful herds, all that a good agrarian state needs
to be fully functional.
The visual repertoire, despite its variations, was more limited than the
textual; but I would argue that paired tropes of grain and flax, male
and female sheep/goat/cattle, domesticated plant plus animal, along
with libations into palm plants and flowing streams, all serve iconi-
cally to reference the same semantic range of the productivity of the
land. Whether this selection is a question of visual economy, with signs
standing for an understood whole, and/or whether it was a symptom
of the separation of the urban community/cult from the rural sphere
(as per Bernbeck and Pollock 2002: 186), thereby reducing the range
of productivity to the level of symbol, the fact remains that the city
was as dependent on agrarian production as was the countryside. The
key to the power of these motifs in the Mesopotamian universe lies
in the very phenomenon that has led to their dismissal by Western
analysts: repetition. If one of the primary characteristics of the fourth
millennium was an “unprecedented focus on duration” (Bernbeck and
Pollock 2002: 189–190), then repeating bands of plants and animals,
especially in a frieze around a circular vessel, represents precisely that
which endures and recurs, that which is not subject to temporal vari-
ability or limitation.
Nor is it surprising that the positive is what is represented visually,
since the objects on which such representations occur are largely cul-
tic, reflecting not so much the “real” natural world, and surely not its
inversion, but rather an “ideal” world order. The works therefore articu-
late the desired best-case scenario, in a repeating, hence infinite, sequence,
leaving the negative for historical accounts and the paradigmatic “Lam-
entation” literature. At the same time, it is important to emphasize that
one must see the visual repertoire as no less rhetorical than the textual.
Abundance is both necessary and desired; and, as on the Warka Vase,
the luxuriance of natural production is what undergirds the ordered
cosmos as well as the urbanized city-state. In this universe, an “icono-
graphy of abundance” lends itself to a “cultural rhetoric of abundance,”
with the agricultural and pastoral landscape as much a component in
the state apparatus, and as political, as the urban landscape.
If (as per A. Smith 2003: 10) the land provides the ground for a
culture’s values and aesthetic response as well as its practical produce,
then what becomes interesting is precisely which aspects of the natural
landscape a given culture or period selects for representational and
symbolic purposes. For the Tigris-Euphrates Valley, from the earliest

phases of the Sumerian city-state, the underlying concept of desired
abundance in the natural world was clearly articulated in text and image
through the representation of often-repeated domesticated plants and
animals, the rivers and the life they sustained, all of which remained
a constant theme for millennia.
In short, what modernity has relegated to “mere ornament” was in
fact highly charged, through the very principle of repetition, to stand as
a sign of eternal and unending productivity—provided by the gods as
part of the cosmic order and actualized through the ruler. In this view,
visual rhetoric no less than verbal rhetoric was likely to have been a
deliberate strategy of the center, and so should not be underestimated
as an artifact of the early state.
Acknowledgments
I am grateful to Jeffrey Hamburger, Robert Hunt, Ann Kuttner, Naomi Miller, Alex
Nagel, Nicholas Postgate, John Russell, Gil Stein, Piotr Steinkeller, and Elizabeth Stone
for stimulating conversations on this topic. Robert Hunt also gave generously of his
expertise and critical acuity in a reading of an early draft.
Bibliography
Abusch, Tzvi. 1995. “Ishtar,” in K. van der Toorn et al., eds., Dictionary of Deities and
Demons in the Bible. Leiden: E. J. Brill, pp. 848–856.
Adams, Robert McC. 1960. “The Origin of Cities,” Scientific American 606, September,
pp. 3–10.
——. 1965. Land Behind Baghdad. Chicago & London: University of Chicago Press.
——. 1981. Heartland of Cities: Surveys of Ancient Settlement and Land Use on the Central
Floodplain of the Euphrates. Chicago & London: University of Chicago Press.
Adams, Robert McC. and Hans J. Nissen. 1972. The Uruk Countryside: The Natural Setting
of Urban Societies. Chicago & London: University of Chicago Press.
Algaze, Guillermo. 1993. The Uruk World System: The Dynamics of Expansion of Early
Mesopotamian Civilization. Chicago & London: University of Chicago Press.
——. 2004. “Trade and the Origins of Mesopotamian Civilization,” Biblioteca Orientalis
61:5–20.
Amiet, Pierre. 1980. La glyptique mésopotamienne archaïque, 2nd ed. Paris: CNRS.
Bahrani, Z. 2002. “Performativity and the Image: Narrative, Representation, and the
Uruk Vase,” in E. Ehrenberg, ed. Leaving No Stone Unturned: Essays on the Ancient Near
East and Egypt in Honor of Donald P. Hansen. Winona Lake: Eisenbrauns, pp. 15–22.
Becker, Andrea. 1993. Uruk: Kleinfunde I—Stein [Ausgrabungen Uruk-Warka, 6]. Mainz
am Rhein: von Zabern.
Bernbeck, R. and S. Pollock. 2002. “Reflections on the Historiography of 4th Millen-
nium Mesopotamia,” in A. Hausleiter et al., eds., Material Culture and Mental Spheres
[AOAT 293]. Münster: Ugarit-Verlag, pp. 171–204.
Boehmer, R. M. 1965. Die Entwicklung der Glyptik während der Akkad-Zeit. Berlin: De Gruyter.
Canby, J. V. 2001. The “Ur-Nammu” Stela [University Museum Monograph 110]. Phila-
delphia: University of Pennsylvania Museum.
Childe, V. Gordon. 1951. Man Makes Himself. New York: Mentor Books.
Cohen, Andrew. 2003. “Introduction to the Culture of Barley,” paper given at the Annual
Meetings of the American Schools of Oriental Research, Atlanta, GA, 11/03.
Cohen, Mark. 1988. Canonical Lamentations of Ancient Mesopotamia, Vol. II, Potomac
MD: CDL Press.
Crawford, Harriet. 1985. “A Note on the Vegetation on the Uruk Vase,” in J. N. Postgate,
ed., Bulletin on Sumerian Agriculture, vol. II. Cambridge/Warminster: Aris & Phillips,
pp. 73–76.
Dolce, Rita. 1989. “Note per una reconsiderazione di alcune strutture de un magazzina-
mento e de lavorazione di derrate nel III e nel II millenio in Mesopotamia e in
Siria,” in R. Dolce & C. Zaccagnini, eds. Il pane del Re: Accumulo e Distribuzione dei
cereali nell’oriente antico [Studi de Storia Antica 13]. Bologna: CLUEB, pp. 17–48.
Edzard, Dietz Otto. 1997. Gudea and his Dynasty [Royal Inscriptions of Mesopotamia:
Early Periods, Volume 3/1], Toronto: University of Toronto Press.
Ertuğ, F. 2000. “Linseed oil and oil mills in central Turkey: Flax/Linum and Eruca,
important oil plants of Anatolia,” Anatolian Studies 50:171–185.
ETCSL: Electronic Text Corpus of Sumerian Literature, Oxford University. Website:
www-etcsl.orient.ox.ac.uk.
Flückiger-Hawker, E. 1999. Urnamma of Ur in Sumerian Literary Tradition [OBO 166].
Fribourg: University Press.
Gansell, Amy. 2004. “Personal Adornment in the 3rd millennium Burials at Ur: Roles
and Interpersonal Dynamics.” Paper given at the Annual Meeting of the American
Schools of Oriental Research, San Antonio TX, 11/04.
Gleason, Kathryn L. 1994. “To Bound and to Cultivate: An Introduction to the Archae-
ology of Gardens and Fields,” in N. F. Miller & K. L. Gleason, eds., The Archaeology
of Garden and Field. Philadelphia: University of Pennsylvania Press, pp. 1–24.
Groenewegen-Frankfort, H. A. 1987 (1951). Arrest and Movement. Cambridge: Harvard

University Press.
Hall, Mark G. 1985. A Study of the Sumerian Moon-god, Nanna/Suen. Unpublished PhD.
dissertation, University of Pennsylvania. Philadelphia.
Hansen, Donald P. 1975. “Frühsumerische und frühdynastische Flachbildkunst,”
Propylaean Kunstgeschichte, vol. 14, ed. W. Orthmann. Berlin: Propylaean Verlag, pp.
179–193.
Heinrich, E. 1936. Kleinfunde aus den archaischen Tempelschichten in Uruk. Berlin: Deutsche
Forschungsgemeinschaft.
Hunt, Robert C. 1987. “The Role of Bureaucracy in the Provisioning of Cities: A
Framework for Analysis of the Ancient Near East,” in M. Gibson and R. D. Biggs,
eds., The Organization of Power: Aspects of Bureaucracy in the Ancient Near East [SAOC
46]. Chicago: The Oriental Institute, pp. 161–192.
——. 2000. “Labor Productivity and Agricultural Development: Boserup Revisited,”
Human Ecology 28/2:251–277.
Jacobsen, Thorkild. 1976. The Treasures of Darkness. NY & London: Yale University Press.
——. 1993. “Notes on the word lú,” in A. F. Rainey, ed., Kinattūt ša dārâti: Raphael
Kutscher Memorial Volume [Tel Aviv Occasional Publications, No. 1], Tel Aviv: Institute
of Archaeology, pp. 69–79.
Kawami, Trudy. 2001. “The Cattle of Uruk: Stamp seals and animal husbandry in the
Late Uruk/Jemdet Nasr Period,” in W. W. Hallo and I. J. Winter, eds., Seals and Seal
Impressions: Proceedings of the XLVe Rencontre Assyriologique Internationale, Part II. Bethesda,
MD: CDL Press, pp. 31–48.
McCorriston, Joy. 1997. “The fiber revolution: Textile extensification, alienation, and
social stratification in ancient Mesopotamia.” Current Anthropology 38:517–549.
Michalowski, Piotr. 1989. The Lamentation over the Destruction of Sumer and Ur [Mesopo-
tamian Civilizations l ]. Winona Lake, IN: Eisenbrauns.
Miller, Naomi. 2000. “Plant Forms in Jewellery from the Royal Cemetery at Ur,” Iraq
LXII:149–155.
Moortgat, Anton. 1969. The Art of Ancient Mesopotamia: The Classical Art of the Near East.
London & New York: Phaidon.
Nissen, Hans J., P. Damerow and R. K. Englund. 1993. Archaic Bookkeeping: Early writ-
ing and techniques of economic administration in the Ancient Near East. Transl. P. Larsen.
Chicago: University of Chicago Press.
Pollock, Susan. 1992. “Bureaucrats and managers, peasants and pastoralists, imperial-
ists and traders: Research on the Uruk and Jemdet Nasr Periods in Mesopotamia,”
Journal of World Prehistory 6:297–336.
Postgate, J. N. 1987. “Notes on Fruit in the Cuneiform Sources,” Bulletin on Sumerian
Agriculture 3:115–144.
—— et al., eds. 1988. Irrigation and Cultivation in Mesopotamia, Parts I & II [Bulletin on
Sumerian Agriculture, vol. IV]. Cambridge/Warminster: Aris & Phillips.
Powell, Marvin. 1985. “Salt, seed, and yields in Sumerian Agriculture. A critique of
the theory of progressive salinization,” Zeitschrift für Assyriologie 75:7–38.
Rothman, Mitchell S. 2004. “Studying the Development of Complex Society: Meso-
potamia in the Late Fifth and Fourth Millennia BC,” Journal of Archaeological Research
12:75–119.
—— ed. 2001. Uruk Mesopotamia and its Neighbors: Cross-Cultural Interactions in the Era of
State Formation. Santa Fe: School of American Research Press.
Sallaberger, Walther. 1993. Der kultische kalender der Ur III-Zeit. Berlin & New York: de
Gruyter.
Schauss, Hayyim. 1973. The Jewish Festivals: History & Observance. New York: Schocken
Books.
Sherratt, A. G. 1980. “Water, soil and seasonality in early cereal cultivation,” World
Archaeology 11:313–330.
Sjöberg, Åke. 1975. “In-nin ša3-gur4-ra,” ZA 65:161–253.

——. 1988. “A Hymn to Inanna and her Self-Praise,” JCS 40/2:165–186.
Smith, Adam T. 2003. The Political Landscape: Constellations of Authority in Early Complex
Polities. Berkeley: University of California Press.
Stein, Gil J. 1999. Rethinking World-Systems: Diasporas, Colonies, and Interaction in Uruk
Mesopotamia. Tucson: The University of Arizona Press.
Strommenger, Eva. 1964. 5000 Years of the Art of Mesopotamia. New York: Abrams.
Szarzynska, Krystina. 1987–8. “Some of the Oldest Cult Symbols in Archaic Uruk,”
Jahrbericht . . . Ex Oriente Lux 30:3–21.
——. 1997. Sumerica: Prace sumeroznawcze, Warsaw: Wydawnictwo Akademickie.
Westenholz, Joan. 1983. “Heroes of Akkad,” JAOS 103:327–336.
Wilkinson, T. J. 2003. Archaeological Landscapes of the Near East. Tucson: University of
Arizona Press.
Winter, Irene J. 1976. Review of H. A. Groenewegen-Frankfort, Arrest and Movement,
re-issued in 1974, in JAOS 78:505–506.
——. 1983. “The Warka Vase: Structure of art and structure of society in the Uruk
Period.” Paper given at the Annual Meeting of the American Oriental Society,
Baltimore MD, March 1983.
——. 2000. “Opening the Eyes and Opening the Mouth: The Utility of Comparing
Images in Worship in India and the Ancient Near East,” in M. W. Meister, ed.,
Ethnography & Personhood: Notes from the Field, pp. 129–162. Jaipur and New Delhi:
Rawat.
——. 2003. “Ornament and the ‘Rhetoric of Abundance’ in Assyria,” Eretz-Israel
27:252–264.
Wright, Henry T. 1994. “Pre-state political formations,” in Gil Stein & Mitchell
Rothman, eds, Chiefdoms & Early States in the Near East: The Organizational Dynamics of
Complexity. Madison WI: Prehistory, pp. 67–84.
Zettler, Richard L. and Lee Horne, eds. 1998. Treasures from the Royal Tombs of Ur.
Philadelphia: University of Pennsylvania Museum.
Figure 1. Alabaster trough, Uruk period, probably from Warka; height 16.3
cm.; length 1.3 m. (British Museum, London, WAA 120000).
Figure 2. Stone bowl fragment, Uruk period (photograph courtesy Département

des Antiquités Orientales, Musée du Louvre, AO 8842).
Figure 3. Drawing, stele fragment, Uruk period, found Warka; height 79 cm.
(after Becker 1993, #785; Iraq Museum, Baghdad, IM 6842).
Figure 4. Stone bowl, Uruk period, found at Ur; height 5.5 cm. (Iraq Museum,
Baghdad, IM 11959).
Figure 5. Cylinder seal impression, Uruk period (photograph courtesy

Département des Antiquités Orientales, Musée du Louvre, MNB 1906, A25).
Figure 6a & b. Alabaster vase, Uruk period, found at Warka; height 1.10 m.
after restoration (Iraq Museum, Baghdad, IM 19606).
Figure 7a & b. Detail, lower portion, alabaster vase.

Figure 8. Silver vase of Enmetena of Lagash, Early Dynastic III period, found
at Tello (photograph courtesy Département des Antiquités Orientales, Musée
du Louvre, AO 2674).
Figure 9. Cylinder seal impression, Early Dynastic III period (photograph courtesy
Département des Antiquités Orientales, Musée du Louvre, AO 10920).
Figure 10. Jewelry assemblage from PG 1237, Early Dynastic III period, found
at Royal Cemetery, Ur (photograph courtesy The University of Pennsylvania
Museum).
Figure 11. Basalt bowl fragment, dedicated by Enmetena of Lagash to

Nin ursag; height 25 cm. (photograph courtesy Staatliche Museen zu Berlin,
VA 3164).
Figure 12. Fragments Stele of Ur-Namma, Third Dynasty of Ur, second

register of obverse, found at Ur (photograph courtesy The University of
Pennsylvania Museum, CBS 116676) 12.
CHAPTER TWENTY-FIVE
READING RITUAL IN THE ARCHAEOLOGICAL RECORD:

DEPOSITION PATTERN AND FUNCTION OF TWO
ARTIFACT TYPES FROM THE ROYAL CEMETERY OF UR
Since its discovery in 1927, the series of graves designated the “Royal
Cemetery” at Ur have been recognized as among the richest and most
important finds of the ancient Near East, occasioning studies on topics
ranging from chronology and iconography to wealth and social status.
The graves were found at the southeastern end of the later temenos
enclosure that marked the sacred precinct of the moon god, Nanna/
Su’en, not far from the palace of Ur-Namma (ca. 2110 b.c.), first ruler
of the Third Dynasty of Ur (Woolley 1934: 5). The excavator, Sir Leon-
ard Woolley, recognized early that there were burials of more than a
single period in the area (Woolley 1934: 20ff.); subsequently confirmed
by H. J. Nissen’s study of the typology and relative chronology of the
graves (Nissen 1966).
The “Royal Cemetery” includes over 500 graves with significant
assemblages (Schwartz 1986: 54). The vast majority, both those des-
ignated “royal” and the so-called “private graves,” have been identi-
fied as belonging to the later part of the Early Dynastic Period, ED
IIIA, dated to ca. 2600 b.c. There are also a number of later burials
belonging to the Akkadian period that generally, but not consistently,
lie higher in the fill. The ceramics of the Early Dynastic burials have
served as type-markers for other assemblages (e.g., Kish: Moorey 1970,
1978). Internal to Ur, the studies of Susan Pollock have tested Nissen’s
chronology, and attempted further to measure a variety of social fac-
tors, such as relative economic status and embedded attitudes toward
gender and the provisioning of the dead (Pollock 1983, 1991). One of
the questions asked by Pollock, as had been previously asked by Rathje
(1977) and Moorey (1977), was whether the artifactual assemblage could
* This article originally appeared as “Reading Ritual in the Archaeological Record:

Deposition Pattern and Function of Two Artifact Types from the Royal Cemetary of
Ur”, in Fluchtpunkt Uruk: Archäologische Einheit aus methodischen Vielfalt, Schriften für Hans Jörg
Nissen, H. Kuhne et al., eds., Espelkamp: M. Leidorf, 1999, pp. 229–56.
228 chapter twenty-five
be used to ascertain the specific social roles of the individuals buried

in the individual graves? Other approaches have recently been taken
by Michelle Bonogofsky (1994) and Silvia Maria Chiodi (1994b), who,
in preliminary studies of the grave goods accompanying burials, have
used literary evidence to suggest that at least some of the material
could have represented not just the provisioning of the deceased, but
gifts intended for identifiable inhabitants of the Underworld;1 and by
Luc Bachelot (1992), who has suggested that grave goods provision the
deceased to keep the otherwise-potentially malevolent spirit fixed in its
proper domain (see also on this, Cooper 1991: 28). Presently, Andrew
Cohen is pursuing the broad cultural context of mortuary ritual in the
Early Dynastic Period, particularly that associated with elite graves and
the political process of state formation (Cohen, forthcoming).
In many ways, the present paper dovetails closely with the work of
Cohen. It is my suggestion that there may be untapped sources of evi-
dence in the grave goods—not only in presence::absence (as Beate Salje
(1996) has suggested for a class of late Babylonian beakers/drinking sets
she believes to have been uniquely associated with burial cults), but in
the actual depositional patterns of the grave goods. I argue that such
patterns may in fact reflect ritual acts connected with the moment of
burial itself. I should like to present this study as an inquiry into the
potential of such evidence, in keeping with the nature of the present
volume as a series of reflections on scholarly activities, raising questions
of methodology as well as the interpretation of data.
To date, most reconstructions of ritual activity in the ancient Near
East have been grounded in either textual evidence or iconographic
analysis: the reading of explicit narrative or explicitly illustrative imag-
ery. One can, however, begin from a different set of propositions more
relevant to archaeologists. First, following the arguments of Michael
Schiffer and others (e.g., Schiffer 1996/1987) that identifiable factors
account for the make-up of the archaeological record in general, when
it is possible to demonstrate pattern in the archaeological record, some
sort of meaningful behavior has gone into producing that pattern. And
1
In this, Bonogofsky has relied heavily on literary texts such as the account of the
“Death of Ur-Nammu,” in which the king is described as presenting a wide array
of gifts, including artifacts not unlike those found in the Royal Cemetery graves, to
notables upon his arrival in the Underworld. It is assumed, therefore, that the text of
the end of the third or early second millennium b.c. would be applied to the social
and cultural situation some 600 years earlier.
reading ritual in the archaeological record 229
second, that ritual, which in the abstract is charged with transformative

power (Watson 1988: 4) and the integration of ethos and world-view
(Bell 1992: 26), and in practice is marked as behavior by “conventional
actions performed at regular intervals fixed by clock, calendar, or cir-
cumstance . . . .” (Rappaport 1971, emphasis mine), constitutes one of the
cultural activities that can, by virtue of its conventions, generate patterns
and thus reveal meaning. Ritual performance would then be understood
as one among many factors leaving material traces as depositional phe-
nomena—whether in the form of organic materials retrievable through
scientific analyses, and/or in the form of artifactual remains.
The reflections that follow may be divided into two sections: those
related to patterns observable from the deposition of specific grave goods
within burials that could permit us to read ritual activity from the
archaeological record; and those related to possible function, buttressed
by cross-cultural analogy as a tool in the interpretation of grave goods as
the residue of such ritual activity. Implicit throughout is the assump-
tion that the material assemblage of a tomb is to be distinguished from
that of a living space, its artifacts susceptible not merely to typological
classification or analysis of value, but also likely to reflect symbolic or
culturally meaningful behavior. Thus, it may be possible to read more
symbolic or culturally meaningful behavior from the archaeological
record of burials than has been understood to date.
Deposition of Grave Goods
The issue of ritual performance in general has long engaged anthropolo-

gists, and is currently the subject of a good deal of scholarly attention
(Smith 1987; Cannadine and Price eds., 1987; Kertzer 1988; Cederroth
et al. 1988; Bell 1992); while mortuary practices and attitudes toward
death have frequently engaged both archaeologists and anthropologists
(Saxe 1970; Binford 1971; Metcalf 1981; Humphries and King 1981;
Bloch and Parry 1982; Scarre 1994). The field of the ancient Near East
has not been slow to pick up on these interests (e.g., Alster 1980; Winter
1992; Campbell and Green 1995; Bremer et al. 1994; Chiodi 1994a
and 1994b; Scurlock 1995). But seldom have the two, ritual and the
archaeology of mortuary practice, been explicitly connected. Mortuary
practice is most often confined to descriptions of tomb construction,
burial position, provisioning and accompanying grave goods. Dan Potts
(1997), for example, in his recent discussion of Mesopotamian burial
tradition(s), makes reference to the physical contents of burials, speaks

of textual evidence concerning the netherworld and the need for food
offerings, plus commemorative activities after death, but does not pursue
the topic of ritual performance at the time of death/interment; Jean
Bottéro (1998), in a broad view of Mesopotamian religion, discusses
death in terms of spirits/ghosts of the deceased and the netherworld, but
includes nothing on funerals.2 Notable exceptions are articles by Lynn
Bright (1995) in which ritual practice—rites, paraphernalia, ideology
and religion—is included, along with grave offerings, tomb typology,
etc. in her model for the study of the “archaeology of death” in gen-
eral; by John C. Barett (1996), in which he defines the “archaeology of
ritual” as comprised of “actions, not things;” and by Jo Ann Scurlock,
in an essay on Mesopotamian attitudes toward death and the afterlife
(1995), in which the author distinguishes between burial rites, post-burial
mourning rites and the full-blown post-interment cult of the dead that
includes continuous offerings. The dissertation of Cohen, cited above,
will pursue this issue further, putting together a wide range of textual
and archaeological evidence toward a nuanced understanding of Meso-
potamian funerary practices and their socio-cultural implications in the
third millennium, b.c.e.—comparable to studies already undertaken
with respect to other archaeological assemblages (e.g., that of ancient
China: Watson 1988; Watson and Rawski 1988; Rawson 1993 and
forthcoming; and the articles contained in Kuwayama 1991) and as
associated with a number of living cultural traditions (e.g., Danforth
1982; Metcalf and Huntington 1991).
A number of prior conditions contribute to the likelihood that
there would have been some ritualized behavior associated with the
2
See especially, Potts 1997: Ch. X (Mortuary Practice), 220–235. Similarly, For-
est 1983: especially 144–147, where he discusses the symbolic character of funerary
practices by inquiry into “traits” such as the disposition of the body, type of tomb,
furniture/grave goods, locale of the tomb, but nothing on traces of ritual performance;
also Byrd and Monahan 1995: 251–287, in which, despite the promising title, burial
data include interment type, grave construction, body orientation, and grave goods, not
ritual, and are primarily used to elucidate social status. A similar range of variables is
considered by Carter and Parker 1995: 96–117, and indeed in almost all of the articles
contained in that volume (e.g., Baker, 1995: 209–220). Carter and Parker (1995: 103) cite
Ian Morris (1992: 24–29) for the aspects of data involved: burial typology, chronology,
grave goods/context, spatial arrangement of burials to settlement, demographics. They
do not discuss any evidence for ritual performance associated with burial, although
Morris himself (1992: 11), notes that ancient ritual can be studied from a variety of
sources, amongst which: “the material remains of rituals.”
interments in the Royal Cemetery. First, the conclusion presented by

Moorey (1977) that at least some of the burials in the “Royal Cem-
etery” belong to individuals associated with the cult of Nanna/Su’en
makes it not unlikely that cultic (i.e., ritual) practice, known to have
played a significant role in the daily life, would equally find expression
at the juncture between the domain of the living and the domain of
the dead. Second, evidence presented by Weadock (1975) that by the
early second millennium there was a well-established cult commemorat-
ing deceased entu-s, priestesses of the cult of Nanna, led her to posit
an earlier commemorative cult as well, making the probability even
higher that cultic acts would also be performed at the time of inter-
ment. Third, the fact of the human entourage accompanying some of
the major burials, especially PG/1237, “The Great Death-pit,” already
indicates a choreographed “event” at the time of burial (Woolley 1934:
113–124; Chiodi 1994b), making ritual activity even more likely.3 And
finally, given the malevolent power that can be shown to have been
associated with dissatisfied ghosts of unburied or untended dead (Cooper
1991; Abusch 1995), logic suggests that the well-attested post-interment
commemorative rituals would have been preceded by at least some
appropriate grave-side protocols.
This chain of reasoning leads to a different position from that of
Jonker (1995: 192), who initially noted in her discussion of the Death
Cult and family memory, that “the relationship between the (post-inter-
ment) rituals and real funerals is difficult to establish;” and, because
“relatively little is known about how people took their final leave of
their dead or how they were buried,” concluded that “there seems to
have been little ritual associated with the moment of death.” What I
would say, in contrast, is that while this moment seems not to have
been recorded in texts, upon which her study was based, this absence
does not preclude such activity, either at the literal moment of death
or subsequently at the burial site.
Issues of ritual intersect with questions of deposition when the
“types” of objects/material remains found in the tomb can be shown
3
To distinguish the “event” itself from activities associated with it, one must press the
distinction between “ceremonial” and “ritual”—a careful analysis of which is beyond
the purview of the present paper. It is not a trivial distinction, however: the one being
defined more by culturally signifying behavior, such as court processions but not nec-
essarily marked by religious significance the other more often attended by specialized
knowledge and practitioners, such as priests, and marked by culturally-sanctioned acts
of mediation between the known world and powers operative beyond its dimensions.
to have been associated with particular activities—i.e., not merely what

(markers of wealth, chronology or identity), but how objects/materials
were deposited/deployed. Two particular vessel-types found in a num-
ber of the Ur burials lend themselves to depositional analysis: shallow
containers called “lamps” by Woolley; and spouted vessels, generally
of metal, referred to as “libation jugs,” usually accompanied by broad,
shallow basins. I shall argue that these vessel-types may represent neither
potential gifts nor personal provisions, but rather the functional residue
of ritual performances associated with the moment of burial.
“Lamps”
A substantial group of small shallow vessels has been identified as
“lamps”, although they show no signs of soot or residue from burning.
The group includes objects of roughly similar shape executed in shell
and alabaster, as well as in gold, silver, copper (Woolley’s metal vessel
Type 115). The shape is characterized by an open channel and bowl
that would hold approximately 6 ounces of liquid by modem measure
(our figs. 1 and 2).4 The shell examples (on the importation of which,
see Aynard 1966 and Edens 1992) were often cut open with some of
the whorls removed in order to enlarge the natural orifice. Woolley
4
See Woolley 1934: Pls. 101 and 102: U.8191 (PG/143), U.8198, and U.8313,
discussed p. 283, = “shells cut to serve as lamps”, with incised bird-beak decoration
around the point of issue, or with bird-head protome and inlaid collar; Pl. 163: U.10004,
silver, from PG/755, attributed to Meskalamdug; Pl. 164: U.10451, discussed p. 282,
from PG/800, tomb of Pu’abi, gold, with shallow bowl and spout; Pl. 170: U.9364,
discussed p. 293, silver, found embedded in a group of similarly-shaped and nested
copper vessels from an unnamed grave that were badly corroded and dissolved away
in order to preserve the silver example; Pl. 173: U.10886, from PG/800, silver, shallow
angular “bowl” and open spout; and Pl. 182: U.10746 from PG/871, discussed pp.
183 and 377 and U.11795 from PG/1134, discussed p. 377, both of alabaster (white
calcite). See also references in the catalogue to unillustrated or unregistered objects of
the class: shell “lamps” from tombs PG/36, 71, 149, 159(2) 169, 177, 300, 324, 335,
(U.8679), 350, 353, 357, 379, 391 (U.8661), 543, 677, 1076 (U.11597), 1180, 1237,
1277 (U.12167), 1321, 1325, 1329, 1337, 1382, 1537, 1591, 1609, 1618, 1625, 1642,
1648, 1650, 1655; silver “lamps” of metal Type 115, from PG/429 (U.8902), 755
(U.10024), 789 (U.10974), 800; copper “lamps” of Type 115 from PG/202, 245, 250,
263, 317, 333, 337, 758, 778, 789, 800, 1038, 1054, 1130, 1236, 1312, 1320, 1321,
1327, 1385, 1392, 1407, 1641; and an alabaster “lamp” from PG/1266 (U.12135).
Without going through the field registers, therefore, one comes to a count of some 72
vessels from 62 different graves: 38 of shell, 23 of copper, 7 of silver, 3 of alabaster
and 1 of gold. The metal examples constitute Müller-Karpe’s “Form-type 8,” and are
also known from Tello, Kish, Tell Asmar and Abu Salabikh in this Period (1993:36–46
and Pls. 9–19).
suggested that these represented the original form from which the
similarly shaped vessels in gold, silver and copper, “so common in the
cemetery,” had been copied. Their designation as lamps was surely due
to a similarity in shape to later Palestinian and Roman lamps, with
the trough/channel serving as a bed for a wick (e.g., Amiran 1969: Pl.
59)—a connection that had been made by Banks at Bismaya (ancient
Adab) a generation earlier, when he, too, found cut conch-shells along
with other Early Dynastic period objects in a temple dump (1912: 271),
and has been followed more recently by Homès-Fredericq (1980–83)
and Scurlock (1995). However, Woolley himself acknowledged that the
shells “exhibited no sign of burning in the trough-like tip where the wick
would naturally lie,” and observed judiciously “that this is against the
identification of the metal vessels as lamps.” Indeed, he noted that the
shape could have been used “for some kind of libation, or for pouring
unguents over the deceased” ( Woolley 1934: 283).
Nevertheless, in the end, Woolley concluded that the shape suggested
“a lamp rather than anything else;” and from that point on, in cap-
tions and in discussion, the objects were generally designated as lamps
without benefit of quotation marks or qualification. The fact that they
show no sign of burning is reconciled by the rationale that there was
“no reason to suppose that the lamps would have been lighted in the
tomb, or that a used lamp would have been a better offering than a
new one.”
The lack of evidence of burning in these objects gives me greater
pause than it did Woolley, as it has equally concerned Postgate (1977:
273) and Müller-Karpe (1993: 44).5 If, on the one hand, “new lamps”
were intended to be saved in pristine condition for use in the Neth-
erworld, for example, then a supply of oil, some associated vessel, or
strategic placement—in a niche, say—should be observable; and in any
case, this projection of an unused lamp, or indeed any lamp, is unsup-
ported by textual data where descriptions of descent into the world of
the dead have been preserved (e.g., Kramer 1991a). If, on the other
hand, “lamps” had been incorporated into interment rituals, then they
should have been lit, and show signs of usage.
5
Two examples of conch shells have been found in ED graves 1 and 224 at Abu
Salabikh as well, and neither shows any sign of burning (Susan Pollock, personal com-
munication, gratefully acknowledged).
Woolley did suggest that some rites were performed at the time of
and immediately after burial. For the former, his discussion is largely
confined to the probability of food offerings in vessels included in
virtually all graves, or to how the sedation/poison was likely to have
been administered in the various death pits containing an entourage
of individuals accompanying the deceased. For after-burial rites, he
noted that in certain of the grave shafts there was evidence of surfaces
bearing traces of activity within the fill: in some cases, the deposit of
what he understood to be food offerings (e.g., PG/1050, 1054, 1237);
in others, evidence of what he called “drink offerings poured to the
dead” (Woolley 1934: 36–7; the general practice discussed more recently
by Bayliss 1973; Cassin 1982; Tsukimoto 1985). Woolley also noted
drains and pipes in several of the more important tombs, including
PG/1054, 789, 337 and 1237, which were thought to be conduits for
these liquid offerings, either to the burial itself below the fill, or directly
to the Underworld.
In addition, Woolley noted how frequently the deceased seemed to
have been buried with an open bowl of either copper or limestone/cal-
cite in the hand—particularly in the “Great Death Pit,” where at least
50 were found associated with the 74 bodies. It was suggested that each
individual probably held such a bowl, the missing vessels accounted for
by decay—limestone reduced to patches of white lime, copper reduced
to green powder (Woolley 1934: 120–21, 134, 150, 166).
If one understands the provisioning of the deceased as “patterned
behavior set by convention,” then, by our definition above, it may be
viewed as consistent with ritual activity, just as the deposition of offerings
and the performance of libations in the shaft as the grave was filled
in would be seen as ritual closure mirroring post-interment rites. And
if then the bowls in hand imply some expectation that the deceased
would in fact be viewed as provisioned, or feed him/herself, or perform
some act following interment, then these observations taken together
contribute to the likelihood that there would indeed have been some
ritual activities at the time of interment and so set the stage for an
inquiry into the objects designated as “lamps.”
At least 72 such objects are recorded in the final publication (see
note above), associated with at least 62 different graves. Although this
represents nowhere near the entire sample of Early Dynastic graves, nor
the same quantitative presence as the open bowls, it is still an impres-
sive sample—over one-third of the graves containing precious metal
(Schwartz 1986: 54). Unfortunately, a number of the graves in which
these objects were found were disturbed in antiquity (e.g., PG/789,

1236), and in some cases, although the presence of a “lamp” is noted
in the catalogue, the object is neither described in Woolley’s narrative
nor illustrated in tomb plans (viz., PG/779, 1618). Therefore, it is
not possible to produce a satisfying statistical or narrative description
of either associations or deposition from the published report alone.6
Nevertheless, some observations can be made that suggest patterning
in deposition, which in turn may help to shed light on their function,
or at least lead to a range of possibilities plus suggestions for data-
gathering in the future.
1. Above the floor, in fill: In at least three cases an object of this type
was found not directly upon the floor or a surface within the burial, but
in the upper fill, however in such a context that the deposition seems
undisturbed. In PG/789, a silver vessel of Type 115 was found above the
floor level of the burial pit, along with a silver bowl and a white calcite
vase. Woolley says explicitly that it “may have been put in as offerings
as the filling-in began” (1934: 71). In the discussion related to PG/871,
thought to be slightly later than the main body of Early Dynastic buri-
als, Woolley also noted that a white calcite [ X ], was found in the fill
above the burial. A second calcite vessel, associated with PG/1134, was
similarly found in the fill of the shaft, about a meter above the body
(Woolley 1934: 377–8). With respect to the object in PB/871, Woolley
first speculated that the placement could have been accidental, then
observed that its style was consistent with the date of its associated
grave, so therefore no disturbance of an earlier grave could account
for it as a floating object; he finally came to the conclusion that it must
have been “an offering flung into the grave-shaft while it was being
filled with earth—a practice of which we have several instances . . . ”
(Woolley 1934: 183). While he does not speculate further, the choice
of descriptive vocabulary—“flung into the grave-shaft . . .”—suggests
a relatively random and last minute act, despite Woolley’s references
to other similar instances. Given his reading of other ritual activities
related to libation pipes and drains in fill above various burials, noted
6
Of the 62 tombs, the “lamps” of only 16 are discussed or plotted: tbs. 263, 337,
543, 755, 789, 800, 871, 1054, 1130, 1157, 1236, 1237, 1266, 1312, 1407, 1648.
[Field records for the remaining 46 should be examined, which is beyond the scope
of the present study: e.g., tbs. 36, 71, 143, 149, 159, 169, 177, 202, 245, 250, 300,
317, 324, 333, 335, 350, 353, 357, 379, 391, 429, 677, 758, 778, 779, 1038, 1076,
1180, 1277, 1320, 1321, 1325, 1327, 1329, 1337, 1385, 1392, 1537, 1591, 1609, 1618,
1625, 1641, 1642, 1650, 1655.]
earlier, the casualness attributed to the deposition of these objects is

difficult to understand. It seems far more likely that, like the evidence
for the repeated pouring of libations, these objects in the upper shaft
or fill relate to some specialized, ritual activity performed in the process
of covering the burial.
2. On or immediately adjacent to the body: Müller-Karpe notes 10 instances
in which objects of the type under discussion were in direct proximity to
the skeleton of the deceased, including examples from the A cemetery
at Kish and from Abu Salabikh (Müller-Karpe 1993: 44; Martin et al.
1985: 26–30, Grave 1). In PG/800 (grave attributed to Pu’abi), a silver
container of this type was lying across the body on the pelvis (Woolley
1934: 89 and Pl. 36). In PG/755 (tomb associated by inscribed mate-
rials with Meskalamdug, our fig. 3), a gold [ X ] was found within the
coffin, along with a gold bowl, near the right shoulder of the deceased;
at his knees another two silver examples, one within the other (Wool-
ley 1934: 155–160, Fig. 35 and Pl. 163). None was found among the
vast array of vessels and weapons outside the coffin, so, as in the case
of PG/800, the placement in proximity to the body is likely to have
been purposeful. In PG/263, a clay coffin, a copper [ X ] of Type 115
was found lying directly with a white calcite vase at the elbow of the
deceased, while at the hand was a small copper bowl (Woolley 1934:
150 and Fig. 29). In PG/1312, also within the coffin, a copper [ X ] was
found “in front of the body”—but from plan, actually, just above the
ribcage between the shoulder and the right hand (Woolley 1934: 173
and Fig. 48; our fig. 4); and in PG/1407, a copper [ X ] was discovered
in front of the forehead of the skeleton (Woolley 1934: 177 and Fig.
52). Finally, in PG/543, a wickerwork coffin, a cut conch was placed
near the feet (Woolley 1934: 152 and Fig. 31); while in PG/1266, a
simple pit, an [ X ] of white calcite is the only object set at the feet of
three bodies (Woolley 1934: 172 and Fig. 47).
This pattern of direct association of our vessel-type with the body is
even found in the two wealthiest and best preserved burials: PG/800
and PG/755, which also included untouched skeletons and a full array
of grave goods. From the placement, two possibilities emerge: one, that
some activity was performed with/through the vessel by someone at the
time of interment, then the object set down/left with the deceased; or
two, that it was placed unused in the hand of the deceased, and has
since fallen close to the body. Placement at the feet in two instances
would seem to favor the first possibility over the second, as would the
far more numerous instances in which it is a limestone or metal bowl
that is found in immediate proximity to the hands, suggesting that, if any-

thing, it was the holding of a bowl that was significant (e.g., PG/1130,
Woolley 1934: 166 and Fig. 42).
3. Privileged north or northwest location: In eight cases of undisturbed
burials, there seems to have been a privileging of the north or northwest
for placement of these objects, and often when not, it is to be found in
whatever direction the feet are oriented. In PG/337, a copper vessel
of this type was found in the NW corner, inside a copper bowl (Wool-
ley 1934: 44 and Fig. 1). In PG/800, along with the silver [ X ] lying
across the skeleton, a second silver [ X ] was placed in the NW corner
of the tomb chamber (Woolley 1934: 91 and Pl. 36). In PG/1648, a
large conch “cut as a lamp” was found along the NW wall, although
the tomb had been looted Woolley 1934: 134 and Fig. 26). Similarly, in
PG/789, the second of two copper [ X ] was found along the NW wall,
but in a what was probably a disturbed context (Woolley 1934: 71). In
PG/1054, a copper [ X ] was found with Burial “C” in the northern
corner of the coffin (Woolley 1934: 99–100 and Fig. 15). In PG/1130,
a pit containing a wooden coffin holding the skeleton of the deceased,
a copper [ X ] was found outside the coffin grouped with and above
a cylindrical copper pot and ostrich-egg vase in the northern corner
(Woolley 1934: 166 and Fig. 42; our fig. 5). In PG/1407, where a copper
[ X ] was found at the forehead of the skeleton, its position is toward
the northern corner (Woolley 1934: 177 and Fig. 52). And in the “great
death pit” of PG/1237, three of the four conches are aligned along
the northern edge of the pit (Woolley 1934: Pl. 71).
This positioning is sufficiently consistent that, despite the small sample
size, it can be suggested to reflect a meaningful deposition pattern.
4. Association with vase and/or bowl: In at least nine instances, a Type
115-shaped object was found along with an open bowl, a vase, or
both. As noted above, in PG/789, the silver [ X ] (U.10463) found in
the fill above the pit floor was discovered along with a silver bowl and
a white calcite vase. In PG/800, a white calcite vase was found on the
bier with the lady Pu’abi, along with our vessel; while in PG/1054,
the copper [ X ] is found with a copper bowl and two vases of white
steatite. In PB/1648, the conch “cut as a lamp” that was found along
the NW wall of the tomb chamber, was adjacent to a hemispherical
copper bowl (although the tomb had been looted, the coffin largely
empty, and this may not have been the original position or association
of the artifacts). In PG/263, the copper [ X ] was found lying with a
white calcite vase within the clay coffin, at the elbow of the deceased.
In PG/543, the wickerwork coffin, the cut conch, here designated “to
serve as lamp or ladle,” was found along the northern edge, along with
a copper bowl. In PG/1130 (our fig. 5), the copper [ X ] found outside
the wooden coffin, in the northern corner of the pit, was one of a group
of artifacts including weapons; in immediate proximity to it lay a
copper pan/bucket and ostrich-egg vase. In PG/1312 (our fig. 4), the
only other grave goods within the coffin along with the copper [ X ]
set on the body were the remains of a headdress near the head, an
ostrich-shell in the western comer and a calcite vase in eastern corner.
In Pg/1407, the copper [ X ] placed directly in front (to north) of the
skeleton’s forehead was set inside a white calcite bowl.
The case of the “Great Death Pit,” PG/1237, is more complex.
Woolley’s description of the attendants’ bodies includes four instances
in which a cut conch shell is recorded in association along with a shal-
low bowl (Woolley 1934: 117–121 bodies no. 10, 12, 20, 66). While at
first this seems a promising association, it is later qualified by Woolley’s
observations that the bodies were so close together, often overlapping,
that one cannot be certain that artifacts were associated with the correct
body; and although some bodies are described as associated with both
copper and limestone bowls (e.g., nos. 25, 27, 32, 43, 44, 50, 53), in all
probability, each body was provided with just a single bowl, whether of
stone or of metal (Woolley 1934: 120–121). Within this burial context
the four conches seem to be in free variation: indeed, those associated
with bodies nos. 10, 12 and 20 may not have been associated with the
actual bodies at all, but rather clustered in proximity to the northern
edge of the pit; the one with body no. 66 is set at the western edge
(Woolley 1934: Pl. 71). In any event, the sample differential between
the large number of open bowls and the small number of Type 115
objects does not permit us to draw any conclusions about their mean-
ingful association.
The other cases are, however, more suggestive—particularly PG/789,
where the Type 115 vessel is found together with both open bowl and
vase, and nothing else, in fill above the grave, where some sort of
ritual activity is likely. A pattern of deposition emerges that at the very
least suggests a more purposeful gesture than Woolley’s “flung” lamp.
Whether in grave shafts or in the burial itself, these objects seem to
have been placed with enough consistency to argue for the “patterned
behavior” that belies meaningful activity—a necessary, if not sufficient
condition for reading ritual activity. Proximity to the body and/or
association with some aspect of north or northwest positioning is further
suggestive. And finally, immediate association with other vessel types,

one open, one closed, could imply a functional constellation.
To summarize: of the sixteen burials published with sufficient descrip-
tion to locate conches or metal objects of Type 115, three show the
vessels in the upper fill. Of the thirteen burials remaining, seven show
the vessel lying across the body, in immediate proximity to the body, or
at the feet. Eight burials privilege the north/northwest for placement,
whether within a coffin or in the chamber; one of which combines
proximity to the body with placement in the north corner. All such
vessels that are not in direct contact with the body of the deceased
maintain a consistent pattern of orientation. And nine of the sixteen
contexts show association with open bowls and/or small vases.
“Libation Jugs”
A series of relatively tall, long-necked metal vessels with narrow upward-
pointing spouts issuing from the body—Types 82, 83, 84, depending
upon the proportions of body-to-neck—have been designated “libation
jugs,” indicative of a ritual pouring of liquids (Woolley 1934: 302 and
Pl. 238; = Müller-Karpe 1993: 13–17, “Form-type 2,” and Pls. 1–6,
including additional examples from Khafaje Sin Temple IX, Kish and
Tello; and see also discussion in Spycket 1968: 21). Of these, Woolley
considered Type 84 (our figs. 6 and 7) the “regular vessel for libations.”
As an illustration of this function, he referred to a large engraved shell
plaque found at Ur, on which a nude male (presumably a priest) holds
just such a vessel in one hand, as if about to pour an oblation before
a standard held in a bull-footed base (Woolley 1934; 282 and Pl. 102,
unfortunately found on the surface, but stylistically ED III in date). To
this may be added an Early Dynastic III limestone wall-plaque from
Tello (our fig. 8) that shows a libation being poured from a spouted jar
of the same type into a concave-sided goblet or vase containing a min-
iature palm tree set up before the goddess Ninhursag (and see additional
examples in Müller-Karpe 1993: Pls. 170–171). From the plaques, there
is no question that the vessel can be used for ritual libation, directly
connected to the worship of deities and/or their emblems.
In discussing the Type 84 vessel, Woolley further noted that it was
“always associated with a paten”—a shallow, broad dish or open bowl
with raised rim (his metal vessel Type 32—see 1934: Pl. 234 for typo-
logy). In all, eight such spouted jugs are recorded in the Royal Cemetery
catalogue and seven Type 32 open bowls (Woolley 1934: 520–521). In
PG/1054, for example, the “set” is found within the burial chamber; the
body is oriented West (head)-to-East, with the vessels, both of copper,
to its north, just on or over the hip (Woolley 1934: Fig. 17, No. 13). In
PG/800, the tomb of Pu’abi, three separate “sets” were noted. At the
north end of the “Death-Pit” dromos of the tomb, just before the wall
of the tomb chamber itself, a collection of 15 silver fluted tumblers were
found with a spouted silver vessel of Type 83 and a silver paten, and
almost immediately adjacent, a collection of copper vessels repeating
the group of silver. Then, within the tomb chamber, a silver “libation
jug” of Type 84, “to which was fastened by corrosion a silver fluted
tumbler and a silver paten of Type 32,” lay among a group of vessels
at the foot of the bier, half-way between the north and the east corners
(Woolley 1934: 81–2 and 90 and Pls. 36 and 171).
In PG/580, two silver jugs of Type 84 were found near an ox skull
in the northwest half of a pit that was not the tomb chamber itself,
but may have been part of a large dromos (Woolley 1934: 46–50 and
Fig. 4). The copper version from PG/1236 was unfortunately found
in the very disturbed tomb chamber, with just a few fragments of
human bone and beads scattered about the floor, so nothing about its
deposition is of use for analysis (Woolley 1934: 113). It is tantalizing,
however, that in tomb 755, attributed to Meskalamdug, a silver vessel
of Type 84 was discovered (along with a copper “tray” of Type 30 that
was a larger version of the “paten”) tucked into the northern corner
of the tomb (Woolley 1934: 159; and see plan, our fig. 3, and vessel
set, our fig. 6).
Thus, while it has been shown by direct illustration that these ves-
sels may have been used for libation, there is nothing in their immediate
proximity in the tombs that would correspond either to divine emblem
or image as on the plaques. Functionally, the spouted jugs can be imag-
ined to have been put to a number of alternate uses, precluding any
specific allocation without further inquiry. Indeed, in both cases, that
of the “lamps” and that of the “libation vessels,” I would argue that
sufficient ambiguity exists to question the functions assigned to each.
Moreover, sufficient consistency in deposition patterns plus associated
objects exists to further suggest that patterned behavior could account
for their placement. Both types, therefore, warrant further discussion
of their possible function(s), as well as possible explanations for their
presence in the tombs.
Cross-Cultural Analogies and Function
Conches and Metal Vessels of Type 115—Reclassified as Pouring Vessels

If the objects under discussion were “lamps”, and their aim was to
provide light at some future point—particularly as they contain no signs
of prior usage—then the accompanying closed vessels should most likely
have contained some sort of oil/fuel; and it is to be lamented that the
analysis of soil samples from the interiors of vessels was not a standard
procedure in excavations of the 1920s (on which, see Ellison et al., 1978).
However, one would also expect that if light were important—e.g. in
passage to the dark netherworld—many more burials would have been
so provisioned. One way out of this conundrum would be to suggest
that containers for the same purpose could also have existed in a mate-
rial originally present but not preserved in the archaeological record;
this argument falters, however, since the most likely non-elite material
for a lamp would surely have been the impermeable, non-consumable
clay, as in the Roman period, and to my knowledge, no clay vessels of
this type have been found.
If not “lamps,” then what? Evidence points to Woolley’s second
alternative: a vessel intended to channel the flow of some liquid, hence
his “ladle,” or even a more purposeful pourer. The open end would
serve not for the laying of a wick, but either as a handle or as a spout
for the liquid.
An inlaid panel from the front of a lyre found in PG/789 provides
strong support for such an interpretation (Woolley 1934: Pl. 105; our
fig. 9). At the top, a nude “hero” embraces two rampant human-headed
bulls. In the lower registers, various anthropomorphized animals engage
in human activities: at the bottom, a gazelle holds two goblets, while
a scorpion-man walks to the left; in the register above, an ass plays a
large floor-lyre, accompanied by other animal-musicians on systrum
and, presumably, voice. In the third register from the bottom, just below
the hero and bulls, a jackal walks to the left, dagger in his belt, carry-
ing a table piled with two animal heads and a haunch; he is followed
by a rampant lion bearing a tall vase with wicker cover and carrying
handle in his left hand.
It is the object held in the lion’s right hand that is of interest in the
present context. Round-bellied, with a tapering shallow tip, it sits on
the lion’s outstretched palm. In conjunction with the vase, this object
is surely intended to be the recipient of the contents of the vase; it
would have held a relatively small quantity of those contents at any

one moment, and the size plus narrow end would permit the object’s
contents to be poured onto a targeted area with relative precision. In
short, it is not in itself a primary container, but rather task-related—as
today, one would first pour olive oil from a storage container into a
measuring cup, and from there into a pot—and indeed, as a comple-
ment to the table of meat carried by the jackal, the substance to be
transferred from the tall jar would more likely have been some sort of
liquid than dry, uncooked stores.7
The scale of the object in relation to the lion’s “hand,” along with
the rounded shape, replicates closely the conch-shells that Woolley
suggests were the prototype for his Metal Vessel Type 115. I would
suggest that what we have illustrated on the lyre plaque is precisely
one of the objects under discussion, strongly supporting their use in
the dispersal of liquid (this function and parallel is noted as well by
Müller-Karpe 1993: 45).
At issue is whether the type represents merely a utilitarian “ladle”,
or dipper, as per Woolley, or something more; and this hinges upon
whether one can argue that the procession of table-plus-vase, or indeed
the whole context of activities depicted on the lyre panel, implies a
banquet/meal or ritual activity associated with “offering”—one that
would entail some sort of cultic libation or ceremonial pouring.8 Yet
a third possibility would be that the same object type could double as
a functional tool in non-ritual contexts and as a ritual implement in
ceremonial contexts (on which, more below).
At present, there is not sufficient internal evidence to favor one
reading of the scene over another, although the ceremonial role of
the lyres themselves and the heroic iconography at the top of the
inlaid panel might be said to privilege a more ritual presentation of
comestibles. What is clear is that there is sufficient evidence to suggest
that we abandon the designation “lamp” in favor of “pouring-vessel”
7
See also the Hebrew Bible book of Numbers 29:1–6, where offerings are enumer-
ated for the New Year sacrifice. Animal offerings of bull, ram and lambs are followed
by choice oils, “each with its libation, as prescribed”—i.e., the solid (meat) and the
liquid (oil).
8
The ritual nature of the “banquet”-seals from the Royal Cemetery, if associated
with the priestesses of Nanna, could also be argued. In addition, in a later, Middle Assyr-
ian context, ritual offering in the temple of Ištar included a sacrificial animal, a meal
of bread-cake, and a libation of beer (see Oppenheim 1966: 253 re KAR 139).
associated with liquids, although equally in quotation marks pending

further confirmation.9
It is possible to further privilege the function of the conch shells
for purposes of libation by analogy with other religious practices, par-
ticularly among Vaisnavite Hindus in India in the celebration of both
home and temple rites (Goudrian 1970; Parpola 1983).10 In the temple
during the first worship of the day, a conch is used to pour a libation at
the invocation, two others are used to bathe the feet and body of the
divine image, and yet another provides drinking water. What is more,
in one South Indian temple at least, the four conches used in this ritual
are then placed “successively to the north of each other” (Goudrian
1970: 191). In the Indian case, the north is an auspicious direction; the
Mesopotamian association with this direction we shall pursue below.
Of interest is the analogous situation of careful placement with respect
to a meaningful cardinal direction.
Most important, in the North Indian Rādhārama a Temple to
Krishna in Vrindaban, access to the biannual bathing ceremony of the
deity of which I here gratefully acknowledge, the conch is the vessel that
receives the five principal liquid substances used in the ritual. Primary
amongst these substances is milk, but also pure water, honey, curds
and clarified butter—in all cases transferred from a large container to
a spouted pot of about a half-gallon capacity, and from there, before
the deity, into the conch from which the liquid is then poured over the
image (our figs. 10 and 11). Not only is the conch used for pouring,
then; it is also the conveyor of a range of different liquid substances.
9
Scurlock 1995: 1884 does note that there were niches for funeral lamps in later,
Neo-Assyrian graves from Assur and Nimrud; but then she refers to “actual examples
of such lamps found at Ur,” by which she presumably means the vessels presently under
consideration, the identification of which is being questioned. It would be interesting
to pursue the question of light/lamps on its own terms to see just when and what
evidence now becomes available.
10
The issue of the validity and ultimate value of cross-cultural analogy cannot be
discussed extensively here, but has been taken up in a separate study: I. Winter, “Images
in Worship: India and ancient Mesopotamia, the legitimacy and utility of cross-cultural
analogy,” forthcoming—originally given as a paper in the seminar on “Pilgrimage, Art
and Ritual: Ethnography and Art History,” organized by Michael W. Meister for the
Center for the Advanced Study of India at the University of Pennsylvania, Fall 1995.
Clearly, to reach some 4000 years and 2000 miles needs a different sort of justification
from that presented by Simpson (1995), which compares historical and present Islamic
burial practice with pre-Islamic practice in the Near East. Suffice it to say that one
must construct an argument based upon systemic comparability, with all of the caveats
regarding the limits of such analogies articulated, when moving across geographical
as well as cultural and historical boundaries.
The functional analogy in the utility of conches for pouring, the objects
themselves coming from a water environment, is clear; their symbolic
association with a positively charged cardinal direction at least suggestive
of the non-random association between object and placement. These
practices from India may then be put together with later third millen-
nium textual references to post-death commemoration rituals, referred to
as ceremonies of “name calling,” which apparently combined libations
of water [often through pipe or drain] into the netherworld with the
speaking of the name of the deceased (Bayliss 1973; Perlov 1980; Tsu-
kimoto 1985). The existence of such ceremonies adds weight, I believe,
to the suggestion that Woolley’s Type 115 object is not only not a lamp,
but is considerably more than a dipper or ladle. It allows us to further
suggest that the presence of such objects in the tombs could signal a
liquid poured at the time of interment, the pouring vessel then placed
on or near the body, or in the appropriate cardinal direction/corner,
when the ritual action was completed. For the pieces found higher up in
the shaft, the gesture involved could have been one related to the ritual
pouring of a libation once the grave was covered, but before the shaft
was completely filled, and then the purposeful deposit of the pouring
vessel on the performance surface in the tomb shaft.
These patterns—especially the association of the type with other ves-
sels in the “ritual locus” of fill above the burials, and the number of
examples found in direct proximity to the buried body—suggest an
association with ritual acts as distinct from mere table service. Such
“final gestures” fit our definition of patterned behavior as entirely
appropriate to ritual performance in general, and to funerary ritual in
particular. In sum, a ritual use for our conches and congeners would
be entirely consistent with the patterns apparent in the depositional
history of the objects.
I would stress that the observable patterns in deposition form the basis
of any reconstruction of these object-types as pertaining to funerary
and/or burial ritual. In most cases, however, the observation serve as
necessary but not sufficient conditions for reconstruction. It is when the
imagery on the PG/789 lyre and the cross-cultural analogy with India
are placed together with the deposition history of the objects that such a
reconstruction becomes not only possible, but compelling.
Spouted Jars, Metal Vessels of Type 82–84—Recast as Multi-Purpose

Pouring Vessels
The large spouted vessels found in a number of the Royal Graves pro-
vide a parallel, but slightly different data set to the conches and their
congeners. For the jugs, each of the designated findspots and associa-
tions suggests a slightly different function. Although a smaller sample,
the distribution in several cases seems to replicate the placement of
the “ladles/libation vessels” discussed above. The spouted vessel of
PG/1054, for example, is found in close proximity to the body of the
deceased; just as the vessel-with-paten of Meskalamdug in PG/755 is
positioned in the north corner (our figs. 3 and 6), once again point-
ing to some privileged relationship between pouring vessels and that
cardinal direction.
Woolley’s association with libation was certainly appropriate, given
the imagery on the shell plaque from Ur and the additional imagery
on the Tello plaque, noted above. However, the plaques and related
images (Müller-Karpe 1993: Pl. 170, nos. 1–2, 4–5, 8) all suggest ritual
activity before or within a temple or sacred space related to divine
worship. In the graves, to our knowledge, there was no divine image
or emblem present, so any “libation” performed would have to have
been in relation to the deceased. Otherwise, some alternative action
requiring pouring must be considered, as well as the possibility that it
might have provisioned the deceased but was not necessarily used at the
time of interment. In the latter case, however, it is harder to understand
the positioning so close to the body in PG/1054.
The consistency of association with an open bowl, yet isolation
together near the body in PG/1054 or in the northern corner in
PG/755 (our fig. 3), suggests that these two vessels functioned as a unit.
On the shell plaque from Ur, the priestly libator actually holds a small
dish or saucer in his right hand, while extending the spouted vessel in
his left toward the divine standard. This does not automatically imply
that some aspect of collection from the container to the open bowl is
intended, however. Oppenheim has noted that in a later, Middle Assyr-
ian ritual from Assur, food in the form of crumbled bread is offered on a
shallow bowl along with a libation made by pouring beer from a goblet
into a kirru-vat, and such food is also fed to the spirits of the dead in
mortuary rites (1966: 257–258, citing CAD E: e emmu; see also Heimpel
1987). The practice of presenting (and in the visual arts, showing) both
food and liquid containers, one in each hand, may have had its origin
in earlier periods. In any case, the association of spouted jar and open
bowl does not preclude the use of the jar in libation.
What is not certain is whether this represents the sole function of
the vessel type or the set. On several cylinder seals from the cemetery
at Ur on which a banquet scene appears in the upper register, the
lower register includes either a procession of figures toward a large
table or architectural facade, one of whom carries just this sort of jar
in his left hand (e.g., Woolley 1934: Pl. 200, Nos. 98, 102 = our fig. 12,
redrawn to provide narrative unity), or a servant holds the vessel before
a seated figure (Müller-Karpe 1993: Pl. 170, No. 9). In each of these
cases, the same individual also carries an open bowl in his right. In
the first two seals, this personage follows immediately behind the one
individual in the procession with clasped hands, who is presumably the
principal worshipper/participant, and it is not inconceivable that the
jar plus bowl are somehow related to his anticipated ritual or votive
act (on which, see below); however in the third seal, the ewer and bowl
are held before a seated figure who tilts a cup back toward his head as
if imbibing, which opens the possibility that the “set” can be directly
involved with servicing a banqueter.11
Such a function would be consistent with the fact that the three
spouted vessels of PG/800 are all associated with tumblers as well
as open bowls. It is possible that here the spouted jugs were used for
pouring some liquid to be drunk from the tumblers—comparable to
the Babylonian trinksets discussed by Salje (1996) and known to have
existed from as early as the Jemdet Nasr Period (Müller-Karpe 1993:
283). It is equally possible that the spouted jars were used to fill the
tumblers for the pouring of some liquid into the grave by mourners—the
designation of the latter as “drinking vessels” would then be a massive
projection from our own typology, since their shape corresponds quite
closely to that of the vessel held by Ur-Namma in his libation of a plant
before a seated deity on the great stele of the Ur in period some 500
years later, and even to similarly-used vessels of the Akkadian period
(Braun-Holzinger 1989: 2). It must be noted that the careful stacking
of these tall, narrow “tumblers” in sets of five, as reported by Wool-
ley, might argue more for provisioning and against individual action,
11
Twelve spouted jars have been found in the late ED/Akkadian palace at Ebla,
and have been studied recently by Stefania Mazzoni, who classifies them within the
set “drinking vessels” (1994: 245–276, esp. 250–251, No. 9); similarly Frances Pinnock
for the Ur vessels (1994: 15–26, esp. Pl. lc&d)
unless they were subsequently collected, ordered, and deposited. In

either case, however, the associated spouted vessel would function as a
conveyor, and not for libation.
Thus, deposition, imagery and association, when taken together,
suggest that the spouted vessel is likely to have been subject to multiple
uses in multiple venues: as a libating instrument in religious (temple)
ritual before deities and/or emblems, as well as possibly in funerary
activity; and in the filling of other containers, including drinking vessels.
At issue is whether there is evidence to support the suggestion that its
presence in some tombs would indicate ritual action associated with
funerary practice, rather than mere provisioning.
Once again, analogous practice in India may be demonstrated, where
related vessel shapes, slightly more bulbous of body, but with similar
upward-directed spout, are not uncommon in ritual performance. In
the Vaisnavite ritual bathing of the divine image referred to above, it
is just such a vessel that contains the bulk liquid that is then transferred
to the conch for immediate libation over the deity (see fig. 10), thereby
supporting the function of our vessel-type as a conveyor of liquid to
another targeted vessel. Yet in the Indian case, too, the same vessel is
used for multiple tasks. The vessel is also used to pour pure, ceremoni-
ally collected water over the hands of those priests and lay persons of
the priestly lineage who have been engaged in the ritual (fig. 13). And
at the end of the ritual bathing process, the vessel is in turn used to fill
the drinking cups of the devoted spectators with the nectars that have
been collected and mixed after bathing the god (fig. 14).
It is the possibility of use for hand-washing on which I would dwell
briefly, for it provides yet another possible function for our spouted
jar. In addition, it is here that the associated paten may be significant,
as its size in relation to the spouted jar is considerably larger than the
hand-held saucer/bowl shown on the Ur shell plaque or the seals, and
comes considerably closer to the sort of copper basin-and-ewer sets for
washing the hands known from Old Kingdom Egypt (Dayagi-Men-
dels 1989: 17). Used alone, the vessel of Type 84 in early Sumerian
Mesopotamia could well have served for pouring into some other con-
tainer—either as a libation vessel itself, issuing into the potted plant of
the goddess Ninhursag on the Tello plaque, or, as in the Indian case,
to fill a conch/smaller libating vessel. But in association with a basin or
open saucer/bowl (e.g., our fig. 12), it suggests that the activity involved
included both pouring from the spouted jug AND catching whatever
was being poured as well, either for purposeful redirection or merely
to avoid spillage. This would be especially true if the purpose of the

ritual act were realized in the pouring, rather than in the dispersal of
the liquid.
One such case would be if something material—objects or body-
parts—were placed in the flow of liquid for “washing,” the liquid then
carried off in the open basin, as in the Indian situation noted just
above, where the laving of hands is the desired act, with the object of
the flow and the liquid resulting from the flow not appropriate to be
casually spilled. A similar act is performed at the beginning of some
Jewish ritual meals, where a two-handled pitcher or spouted water jar
and basin is carried around to participants, and water poured over the
hands. The hands then occupy intermediary space between the jug
containing the cleansing water and the basin that will catch the flow;
the purpose of the “set” is to ensure capture of that flow within an
interior, domestic space.
It is noteworthy as well that the practice of observant Jews also
entails washing of the hands upon leaving the cemetery after partici-
pating in a funeral, or upon re-entry to the home after burial services,
where a spouted water jar or pitcher and basin are set at the door for
mourners.12 This practice follows from Biblical law that contact with
the dead is polluting (e.g., Numbers 19:10–13). In the Talmud, this is
referred to as “corpse-uncleanness” (Neusner 1992: 178; 1993: 53);
and therefore, ritual cleansing—“by water of sprinkling” (Numbers
31:19–24)—is required upon return to the “normal” world (see dis-
cussion in Haran 1985: 11, 138, 176; the practice is also attested for
funerals in modern Greece, Danforth 1982: 42–43).
It is this association of death with pollution, and the relationship
between funerary practice and the need for ablution/purification, to
which I would point as one possible explanation for these sets in the
graves at Ur. In Ugaritic ritual, the act of washing the hands precedes
participation in ritual sacrifice (Levine 1963: 105). Evidence certainly
does exist in the Sumerian textual repertoire as well to suggest that
hand-washing was ritualized: a šu-lu —literally “hand-washing”—
ceremony is attested in relation to royal activity in the temple at Ur
in the later, Isin-Larsa period, and is suggested by Jan Wilson to have
12
Referred to as netilat-yada’im, “the cleansing of hands,” in Cecil Roth, ed., 1971;
Encyclopedia Judaica, Vol. 12: 998–999; also, Vol. 4; 1520, “burial;” Vol. 13: 1406, “purity
and impurity.” See also Lamm 1969.
been intimately connected with purification and sanctification, although

not directly associated with funerary activity (1994: 8–14).13 Wilson
further notes that it was the sanctification achieved through the šu-
lu ceremony that then enabled individuals, particularly the ruler, to
enter the presence of the deity (Wilson 1994: 36), just as the converse
is lamented over the destruction of Nippur, when the temple has been
defiled, and the šu-lu rituals not properly performed (Kramer 1991b:
16). Finally, Wilson cites a text indicating that lustration/hand-wash-
ing was performed both by and for priests/priestesses, always related to
purification (1994: 37), and would also be consistent with the seal from
Ur noted above (fig. 12), on which the individual with jar and bowl
follows directly behind the principal figure in procession.14
All of the above circumstances can be projected for funerary prac-
tice. The ritual “hand-washing” that marked “the transition from the
profane outer areas into the holy inner portions of the temple” can be
seen as an analogue both to a need for ritual washing of the corpse
before the deceased undertakes his/her journey to the Netherworld,
and to a need for participants—priests and/or mourners—to ritually
cleanse themselves as they re-enter the normal world, leaving the dead
behind. And indeed, both of these factors are attested for later Islamic
practice in the Near East, which includes elite associations of ewer and
basin (al-Qaddūmī, 1996: 323–324) and their association with funerary
activity, extending even to the representation of a ewer and bowl on
adult tombstones in western Iran and southeast Turkey as a washing set
(Simpson 1995: 248). Ritual libation at interment is further attested for
Greek burials from the Byzantine period through to modern times, both
13
Wilson (1994) discusses a text of Rim-Sin of Larsa, line 38: and notes, p. 14
that it is also attested in Tablet IX, 88–95 of Šurpu: “Incantation: Your hands are
washed. . . . you are holy; your hands are washed, you are pure.” The practice clearly
goes on into the Late Babylonian period, as shown recently by Cagirgan and Lambert
(1991–3: 95, Obv. II: 55–6), in which a ritual performed before the gate of Esagila
includes “water for the hands” [mê-qātē] of the god Bel; indeed, it is suggested (Wil-
son 1994: 101) that a cultic utensil specific to the “giving of water for the hands,”
normally residing in the cella, may have been deified. [H. G. Güterbock tells me,
personal communication, that there is also a reference in Hittite text to a “vessel for
washing the hands . . .”, NIG3.ŠU.LUH.HA]. See also Maul 1994: 136 and 144, re
hand-washing rituals.
14
In Jewish tradition, it is males of the Levite lineage who pour the water for
the cleansing of the priest/rabbi, which has its expression in visual iconography as
well—see Encyclopedia Judaica vol. 15, 1972: 1231: “tombstones,” for a Levite tombstone
on which a ewer is depicted.
by observed practice and by pouring vessels found left in archaeological

contexts after use (Danforth 1982: 42; Sanders 1998).
For both vessel types, then, the conches and their metal relatives plus
the tall jars, some association with libation and/or cleansing is thus
both logical and consistent with Mesopotamian practice. Unfortunately,
despite several recent studies concerned with identifying the Sumerian
terms for specific vessel shapes (Braun-Holzinger 1989; Selz 1996), it
is not possible to attach names to our vessels; nor, without further tex-
tual evidence and/or chemical analysis (as per Badler et al. 1996 for a
spouted vessel of the Uruk period), is it possible to determine just what
might have been poured from either vessel type. For hand-washing, of
course, water is evident; if a libation was to have been poured, a variety
of liquids could have been used. The multiple substances used in the
bathing of the Krishna image in Vrindaban have been noted above.
The Hebrew Bible records drink offerings to a variety of non-Israelite
gods ( Jeremiah 19:13; 32:29; 44:17) and speaks of libations of water,
blood and wine within Israelite sanctuaries and in the outer court of
the Temple in Jerusalem (e.g., I. Sam. 7:6; discussion in Haran 1985:
216–217). In funerary contexts, the ancient Egyptian sem-priest purifies,
censes and libates; the libations consisted essentially of milk, but water is
also attested (Milde 1994: 18–19). Water libations are offered at Bedouin
graves in Jordan to this day (Simpson 1995: 249). A Hittite royal funeral
is described in which wine is libated to the dead king on the first day,
and then the jug broken; and on the second day, beer, wine, and an
unidentified walhi-drink are poured (Van den Hout 1994: 59–60). It is
unfortunate that the extant descriptions of Early Dynastic burials in
Mesopotamia seem to record grave goods, not processes (Foxvog 1980;
Tsukimoto 1980; Steinkeller 1980); however, texts of later periods, in
both Sumerian and Akkadian, speak of beer, oil, honey, wine, milk and
clarified butter in association with temple libations, ritualized sprinkling
and/or offerings (Oppenheim 1966: 251; Borger 1971: 72–80; Cagirgan
and Lambert 1991–1993: 95; Kramer 1991b: 14; Michalowski 1994:
32–3; Bottéro 1998: 241, 249–261, 290; PSD A1: a. A3, “libation or
ablution water”; CAD P files: parsu: rite/ritual). The frequency of beer
and, especially, water in the post-interment libations related to the cult
of deceased ancestors and the appeasement of the e emmu (e.g., PSD
A1: a-si-ga; PSD B: bal D3; Bayliss 1973: 118, Wilcke 1988: 252–253)
would make them the most compelling alternatives.
A different and tantalizing possibility arises for the conches. Müller-
Karpe cites a reference in Leo Oppenheim’s conchology study to a spe-
cific shell for oil, išqilat šamnim (1993: 45 re Oppenheim 1963: 407–12).
To this may be added Neo-Assyrian and Babylonian references to the
(royal) body being anointed with oil at death, and to oil or unguents
in a text actually found in a burial (Bottéro 1982: 398), as well as
mention in a Mari text of two kinds of oil sent to a grave for use in
anointing or as incense (McGinnis 1987: 4, 8). The chain of linkages is
further strengthened by two texts that probably date to the end of the
third millennium or beginning of the second, thus far closer in time
to Early Dynastic Period practices. In the first, the Sumerian version
of Gilgameš and the Netherworld, reference is made to “sweet oil from a
jar,” that would be rubbed upon a corpse (Shaffer 1963: l. 187). In the
second text, The ‘Traveller’ and the ‘Maiden’, it is made quite explicit that
the body is anointed with “sweet(-smelling) oil” (Kramer 1977: l. 46).15
The foregoing association of shells with oil, and of oil with the actual
body of the deceased, seems particularly compelling given the number
of instances in which the conches and their derivatives were found in
immediate proximity to the body in the Ur graves. Furthermore, their
explicit identification with the royal body in the later texts could imply
that status was a factor in the performance of anointment rites, which
could in turn explain why such artifacts/traces have not been found in
every grave of the “Royal Cemetery” (as has been observed in other
cultural contexts as well, cf. Little et al. 1992).
Equally compelling is the association within the text of The ‘Traveller’
and the ‘Maiden’ of a water libation poured for the deceased immedi-
ately preceding anointment of the corpse with oil. If the reading of
Cohen is correct, these acts, along with the cleaning of the body and
its dressing, were all part of the care of the body prior to burial (cf.
ll. 42–47, esp. 1. 45).16 Thus, the two vessel types discussed here may
15
Gilgameš l. 187: i3-du10-ga bur-ra; Traveller l. 46: u5-ze2-ba-mu e2-gar8 mu-un-na-
šeš2. For these readings, and for direction to the texts in question, I am indebted to
Andrew Cohen. He understands the whole passage of Gilgamesh’s instructions to
Enkidu, ll. 182–199, as a series of things one normally does to a corpse, hence what
Enkidu should NOT do, lest he, too, be mistaken as a dead spirit and therefore not
be able to return. With Kramer, Cohen also sees the “‘Traveller’ and the ‘Maiden’ ” as an
exchange between an unmarried maiden and another young woman; however, where
Kramer believes the deceased lover will visit his betrothed as a ghost, Cohen under-
stands his return as that of a corpse. A further discussion of each will be presented
in his dissertation, cited above. It is also interesting to note that the term used for the
oil container, bur, has been shown to be of stone, which would correspond well to the
alabaster jars found often in association with the conches (see above).
16
“‘Traveller’”, l. 45: a ib2-ta-de2 ki in-de2 ba-ab-nag.
actually have functioned in sequential ways in the rituals attendant

upon formal burial.
For the north/northwest placement of our vessels, one would want
to pursue Mesopotamian associations with those directions systemati-
cally, which is beyond the scope of the present paper; but some hint
may be provided by Wayne Horowitz’ recent study of Mesopotamian
cosmic geography (1998) and Thorkild Jacobsen’s study of the sign lil2
in the god-name En-lil, the primary meaning of which is “wind” (1989).
Horowitz (1998: 195–6) notes that each of the four compass points had
associated winds, but that the prevailing winds actually came from the
northwest and southeast, so that “north” and “northwest” constituted a
conceptual unity. The association of winds and cardinal directions had
been noted earlier by Jacobsen (1989: 267–276), who further showed
that the winds were themselves associated with both deities and demons.
He cites a late bilingual text from Nineveh (K3372+: 4–10) that may
be relevant to the Ur context, in which the Sumerian version translates:
“The evil wind demons come out of the netherworld, for funerary
offerings and libations of water, they come out of the netherworld;”
the Akkadian version, interestingly, substitutes the word “grave” for
netherworld!17 Jacobsen speaks specifically of the chilling winter winds
of the desert that come from the northwest causing illness and bringing
evil—that very northwest wind to which Woolley himself referred in
describing the piercing cold at the site of Ur in December and Janu-
ary (1934: 2). If one could ascertain further the associations with these
directions, one would be in a better position to explain the depositional
predilections at Ur; but for the present, one is led to suppose that the
placement has been purposefully chosen with respect to some larger set
of associations and the specific function of the vessels in mind.
In closing, I would stress that in the foregoing “reflections,” my pur-
pose has been neither to present primary data, nor to move toward a
substantive conclusion. Rather, the goal has been to point to several
issues of importance with respect to the interpretation of archaeological
data, using as test cases two vessel types found in the Royal Cemetery
burials. The dissertation of Andrew Cohen should provide a nuanced
integration of Early Dynastic burial customs within its particular socio-
17
Sum: lil2-la2-e-ne ul-a-meš urugal-la-ta im-ta-e3-a-meš
ki-se3-ga a-de2-am3 urugal-la-ta im-ta-e3-a-meš
Akk: za-qi2-qu lem-nu-ti iš-tu qab-rim it-ta- u-ni
a-na ka-sa-ap ki-is-pi u na-aq me3-e iš-tu qab-rim MIN
cultural surround, including the importance of ritual performance as

part of that surround. What has been suggested here is that circum-
stantial evidence and other cultural practices—such as the “Great
Death-pit” sacrifices and post-interment rites—argue for the likelihood
of ritual acts associated with burial as well. It is further suggested that
depositional patterns constitute a necessary, if not sufficient, condition
for the patterned behavior that would have been associated with the
enactment of ritual at the burial site, and can themselves constitute
evidence for such performance. Indeed, such patterns may offer a rare
hope of transcending the skepticism of those who worry about our
ever being able to recover “actions, not things” (Barrett 1996). Another
such hope may be found in “ethnographic analogy,” that window into
the ancient world from present or parallel cultural practice. Despite its
often-demonstrated limitations in never offering complete explanations
of ancient behavior (cf. Scarre 1994: 81), analogy used with caution
can serve to stimulate or even buttress the pursuit of explanation within
the cultural domain under scrutiny.
In other words, unless deposition is considered (and is in future
accompanied by scientific analysis of adjacent soils and vessel contents),
and without the elicitation of striking parallel praxes, it may never
be possible to determine whether a given spouted jar, to cite but one
example, reflects symbolic provisioning for the deceased in an afterlife
of continuing need; provisioning for a specific single “banquet”-meal
or gift associated with the entry of the deceased into the netherworld;
a relatively passive ritual offering to (the soul of ?) the deceased at the
time of burial; or a remnant of a more active ritual performance, includ-
ing libation or hand-washing, at the time of burial. In attempting to
argue for one or another alternative, it is my view that all possibilities
for which there may be circumstantial evidence need to be considered,
and competing hypotheses annihilated, before we can hope to develop
informed constructs of ancient Mesopotamian society, its beliefs and
practices. Otherwise, unsupported assumptions or merely not-discon-
firmed hypotheses will tend to stand as if confirmed, and we will neither
approach the truth of ancient practice nor work out the methodologies
necessary for dealing with degrees of certainty approaching confirma-
tion. I would further submit that uncertainty itself is constructive, since
unresolved ambiguities may serve as stimuli for the development of
future research.
In the meantime, there remains the particular case of the conches,
their metal and stone derivatives, and the tall spouted vessels of the
Royal Cemetery. What I would hope to have shown, in line with

Postgate (1977) and Müller-Karpe (1993), is that there is no evidence
that the first group represents lamps, and some evidence to suggest
that instead they served as receptacles for liquid—dippers/ladles or
ritual pouring vehicles. For the second group, while there is good
evidence that they were at some points used for libation, there seems
to be further evidence to suggest that they were multi-purpose vessels
that could be used for the transfer of liquid under differing conditions
and for different targets. And finally, while each group may have had
practical uses in some contexts, there is strong evidence that they also
had ritual functions that are not inconsistent with their placement in
the tombs. Indeed, the importance of ritual as a consolidating and
symbolic activity at key social/biological/calendrical junctures in gen-
eral (Lawson and McCauley 1990: 121; Scheffler 1997: 129), and the
variety of ritual practices in non-funerary domains in Mesopotamia in
particular—especially those related to post-interment offerings to the
deceased and the intimation that they actually began as grave-offerings
(Tsukimoto 1985: 232)—should make it incumbent on the scholar to
prove there was NOT such activity at the time of burial!
This does not deny the possibility that some of the grave goods
accompanying burials in the Royal Cemetery of Ur might have been
intended as gifts for the high-ranking deities and heros of the Nether-
world; nor the possibility that some of the vessels might have been left
filled with food as offerings to the deceased or for an envisioned funeral
banquet. It is not unlikely that individual “types” of grave goods could
have served a multiplicity of purposes, and therefore their presence
would not automatically suggest a single use. Nor do I wish to gloss
over the fact that the particular vessel types under discussion are not
present in every burial; hence one cannot claim for them normative
patterns of usage, but rather seek explanation in the status/role of
grave occupants, beyond the scope of this study. Nevertheless, for the
present, I believe there is sufficient indication from functional analysis,
deposition patterns in the graves, cultural context, and cross-cultural
analogy to suggest that some sort of ritualized performance involving
libation and/or anointment, possibly directly onto the corpse, and/or
hand-washing would have been entirely consistent with what we know
of Mesopotamian religious practice and attitudes toward death in the
Early Dynastic Period.
Acknowledgments
The substance of this paper was first put together as a brief communication to
archaeological colleagues of the Near East “Marching and Chowder” Society in Phila-
delphia, 1990, but was retracted in the face of an overfull program, and ultimately
presented to that group in 1994; I profited greatly from the ensuing discussion. It is
a further pleasure to acknowledge the work of the eight students who participated in
an undergraduate seminar on the Royal Cemetery of Ur, taught in the Ur gallery of
the University Museum, during my last semester at the University of Pennsylvania in
the Spring of 1988. I should also like to thank Michelle Bonogofsky for making her
MA Thesis available to me; Daniella Bar Yosef for bibliography on shells; Richard J.
Israel for perspective on Jewish mourning and the cleansing of hands; Guy Sanders for
confirmation of Byzantine burial practice; Rochelle Kessler for the drawing of fig. 12;
Jessica Rawson for bringing the issue of “sets” in early Chinese burial deposits to my
attention; Elizabeth Stone for bibliography on archaeological analogy; and Robert Hunt
for many discussions on cross-cultural comparison. I am especially grateful to Pandit
Srivatsa Goswāmī and the priestly lineage of the Rādhārama a Temple in Vrindaban
for continued acts of generosity and privileged access to temple ritual and practice;
and to Andrew Cohen for many stimulating conversations and exchanges of ideas on
ritual and related topics. I was not able to consult Michael Müller-Karpe’s masterful
Metallgefässe (1993) until after a draft of the present text was completed, and was both
dismayed and delighted to discover that we had come to similar conclusions on many
of the questions of function raised here. I have added references where appropriate,
deciding not to pull my own paper from publication, since the focus on only two vessel
types, on deposition and ritual activity associated with interment, and on cross-cultural
analogy, does provide different emphasis.
Bibliography
Abusch, Tzvi. 1995. “E emmu”. In K. van der Toorn et al., eds., Dictionary of Deities
and Demons in the Bible, 588–594. Leiden: E. J. Brill.
Alster, Bendt, ed. 1980. Death in Mesopotamia: Papers Read at the XXVIe Rencontre Assyri-
ologique Internationale. Mesopotamia 8. Copenhagen: Akademisk Forlag.
Amiran, Ruth. 1969. Ancient Pottery of the Holy Land. Jerusalem: The Israel Museum.
Aynard, J. M. 1966. “Coquillages mésopotamiens”. Syria 43: 21–37.
Bachelot, Luc. 1992. “Iconographie et pratiques funéraires en Mésopotamie au troisième
millénaire av. J.-C.” In D. Charpin and F. Joannès, eds., La circulation des biens, des
personnes et des idées dans le Proche-Orient ancien, 53–67. Paris: Éditions Recherche sur
les Civilisations.
Badler, Virginia R., Patrick E. McGovern and Donald L. Glusker. 1996. “Chemical
Evidence for a Wine Residue from Warka (Uruk) Inside a Late Uruk Period Spouted
Jar.” Baghdader Mitteilungen 27: 39–43.
Baker, Heather. 1995. “Neo-Babylonian Burials Revisited.” In S. Campbell and
A. Green, eds., The Archaeology of Death in the Ancient Near East, 209–220. Oxford:
Oxbow Books.
Banks, E. J. 1912. Bismaya: The Lost City of Adab. New York: Putnam’s.
Barrett, John C. 1996. “The Living, the Dead, and the Ancestors: Neolithic and Early
Bronze Age Mortuary Practice.” In R. Preucel and I. Hodder, eds.. Contemporary
Archaeology in Theory, 394–412. Oxford: Blackwell.
Bayliss, M. 1973. “The Cult of Dead Kin in Assyria and Babylonia.” Iraq 35:
115–125.
Bell, Catherine. 1992. Ritual Theory, Ritual Practice. Oxford: Oxford University Press.
Binford, Lewis R. 1971. “Mortuary Practices: Their Study and their Potential.” In
J. A. Brown, ed., Approaches to the Social Dimensions of Mortuary Practices, 6–29. Memoirs
of the Society for American Archaeology 25. Washington: Society for American
Archaeology.
Bloch, Maurice and Jonathan P. Parry, eds. 1982. Death and the Regeneration of Life.
Cambridge: Cambridge University Press.
Bonogofsky, Michelle A. 1994. Gifts for the Netherworld: Objects Excavated from the
Royal Cemetery at Ur. M.A. Thesis, University of California at Berkeley.
Borger, Rykele. 1971. “Das Tempelbau-Ritual K48+.” Zeitschrift für Assyriologie 61:
72–80.
Bottéro, Jean. 1982. “Les inscriptions cunéiformes funéraires.” In G. Gnoli and J.-P.
Vernant, eds., La mort, les morts dans les sociétés anciennes, 373–406. Cambridge: Cam-
bridge University Press.
——. 1998. La plus vieille religion: en Mésopotamie. Paris: Gallimard.
Braun-Holzinger, Eva. 1989. “REC 447.la2 = Libationsbecher.” Zeitschrift für Assyrio-
logie 79: 1–7.
Bremer, Jan M., Th. P. J. van den Hout and R. Peters, eds. 1994. Hidden Futures: Death
and Immortality in Ancient Egypt, Anatolia, the Classical Biblical and Arabic-Islamic World.
Amsterdam: Amsterdam University Press.
Bright, Lynn. 1995. “Approaches to the Archaeological Study of Death with Particular
Reference to Ancient Cyprus.” In S. Campbell and A. Green, eds., The Archaeology
of Death in the Ancient Near East, 62–74. Oxford: Oxbow Books.
Byrd, Brian F. and Christopher M. Monahan. 1995. “Death, Mortuary Ritual, and
Natufian Social Structure.” Journal of Anthropological Archaeology 14: 251–287.
CAD. Assyrian Dictionary (Chicago Assyrian Dictionary). Vol. P: parsu, files. (Glückstadt:
J. J. Augustin Verlagsbuchhandlung).
Cagirgan, G. and W. G. Lambert. 1991–93. “The Late Babylonian kislīmu Ritual for
Esagil.” Journal of Cuneiform Studies 43–45: 89–101.
Campbell, Stuart and Anthony Green, eds. 1995. The Archaeology of Death in the Ancient
Near East. Oxford: Oxbow Books.
Cannadine, David and Simon Price, eds. 1987. Rituals of Royalty: Power and Ceremonial
in Traditional Societies. Cambridge: Cambridge University Press.
Carter, Elizabeth and Andrea Parker. 1995. “Pots, People and the Archaeology of
Death in Northern Syria and Southern Anatolia in the latter half of the Third
Millennium BC.” In S. Campbell and A. Green, eds., The Archaeology of Death in the
Ancient Near East, 96–116. Oxford: Oxbow Books.
Cassin, Elena. 1982. “Le mort: valeur et représentation en Mésopotamie ancienne.”
In G. Gnoli and J.-P. Vernant, eds., La mort, les morts dans les societés ancienns, 355–372.
Cederroth, Sven et al., eds. 1988. On the Meaning of Death: Essays on Mortuary Rituals
and Eschatological Beliefs. Acta Universitatis Uppsaliensis. Stockholm: Almqvist &
Wiksell.
Chiodi, Silvia Maria. 1994a. Le concezioni dell’oltretomba presso i Sumeri. Memorie, Serie
IX, Vol. IV, Fasc. 5. Roma: Accademia Nazionale dei Lincei.
——. 1994b. “Le tombe reali di Ur e il problema storico-religioso delle sepolture
multiple,” OCNUS 1994: 11–22.
Civil, Miguel. 1987. “KBo 26 53 and Funerary Personnel,” N.A.B.U. 1987: 4–5.
Cohen, Andrew. Forthcoming. Death Rituals and the Development of Kingship in
Early Dynastic III Period Mesopotamia. PhD. thesis, Bryn Mawr College.
Cooper, Jerrold S. 1991. “The Fate of Mankind: Death and Afterlife in Ancient
Mesopotamia.” In H. Obayashi, ed., Death and Afterlife: Perspectives of World Religions,
19–33. New York: Greenwood Press.
Danforth, Loring M. 1982. The Death Rituals of Rural Greece. Princeton: Princeton
University Press.
Dayagi-Mendels, Michal. 1989. Perfumes and Cosmetics in the Ancient World. Jerusalem:
The Israel Museum.
Edens, Christopher. 1992. “Dynamics of Trade in the Ancient Mesopotamian ‘World
System’.” American Anthropologist 94: 118–139.
Ellison, Rosemary, Jane Renfrew, Don Brothwell and N. Seeley. 1978. Some Food
Offerings from Ur, Excavated by Sir Leonard Woolley, and Previously Unpublished.”
Journal of Archaeological Sciences 5: 167–177.
Forest, Jean-Daniel. 1983. Les pratiques funéraires en Mésopotamie du 5e millinaire au début du
3e. Paris: Editions Recherche sur les Civilisations.
Foxvog, D. A. 1980. “Funerary Furnishings in an Early Sumerian Text from Adab.”
In B. Alster, ed.: Death in Mesopotamia: Papers Read at the XXVIe Rencontre Assyriologique
Internationale, 67–76. Mesopotamia 8. Copenhagen: Akademisk Forlag.
Goudrian, T. 1970. “Vaikhānasa Daily Worship: According to the Handbooks of Atri,
Bh gu, Kāsyapa, and Marīci.” Indo-Iranian Journal XII: 161–215.
Gnoli, Gherardo and Jean-Pierre Vernant, eds. 1982. La mort, les morts dans les societés
anciennes. Cambridge: Cambridge University Press.
Haran, Menahem. 1985. Temples and Temple-Service in Ancient Israel. Winona Lake:
Eisenbrauns.
Heimpel, Wolfgang. 1987. “Libation.” Reallexikon der Assyriologie VII: 1–5. Berlin: Walter
de Gruyter.
Homès-Fredericq, D. 1980–83. “Lampe.” Reallexikon der Assyriologie VI: 460–463. Berlin:
Walter de Gruyter.
Horowitz, Wayne. 1998. Mesopotamian Cosmic Geography. Winona Lake: Eisenbrauns.
Humphries, Sally C. and Helen King, eds. 1981. Mortality and Immortality: The Anthropol-
ogy and Archaeology of Death. New York: Academic Press.
Jacobsen, Thorkild. 1989. “The lil2 of dEn-lil2.” In H. Behrens, D. Loding and M. T.
Roth, eds.: Dumu E2-dub-ba-a: Studies in Honor of Ake W. Sjöberg, 267–276. Philadelphia:
University Museum.
Jonker, Gerdien. 1995. The Topography of Remembrance: The Dead, Tradition, and Collective
Memory in Mesopotamia. Leiden: E. J. Brill.
Kertzer, David. 1988. Ritual, Politics and Power. New Haven: Yale University Press.
Kramer, Samuel N. 1977. “The GIR5 and the ki-sikil: A New Sumerian Elegy.” In
Maria de J. Ellis, ed.: Essays on the Ancient Near East in Memory of Jacob Joel Finkelstein,
pp. 139–142. Hamden, CT.: Archon.
——. 1991a. “The Death of Urnammu.” In M. Mori, ed.: Near Eastern Studies Presented
to H.I.H. Prince Takahito Mikasa on the Occasion of his Seventh-Fifth Birthday, 193–214.
Wiesbaden: Harassowitz.
——. 1991b. “Lamentation over the Destruction of Nippur.” Acta Sumerologica 13:
1–26.
Kuwayama, George (ed.). 1991. Ancient Mortuary Traditions of China. Los Angeles: LA
County Museum of Art.
Lambert, Wilfred G. 1980. “The Theology of Death.” In Alster, ed. Death in Mesopotamia:
Papers Read at the XXVIe Rencontre Assyriologique Internationale, 53–66. Mesopotamia 8.
Copenhagen: Akademisk Forlag.
Lamm, Maurice. 1969. The Jewish Way of Death and Mourning. New York: J. David.
Lawson, E. Thomas and Robert N. McCauley. 1990. Rethinking Religion: Connecting
Cognition and Culture. Cambridge: Cambridge University Press.
Levine, Baruch. 1963. “Ugaritic Descriptive Rituals.” Journal of Cuneiform Studies 17:
105–111.
Little, Barbara J., Kim M. Lamphear and Douglas W. Owsley. 1992. “Mortuary Display
and Status in a 19th century Anglo-American Cemetery in Manassas, Virginia.”
American Antiquity 57: 397–418.
Martin, Harriet P., Jane Moon and J. Nicholas Postgate. 1985. Abu Salabikh Excavations,
Volume 2: Graves 1 to 99. Hertford: British School of Archaeology in Iraq.
Maul, Stefan M. 1994. Zukunftsbewältigung: Eine Untersuchung altorientalischen Denkens anhand
der babylonisch-assyrischen Löserituale (Namburbi). Mainz: Philipp von Zabern.
Mazzoni, Stefania. 1994. “Drinking Vessels in Syria: Ebla and the Early Bronze Age.”
In L. Milano, ed.: Drinking in Ancient Societies: History and Culture of Drinks in the Ancient
Near East, 245–276. Padua: Sargon srl.
McGinnis, J. 1987. “A Neo-Assyrian Text Describing a Royal Funeral.” State Archives
of Assyria Bulletin 1(1): 1–12.
Metcalf, Peter. 1981. “Meaning and Materialism: The Ritual Economy of Death.”
Man 16: 563–578.
Metcalf, Peter and Richard Huntington. 1991. Celebrations of Death: The Anthropology of
Mortuary Ritual. 2nd ed. Cambridge: Cambridge University Press.
Michalowski, Piotr. 1994. “The Drinking Gods: Alcohol in Mesopotamian Ritual and
Mythology.” In L. Milano, ed.: Drinking in Ancient Societies: History and Culture of Drinks
in the Ancient Near East, 27–44. Padua: Sargon srl.
Milano, L., ed. 1994. Drinking in Ancient Societies: History and Culture of Drinks in the Ancient
Near East. Padua: Sargon srl.
Milde, Henk. 1994. “ ‘Going Out into the Day:’ Ancient Egyptian Beliefs and Practices
Concerning Death.” In J. M. Bremer; Th. P. J. van den Hout and R. Peters, eds.:
Hidden Futures: Death and Immortality in Ancient Egypt, Anatolia, the Classical Biblical and
Arabic-Islamic World, 15–35. Amsterdam: Amsterdam University Press.
Moorey, Peter R. S. 1970. “Cemetery A at Kish: Grave Groups and Chronology.”
Iraq 32: 86–128.
——. 1977. “What Do We Know about the People Buried in the Royal Cemetery?”
Expedition 20: 24–40.
——. 1978. Kish Excavations, 1923–1933. Oxford: Clarendon Press.
Morris, Ian. 1992. Death, Ritual and Social Structure in Classical Antiquity. Cambridge:
Cambridge University Press.
Müller-Karpe, Michael. 1993. Metallgefässe im Iraq I. Prähistorische Bronzefunde,

Abteilung II, 14. Stuttgart: Steiner Verlag.
Neusner, Jacob, transl. 1992. The Talmud of Babylon XIV.C: Tractate Ketubot, Chs. 8–13.
Brown Judaic Studies 260. Atlanta: Scholars Press.
——. 1993. The Talmud of Babylon IV.B: Pesahim, Chs. 2 and 3. Brown Judaic Studies
282. Atlanta: Scholars Press.
Nissen, Hans J. 1966. Zur Datierung des Königsfriedhofes von Ur: Unter besonderer Berucksichtigung
der Stratigraphie der Privatgräber. Bonn: Habelt.
Oppenheim, A. Leo. 1963. “Mesopotamian Conchology.” Orientalia N.S. 32: 407–
412.
——. 1966. “Analysis of an Assyrian Ritual (KAR 139).” History of Religions V:
250–265.
Parpola, Asko. 1983. Grant Proposal to the Government of India: “The Domestic
Rituals of the Jaiminiya Brahmins.” Unpublished.
Penglase, Charles. 1995. “Some Concepts of Afterlife in Mesopotamia and Greece.”
In S. Campbell and A. Green, eds.: The Archaeology of Death in the Ancient Near East,
192–195. Oxford: Oxbow Books.
Perlov, Boris. 1980. “Families of the Ensis Urbau and Gudea and their Funerary Cult.”
In B. Alster, ed.: Death in Mesopotamia: Papers Read at the XXVIe Rencontre Assyriologique
Internationale, 77–81. Mesopotamia 8. Copenhagen: Akademisk Forlag.
Pinnock, Frances. 1994. “Considerations on the ‘Banquet Theme’ in the Figurative Art
of Mesopotamia and Syria.” In L. Milano, ed.: Drinking in Ancient Societies: History and
Culture of Drinks in the Ancient Near East, 15–26. Padua: Sargon srl.
Pitard, Wayne T. 1996. “Care of the Dead at Emar.” In M. W. Chavalas, ed.: Emar:
The History, Religion, and Culture of a Syrian Town in the Late Bronze Age, 123–140.
Bethesda: CDL Press.
Pollock, Susan. 1983. “Chronology of the Royal Cemetery of Ur.” Iraq 57: 129–158.
——. 1991. “Of Priestesses, Princes and Poor Relations: The Dead in the Royal
Cemetery of Ur.” Cambridge Archaeological Journal 1: 171–189.
Postgate, J. Nicholas. 1977. “Excavations at Abu Salabikh 1996.” Iraq 39: 269–299.
Potts, Daniel T. 1997. Mesopotamian Civilization: The Material Foundations. Ithaca: Cornell
University Press.
PSD. The Sumerian Dictionary of the University Museum of the University of Penn-
sylvania. (Philadelphia Sumerian Dictionary): Philadelphia: University Museum.
——. Volume A1: a A3: “libation or ablution water”.
——. Volume B: bal D3; “to pour out”.
al-Qaddūmī, G. al-Hijjāwī, transl. 1996. Book of Gifts and Rarities (Kitāb al-Hadāyā wa
al-Tuhaf ). Harvard Middle East Monographs XXIX. Cambridge: Harvard Uni-
versity Press.
Rappaport, Roy A. 1971. “Ritual, Sanctity, and Cybernetics.” American Anthropologist
73: 59–76.
Rathje, William L. 1977. “New Tricks for Old Seals: A Progress Report.” In M. Gib-
son and R. D. Biggs, eds.: Seals and Sealing in the Ancient Near East, 25–32. Bibliotheca
Mesopotamia 6. Malibu: Undena.
Rawson, Jessica. 1993. “Ancient Chinese Ritual Bronzes: the Evidence from Tombs and
Hoards of the Shang and Western Zhou Periods.” Antiquity 67 (257): 805–823.
——. Forthcoming. “Ancient Chinese Ritual as seen in the Material Record.” In
J. McDermott and J. Laidlaw, eds.: Chinese Court Ritual, 20–49.
Roth, Cecil, ed. 1971. The Jewish Encyclopedia. Jerusalem: Keter Publishing House.
Salje, Beate. 1996. “Becher für die Toten? Bronzene Trinksets und ihre Bedeutung im
Grabkult.” Baghdader Mitteilungen 27: 429–446.
Salles, Jean-Francois. 1995. “Rituel mortuaire et rituel social à Ras Shamra/Ougarit.”
In S. Campbell and A. Green, eds.: The Archaeology of Death in the Ancient Near East,
171–184. Oxford: Oxbow Books.
Sanders, Guy. 1998. “The Other Dark Age: 300 Years of Corinth between Anastasius I
and Michael II,” Paper delivered at the Archaeological Institute of America Annual
Meetings, Washington DC, 28 December 1998.
Saxe, Arthur A. 1970. The Social Dimensions of Mortuary Practice. Ann Arbor: University
Microfilms.
Scarre, Chris. 1994. “The Meaning of Death: Funerary Beliefs and the Prehistorian.”
in C. Renfrew and E. B. W. Zubrow, eds.: The Ancient Mind: Elements of Cognitive
Archaeology, 75–82. Cambridge: Cambridge University Press.
Shaffer, Aaron. 1963. Sumerian Sources of Tablet XII of the Epic of Gilgameš. Philadelphia:
Ph.D. thesis. University of Pennsylvania.
Scheffler, Israel. 1997. Symbolic Worlds: Art, Science, Language, Ritual. Cambridge: Cam-
bridge University Press.
Schiffer, Michael B. 1996. Formation Processes of the Archaeological Record (1st edition 1987).
Salt Lake City: University of Utah Press.
Schmidt, Brian B. 1996. “The Gods and the Dead of the Domestic Cult at Emar: A
Reassessment.” In M. W. Chavalas, ed.: Emar: The History, Religion, and Culture of a
Syrian Town in the Late Bronze Age, 141–163. Bethesda, MD: CDL Press.
Schwartz, Glenn M. 1986. “Mortuary Evidence and Social Stratification in the Ninevite
V Period.” In H. Weiss, ed.: The Origins of Cities in Dry-Farming Syria and Mesopotamia
in the 3rd Millennium B.C., 45–60. Guilford, CT: Four Quarters Publishing Co.
Scurlock, Jo Ann. 1995. “Death and the Afterlife in Ancient Mesopotamian Thought.”
In J. Sasson, ed.: Civilizations of the Ancient Near East, Vol. III, 1883–1893. New York:
Macmillan.
Selz, Gebhard. 1996. “ne-sag, bur-sag und gu2-ne(sag-ga): ZU zwei Gefäßbe-
zeichnungen, ihren Bedeutungsentwicklungen und einem sumerischen Wort für
(Gefäß)schrank.” Studi Epigrafici e Linguistici sul Vicino Oriente antico 13: 3–8.
Simpson, St. John. 1995. “Death and Burial in the Late Islamic Near East: Some
Insights from Archaeology and Ethnography.” In S. Campbell and A. Green, eds.,
The Archaeology of Death in the Ancient Near East, 240–251. Oxford: Oxbow Books.
Smith, Jonathan Z. 1987. To Take Place: Toward Theory in Ritual. Chicago: The University
of Chicago Press.
Spycket, Agnès. 1968. Les statues de culte dans les textes mésopotamiens des origines à la 1e.
dynastie de Babylone. Paris: Gabalda.
Steinkeller, Piotr. 1980. “Early Dynastic Burial Offerings in Light of Textual Evidence.”
Paper presented at the AOS Meeting, San Francisco.
Tsukimoto, Akio. 1980. “Aspekte von kispum als Totenbeigabe.” In B. Alster, ed.: Death
in Mesopotamia: Papers Read at the XXVIe Rencontre Assyriologique Internationale, 129–138.
Mesopotamia 8. Copenhagen: Akademisk Forlag.
—— 1985. Untersuchungen zur Totenpflege (kispum) im alten Mesopotamien. Alter Orient und
Altes Testament 216. Neukirchen-Vluyn: Neukirchener Verlag.
Van den Hout, Theo. 1994. “Death as a Privilege: The Hittite Royal Funeral Ritual.”
In J. M. Bremer, T. P. J. van den Hout and R. Peters, eds.: Hidden Futures: Death and
Immortality in Ancient Egypt, Anatolia, the Classical Biblical and Arabic-Islamic World, 37–75.
Amsterdam: Amsterdam University Press.
Watson, James L. 1988. “The Structure of Chinese Funerary Rites: Elementary
Forms, Ritual Sequence and the Primacy of Performance.” In J. L. Watson and
E. S. Rawski, eds.: Death Ritual in Late Imperial and Modern China, 3–20. Berkeley:
University of California Press.
Watson, James L. and Evelyn S. Rawski, eds. 1988. Death Ritual in Late Imperial and
Modern China. Berkeley: University of California Press.
Weadock, P. N. 1975. “The Giparu at Ur.” Iraq 37: 101–128.
Wilcke, Claus. 1988. “König Sulgis Himmelfahrt.” Münchner Beiträge zur Völkerkunde 1:
245–255.
Wilson, E. Jan. 1994. “Holiness” and “Purity” in Mesopotamia. Alter Orient und Altes
Testament 237. Neukirchen-Vluyn: Neukirchener Verlag.
Winter, Irene J. 1992. “ ‘Idols of the King’: Consecrated Images of Rulers in Ancient
Mesopotamia.” Journal of Ritual Studies 7: 13–43.
Woolley, C. Leonard. 1934. Ur Excavations II: The Royal Cemetery. Publications of the
Joint Expedition of the British Museum and the Museum of the University of
Pennsylvania to Mesopotamia. London: The British Museum.
Figure 1. Conches, PG/755, Royal Cemetery, Ur. [ U.8191 and 8198;

Photo courtesy The Trustees, The British Museum].
Figure 2. Conch with bird-head finial and inlay, PG/1237, Royal Cemetery,
Ur. [ U.8313; Photo courtesy The Trustees, The British Museum].
Figure 3. Plan of Tomb PG/755, Royal Cemetery, Ur.

Figure 4. Plan of Tomb PG/1312,

Royal Cemetery, Ur.
Figure 5. Plan of Tomb PG/1130,

Royal Cemetery, Ur.
Figure 6. Spouted silver jug and copper basin, PG/755, Royal Cemetery,
Ur. [U.10035, 10036; Photo courtesy The Trustees, The British Museum].
Figure 7. Spouted silver jug, PG/800, Royal Cemetery, Ur. [ U.11837;

Photo courtesy The University Museum, University of Pennsylvania].
Figure 8. Limestone plaque, showing libation before the goddess Nin ursag,
Tello. [AO 276; Photo courtesy Département des Antiquités Orientales,
Musée du Louvre].
Figure 9. Detail, Inlaid plaque from front of Lyre, PG/789,

Royal Cemetery, Ur. (CBS 17694; Photo courtesy The University
Museum, University of Pennsylvania).
Figure 10. Performance of bathing ritual (abhishēkha) of image.

Rādhāramana Temple, Vrindaban, May 1991: conch used in libation.
Figure 11. Performance of bathing ritual (abhishēkha) of image.

Rādhārama a Temple, Vrindaban, May 1991: conch and spouted jug.
Figure 12. Lower register of “banquet” seal impression, PG/1749, Royal

Cemetery, Ur. Ritual procession to temple or altar [ U.14493A; drawing by
Rochelle Kessler].
Figure 13. Ritual cleansing of hands, Rādhārama a Temple, Vrindaban,

May 1991: spouted jug.
Figure 14. Distribution of nectar to devotees after bathing ritual of image

(abhishēkha). Rādhārama a Temple, Vrindaban, May 1991: spouted jug.
CHAPTER TWENTY-SIX
“SURPASSING WORK”:
MASTERY OF MATERIALS AND THE VALUE OF
SKILLED PRODUCTION IN ANCIENT SUMER
In 1985, with the publication of Materials and Manufacture in Ancient

Mesopotamia, Roger Moorey noted the potential of textual evidence in
contributing to our understanding of ancient Mesopotamian mate-
rial culture (Moorey 1985: x). Although he himself made a conscious
decision to focus on the archaeological record (1985: vii, x–xi), it was
clearly not possible to avoid texts altogether. Throughout the introduc-
tion to this important volume, he pointed to the need to integrate the
material remains with relevant documentary evidence, especially with
respect to the socio-economic context in which works were produced
and consumed (e.g. 1985: ix).1
This is no less true when considering the socio-cultural context of
value—properties deemed worthy of positive evaluation when assessing
the overall worth of materials and individual specimens of material
culture. Similarly, in a study dealing with aesthetics for the anthropolo-
gist, Warren d’Azevedo (1958) observed that every (artistic) object has
two social co-ordinates: the context of production and the context of
appreciation. What has become increasingly clear is that if we are to
pursue the meaning(s) attached to major works of material culture in
antiquity, and in particular meaning(s) attached to appreciation, this can
only be accessed with reference to a combination of evidentiary sources
that include both the archaeological and the textual record.
* This article originally appeared as “Surpassing Work: Mastery of Materials and

the Value of Skilled Production in ancient Sumer,” in Culture through Objects: Ancient Near
Eastern Studies in Honour of P. R. S. Moorey, T. Potts, M. Roaf and D. Stein, eds. Oxford:
Griffith Institute, 2003, pp. 403–421.
1
It is a privilege to be able to present this tentative inquiry into the attributes of
and values attached to works of various media as a tribute to Roger Moorey who has,
throughout his career, contributed complex questions, and oftentimes answers, to our
field of discourse. The following bibliographical abbreviations are used in this article:
CAD The Assyrian Dictionary of the University of Chicago. Chicago: The Oriental Institutute.
PSD The Sumerian Dictionary of the University Museum of the University of Pennsylvania.
Philadelphia: University of Pennsylvania.
272 chapter twenty-six
Moorey chose to begin his own introduction to ancient Mesopota-

mian materials with a quote from Leo Oppenheim (1978: 646) to the
effect that the notable lack of attention to technological achievements
in the past was directly proportionate to a lack of scholarly interest in
material culture. Indeed, Oppenheim’s own manuscript on material
culture of ancient Mesopotamia languishes unpublished to this day,
although happily, the scholarly environment has changed significantly
since Oppenheim wrote: several conferences directly concerned with
approaches to the material production of societies have taken place
(for example, Appadurai ed. 1986; Lubar and Kingery eds. 1993) and
theoretical attention to the role of material culture in socio-political,
economic and symbolic systems has significantly altered the ways in
which we view the archaeological landscape (Giddens 1992; Helms
1993). In addition, with the continuing publication of the textual
corpus in both Sumerian and Akkadian (e.g. Frayne 1990: Edzard,
1997), and the appearance of a number of archival studies that have
provided information on specific aspects of craft production (Loding
1974, 1981; Neumann 1987; Mieroop 1987), it is now possible for the
non-philologist to have access to vocabulary for, descriptions of, and
context of usage appropriate to the materials and works so masterfully
presented by Moorey—both in 1985 and subsequently in 1994. Inter-
estingly, the latter study, whether dealing with resource procurement or
craft production, does place considerably greater reliance upon textual
evidence than the former.
Nevertheless, a synthetic study of the place of different artisan tradi-
tions within Mesopotamian culture that exploits equally the artefactual
and the textual evidence has yet to be written. In the paper that follows,
I should like to pursue one aspect of this subject: the ‘economy of value’
beyond the socio-economic—that is, the environment of (aesthetic)
appreciation—in which early Mesopotamian production flourished. A
pioneering study in this direction was undertaken by Jack Sasson for
artisans and artefacts from Mari (Sasson 1990), which laid out several
avenues for the pursuit of meaning and value; and at his behest, I
myself began an inquiry into these issues (Winter 1995). In the present
study, I would like to expand a bit on the question of value associated
with qualities usually tied to production: mastery, embellishment and
claims of ‘perfection’—particularly with respect to the period from the
mid-third to the early second millennium b.c.2 Once Sumerian works
2
That the materials used were also perceived as possessing ‘value’, both symbolic
“surpassing work” 273
of the Early Dynastic through the Ur III period are put together with
Sumerian textual references to similar or related works, it becomes
clear that a high degree of mastery is not only evident retrospectively
as we examine the objects (consistent with works we would designate
‘art’ in our terms); the attributes noted above—mastery, embellishment
and perfection—along with the resultant high-quality works that bore
these attributes were also celebrated in their own time.
It will come as no surprise that a complex culture with a specialist
labour force and sophisticated material production will have developed
a complex vocabulary for the making of works and the assessment
thereof. Fortunately in the textual record we may distinguish a variety
of verbs in Sumerian, along with their Akkadian counterparts from
lexical lists and bilinguals, that differentiate drawing (Sum. ur, šab;
Akk. e ēru), carving (Sum. bal, gul; Akk. naqāru), mounting in precious
metal (Sum. gar; Akk. u uzu), and building (Sum. du3; Akk. ra āpu).
These verbal forms correspond to the production of classes of works
referred to in text: the production of images in paint, relief, and/or
free-standing sculpture; of metal vessels, divine and royal insignia,
and ornaments; and of buildings and their appointments. There also
occur more general verbs that refer to the act of creation/production
(Sum. dim2 ak, kin; Akk. banû, bašāmu, patāqu, meaning ‘to make, cre-
ate, fashion, form’) and substantives that suggest formal plans and/or
execution according to a plan (Sum. giš. ur; Akk. šute bû, particularly
with respect to architecture).
This last is evident in the written account by Gudea of Lagash of the
building of the Eninnu temple for the god Ningirsu, where the god Ea/
Enki, known in later Akkadian texts as bêl nēmeqi, ‘Lord of Wisdom/
Skill/Craft’, is presented as overseeing the plan for the temple Gudea has
designed. We are told explicitly that ‘Enki “straightened out” the plan
of the temple for him, e2-a den-ki-ke4 giš- ur-bi si mu-na-sa2 (Edzard
1997: 80: Cyl. A. xvii:17); and it is assumed to be this ‘plan’ that Gudea
holds on his lap in the well-known Statue ‘B’ from Tello (fig. 1).3
and material, goes without saying (see, for example, Dijk 1983: Ross 1999: Winter
1999), but that is not the subject of inquiry here.
3
On the giš-ur as wooden writing board on which a ‘plan’ or anything else would
be drawn, see also Veenhof 1995; and see also CAD G: 101, giš uru. Dolce (2001: 372)
has recently argued that actual architectural plans incised on clay tablets are more likely
to have been records of existing structures than working drawings for construction. For
the verb ak, see now PSD AIII: 72–73, ak 3, ‘to make, fashion, construct;’ also 120–123:
lexical section, for meanings, ‘to form, mold, erect, set up.’ Note also that with respect
to the production of human and divine statues, the verb most frequently employed is
The lexicon for acts of artisan production and for a variety of finished
products (e.g. Foxfog 1980) would have little utility for us in establishing
value as distinct from process, were it not that often descriptions of work
employing such verbs also include references, to the skill or wisdom
employed in carrying out the endeavour or as manifest in the finished
work. An elaborate vocabulary exists for conveying this expertise, both
in Sumerian and Akkadian: Sum. nam-ku3-zu, Akk. nēmequ, literally
‘knowledge(ably), skilful(ly), expert(ly-made)’; Sum. galam, Akk. nakliš,
nikiltu, ‘masterful, artful, ingenious’. These terms can be employed
metaphorically, as in a Sumerian love song, in which the beloved is
compared to expertly-crafted works of art:
nig2 nagar ku3-zu dim2-ma-mu
My (awesome) thing, made by the skilled carpenter;
tibira ku3-zu kin aka-a-mu

My (great) work fashioned by the skilled metal/inlay-worker.4
The same vocabulary is also employed by rulers who claim divine
merit or historical credit for the production of valued works, pointing
not dim2, but rather tu(d), signifying creative ‘birthing’ as opposed to more material
manufacture (e.g. Early Dynastic period Urnanše text (Steible 1982: 89 Urn. 24, 2.2)
Gudea Statue ‘B’ (Steible 1991: 170 7.13; Enmerkar and the Lord of Aratta (Cohen
1973: 318), cited in PSD AIII, 166 and see discussion in Winter 1992: 21–3).
4
Alster 1985: 131, 133 r. I:9–10, who translates the same passage: ‘piece (of art)
shaped by the skilled carpenter, my (beloved) manufactured by the skilled coppersmith’
[cited also in Ross 1999: 1065 and translated: ‘my thing fashioned by the expert car-
penter, my (thing) worked by the expert inlayer’]. The two lines are in direct apposition
to one another: kin and nig2 as paired products, synonyms for made objects; verbs of
making, dim2 and ak, similarly paired. While the literal meaning of ku3 is ‘pure’, it
is suggested that in this context ku3-zu, as in nam-ku3-zu, conveys perhaps a ritually
pure, but surely also an appropriately skilled or expert, worker/craftsman. For tibira
see Flückiger-Hawker’s discussion of this as a craftsman working in metal and wood,
possibly a cabinet maker (1999: 174; reference also in Dumuzi and Inanna 47, cited
in PSD AIII, 90: ak 8: tibira-ku3-zu kin ak-a-mu, ‘my clever tibira-craftsman, who
executes the work . . .’). I wonder whether it might not convey one who works in several
media or materials at once, particularly characteristic of the various inlaid works from
the Royal Cemetery of Ur—for example, the musical instruments (e.g. fig. 2), where
gold and other precious metals are combined with precious stones, such as lapis lazuli
and carnelian, as well as shell, and this is indeed how Ross has translated the term
(see also the text of ‘Enki and the World Order’ [cited in Ross 1999: 1082, ll. 408], in
which the goddess Nin-mu(g) is called the ‘inlay worker of the land,’ tibira kalam-ma).
For emphasis in this context, I have included ‘. . . (awesome) thing . . .’, as this is indeed
how the term nig2 is inflected in many usages; then, in parallel, ‘. . . (great) work . . .’
corresponds to the value implied by the lover for her beloved in the text.
to their own, often divinely inspired, ingenuity or to the skills of crafts-

men employed in their undertakings. Instances also abound in which
adjectives denoting skill, mastery and/or ingenuity are attached directly
to the substantive work being commemorated. A text attributed to Ur-
Namma of Ur, for example, describes his construction of the sanctuary
Ekišnugal, speaking of it as galam, ‘skillfully built’ (Flückiger-Hawker
1999: 266–7 Urnamma E/F, l. 3'), while an Old Babylonian version of
the Sumerian myth Enki’s Journey to Nippur provides actual details of a
skillfully-made temple facade, as we are told that the god built his house
of precious metal and lapis lazuli, ‘its artful facade emerging from the
Abzu,’ muš3-ku3 galam du11-ga abzu-ta e3-a (cited in Ross 1999, 1090).5
In another text, dating to the Early Dynastic Period, Enannatum I of
Lagash tells us that he built the Ibgal for Inanna, decorating it and
furnishing it, and making it ‘surpassing (diri) over the land’.6
The whole array of terms is called forth in a single Sumerian text
describing a statue of silver and a precious stone that Abî-Sarē, the
king of Larsa, has had made. We are first told that in the past no
such statue had existed, but that the god Enki gave to the ruler the
great intelligence to do a surpassing job (Frayne 1990: 122: i 28'–31';
ii l'–3'—diri geštu2-ma nig2-nam-ma diri-ga). Finally, the work itself
is assessed, and the terminology employed comes as close as can be to
the Western concept of ‘master work’:
nam-ku3-zu ak . . . it is expertly fashioned;
me-dim2-bi3 this work (is)
me-dim2-ma diri-ga (a) surpassing work,
nig2-ar-eš dib-b[a] a thing of (beyond?) praise.7
While one might think of these words as referring more to reception
than to a description of manufacture, what I would call attention to
is the emphasis on skill and expertise in the process of making that is
recorded as part of the value of the end product. Surely, when one
5
See Steinkeller 2002: 360, n. 7 for a different reading: su galam du11-ga abzu-ta
e3-a, ‘it is (like) a skillfully made head-band (reading muš3-ku3 as su ) rising from the
Abzu. For the complexity of the meaning of muš3 in architecture, see discussion by
Sjöberg (Sjöberg and Bergmann 1969: 17, 53 and 55), where it is often translated as
‘crown’, although facade or upper works is probably preferable.
6
Steible 1982: 187 En. I, 10: 1.9. The text can be read: ‘higher than (or surpass-
ing) the mountains’, kur-kur-ra mu-na-diri-ga-a, but the context suggests the less literal
reading.
7
Frayne 1990: 122 Abî-Sarē 1:ii.6'–10'.
is confronted with works of quality from the archaeological record,

especially works in which various materials have been skillfully joined
to produce striking representations, such as the rearing goats or various
of the musical instruments from the Royal Cemetery of Ur (e.g. fig. 2),
this is a vocabulary we would be comfortable with today.
In the various city-states of ancient Sumer, including Lagash, Ur
and Isin, such works were often deemed of sufficient import that their
construction and/or dedication became the signature event in a num-
ber of regnal year-names. For example, Gudea named his third year
‘the year the lyre, “Great Predator of the Land”, was fashioned’; one
of his successors, Ur-Baba, named a year after the construction of
the temple of the goddess Baba; Ur-Namma of Ur named his sixth
year after the construction of the Ninsun temple in Ur; his son Šulgi
named his third year the year in which ‘the chariot of Ninlil was built’,
his fourth the year in which the foundation of the temple of Ninurta
was laid, and his sixteenth the year in which the bed of the goddess
Ninlil was made; Amar-Su’en of Ur called his third year ‘the year the
king . . . joyfully made the divine throne of Enlil’; Išbi-Erra of Isin called
his sixth year ‘the year a bed for Inanna was fashioned’, his eighteenth
‘the year . . . the king fashioned a great emblem for Enlil and Ninurta’,
his 20th ‘the year . . . the king fashioned . . . the great emblem of Inanna’
and his 25th ‘the year . . . the king fashioned a dais/throne for Ninurta;’
and finally, his successor, Šu-ilišu, commemorated the fashioning of a
great emblem for Nanna, an exalted throne for An, a dais for Ninisin,
a magur-boat for Ninurta, and a dais for Ningal in his years 2, 5, 7,
8, and 9, respectively.8 In the later reigns of the Isin Dynasty, more
graphic description is provided, as, for example Bur-Sîn’s year ‘D’: ‘Year
Bur-Sîn the king fashioned for Nin-Isin a copper stand . . . (representing)
an overflowing stream and a lofty copper platform for offerings,’ or his
year ‘E’: ‘Year Bur-Sîn the king made for Ninurta . . . a three-headed
gold mace with heads of lapis-lazuli as a great emblem for Ninurta;’
his successor Lipit-Enlil called his year ‘A’: ‘Year Lipit-Enlil the king
fashioned for the temple of Enlil a large golden vase with handles
called “Enlil is exalted” ’; and, most important, Enlil-bani’s year ‘J’
is called: ‘Year Enlil-bani the king fashioned a couch decorated with
gold and silver, a work for the great sanctuary of (the god) Enki’—mu
8
See PSD ‘B’: 75: balag A; also Sigrist and Gomi 1991: 317, 318, 319, 321 and
325; Sigrist 1988: 13, 17, 19, 22 and 23.
d
en-lil-ba-ni lugal-e gišgu-za zag-be2-us2 ku3-sig17 ku3-babbar-ta kin-gal-eš
d
en-ki-ra mu-na-dim2.9
These year names provide a glimpse of the value attached to, and the
precious materials and investment in craftsmanship likely to have been
characteristic of, the works commemorated. Although they represent
but selected examples, the message is clear: sharing the calendar with
other important markers of a ruler’s reign such as victory in battle,
the dedication of canals, road construction, the introduction of dei-
ties into new temples and the installation of en-priestesses and priests,
we must conclude that the objects mentioned in the year names were
perceived as works of great value. And just as the statue made for Abî-
Sarē of Larsa, discussed above, we would expect works like the couch
for Enki mentioned by Enlil-bani and the other objects designated in
year-names to have been described in more discursive texts with similar
designations—‘surpassing works,’ ‘expertly fashioned,’ and ‘worthy of
praise.’ Such designations would be consistent also with references to
thrones and daises for the gods Nanna, Ningal and Utu recorded by
Warad-Sin of Larsa, and a rock crystal vase with gold rim and silver
base recorded by Rim-Sin I of Larsa (Frayne 1990: 220–2 and 306),
all of which were made of precious materials and clearly highly valued
by their commissioning rulers.10
9
Sigrist 1988: 30, 31 and 34. See also discussion by Selz 1997: 198, n. 197 and passim
of the god-like status of statues and other cultic objects mentioned in year names.
10
Of the crafts employed in the making of such works, we have several enumerations
(for example, in the myth of Inanna and Enki), where carpentry, inlay work, leather
work, metal-smithing, architecture and reed work are mentioned in sequence (Farber-
Flüge 1973: 22–3, I. ll. iii. 8'–11' and iv. 10'–17', cited in Ross 1990: 1042). Studies such
as Loding (1974) and Van De Mieroop (1987), as noted above, have begun to look at
particular third and second millennium workshop archives in terms of craft activities;
indeed, their works and others are summarized in Moorey (1994: 14–7). More needs to
be known, however, about any hierarchy of Sumerian craftsmen producing such works,
before one can get at a sense of expertise and ‘quality’ in manufacture. Sasson (1990:
22–23) implied such a hierarchy when he noted distinctions in the Akkadian texts from
Mari differentiating some craftsmen as ‘competent’ and others as ‘reliable/experienced’
(see also Moorey 1994: 15). In later, particularly Neo-Assyrian, texts craftsmen are occa-
sionally referred to as ummânu, literally ‘master’ or ‘expert’ (on which, see Parpola 1983:
270, who identifies them as ‘highly trained experts of specific crafts’; see also Parpola
1983: xx and Ataç 2000 for the association of representative ‘experts’ with scribal and
sacerdotal practice; also PSD AIII, 108: dub-sar umum ak, ‘expert scribe’). Lanfranchi
and Parpola (1990: 50: No. 56) further discuss an Assyrian text describing architectural
work in which ‘master builders’ are distinguished from juniors/apprentices. Loding
(1974: 142–4) proposed Sum. giš-kin-ti as an Ur III equivalent for ummânu; Thureau-
Dangin (1921, cited in Ross 1999, 1069) noted Sum. lu2dumu-meš um.man, ‘experts,’ in
a late bilingual text on the Akitu-festival; and Glassner (1995: 1815) cites UM.ME.A =
Often, judgments of quality are inherent in the addition of terms of

value to the work, as in Sum. me-dim2-sa6 or kin-sa6-ga, as markedly
‘good’ or ‘auspicious’ work.11 The sense behind that distinction, along
with the underlying qualities expressed by the Sumerian terms galam
and nam-ku3-zu, comes quite close to a ‘work of art’ in our terms, or at
least in comparable terms to the Medieval European notion of expert
production: that is, things expertly, masterfully, artfully, knowledge-
ably made and valued accordingly.12 Thus, it is not uncommon in the
Sumerian temple hymns to speak of temples built, or established, in an
artful fashion. For example, in the hymn to the temple of Nin-isina in
Isin, it is declared that ‘its interior is built in an artful fashion’, ša3-bi
ummânu as ‘a possessor of specialized knowledge or craft, scribe, artist, artisan’. In

the late bilingual, the generic reference to (expert) craftsmen comes immediately
after the designation of individual craft specialists, jewellers and workers in precious
metal (lu2ku3-dim2), carpenter (lu2nagar) and weaver (lu2uš-bar). Other texts, such as the
Gudea Cylinder A, the Curse of Agade and the Lugalbanda Epic, provide terms for
lapidaries and coppersmiths (za-dim2, simug) (Edzard 1990: xvi, 25–30, cited by Ross
1999: 830; Ross 1999: 1084 l. 41, 1094 ll. 409–410; discussion also in Loding 1974:
271–275); and for sculptor (lu2alam), see PSD AIII: 170: alam. Such expert craftsmen
were clearly valued sufficiently that they are expressly mentioned in Lugalbanda as hav-
ing been transferred to Uruk after the destruction of Aratta, while earlier, as reflected
in Enmerkar and the Lord of Aratta, craftsmen from Aratta were requested to be
sent to Sumer along with precious raw materials such as gold, silver and lapis lazuli
(Cohen 1973; Zaccagnini 1993: 38). Several Old Akkadian references to a ‘lieutenant
of the metalworkers,’ nu-banda2 lu2ku3-dim2, seem to suggest some ranked hierarchy or
achieved status within individual craft specializations, at least in that period (see Ross
1990, 894 and 1020). In Gudea Cylinder A, the combination of terms sanga simug
has led to the translation: ‘chief of the (copper) smiths’ (Edzard 1997: xvi, 29); and one
would want to know more about the status of individuals designated as ugula (Sum.
‘overseer’). Although various craftsmen are documented for the Inanna Temple of Nip-
pur in the Ur III period, including metalworkers in copper/bronze and gold (Zettler,
1992: 226–31), craftsmen tend to be referred to as individuals, without reference to
hierarchy. As noted by Moorey (1994: 16), the textual tradition preserved to us is not
likely to provide a coherent account of an artisan’s craft and there is much we do not
know to date about the organization of workshops, the training of craftsmen and the
actual practice of the various crafts that would have gone into the production of works
referred to as of high value in the public texts we have been citing.
11
Other terms seem also to convey fine/quality work, as in the epic Enlildiriše,
l. 300 (cited PSD AIII, 161: alam 1), where copper statues are referred to as nig2-kala-
ga, ‘mighty things.’ But more than that, there is further indication that manufacture
of certain works, particularly those intended for cultic use, were not just the products
of secular labour. In Enmerkar, for example, mention is made of ‘a golden statue
fashioned on a propitious day’, alam-ku3-sig17-ga u4-du]0-ga tu-da, literally ‘birthed
on a good/sweet day’, suggesting that some sort of ritual selection of the auspicious
attended the consecration.
12
See on this the introduction by J. G. Hawthorne and C. S. Smith to On Divers Arts:
The Treatise of Theophilus (1963: ii, xxxiv), which chronicles various crafts recorded by
a Benedictine monk of the 12th century. Also Koerner 1999, for a discussion of the
power and meaning of ‘making’ in the later Middle Ages and early Renaissance.
galam-kad5-am3 (Sjöberg and Bergmann 1969: TH 30, 380; also TH

15, 187, TH 36, 461).13 Indeed, references of this sort are applied to
the full range of elite works, including temples, cult implements and
statuary.
In this sense, craftsmanship is frequently recognized as ‘artful’ and/or
‘ingenious’, attesting to the skill, inspiration and inventiveness required
as part of the creative process and then perceived as part of the value
inherent in the work. It will be noted in consulting the translations of
a number of scholars that words like Sum. galam or nam-ku3-zu are
frequently rendered as ‘artistic,’ or ‘artistically made’. This I would
avoid, substituting ‘artful’ as a term better conveying the sense of mas-
tery or skill (as at the root of Latin ars, Greek techne, as per Vickers and
Gill 1996) that I believe is implied, without assuming an autonomous
category of ‘art’ work, hence the ‘artistic,’ so far undemonstrated for
Mesopotamia—that is, works valued as members of a class of art as
such, independent of the intended context of use.14
In addition to skill and mastery as general terms of value, there are a
number of references to the quality of ‘embellishment’ or ‘decoration’
per se applied to works—viewed always as a positive attribute. Here,
I would argue even further that the fact of decoration or embellish-
ment is perceived as an inseparable artefact of skilled making. Once
again the vocabulary is well-developed. Verbs such as Sum. šu . . . tag
and šerkan . . . du11, ‘to decorate, adorn’, embellish’ or the quality of
being adorned or decorated (Sum. šerkandi) are frequently included in
descriptions of important works, particularly in the case of elite buildings
and precious, often cultic, objects.15 As examples, both Entemena and
Gudea of Lagash (c. 2400 and 2110 b.c., respectively) make reference
to the decorative programs of the Ningirsu temple in Girsu:
13
Since kad5 has posed somewhat of a problem of interpretation (see Sjöberg &
Bergmann 1969: 122), it may be useful to pursue the meaning through an Akkadian
lexical equivalent to kasāru, which relates to some sort of weaving, or tying of knots,
binding together and also joining in architecture (CAD ‘K’: 258 kasāru 1b), hence the
possible sense of ‘artfully constructed/put together’.
14
This would not preclude an argument that, even without a word for ‘art’, Meso-
potamian works meet the criteria for ‘art’, as discussed by Denis Dutton (2000: 233–6),
including requisite skill, rules of form and fashioning, a critical language of judgement,
consciousness of the ‘special’ nature of the works and of charged experience for both
producers and audiences. In this, Dutton’s enterprise to examine the status of ‘tribal’
works must be expanded to include all ‘others’, that is, non- as well as pre-Western
enlightenment artistic production associated with state organization.
15
For a study of the terms for and forms of geometric designs in Sumerian and
Akkadian that would describe non-figural elaboration, see Kilmer 1990.
guškin ku3-babbar2-ra
šu mu-na-ni-tag
(having built the Antasurra-temple of Ningirsu), with gold and silver he
(Entemena) decorated it.
gu3-de2-a še-er-zi-an-na-ka šu-tag ba-ni-du11

Gudea decorated it (the temple) with the splendour of heaven.16
Other references include the attention paid to separate parts of build-
ings and cultic objects, such as a ceremonial chariot for the god Nin-
girsu, a processional boat and stone and metal cult vessels. Gudea, for
example, described the chariot he had been instructed to make for the
Ebabbar: gišgigir-bi ku3.ne za-gin3-na šu u3-ma-ni-tag, ‘this chariot would
you (Gudea) decorate with silver and lapis’ (Edzard 1997: 73: Cyl. A,
vi:19), while Šulgi of Ur referred to the ceremonial magur-boat he has
constructed as an-gim mul-a še-er-ka-an mi-ni-ib2-du11, ‘decorated
with stars like heaven’ (Klein 1981: 12, 86).17 Another text of the Ur III
period, attributed to the father of Šulgi, Ur-Namma, refers to temple
door-lintels ‘decorated (še-er-ka-an . . . du11) with electrum and pure sil-
ver’, on which ‘the Anzû-bird has spread its talons’ (Flückiger-Hawker
1999: 191 ll. 22–24). This certainly calls to mind the copper/bronze
door lintel plaque recovered from the Temple of Ninhursag at al-
Ubaid, dated to the Early Dynastic Period (fig. 3), masterfully executed
as three-dimensional figures of the Imdugud/Anzû-bird between two
stags, encased within a rectangular frame, which permits us to visualize
what a decorated door lintel would have looked like.18 In another Ur
16
Steible 1982: 219 Ent. 16:2.3–4; Edzard 1997: 86 Cyl. A. xxviii:1–2. I resist
Edzard’s translation of ‘painted’, preferring the more generic rendering of ‘decorated’,
even if the most likely medium of decoration would have been paint, since we cannot
rule out other media, such as carved elements or appliqués of precious metal. See
also Klein 1989 re Ur-Namma and Šulgi making similar claims. Note also that in
Enannatum I’s description of his building of the Ibgal for Inanna, he also spoke of
having decorated it for her with gold and silver, dinanna-ra ib-gal mu-na-du3 . . . ku3-sig17
ku3-babbar2-ra šu mu-na-ni-tag (Steible 1982: 185 En. I 9, ll. 3.3–9), while in the text
known as Enki’s Journey to Nippur, of Old Babylonian date, we are told that the god’s
house/temple was built with silver, coloured with lapis lazuli and ‘lavishly decorated
with gold’, gal-le-eš ku3-sig17-ga šu-tag ba-ni-in-du11 (cited in Ross 1990: 1090).
17
It is interesting that a later Ur III king, Šu-Sin, also had a magur-boat built,
commemorated in the name for his eighth regnal year: mu dšu-den.zu . . . ma2-gur8 ma
d
en-lil2 dnin-lil2-ra mu-ne-dim2 ‘the year Šu-Sin . . . constructed the exalted magur-boat
for Enlil and Ninlil’ (Sigrist and Gomi 1991: 327).
18
That the elaboration of the temple gate constituted an important aspect of its
visual impact is further reflected in the description of the destroyed sanctuary of Ur,
in which ‘the great door ornament of the temple was felled, its parapet destroyed; the
III text, precious stones and metals are used ‘to decorate the statue of
the ruler Šu-Su’en, [ala]m-dšu-den.zu [šu d]u7-du7-de3 (cited PSD AIII: 167:
alam 1.6.4). And finally, a text of the late Ur III ruler, Ibbi-Su’en,
speaks of a gold vessel, likely an ointment jar, that seems to possess all
of the properties we have been discussing:
bur-šakan-ku3-sig17 kin ga-lam kad5! [=pap?]
gu4-alim-muš-ba šeg3-gi6-ni2-IL2 [=gur3?]
še-er-ga-an-du11-ga-bi u6-di nu-til-le-dam
gold šakan-vessel, work (of ) artful/ingenious construction,
bearing wild bulls and serpents,
its decoration of unending admiration19
This last must surely evoke the well-known silver vase of Entemena,
found at Tello (ancient Girsu), which stands upon a copper base (fig. 4).
Around the collar is a dedicatory inscription commemorating the ruler
and his presentation of the vessel to Ningirsu and identifying it as a
gurgur-vessel for monthly offerings to the deity (Steible 1982: Ent. 34
l. 15; Selz 1993; for other cult vessels, see Braun-Holzinger 1989; Selz
1996). It is decorated on the shoulder with incised designs of couch-
ant calves and, around the body, as on the door lintel from al-Ubaid
mentioned above, heraldic motifs of the lion-headed Imdugud/Anzû,
bird of the thunderstorm, his talons sunk into the backs of four pairs
of animals—caprids, lions, bull calves and more lions. Moorey (1985:
115) has aptly called this vase ‘the outstanding silver object to have
survived from Early Dynastic Mesopotamia’. That such a vessel would
have been viewed in its own time as representing masterful/ingenious
work, and consequently, would have been appreciated as an object of
unending admiration, seems entirely appropriate. What is important
for our purposes is the linkage established in the text, as reflected in
the actual the silver vase, between quality of work and response: from
masterful and ingenious object, to its decoration and the resultant
unending admiration.
wild animals that were intertwined on its left and right lay before it like fallen warriors’
(Michalowski 1989: 63 ll. 420–2).
19
Steible 1991: 286: Ibbi-Su’en 9/10 ll. 17–22; also presented in transliteration
with translation in the dissertation of Mark Hall (1985: 123); see also discussion in
Selz 1996. Note that adornment is also referenced metaphorically in poetry, as in the
love song ‘Dumuzi and Inanna’ (Sefati 1998: 129–130 l. 32), in which the beloved is
alluded to as ‘alabaster figurine, adorned with the lapis diadem, sweet is your allure,’
dim3-giš-nu11-gal *su -za-gin3 keš2 i-li-zu ze2-ba-am3.
A similar correlation can be found in a text of the reign of Rim-Sin

of Larsa (Frayne 1990: 303 ll. 23–24), in which the ruler speaks of a
diorite water vessel or trough he has dedicated to the goddess Inanna,
which is deemed to be:
šu-tag-ga-še3 tum2-ma
nig2 u4-bi-ta nin-igi-du-mu-ne ba-ra-an-dim2-ma-a
fitting for adornment,
a thing no previous queen had fashioned
What may be concluded from the above references and discussion is
that ornament, or embellishment, is essential—not merely additive—to
the value and efficacy of the final product. This is entirely in keeping
with what was described by A. K. Coomaraswamy (1946: 85 and 96)
for the traditional arts of Sanskritic India, in which ‘ “ornament” and
“decoration” are, properly speaking, integral factors of the beauty of
the work . . ., essential to their utility . . . and their perfection as such’.
Similarly, it is clear from the Sumerian sources that objects receive
the crowning degree of praise when they are deemed to have been
‘perfected’ (Sum. šu . . . du7) through their craftsmanship and embel-
lishment. One such instance is to be found in the hymn known as
Ur-Namma A, ‘The Death of Ur-Namma’, in which the Ur III ruler
is said to bring gifts to various deities of the Netherworld. The gifts
include a bur-šagan šu du7-a, ‘a šagan-vessel of perfect make’ and
gil-sa šu du7-a, ‘perfectly-wrought jewellery’ (Flückiger-Hawker 1999:
118–120 ll. 97 and 106).20
Such an appraisal of ‘perfection’ for finely crafted works fits well
with the insistence of anthropologist Franz Boas in his fundamental
study of non-Western art (1955: 10–11) that ‘the judgment of perfec-
tion of technical form is essentially an esthetic judgment . . . Since a
perfect standard of form can be obtained only in a highly developed
and perfectly controlled technique, there must be an intimate relation
between technique and a feeling for beauty.’ Underlying Boas’ argument
is an insistence upon the principle that in the value systems and
(aesthetic) responses of many cultures around the world, one cannot
20
The Susa version of this same text further describes such a piece of jewellery: ‘a
pin made of gold and silver, whose head/finial (is) that of a bison’ (Flückiger-Hawker
1999: 161 l. 98'). This last corresponds well to actual objects found in the Early Dynastic
graves at Ur (e.g. Woolley 1934: pl. 145 (gold pin with lapis head) and pl. 231 (cop-
per/bronze pin with horned head)) and one rather wishes intact graves of the rulers
of the Third Dynasty of Ur had been preserved as well.
separate perception and/or judgment of form from an underlying

conception of the ‘class’ to which the work belonged and a standard
of measure according to which the immediate example’s place in the
universe would be assessed.
Clearly, in formulating this position, Boas was in dialogue with West-
ern aesthetic theory, such as that put forward by Immanuel Kant in the
late 18th century. Kant’s inquiry into the nature of aesthetic judgment
quite explicitly precluded such a relationship between a ‘pure’ judge-
ment of beauty and an assessment of technique, or perfection. In his
Third Critique, for example, Kant stated that ‘the judgement of taste is
entirely independent from the concept of perfection’, since to include
such a concept would be to necessarily introduce a prior concept of
the object’s intended utility, which, he claimed, must remain separate
from any response to its ‘beauty’ (Kant/Guyer 2000: 111 § 15). I have
attempted to deal elsewhere with the futility of such a separation for
the art of non-Western or even pre-modern Western cultures (Winter
in press), specifically for that of ancient Mesopotamia; but I insist upon
it here, because I would argue that the claims of early Mesopotamian
rulers that they were responsible for the production of elaborately embel-
lished, skilled or masterful works is directly linked to the description of
the various attributes of such works, which then occasioned an aesthetic
response. What is remarkable is precisely the fact that, as we have seen,
Sumerian rulers thought it necessary to stress the process of making
and the attributes of skilled-production as part of the larger picture of
aesthetic and cultural value.
In conclusion, ascriptions of mastery, embellishment and perfection,
insofar as these properties were asserted to be manifest in individual
works, served to confer ‘quality’ upon the finished product. In claim-
ing these properties for works of royal patronage, rulers may be seen
to anticipate/beg the question of the audience’s response for their
own rhetorical purposes; however, in the absence of more objective
contemporary critical records, we must see those properties related
to production as inseparable from, and part of, the overall stimulus
leading to aesthetic response itself. This comes close to what Kant
ultimately defined as ‘adherent’ or ‘dependent’ beauty (Kant/Guyer
2000: 114–116 § 16), a property that, unlike ‘pure’ beauty, could be
measured against a standard of perfection. For ancient Sumer, then,
mastery, embellishment and perfection not only constitute perceptible
affective properties that contributed to aesthetic experience at the high-
est cultural level; they also represent necessary attributes of works that
qualify as ‘great’.
References Cited
Alster, B. 1985. Sumerian Love Songs. Révue d’Assyriologie et d’Archéologie Orientale 79:
127–59.
Appadurai, A. ed. 1986. The Social Life of Things: Commodities in Cultural Perspective.
Ataç, M.-A. 2003. Visual Decorum and the Agency of Scribal and Sacerdotal Per-
sonnel in the Production of the Neo-Assyrian Palace Reliefs. Ph.D. dissertation,
Harvard University.
Azevedo, W. d’ 1958. A structural approach to esthetics: Toward a definition of art in
anthropology. American Anthropologist 60: 702–14.
Boas, F. 1955. Primitive Art. New York, Dover.
Braun-Holzinger, E. A. 1989. REC 445.LÁ = Libationsbecher. Zeitschrift für Assyriologie
und Vorderasiatische Archäologie 79: 1–7.
Cohen, S. 1973. Enmerkar and the Lord of Aratta. Unpublished PhD dissertation.
University of Pennsylvania.
Coomaraswamy, A. K. 1946. Figures of Speech or Figures of Thought: Collected Essays on the
Traditional or ‘Normal’ View of Art. London: Luzac.
Dijk, J. van 1985. Lugal ud me-lám-bi Nir-gál. Leiden: Brill.
Dolce, R. 2000. Some architectural drawings on clay tablets: Examples of planning
activity or sketches?. In Proceedings of the 1st International Congress on the Archaeology
of the Ancient Near East, Rome, 18–23 May, 1998 (P. Matthiae, A. Enea, L. Peyronel
and F. Pinnock, eds.). Rome: Università degli Studi di Roma, La Sapienza, Vol. 1,
365–95.
Dutton, D. 2000. ‘But they don’t have our concept of Art’. In Theories of Art Today
(N. Carroll, ed.). Madison: University of Wisconsin Press, 217–40.
Edzard, D. O. 1997. Gudea and his Dynasty. Royal Inscriptions of Mesopotamia. Early
Periods. Vol. 3/1. Toronto, University of Toronto Press.
Farber-Flügge, G. 1973. Der Mythos ‘Inanna und Enki’ unter besonderer Berücksichtigung der
Liste der me. Studia Pohl 10. Rome: Biblical Institute Press.
Flückiger-Hawker, E. 1999. Urnamma of Ur in Sumerian Literary Tradition. Orbis Biblicus
et Orientalis 166. Fribourg, Switzerland: Fribourg University Press.
Foxfog, D. A. 1980. Funerary furnishings in an early Sumerian text from Adab. In Death
in Mesopotamia (B. Alster, ed.). Mesopotamia 8, Copenhagen: Akademisk Forlag, 67–75.
Frayne, D. 1990. Old Babylonian Period (2003–1595 B.C.). Royal Inscriptions of Meso-
potamia, Early Periods Vol. 4. Toronto: University of Toronto Press.
Giddens, A. 1992. Sociology and the Explanation of Human Behaviour. In Human
Societies: An Introductory Reader in Sociology (A. Giddens, ed.). Cambridge: Polity Press,
362–65.
Glassner, J.-J. 1995. The uses of knowledge in ancient Mesopotamia. In Sasson ed.
1995: Vol. 3, 1815–23.
Hawthorne, J. G. and Smith, C. S. transl. 1963. On Divers Arts: The Treatise of Theophilus.
Chicago: University of Chicago Press.
Helms, M. W. 1993. Craft and the Kingly Ideal: Art, Trade and Power. Austin: University
of Texas Press.
Kant, I. 2000. (1790) Critique of the Aesthetic Power of Judgment, (P. Guyer, ed. and translator).
Kilmer, A. D. 1990. Sumerian and Akkadian names for designs and geometric shapes.
In Investigating Artistic Environments in the Ancient Near East (A. C. Gunter, ed.). Wash-
ington, DC: Smithsonian Institution, 83–91.
Klein, J. 1981. Three Šulgi Hymns. Sumerian Royal Hymns Glorifying King ulgi of Ur. Bar Ilan
Studies in Near Eastern Languages and Cultures. Ramat Gan: Bar Ilan University.
——. 1989. Building and Dedication Hymns in Sumerian Literature. Acta Sumerologica
11: 27–67.
Koerner, J. 1999. Factura. Res 36: 5–19.

Lanfranchi, G. and Parpola, S. 1990. The Correspondence of Sargon II, Pt.2: Northern and
Northeastern Provinces. State Archives of Assyria 5. Helsinki: University of Helsinki
Press.
Loding, D. 1974. Craft Archive from Ur. Ann Arbor: University Microfilms.
——. 1981. Lapidaries in the Ur III Period. Expedition 23: 6–14.
Lubar, S. and Kingery, W. D. eds. 1993. History from Things: Essays on Material Culture.
Washington and London: Smithsonian Institution.
Michalowski, P. 1989. The Lamentation over the Destruction of Sumer and Ur. Winona Lake:
Eisenbrauns.
Mieroop, Marc Van De 1987. Crafts in the Early Isin Period. Orientalia Lovaniensia
Analecta 24. Louvain: Departement Orientalistiek.
Moorey, P. R. S. 1985. Materials and Manufacture in Ancient Mesopotamia: The Evidence of
Archaeology and Art. Metals and metalwork, glazed materials and glass. BAR International
Series, 237. Oxford: British Archaeological Reports.
——. 1994. Ancient Mesopotamian Materials and Industries: The Archaeological Evidence. Oxford:
Clarendon Press.
Neumann, H. 1987. Handwerk in Mesopotamien, Schriften zur Geschichte und Kultur
des Alten Orients 19. Berlin: Akademie Verlag.
Oppenheim, A. L. 1978. Man and Nature in Mesopotamian Civilization. Dictionary of
Scientific Biography 15. New York: Scribner, Supplement 1, 634–666.
Parpola, S. 1983. Letters to Assyrian Scholars, Part II: Commentary and Appendix. Alter Orient
und Altes Testament 5/2. Neukirchen-Vluyn: Neukirchener Verlag.
Parrot, A. 1948. Tello, vingt campagnes de fouilles (1877–1933). Paris: Editions Michel.
Ross, J. C. 1999. The Golden Ruler: Precious Metals and Political Development in the Third
Millennium B.C. Near East. Ann Arbor: University Microfilms.
Sasson, J. M. 1990. Artisans . . . Artists: Documentary perspectives from Mari. In Investi-
gating Artistic Environments in the Ancient Near East (A. C. Gunter, ed.) Washington, DC:
Smithsonian Institution, 21–7.
Sasson, J. (ed.) 1995. Civilizations of the Ancient Near East. New York: Scribner.
Sefati, Y. 1998. Love Songs in Sumerian Literature: Critical Edition of the Dumuzi-Inanna Songs.
Ramat Gan: Bar Ilan University Press.
Selz, G. 1993. Beobachtungen zur ‘Silbervase’ des Entemena. Aula Orientalis 11:
107–11.
——. 1996. NE-SAG, BUR-SAG und GU2-NE(-SAG-GA2): Zu zwei Gefässbe-
zeichnungen, ihren Bedeutungsentwicklungen und einem sumerischen Wort für
(Gefäss)schrank. Studi Epigrafici e Linguistici 13: 3–8.
——. 1997. ‘The Holy Drum, the Spear, and the Harp’: Towards an understanding
of the problems of deification in third millennium Mesopotamia. Sumerian Gods and
their Representations, (I. L. Finkel and M. J. Geller, eds.). Cuneiform Monographs 7.
Groningen: STYX: 167–213.
Sigrist, M. 1988. Isin Year Names. Institute of Archaeology Publications, Assyriological
Series Vol. 2. Berrien Springs, MI: Andrews University Press
Sigrist, M. and Gomi, T. 1991. The Comprehensive Catalogue of Published Ur III Tablets.
Bethesda, MD: CDL Press.
Sjöberg, A. W. and Bergmann, E. 1969. The Collection of Sumerian Temple Hymns. Texts
from Cuneiform Sources 3. Locust Valley, NY: J. J. Augustin.
Steible, H. 1982. Die altsumerischen Bau- und Weihinschriften. Freiburger Altorientalische
Studien 5. Wiesbaden: Steiner Verlag.
——. 1991. Die neusumerischen Bau- und Weihinschriften. Freiburger Altorientalische Studien
9. Stuttgart: Steiner Verlag.
Steinkeller, P. 2000. Stars and stripes in ancient Mesopotamia: A note on two decora-
tive elements of Babylonian doors. Iranica Antiqua 37: 357–69.
Veenhof, K. R. 1995. Old Assyrian i urtum, Akkadian e ērum and Hittite giš.ur. In
Studio Historiae Ardens: Ancient Near Eastern Studies Presented to Philo H. J. Houwink ten Cate
on the Occasion of his 65th Birthday (T. van den Hout and J. de Roos, eds.). Istanbul:
Nederlands Historisch-Archaeologisch Instituut, 311–32.
Vickers, M. and Gill, D. 1996. Artful Crafts: Ancient Greek Silverware and Pottery. Oxford:
Oxford University Press.
Wilson, E. J. 1994. Holiness and Purity in Mesopotamia. Alter Orient und Altes Testament
237. Berlin: Neukirchener Verlag.
Winter, I. J. 1992. ‘Idols of the King’: Royal images as recipients of ritual action in
ancient Mesopotamia. Journal of Ritual Studies 6: 13–42.
——. 1995. Aesthetics and Mesopotamian Art. In Sasson ed. 1995: Vol. 4, 2569–
2580.
——. 1999. The Aesthetic Value of Lapis Lazuli in Mesopotamia. In Cornaline et
pierres précieuses: La Méditerranée de l’Antiquité á l’Islam (A. Caubet, ed.). Paris: Musée
du Louvre, 43–58.
—— In press. Defining ‘Aesthetics’ for non-Western studies: The case of ancient
Mesopotamia. In Art History, Aesthetics, and Visual Culture (M. Holly, ed.).
Woolley, C. L. 1934. The Royal Cemetery. Ur Excavations 2. London: Trustees of the
British Museum.
Zaccagnini, C. 1993. Ideological and procedural paradigms in ancient Near Eastern
long distance exchanges: the case of Enmerkar and the Lord of Aratta. Archiv für
Orientforschung 20: 34–42.
Zettler, R. L. 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. Berlin:
Dietrich Reimer Verlag.
Figure 1. Detail, Gudea Statue ‘B’, found Tello, c. 2110 b.c. Diorite; ht. 93
cm. (Louvre AO2). Photo courtesy Département des Antiquités Orientales,
Musée du Louvre.
Figure 2. Bull-headed ‘Great Lyre,’ from Tomb PG 789, Royal Cemetery of

Ur, c. 2500 b.c. Gold, lapis lazuli, shell on wooden reconstruction; ht. of head
35.6 cm., ht. of plaque 33 cm. (UM B17694. Photo courtesy The University
Museum of the University of Pennsylvania).
Figure 3. Relief of lion-headed eagle and two stags, from lintel of Temple of
Nin ursag, Tell al-Ubaid, c. 2450 b.c. Copper; ht. 106 cm., l. 238 cm. (WAA
114308. Photo courtesy of The Trustees, The British Museum).
Figure 4. Cult vase of Entemena of Lagash, found Tello, c. 2400 b.c. Silver
and copper; ht. 35 cm. (Louvre AO2674. Photo courtesy Département des
Antiquités Orientales, Musée du Louvre).
CHAPTER TWENTY-SEVEN
THE AESTHETIC VALUE OF LAPIS LAZULI

IN MESOPOTAMIA
That lapis lazuli in particular among the precious and semi-precious

stones known from Mesopotamia was accorded considerable value
in antiquity may be inferred from the archaeological record through
association with high-status locii and goods. It is further attested in the
textual record, as has been well-documented in the scholarly literature,
including the catalogue that accompanied the exhibition, Les Pierres
précieuses de l’Orient ancien, held at the Musée du Louvre in 1995, and
several of the papers presented at the related colloquium, Cornaline et
pierres précieuses, that constitute the present volume. Rulers and members
of the elite are buried with objects and ornaments made entirely or
partly of the stone;1 kings send emissaries off to acquire the stone in
foreign trade missions, and hoard the raw material within their palaces;2
deities receive votive gifts and booty of lapis, consisting of items of
personal adornment and cult objects, while their temples are described
as decorated with lapis or shining like lapis.3 These references amply
* This article originally appeared as “The Aesthetic Value of Lapis Lazuli in Meso-
potamia,” in Cornaline et pierres précieuses. La Méditerranée, de l’Antiquité à l’Islam., Annie
Caubet, ed. Paris: Musée du Louvre, 1999, pp. 45–58.
1
For example, the contents of the graves in the Royal Cemetery of Ur: a cup, inlaid
daggar, whetstone, inlaid scepter, numerous items of jewellery with lapis lazuli beads,
and various objects employing inlay that include lapis among the insets, such as gaming
boards, the «standard of Ur,» decorative plaques and hair-curls on protomes of musical
instruments, all of which come from the wealthiest burials, such as PG 789,800 and
1236 (e.g., Woolley 1934, Pls. 87, 91–3, 95, 107, 110, 140, 153, 174).
2
Such a mission forms the core of the epic poem, «Enmerkar and the Lord of
Aratta» (S. Cohen 1973), and is discussed by Guichard 1996, pp. 30–32, where he
also presents a text in which it is recorded that Zimri-Lîm of Mari sent an emissary
to acquire lapis from Larsa. Pinnock, 1988: esp. p. 108, further provides a reference
to 22 kgs of unworked lapis found in the third-millennium palace at Ebla.
3
References cut across the textual categories of literary texts, dedicatory inscrip-
tions, year-name lists and economic texts. One pre-Sargonic Sumerian dedicatory
text on a stone vessel from Nippur records the presentation of lapis lazuli and other
292 chapter twenty-seven
attest to the high value accorded lapis lazuli, which is further supported
by those economic texts that record specific equivalents of lapis in silver.4
What I should like to explore in the present paper, therefore, is not that
this material was valued, but rather which of its various properties were
held in high esteem; in short, how the stone was valued.
To pursue this topic is to attempt to understand better those qualities
that reflect categories of value within Mesopotamian society, thereby
bringing us closer to ancient Mesopotamian thought and experience.
Such an inquiry must include both material properties—color, hard-
ness, sheen—and ascribed properties—that is, amuletic/protective or
associative dimensions; and must therefore take into consideration
not only the visual but also the symbolic aspects of the stone. In the
process, the issue of Mesopotamian «aesthetics» is engaged, which for
present purposes is concerned with the range of properties associated with
materials gotten as booty from Kish to Enlil ( J. S. Cooper 1986, p. 105 = Uk. 4.1),
while Shalmaneser III of Assyria records tribute in lapis lazuli (and carnelian) from
Tyre and Sidon on his bronze gate-bands from Balawat (Grayson 1996, p. 147 = Shal.
III A.0.102.84). Dedicatory gifts that include lapis are mentioned in Sumerian later,
Ur III, texts from Ur (see PSD Al:160, a-ru-a A.2.4). Lapis lazuli ornaments and/or
gold ornaments inlaid with lapis, are also recorded in the inventories of temples or as
intended for specific deities (for example, jewels associated with the god Adad from
Tell Hadad of the Old Babylonian period—for which, see al-Rawi and Black, 1983,
pp. 137–143, or for Shamash in Babylon, Mari and Qatna—Durand, 1990, esp. p. 150
n. 100 and 153–4). Two lapis lazuli seals are actually preserved from Babylon which,
according to their inscriptions, were intended as gifts to the god Marduk, one expressly
to be worn around the neck of the cult statue in the Esagila (Frame 1995, pp. 104–5
= Marduk-zakir-sumi I B.6.7.1), the other bearing three separate inscriptions, the lat-
est by Esarhaddon of Assyria (ibidem, pp. 165–6, = Es. B.6.31.1). In the literary text
known as «Inanna’s Descent,» the goddess Inanna dressed herself in preparation for her
journey to the Netherworld, in a necklace of lapis lazuli beads, and held a lapis lazuli
measuring rod and line in her hand (Sladek, 1974: lines 19 & 25). In the text known
as the «Death of Urnamma» (originally, Kramer 1951; a more recent edition available
to me in the ms. of Steve Tinney, prepared for the Sumerian Dictionary Project of the
University of Pennsylvania, and read with Andrew Cohen over the course of 1995–6,
both gratefully acknowledged), lapis objects figure prominently among the gifts brought
by the king to deities in the Netherworld, including a scepter for Dumuzi, a bracelet
seal and haircomb for the wife of Namtar, a seal for Dimpimekug, a measuring rod for
Ninazimua (see lines 103, 110, 111, 120, 125); and the pediment Ur-Namma claims
to have set up as divine service for the Anunna-gods in his lament over his untimely
death and unjust desserts, is also said to have been constructed of lapis (line 159). For
the temples themselves, see for example, Sjöberg and Bergmann 1969, pp. 47 & 48, re
Hymns 40 & 42 to the temples of Inanna in Ulmaš and Nisaba in Ereš, respectively,
where the shrines themselves seem to be described as of lapis; and ibidem, pp. 22, 27,
& 44, re Hymns 7, 13, & 36, to the temples of Nin ursag in Keši, Utu in Larsa, and
Nergal in Gudua, where the shrines seem to be likened to lapis in their gleaming.
4
See, for example, Westenholz 1975, p. 16 re No. 17, lines 3–4, where one block
of lapis lazuli is said to be worth 13 shekels of silver.
the aesthetic value of lapis lazuli 293
plus sensory and emotional responses to the specific substance and works
composed of the substance under discussion.
The principal component constituting a positive attribute of physi-
cal matter in the Mesopotamian lexicon is the quality of emanated
light, or shine. The importance of light and light-bearing with respect
to the divine in Mesopotamia and also with respect to aesthetic value
has been developed elsewhere.5 In all cases, it is apparently the com-
bination of light-plus-sheen yielding a kind of lustrousness that is seen
as particularly positive and auspicious, so that persons and things that
are holy, ritually pure, joyous or beautiful are generally described in
terms of light.
Since to be bright is valued as positive, one would expect that to
be dark should be correspondingly negative. While this is largely true,
there are some cases in which darkness can also be coded positively,
however these are virtually all instances in which lustrousness is none-
theless manifest. It would appear therefore, that the operative dimen-
sion is not so much light as opposed to dark, but rather luster versus
the lack thereof.
The way in which the dark-hued material of lapis lazuli (Sum. za-
gin3, Akk. uqnû) is employed in a number of texts serves as a perfect
illustration. Throughout history, precious stones of all sorts and values
have been remarked upon for their ability to transmit or reflect light.
From diamonds to agates, sparkle and/or luster are found as descriptive
terms cross-culturally, and ancient Mesopotamia constitutes no excep-
tion. In the heroic epic known as «Lugale», the god Ninurta blesses
haematite by describing the stone as one worthy of respect, «whose
surface reflects the light».6
Of the precious stones known in Mesopotamia, lapis is among those,
like haematite and diorite, that represent special cases by virtue of their
dark hue. Written with the determinative for stone, na4, lapis is often
mentioned in descriptions of luxuriously appointed shrines or artifacts.
In many instances, it is to be understood that the actual material is
being referenced. For example, in the Sumerian temple hymn to Inanna
in Ulmaš cited above, the temple is described as of lapis [e2-dinanna
ku3 na4-za-gin3-na. . . .], while the temple to Nisaba in Ereš is said to
5
Cassin 1968; Garelli 1990, pp. 173–177; Winter 1994, pp. 123–132 and 1995,
pp. 2569–2580.
6
Van Dijk 1983, p. 116, line 500.
be sparkling with lapis [e2 za-gin3 gun3-a]. A dedicatory gift of a king

of Ur is recorded, consisting of a lapis-lazuli weapon inlaid with gold
[giš-tukul-na4-za-gin3 ku3-sig17 gar-ra a-ru-a-lugal], which is comparable
to the daggar actually found in the Royal Cemetary of Ur.7 In the Assyr-
ian version of the Epic of Gilgamesh, when Ishtar attempts to seduce
the hero by offering him a chariot of lapis and gold [narkabti na4.uqnî u
urasi], again it is presumably the preciousness of the material, hence
the value of the gift, that seems to be being emphasized; and when
the seduction is not successful, her instrument of vengeance, the Bull
of Heaven, is described with literal «horns of lapis», the artisans of
Uruk called in to admire their weight, thickness and fashioning once
the Bull has been defeated in combat.8
However, in the Sumerian hymns to the temples of Nin ursag, Utu
and Nergal also cited above, the interiors of the temples are appar-
ently simply likened to lapis, rather than as a reference to their actual
material decoration, and this reading is confirmed by the hymns to
the temples of Utu in Larsa and Sippar, where the sungod is said to
«emerge from a lapis-lazuli sky» [dutu an-za-gin3-ta e2-a]—i.e., the dark
but still bright night sky, from which, precisely the sun emerges to bring
the day!9 In addition, on those occasions when body hair or beards
are referred to, as used in describing various Mesopotamian rulers, or
the son of the king in the text known as «Enmerkar and the Lord of
Aratta», then the designation of a «lapis lazuli beard»10 seems to serve
as does «ebony» in many modern languages when applied similarly:
that is, not as the literal material, but metaphorically, as a dark, rich,
lustrous hue, for which the material name comes to be used as a color
name. In the case of lapis, the color signified is thus not the «blue»
which we attribute to it today, but the deep «blue-black» of some lapis
sources, more approriate to hair-color. This would accord well with
Landsberger’s observation in his study on colors that Akkadian uqnû
7
See Pennsylvania Sumerian Dictionary (henceforth PSD) Al:160, a-ru-a, citing
UET 3, 378, pp. 1–2.
8
Kovacs 1989, pp. 51 and 56 = Gilg. VI: 10–11 and VI, pp. 160–162.
9
See Sjöberg and Bergmann 1969, pp. 27, 45 and discussion 87, re Temple Hymns
13 and 37. My reading differs from that of Sjöberg, however, in that he translates it
merely as «bright sky». The understanding of the «lapis-lazuli sky» as referring to the
night sky is further supported by its clear association in Egypt with the «ciel nocturne»
(Sydney Aufrère, personal communication, gratefully acknowledged).
10
E.g., su6-na4-za-gin3, Enmerkar and the Lord of Aratta 209 & 527, cited PSD
Al:153, a-ri-a.
does not correspond to our «blue», but rather to «dark,» and has as
well polished or light-reflecting attributes.11
With such a reading, the description of the sungod Utu as literally
«the one adorned with a lapis-lazuli beard. . . .» should be understood,
as indeed Mark Cohen has rendered it, as «the one adorned with a
gleaming (or dark lustrous) beard»; and when Shulgi of Ur is referred
to as the king with «a lapis-lazuli beard (and) a pure breast» [su6-za-gin3
gaba-ku3-ga], it would be well to translate it: with a (dark) lustrous beard
(and) a (bright) gleaming breast—particularly as the word for «pure» in
Sumerian also carries with it the physical manifestation of that purity
as shine.12 This does not ignore the actual rendering of hair and beards
in stones of lapis lazuli in a number of works of sculpture preserved
to us, such as the beards of the protome bulls on musical instruments
from the Royal Cemetary, just as there are examples executed in ste-
atite as well.13 It simply takes away the necessity of projecting from the
material sculpture to a literal meaning in the texts. Instead, the term
«lapis» would be synonymous with the dark and lustrous, just as silver,
gold or alabaster would be synonymous with the light and lustrous, each
having a positive aesthetic valence;14 and, as noted with respect to the
Shulgi hymn above, the two aspects may be purposely juxtaposed for
contrast in literary texts.
In terms of cultural meaning, one might also suggest that, just as the
beard of the Akkadian ruler Naram-Sîn is longer than anyone else’s
on his stele, suggesting coded signs for more developed masculinity, the
dark color attributed to particular rulers’ beards could also have carried
11
Landsberger 1967, pp. 164–5.
12
See Cohen 1972, line 18, su6.mu2.za.gin3 su3.su3 ur.sag . . ., and Klein 1981, pp.
72–3, re Sulgi D.7. In the late Babylonian bilingual text cited above, the Akkadian
version includes the lapis (color of the) beard is clearly modified by an adjective that
can mean shining as well as pure, elletu. It should be noted that «lapis lazuli beards»
are attributed to many divine and royal personages—see, for example, Jacobsen, 1985,
pp. 65–72, esp. p. 67, ll. 5–6, as attributed to the god Enki.
13
See on this Pinnock 1988, p. 107, citing evidence from Ebla, Mari and Ur. The
interchangeable occurrences of lapis as a color term meaning «dark» with a material
which to us is more «black» similarly occurs in early Sanskrit texts, where one of the
faces of the god Siva is described in one version of the text as being of the color of
lapis, and in another version, is described as the color of «crushed kohl» (Dagens, ed.,
1994, p. 847, note 66).
14
Alabaster, for example, is also among the stones blessed in Lugale (see van Dijk,
1983: 118–9 re lines 515–523). The description of the bright, white stone, whose
material «shines as the light of day» [na4/gis-nu11 su u4-gim kar2-kar2-ra-ke4 = na4/II
ša zumuršu kīma ūme ittananbi u], is, in fact, the perfect material opposite to the dark
luster of the lapis stone.
significance. In more modern times, in the Court of Louis XIV, for

example, abundant dark hair was synonymous with sexual prowess; and
the ascription of a dark lustrous beard hair in Mesopotamia could well
have signaled the same sort of mature maleness as the abundant facial
hair noted just above.15 Indeed, when the god Ninurta is described at
the beginning of the bilingual epic «Lugale», the Sumerian version tells
us that the god bore «a princely lapis lazuli beard» [su6-nun-e a-za-gin3
ru-a], while the Akkadian version leaves out any mention of the lapis,
and instead refers to a «princely beard of bright/pure masculinity»
[ziqnu rubê sa rihûtu elletu].16 The implication is thus strongly supported
that the literary references to lapis lazuli beards carry with them an
extended meaning of rich, well-developed maleness.
When associated with objects, it is equally useful to keep the
metaphoric possibilities of lapis in mind. A reference to the dais of
the goddess Inanna as bara2-za-gin3, in the Sumerian hymn «Innin
Sagurra», may have been intended to convey not merely a literal
pedestal constructed of lapis lazuli, but a «dark, lustrous dais» fit for
the goddess, while its opposite, king Sulgi’s dais in one of the Ur III
period royal hymns, may have suggested not only that it was literally
made of silver, but also that it was «bright and shining,» bara2-babbar-
ra.17 And however complex and elusive the term muš3 as used in the
Sumerian temple hymns may be, when modified by lapis to describe the
Giguna-shrine in the temple of Nin ursag in Keš [gi-gu3-na-zu muš3-
za-gin3], it is probably best translated «Your Giguna, of dark/lustrous
appearance»18 . . . As such, the phrase would echo the point two lines
15
For dark hair under Louis XIV, see Cheng, in press. This adds an extra pos-
sibility to our understanding of the rampant male goats with rosette bush from the
Royal Cemetery, where the use of multiple lappets of lapis lazuli for their coat could
suggest not only an interchangeability with our «black», usually associated with the
wool derived from goat, rather than sheep, hair, but also signal the prime maleness
generally associated with the leader of a herd.
16
Van Dijk 1983, p. 52, re line 10; cited also in CAD Z: 125, ziqnu.
17
See Klein 1981:200–1, re Sulgi A.89 and Sjöberg, Innin Sagurra, p. 201, line
260.
18
See discussion of muš3 in Sjöberg and Bergmann 1969, p. 55, and also in Jacobsen
1985, p. 68. Jacobsen argues for a sense of «crest», when related to mountains, hence
in architectural context a «tower» or high terrace. There do seem to be instances
when the temple itself, or its facade, is referenced, however, so I wonder whether the
architectural term «elevation» might not be the more appropriate term for the specific
architectural feature, and «appearance» the more neutral rendering until the specific
feature can be better identified, particularly given the equation of muš3 with Akkadian
zīmu, as per the discussion by Sjöberg cited above.
earlier in the same hymn, in which it is stated: «The interior is dusky,

the moonlight does not enter», thereby developing the literary theme
of the dark, mysterious, but yet still lustrous temple.19
In short, I would argue that there are special cases when lapis lazuli
is used not in the material sense, but to convey color or dark lustrous
properties that have a positive valence; and in such cases, usage must
be seen as a complement to that which conveys the light and lustrous.
Furthermore, both types of reference must be seen in the larger context
of the positive value placed generally upon sheen and luster, which then
calls forth a positive aesthetic response on the part of an intended or
anticipated audience.
In textual references to jewellery or other adornments, lapis is often
specifically mentioned, and occasionally also accompanied by additional
(semi-)precious stones. Actual examples are readily observed in the
finds from burials in the Royal Cemetary of Ur of the mid-third mil-
lennium b.c., as noted above; but usage plus combination with other
stones is equally attested for later periods, as in a Kassite reference to a
horned crown for the Babylonian god Marduk, made of gold and lapis
lazuli and ornamented with precious stones.20 In shape, the crown was
probably not unlike the sort of divine headgear one sees depicted on
various Mesopotamian royal stelae and cylinder seals. What is missing
from the depictions on stelae and seals is, of course, a sense of the
materials, color and sheen of the actual object. One work of the Neo-
Assyrian period has been recovered, however, that does incorporate
several precious stones, including lapis lazuli, feldspar and magnetite,
in a gold setting: a roundel inscribed with a dedication to the ruler
Tiglath Pileser III (744–727 b.c.), found at Tell Ta’yinat in the ‘Amuq
region of Northern Syria.21 From this roundel, despite dammage due
to time and poor preservation, one can perhaps glimpse a sense of the
19
Sjöberg and Bergmann 1969, p. 22, re TH 7, p. 95, and discussion, p. 73.
20
Text cited CAD A:154, agû A: agê qarni sirati. . . . ša uqnî u hurasi ina qaqqadisu lut
askunuma ina UGU SAG agîšu NA4 . . . lu aškunuma NA4 . . . ina UGU agî[šu] lu uza’inu[ma].
21
= T. 3264, currently on display at the Oriental Institute, University of Chicago,
and among the materials to be published as part of the small finds of the Iron Age
from the Chicago expedition to the ‘Amuq, assigned to W. Orthmann. For the inscrip-
tion, see note by Brinkman, 1977, p. 62. Of course, Assyrian jewellery such as that
discovered at Assur and more recently at Nimrud also can combine several stone-types
in a single piece; however they are generally used as separate components. What
interests one in the Ta’yinat roundel is the setting of several stones onto the surface of
the metal disk.
multi-colored aspect of Marduk’s crown.22 In addition, the use of the

term šalummatu to describe the overall impact of the finished crown in
the Kassite text, generally translated as «awesome splendor»,23 further
provides a sense of the attributed affect of the combination of gold,
lapis and other stones. In this respect, it is of particular interest that in
the epic «Lugale», the poem joins agate, chalcedony, cornelian, lapis,
jasper, «pearls», and possibly garnet [Akk. dusu, hulalu, sandu, uqnû,
abašmu, egizangu, girim ilibu = Sum. na4/du8-si-a, nir2, gug, za-gin3, amas-
pa-e3, e-gi-za3-ga, gi-rin2- i-li-ba] in a single group of stones blessed
by the god Ninurta, and accords them the divine judgment that they
may be favored (literally, chosen «for honey and wine»), and «duly set
in gold».24
There is also some evidence that stones were considered to have had
physical properties that rendered them protective, and hence could serve
amuletic as well as visual ends. This was clearly the case in ancient
Egypt, where in a magical text of the Middle Kingdom it is recorded
that blue, red and «green» beads of lapis, jasper and turquoise are
hung on the neck of a new-born, accompanied by incantations that
make them potent protection against illness.25 And it is equally the case
22
Whether the juxtaposition of stones also had symbolic or amuletic significance is
uncertain. However, based upon the Sanskrit tradition of nine precious or semi-precious
stones, navaratna, which when worn together or placed together in foundation deposits
were considered to be particularly auspicious (see Ghose, ed., 1992, p. 69), one must
leave the question of meaning open—particularly as at least two of the later Akkadian
magico-medical texts referring to the protective properties of precious stones include
specific references to the wearing of 9, or in one case, 10, different stones (see Chicago
Assyrian Dictionary (henceforth CAD) A1: 39, abašmû re KAR 213 i 4).
23
See on this, Cassin 1968, passim, and CAD Š1: 283, šalummatu, for an example
where a copper double-axe is said to possess the same «awesome splendor» as is
ascribed to Marduk’s crown.
24
Akk. ina hurasi kiniš likannukunuti = Sum. ku3-ga zi-de3-eš mi2 [e2]e-en-ze-en (and
variant, ku3-gi-ta mi2zi-di3-eš e2-en-ne-en-ze2-en)—see Van Dijk 1983, pp. 120–122, re
lines 531–546. [For the term kiniš, see CAD K: 385, where another late bilingual text
refers to a divine emblem (a crescent) «fashioned/perfected with due/true care»: ina
taqniti kiniš šuklul = šu-du7 mi2-zi-de3-eš dug4-ga, while a second text is cited that refers
to an image of the god Shamash correctly/duly prepared: alam DN......kiniš ukanni—in
other words, it serves as an emphatic to stress the verb that follows; and when related to
verbs of action, as here for the manufacture of cult or luxury objects, the usage affords
a glimpse of the seriousness and/or correctness of the undertaking.] In the contexte
of Lugale, it is all the more interesting that this same range of stones often appears
grouped in late Neo-Assyrian lists of precious stones from palace and temple admin-
istrative texts from Nineveh (Fales and Postgate, 1992: e.g. #’s 83, 85, 86, 118).
25
This text, Berlin Papyrus 3027, was published by Adolf Erman 1901, p. 9 re
1,1–4; a more recent edition has been prepared by Bernard Mathieu, université Paul-
from ancient through modern India, where stones have both auspicious
meaning and medicinal uses, and are believed to «radiate certain ener-
gies and transmit them to the wearer».26
In Mesopotamia, it has long been suspected that beads and cylinder
seals of particular materials carried amuletic as well as functional value.
Edith Porada cites a text that would make this clear: «If someone carries
a seal of lapis lazuli, he will have protection and his god will rejoice in
him . . .»27 Seals of other stone-types are specifically cited as to be worn
or used for medical cures/protection in magical-medical incantations;28
and this could well help to explain the prevalence of lapis seals in
high-status contexts, as, for example, in the Royal Cemetery burials,
in the 2nd millennium palace deposit from Mycenean Thebes, and
as referred to in some Mari texts.29 If so, it would once again parallel
Egyptian evidence, as recorded in the papyrus cited just above, where
it is stated explicitly in a number of incantations related to the health
of the mother: «This seal is . . . your protection.»30
Lapis is also attested as a «neck-stone», of the sort worn by Naram-
Sîn on his stele—beads that were often inscribed to a particular deity,
and were explicitly stated to provide protection.31 One important
example is that of the lapis cylinder seal intended for the neck of the
cult image of the god Marduk in Babylon, referred to above (endnote 3).
Other stones have also been mentioned in this regard, particularly a
mottled «ašnan-stone» cited by Sennacherib.32
Finally, in an recent study by Nicholas Postgate on the classification of
various stone-types in Mesopotamia, the author remarks upon the pres-
ence of lapis, carnelian and other stones on the body of a burial below
the floor of a private house at Abu Salabikh,33 while lapis, along with
gold and silver, and occasionally accompanied by carnelian and other
materials, is also attested for over two millennia as part of foundation
Valéry, Montpellier. I am grateful to Sydney Aufrére for the initial reference and to
Dr. Mathieu for his generosity in making his edition available to me.
26
Ghose 1992, pp. 68, 71.
27
Porada 1981, pp. 7–8, quoting Oppenheim.
28
Finkel 1980, p. 47.
29
For the Theban seals, see Porada 1981: 1–70; the Mari test (ARM VII: 248, from.
Room 110) is cited in an unpublished paper by Clemens Reichel, «The Administrative
Use of Sealing in the Mari Palace».
30
Erman 1901, p. 41 re 2,1–3,3.
31
For the neck stone, see Winter 1996, pp. 11–26.
32
The ašnan-stone is cited by Reiner, 1995, pp. 119–120.
33
Postgate 1997, p. 205, citing Martin, Moon and Postgate 1985.
deposits for religious buildings.34 These last references, stones laid upon
the body in burial or placed in foundation deposits, might seem to
reflect merely the high status and economic value accorded lapis—like
being buried in one’s finest garments!—were it not for texts that seem
to suggest additional semantic values attached to the stone as well.35
In Assyrian texts preserving medical incantations, lapis is invoked
as pure, the counter-substance to cure conditions such as impotence.36
In addition, actual pieces of lapis (and jasper) seem to have been laid
on or set around the neck of an ill individual as part of curing rituals,
while various minerals threaded on red wool are tied around the hips
of pregnant women as protection.37 And stones, including jasper and
lapis, are apparently worn to dispel sorcery.38 In this regard, it is even
conceivable that when Inanna adorned herself for her descent to the
Netherworld by hanging lapis lazuli beads around her neck,39 she was
not only putting on all her finery, but actually girding herself against
a variety of possible evils by the wearing of amuletic protection.
In a recent article, Alasdair Livingstone has remarked upon the
relatively unexplored area of «stones, their individual characters, and
the tying on of amulets,» citing those few scholarly works that have
34
See Frankfort 1935, p. 36, for the Temple Oval at Khafajeh; Parrot 1956:18 for
the Early Dynastic Ishtar Temple at Mari. Postgate, in the article cited above, p. 211,
notes this Mari reference, in addition to later examples that reflect «a long tradition»
through the Neo-Assyrian period: i.e., Sargon II of Assyria (722–705 b.c.) included
inscriptions on gold, silver, copper, tin, lead, lapis, and limestone in the foundations for
his new capital at Khorsabad; his son, Sennacherib, included silver, gold, carnelian,
lapis, agate and chalcedony in his foundations for a new shrine.
35
Certainly in Sanskrit treatises on building, the use of lapis and other stones in
foundation deposits is explicitly deemed auspicious (see on this, Dagens, ed., 1994, Ch.
12, pp. 121–150, esp. 127). In addition, minerals and their associated colors are charged
with symbolic values, particularly associated with the elements and with specific divinities
(Sabarathinam, 1995, pp. 47–65). At least it is possible that in Mesopotamia as well,
what we tend to see as mere references to valuable materials could instead represent
very purposeful selections of materia with associative instrumental agencies.
36
Biggs 1967, p. 28, Text 11: 2.
37
Maul 1994, pp. 142–7, re line 159, Namburbi. See also, Finkel 1980, pp. 47–51,
for the use of various minerals threaded on red wool specifically to protect pregnant
women.
38
See on this, CAD Al: 39, abašmû (= jasper), where it is noted that the stone was
used as a charm against specific diseases, and/or with various other stones, to dispel
sorcery; see also, Reiner, 1995, on the magical properties of various stones.
39
na4-za-gin3-di4-di4-la2 gu2-na ba-an-la2. For discussion of lapis beads from Ebla
and their significance in general, see also Pinnock, 1988: esp. 127–8.
pursued the nature of certain stones and their use in ritual practice.40
He discusses a Late Babylonian text from Sippar that prescribes, as
part of ritual activity for the beginning of the month of Nisannu, the
libation of red beer before Shamash and the binding of a stone of
lapis lazuli upon the finger (or fist), and sees the stone’s role as actively
implementing the communication between suppliant and superior.41
Such a reconstruction of the instrumental function of certain stones,
lapis among them, would fit well with the picture I am attempting to
reconstruct here—one which draws upon both archaeological evidence
and textual references to impute purpose and meaning to the use of
lapis lazuli in certain circumstances.
One final area of meaning may be associated with gender. From
the archaeological record, it may be observed that lapis seems to have
been combined frequently with carnelian. Beads of carnelian, presum-
ably coming from the Indian subcontinent, were apparently reaching
Mesopotamia as part of the Indus Valley trade in the early third mil-
lennium, just as the lapis itself was conveyed across the Iranian plateau
from its sources in the mines of Afghanistan.42 By the time of the Royal
Cemetery in the mid-third millennium, lapis and carnelian were strung
together as necklaces, earrings and other ornaments.43 It is certainly
possible that it was simply the juxtaposition of the two colors—blue
and red—that afforded visual pleasure to the ancient Mesopotamian.
It would appear from the graves in the Royal Cemetary that there was
no strict separation of materials according to sex, since a lapis and
gold daggar was associated with a male burial, while a considerable
amount of lapis jewellery is associated with female burials. Similarly,
gifts presented by Urnamma in the Netherworld include lapis objects to
both male and female divinities. Nevertheless, certain types of objects
40
Livingstone 1993, pp. 97–113, esp. 197, citing Lambert 1980, pp. 77–83, and
George 1986, pp. 133–146.
41
Livingstone 1993, pp. 107–8, re BM 46553, ll. 4’–9’.
42
For the carnelian, see During-Caspers 1972; Reade 1979; Inizan 1991, pp. 123–
131. That this eastern trade continued well into the Akkadian period is apparent from
mention in a letter of the period, which refers to 340 pieces of carnelian and 4 long
stones of a material coming from Elam, cited in Foster 1983, p. 161 (= YBC 16676),
for which reference I am grateful to Tonia Sharlach. For the lapis trade, see Hermann
1968, pp. 21–57; Sarianidi 1971, pp. 12–15; Pinnock 1988; and von Rosen 1990. See
also the overviews presented in Potts 1994, esp. Chapter 5: Stones and Stonework, pp.
177–216, as well as in Casanova 1995, pp. 15–20 and Inizan, 1995, pp. 21–24.
43
See, for example, Woolley 1934: Pls. 133–5, and the recent discussion by Bahrani
1995, pp. 1635–1646.
do clearly seem to have had gendered associations, as in an Old Baby-

lonian bilingual text to facilitate childbirth that directs the mother if
she wants a son, to look at a weapon, a daughter, to look at a comb,
which would correspond well to the scepter of lapis given by Urnamma
to Dumuzi, while a comb, explicitly called «symbol of womanhood»
[nig2 nam-munus-a], was given to a female divinity.44
A case can also be made that in some instances, lapis may have been
coded as «male», carnelian as «female». In the same Old Babylonian
bilingual incantation cited just above, the unborn child is likened to
a boat filled with carnelian or lapis. The text goes on to state that it
is not known «if it is carnelian or lapis»,45 implying thereby that the
gender of the child will only be known at birth. This formulation occurs
elsewhere as well,46 and the collected references seem to suggest as
association between the feminine sāmtu, carnelian, and the masculine
uqnû, lapis.47 Should this indeed be the case, one would be forced to
consider whether the careful combination of lapis and carnelian beads
in the Royal Cemetery at Ur was not a coded reference to male-female
union and hence fertility (dare I say «afterlife»?), no different from the
processions of alternating male and female animals on the bottom
register of the alabaster vase from Warka, for example, or the paired
rosettes and goats that occur frequently on other types of objects
from the Royal Cemetery. One might even be led to speculate on the
meaning of the pairing of lapis and carnelian fruits on the tree seen
by Gilgamesh in the course of his travels, when he is in search of the
«plant of life,» or on the color pairings found in «Inanna’s Descents,»
when Dumuzi is described as clothed in a red garment, playing upon
a lapis lazuli flute in one context, then as provisioned with a carnelian
ring and a lapis flute in another.48
44
Cohen 1976, re lines 53ff, as compared to «Death of Urnamma», lines 103,
111.
45
Ibidem, lines 10ff. The text and its genre is discussed by Gertrud Farber, 1984, pp.
311–316. I am very grateful to Tonia Sharlach, Harvard University, for having pointed
out these texts and their attendant associations to me.
46
Farber 1984, note 6 and T. Sharlach, personal communication, re YOS
11.85.9’.
47
Cited CAD S: 121b, sâmtu A. Note that in ancient Egypt as well, lapis is classified
as a «male»-blue (Aufrère, 1991, p. 422).
48
Kovacs 1989, pp. 78–9, re Tablet ix, lines 280ff; Sladek 1984, lines 129 and 136,
resp. Indeed, Frances Pinnock has suggested (1993, p. 151, following Kramer) that
a special relationship may also have existed between the goddess Inanna and lapis,
beyond her wearing of the stone in the Descent text. If so, this would be interesting,
In conclusion, while there is little question that lapis lazuli was highly
valued throughout the historical sequence in ancient Mesopotamia,
it is nonetheless useful to pursue the specific qualities associated with
the stone. When the instructions are given in «Inanna’s Descent to the
Netherworld» not to let precious lapis be worked with the (ordinary)
stone-cutter’s stones, nor to let valuable boxwood be chopped up with
the (ordinary) carpenter’s lumber,49 the sense is clear that lapis is per-
ceived as a commodity beyond the commonplace, and is to be kept
separate from the commonplace. And when a story such as the «Myth
of Etana» is set in a period «(when) the scepter was not (yet) deco-
rated with lapis lazuli»,50 the implication is that presence of the stone
is used to signify the accumulation of wealth from long-distance trade
and the centralization of government associated with early urbanism,
hence its absence indicates an era prior to such developments. Finally,
lapis further shares a special status with a number of other precious
stones, particularly carnelian, in that it is used in the most elite royal
commissions and highly valued as a component in finished work: as,
for example, when a statue of silver and carnelian made by the ruler
Abī-Sarē of Larsa is described, then assessed as «a surpassing work, a
thing of (beyond) praise».51
The Mesopotamian reaction to such master works recalls Paul
Valéry’s response to French architecture, as «de très beaux monuments
qui émerveillent les yeux»—a formulation that has its counterpart in
both Sumerian and Akkadian, with respect to phenomena producing
wonder and admiration.52 Works like the sacred lapis-and-gold crown
for Marduk certainly provoked such responses; while the association
of lapis with «dark luster» is likely to have carried with it a sense of
divine mystery as well as visually pleasing sheen. What is also clear
from the appearance of lapis in so many magico-medical texts is that
the stone was thought to possess magical properties in addition to its
especially given the complexly gendered identity of the goddess. That certain deities
were associated with particular stones and/or colors would not be suprising, given
such associations in both early Sanskrit and ancient Egypt texts. For the Egyptian case
(Aufrère 1991, p. 252), it is the goddess Isis who is specifically identified with lapis
lazuli, as Hathor is with turquoise.
49
Sladek 1974, line 45, and repeated in lines 54, 62 and 187.
50
See the OB Etana text, cited in CAD , pp. 132–3, epēru: ha um uqniam la
aprat.
51
Frayne 1990, p. 122 (=A-S 1:ii.6’–10’), me-dim2-ma diri-ga nig2-ar2-eš dib-b[a].
52
This topic, with the attendant vocabulary, is briefly referenced in Winter 1995,
pp. 2575–2577, and will be further developed in Winter, forthcoming.
positive visual qualities and/or economically valuable associations, and

may well have implied a whole set of additional associations related
to gender and fertility in some contexts. Obviously, not all of these
potential associations are likely to be activated or cognized in every
context; yet it is also the case that a range of associations can be active
in any individual usage. If we are to make progress in understanding
the aesthetic impression made by lapis, or indeed any precious stone,
in ancient times, I would argue that we cannot reduce aesthetics to the
merely physical, visual properties of the stone. Rather, it is essential
to take into consideration the multivalent associations of lapis and its
underlying (sacred) properties—of which the sheen and color were
but the external manifestations, and for which the broader context of
cultural meaning constitutes the necessary ground for both experience
and appreciation.
Acknowledgments
I would like to thank Sydney Aufrère, Andrew Cohen, Bernard Mathieu, Nicholas
Postgate, Tonia Sharlach and Kapila Vatsyayan for fruitful exchanges and for help in
access to materials otherwise not available to me.
Bibliography
al-Rawi, F. N. H., et Black, Jeremy A., 1983, “The Jewels of Adad”, Sumer 39, pp.
37–143.
Aurère, Sydney, 1991, L’Univers minéral dans la pensée égyptienne, Bibliothèque d’étude
CV/1–2, IFAO, Cairo.
Bahrani, Zainab, 1995, “Jewelry and Personal Arts in Ancient Western Asia”, in
J. Sasson, ed., Civilizations of the Ancient Near East, vol. III, New York, pp. 1635–
1646.
Biggs, Robert D., 1967. ŠA. ZI. GA: Ancient Mesopotamian Potency Incantations, Locust
Valley (N.Y.).
Brinkman, J. A. 1977, “Philological Note”, JCS 29, p. 62.
Casanova, Michèle, 1995, «Le lapis-lazuli dans l’Orient antiqua» et «La fabrication
des perles de lapis-lazuli», in Tallon, Françoise, ed., Les Pierres précieuses de l’Orient
ancien, Paris, pp. 15–20 & 45–46.
Cassin, Elena, 1968, La splendeur divine: Introduction à l’étude à la mentalité mésopotamienne,
Civilisations et Sociétés 8, La Haye.
Cohen, Mark, 1972, Balag Composition to Entil, Ph.D. dissertation, University of Penn-
sylvania, Philadelphia.
——, 1976, “Literary Texts from the Andrews University Archaeological Museum”,
Revue d’assyriologie 70, pp. 133–144.
Cohen, S., 1973, Enmerkar and the Lord of Aratta, Ph.D. Dissertation, University of
Pennsylvania, Philadelphia.
Cooper, Jerrold S., 1986, Sumerian and Akkadian Royal Inscriptions I: Pre-Sargonic Inscrip-
tions, New Haven.
Dagens, Bruno, ed., 1994, Mayamatam: Treatise of Housing, Architecture and Iconography,
New Delhi.
Durand, J.-M. 1990 a, «La culture matérielle à Mari (I): Le bijou *húb-til-lá/“gur7-
me”», MARI 6, pp. 125–158.
——, 1990 d, «ARMT III, ARMT VI, ARMT XIII, ARMT XXII», in Tunca, Ö., ed.,
De la Babylonie à la Syrie en passant par Mari, Mélanges J.-R. Kupper, Liège, pp. 149–177.
During-Caspers, E. C. L., 1972, “Etched Cornelian Beads”, Bulletin of the Institute of
Archaeology 10, pp. 83–98.
Erman, Adolf, 1901, Zaubersprüche für Mutter und Kind aus dem Papyrus 3027 des Berliner
Museums, Berlin.
Farber, Gertrud, 1984, “Another Old Babylonian Childbirth Incantation”, JNES 43,
pp. 311–316.
Finkel, Irving L., 1980, “The Crescent Fertile”, Archiv für Orientforschung 27; pp. 37–52.
Foster, B. L., 1983, “Selected Business Documents from Sargonic Mesopotamia”, JCS
35, pp. 147–175.
Frame, Grant, 1995, Rulers of Babylonia from the Second Dynasty of Isin to the end of Assyrian
Domination, Royal Inscriptions of Mesopotamia, Babylonian Periods, no 2, Toronto.
Frankfort, H., 1935, “Oriental Institute Discoveries in Iraq, 1933/4”, OIC 19, Chi-
cago.
Frayne, Douglas, 1990, Royal Inscriptions of Mesopotamia IV: Old Babylonian Period (2003–
1595), Toronto.
Garelli, Paul, 1990, «La conception de la beauté en Assyrie», in Abusch, T., Hueh-
nergard, J., et Steinkeller, P., ed., Lingering over Words: Studies in Ancient Near Eastern
Literature in Honor of William L. Moran, Atlanta, pp. 173–177.
George, Andrew R., 1986, “Sennacherib and the tablet of destinies”, Iraq 48, pp. 133–146.
Ghose, V., ed., 1992, Tirtha: The Treasury of Indian Expressions, New Delhi.
Grayson, A. K., 1996, “Assyrian Rulers of the Early First Millennium b.c.”, Royal
Inscriptions of Mesopotamia, Assyrian Periods 3, Toronto.
Hermann, G., 1968, “Lapis Lazuli: The Early Phase of its Trade”, Iraq 30, pp. 21–57.
Inizan, Marie-Louise, 1991, “At the Dawn of Trade, Cornelian from India to Mesopo-
tamia in the Third Millennium: The Example of Tello”, SAA 11, pp. 123–131.
——, 1995, «Cornaline et agate: production et circulation de la préhistoire à nos jours»,
dans Tallon, Françoise, ed., Les pierres précieuses de l’Orient ancien, Paris, pp. 20–24.
Jacobsen, Thorkild, 1985, “Ur-Nanshe’s Diorite Plaque”, Orientalia 54, pp. 65–72.
Klein, Jacob, 1981, Three Shulgi Hymns, Ramat Gan.
Kovacs, Maureen G., translator, 1989, The Epic of Gilgamesh, Stanford.
Kramer, Samuel N., 1951, “Inanna’s Descent to the Netherworld”, JCS 5, p. 1 sq.
Lambert, Wilfred, G., 1980, “The Twenty-one ‘Poultices’”, Anatolian Studies, 30, pp. 77–83.
Landsbergbr, Benno, 1967, “Über Farben im Sumerisch-Akkadischen”, JCS 21, pp.
139–173.
Livingstone, Alasdair, 1993, “The Case of the Hemerologies: Official Cult, Learned
Formulation and Popular Practice”, in Matsushima, E., ed., Official Cult and Popular
Religion in the Ancient Near East, Heidelberg, pp. 97–113.
Martin, H. P., Moon, J., et Postgate, J. N., 1985, “Graves 1 to 99”, Abu Salabikh Excava-
tions 2, London.
Maul, Stefan M., 1994, Zu Kunftsbewältigung: Eine Untersuchung altorientalischen Denkens
anhange der babylonisch-assyrischen Löserituale (Namburbi), Mayence.
Parrot, André, 1956, Le temple d’Ishtar, Mission archéologique de Mari I, Paris.
Pinnock, Frances, 1988, “Observations on the Trade of Lapis Lazuli in the IIIrd Mil-
lennium b.c.”, in Hauptmann, H., et Waetzoldt, H., ed., Wirtschaft und Geseltschaft von
Ebla, Heidelberg, pp. 107–110.
Porada, Edith, 1981, “The Cylinder Seals found in Thebes in Boeotia”, Archiv für
Orientforschung 28, pp. 1–78.
Postgate, J. N., 1997, “Mesopotamian Petrology: Stages in the Classification of the
Material World”, Cambridge Archaeological Journal 7, pp. 205–224.
Potts, T., 1994, Mesopotamia and the East: An Archaeological and Historical Study of Foreign
Relations c. 3400–2000 B.C., Oxford.
Reade, J. E., 1979, “Early Etched Beads and the Indus-Mesopotamia Trade”, British
Museum Occasional Paper 2, London.
Reiner, Erica, 1995, Astral Magic in Babylonia [Transactions of the American Philosophi-
cal Society 85/iv], Philadelphia.
Rosen, L. von, 1990, Lapis Lazuli in Archaeological Contexts, Studies in Mediterranean
Archaeology, no 93, Jonsered.
Sabarathinam, S. P., 1995, “Agamic Treatment of Mahabhutas in Relation to Mandalas
and Arts”, in Bäumer, B., ed., Prakriti, The Integral Vision, vol. III: The Agamic Tradition
and the Arts, New Delhi, pp. 47–65.
Sarianidi, V., 1971, “The Lapis Lazuli Route in the Ancient East”, Archaeology 24,
pp. 12–15.
Sjöberg, Å., et Bergmann, E., s.j., 1969, The Collection of Sumerian Temple Hymns, Texts
from Cuneiform Sources III, Locust Valley (N.Y.).
Sladek, Ron, 1974, Inanna’s Descent to the Netherworld, Ph.D. thesis, Johns Hopkins Uni-
versity, Baltimore.
Van Dijk, Johannes, 1983, Lugal ud me-lám-bi nir-gal: he récit épique et didactique des travaux
de Ninurta, du Détuge et de la Nouvelle Création, Leiden.
Westenholz, Aage, 1975, Literary and Lexical texts and the Earliest Administrative Documents from
Nippur [Old Sumerian and Old Akkadian Texts in Philadelphia, Part One], Malibu.
Winter, Irene J., 1994, “Radiance as an Aesthetic Value in the Art of Mesopotamia”,
in B. N. Saraswati et al., ed., Art: The Integral Vision. A Volume of Essays in Felicitation
of Kapila Vatsyayan, New Delhi, pp. 123–132.
——, 1995, “Aesthetics in Ancient Mesopotamian Art,” in Sasson, J., ed., Civilizations
of the Ancient Near East, vol. IV, New York, pp. 2569–2580.
——, 1996, “Sex, Rhetoric and the Public Monument: The Alluring Body of Naram-
Sîn of Agade”, in Kampen, N., ed., Sexuality in Ancient Art, Cambridge, pp. 11–26.
Woolley, L. et al., 1934, Ur Excavations II: The Royal Cemetery, London-Philadelphia.
CHAPTER TWENTY-EIGHT
AGENCY MARKED, AGENCY ASCRIBED: THE AFFECTIVE

OBJECT IN ANCIENT MESOPOTAMIA
Ancient Mesopotamia offers a useful test case for Alfred Gell’s ascrip-
tion of agency to works we call “art”: that is, that artworks may be
considered the equivalent of persons, capable of acting on and for
their social universes, both psychologically and physically.1 At the most
literal level, the textual and artifactual record permits identification
of several classes of material objects explicitly intended to exert force
on the world (amulets, apotropaic gateway figures). At a deeper level,
additional classes of material objects were conceived as animate, hence
as having the same agency as living entities. This was achieved largely
through ritual consecration, in a system of belief that allowed for the
transfer of personhood and/or divinity into physical matter. Such ritual
enlivenment is analogous to that found in the practices of Buddhist and
Hindu Asia, for example, where matter, once consecrated, becomes the
locus of divine manifestation, as distinct from the merely representa-
tional that is theologically rationalized (with considerable anxiety) in
Christianity and explicitly denied in Judaism and Islam.
It will be my argument throughout the present paper that one must
distinguish between agency ascribed by the analyst of a given work
from agency marked by cultural practice, and even grammar, within
the originating culture, if we are to fully understand the historical role(s)
accorded the artwork. I stress this point in order to avoid conflating
indigenous and analytical perspectives until and unless they may be
shown to be congruent. I do so also to reduce ambiguity in the semantic
range of the very term “agency.”
The agentive role is grammatically marked in early Sumerian,
particularly of the third millennium b.c., and can be implied in later
Sumerian along with the Semitic dialects of Akkadian, for which
* This article originally appeared as “Agency Marked, Agency Ascribed: The Affec-
tive Object in ancient Mesopotamia,” in Art’s Agency and Art History, R. Osborne and
J. Tanner, eds., Malden & Oxford: Blackwells, 2007, pp. 42–69.
1
A. Gell, Art and Agency: An Anthropological Theory (Oxford, 1998), pp. 9, 66, 96.
308 chapter twenty-eight
relevant texts are preserved from the third millennium to the end of the
first millennium b.c. Ascription of agency in these two linguistic and
cultural traditions of ancient Mesopotamia extends beyond the animate
person to anthropomorphic images, buildings and artifacts; indeed,
in Sumerian, the agentive marker can shift in the course of narrative—
as, for example, when Gudea, ruler of the early city-state of Lagash
(c. 2110 b.c.), recounts his (re)building of the Eninnu, a temple for the
god Ningirsu. In the first part of the narrative, preserved on the clay
artifact known as “Cylinder A,” Gudea himself is indicated as the
agent, instrumental in carrying out the construction and outfitting of
the temple. However, in the succeeding “Cylinder B,” once the building
is complete and has been ritually consecrated, with the deity installed
in residence, the temple itself receives the agentive marker, a morpho-
logical sign that the temple has taken on an active role in emitting its
affective properties to the community. In other instances, images are
expected to take on the animate properties of their referent—attested
for statues of this same Gudea, for example—following specific rituals
that effectively open the sensorium of the image. These rituals result in
the image’s putative ability to respond actively to stimuli such as smell,
taste, hearing, and sight, along with speech and thereby the ability to
engage in direct discourse with the gods.
The Mesopotamian case therefore supports Gell’s construct, by
offering concrete examples of the agency of the material object or
work of art. At the same time, it also gives rise to a critique of Gell:
namely, whether he distinguished sufficiently between agency literally
ascribed within a given folk tradition or grammatically marked as a
coded property of the work, and agency figuratively inferred or meta-
phorically projected for certain artworks under certain circumstances
or in general, largely by analysts. The importance of this distinction is
pursued in the present chapter, in terms of both the locus of agency
and the complementary concept of the “affective” as an alternative for
the second, figurative condition.
The Locus of Agency
“Agency” (i.e., the affective or instrumental force exerted by a source

of energy or action upon a recipient) can be exercised not only by
individuals, but also by social institutions and material objects. An obvi-
ous category of objects considered “affective” would include talismans,
agency marked, agency ascribed 309
amulets, and apotropaic figures or emblems, all of which are thought

to exert power and may be attested in any number of traditions.2
Such works are not only accorded the power to affect their respective
universes, they also participate in an underlying belief system permit-
ting such objects to exercise force. However, these clearly designated
“affective objects” are only a minor subset of works exercising agency
in Alfred Gell’s construct of the role of art in society.
In the preface to Gell’s Art and Agency, Nicholas Thomas makes clear
that, for Gell, any form of “doing” is “theorized as agency.”3 It is the
next step, however, that has significant consequences for the analysis
of the artwork. Thomas glosses Gell’s view that “the anthropology of
art is constructed as a theory of agency or of the mediation of agency
by indexes,” with these indexes then “understood . . . as material entities
which motivate inferences, responses or interpretations.”4 Unfortunately,
the “or” in the previous sentence tends to get lost, particularly by those
who would apply Gell’s study to specific cases, with the artwork itself
accorded autonomous agency, even on those occasions when in effect
it serves largely as the index for an agent lying behind it.
In my opinion, the concept of agency is most useful when a distinc-
tion is made between animate or human agency exerted as physical
action and the impact derived from the affective properties of a work,
which must then be inferred (Gell’s abduction of agency) by the entity
upon whom or which the affect is imprinted. With such a distinction,
in which “affect” is a necessary if not sufficient consideration within
an anthropological theory of art (see below), the artwork under discussion
most often requires the adjectival formulation “distributed agency,” even
when, as Gell notes, the work has been subject to ritual transformation
according it an agency of its own.5
2
See, for example, C. Faraone, Talismans and Trojan Horses: Guardian Statues in Ancient
Greek Myth and Ritual (Oxford, 1992). For the ancient Near East, see E. Reiner, “Plague
amulets and house blessings,” Journal of Near Eastern Studies 19 (1960), pp. 148–55.
3
Gell, Art and Agency, p. ix.
4
Ibid., emphasis mine.
5
The notion of distributed agency Gell seems to have derived from the work of
Marilyn Strathern on the extension of the person through the social processes of mate-
rial exchange (see The Gender of the Gift (Berkeley, 1988), esp. p. 273 where she defines
personal agency as “mindful action”). Gell only got to this part of the discussion in
ch. 7, “The Distributed Person,” well after his chs. 2 and 3 on the “Art Nexus.” Yet
his argument for the extension from animate to distributed agency (e.g., with respect
to consecrated works, Art and Agency, pp. 133–53) needs the former prior to articulation
It is this ritual transformation of the material that empowers the

image “to speak, or to see, or to act, through various culturally-sub-
scribed channels,”6 after which it may be said to exercise its own agency.
In Mesopotamia, the principal sensory organ addressed in the enlivening
process is related to speech, while in Hindu and Buddhist practice it is
related to vision. In all cases, lengthy and complex ritual processes are
engaged to “open the mouth” or “open the eyes,” so that the image
then, within the cultural/belief system, can act upon or interact with
its audience.7 Objects as well can be ritually charged with such power
to channel divine forces, hence be endowed with agency.8
These are special cases of agency, however, dependent upon ritual
transformation of substance. When one shifts to more normative, lan-
guage-based indicators of agency, purposeful action and causal conse-
quence must be able to be indicated for individuals, supra-human forces,
spirits, ancestors, or the divine in virtually all languages and cultures. In
most Indo-European and Semitic languages, agency can be indicated
directly in transitive sentences such as “the king [active subject] built
[transitive verb] the temple [passive object],” or “the king killed the
enemy.” An intransitive sentence permits one to shift subject-object
relations, as in the statements “the temple [now the subject] was built
by the king” and “the enemy was killed by the king, “ while retaining
understanding that it was the king who accomplished the act.9 In the
early Sumerian texts we will look at, by contrast, there is no intransitive
of the latter. The reader must therefore evaluate the robustness of each of Gell’s terms
or concepts as his argument progresses.
6
I. J. Winter, “Idols of the king: royal images as recipients of ritual action in ancient
Mesopotamia,” Journal of Ritual Studies 6 (1992), 13.
7
J. P. Waghorne and N. Cutler, eds., Gods of Flesh, Gods of Stone: The Embodiment of
Divinity in India (Chambersburg, PA, 1985); C. Walker and M. B. Dick, “The induc-
tion of the cult image in ancient Mesopotamia: The Mīs Pî Ritual,” in M. B. Dick,
ed., Born in Heaven, Made on Earth: The Making of the Cult Image in the Ancient Near East
(Winona Lake, IN, 1998); I. J. Winter, “Opening the eyes and opening the mouth:
the utility of comparing images in worship in India and the ancient Near East,” in
M. W. Meister, ed., Ethnography and Personhood: Notes from the Field ( Jaipur and New
Delhi, 2000), pp. 129–62.
8
For Mesopotamia, see G. Selz, “The holy drum, the spear, and the harp: towards
an understanding of the problem of deification in the third millennium Mesopotamia,”
in M. Geller and I. Finkel, eds., Sumerian Gods and their Representations (Groningen, 1997),
pp. 149–94; in general, see R. R. Dipert, Artifacts, Art Works, and Agency (Philadelphia,
1993).
9
For the relationship between active voice and transitive verb, as distinguished from
passive voice and intransitive verb, see The Chicago Manual of Style, 15th ed. (Chicago,
2003), pp. 172–3, 176–7.
construction. In the language of linguistics, subject::object construal

gives way to that of agent::patient—the very vocabulary adopted by
Gell. The agent (i.e., the king) remains constant, as the originator of the
action, and so does the recipient or product of his action, the patient
(i.e., the temple, the enemy).
The building paradigm is the more interesting for our purposes,
because the patient in this case is not another person, but a product
that can be construed as an artwork: the temple. The role of the ruler
as builder has a long history in Mesopotamian texts,10 and the visual
trope of an active ruler engaged in the building process has an equally
long history, from the Sumerian Early Dynastic Period of the mid-third
millennium b.c. to the Neo-Assyrian Period ending in the mid-first mil-
lennium b.c. On the well-known plaque of Ur-Nanše, founder of the
Early Dynastic dynasty at the city-state of Lagash, for example, the ruler
is depicted with a brick- or mud-bearing basket on his head—that is,
as if having actively participated in construction—while in the accom-
panying text, beginning just to the right of the ruler’s head (fig. 1), he
is equally clearly identified as agent, the builder of the temple:
nanše.ur lugal šir.bur.LA . . . . e2-dnin.gir.su mu-du3
Ur-Nanše, king of Lagash . . . built the temple of (the god) Ningirsu11
The tense/aspect of the Sumerian verb tells us that this is a completed,
not an ongoing, act. Depicting the ruler with a basket on his head is in
effect a trope for the class of act commemorated on the plaque. Yet, to
match the verbal epithet, Ur-Nanše is not “carrying the basket” in a
perpetual imperfective, iconic of a “type.” Rather, it is to be inferred
that he “carried” the basket (perfective), a way to visually declare his
having built the temple.12
A later ruler, Ur-Namma of Ur (c. 2110 b.c.), on the obverse of his
stele found within the temple precinct of the moon god Nanna/Su’en,
10
S. Lachenbacker, Le roi bâtisseur: les récits de construction assyriens des origines à Teglat-
phalasar III, Etudes assyriologiques, Cahier 11 (Paris, 1982).
11
H. Steible, Die altsumerischen Bau- und Weihinschriften, Freiburger altorientalische
Studien, Bd. 5 (Wiesbaden, 1982), pp. 82–4.
12
The implications of this correlation for art-historical studies of text::image relations
are not trivial, but cannot be pursued fully here. Alone, the action could be ambiguous,
understood either as ongoing, indicating a royal trope of ruler as perpetual builder, or
as tied to a specific historical act. With the use of the past tense or perfective aspect of
the verb in the inscription, however, the reader is cued to understand that the image,
too, has a temporal aspect. Such instances, in which an associated text exerts authority
over an ambiguous image by limiting its semantic range, merit further consideration.
participates in the same class of activity. The king is here depicted

walking behind a deity, with that same basket and building-tools slung
over his back, to indicate his direct participation in the construction
of the ziggurat complex of the moon god, initiated during his reign
(fig. 2). The culmination of the process is then depicted in the registers
above, through the ritual performance being enacted in the completed
sanctuary.
Ur-Namma’s Neo-Sumerian contemporary, Gudea of Lagash, contin-
ued the building practices of his Early Dynastic predecessor, Ur-Nanše,
by further (re)building of a temple to Ningirsu in the satellite city of
Girsu. Images and texts of Gudea commemorating his undertaking
have been well studied.13 In particular, the seated statue known as
Gudea B preserves for us the large tablet on the ruler’s lap, on which
the plan of the temple and precinct has been engraved, with a stylus
aligned along the left-hand edge (fig. 3). An account of the king, the
statue, and the building project is also inscribed around the physical
mass of the statue, a shortened version of what appears in Gudea’s
lengthy Cylinders A and B.
At the obvious level of action, Gudea is clearly the agent through
whom the project is realized, just as Ur-Nanše and Ur-Namma had
been. However, three additional aspects of this account are useful in
the present context.
First, in the building account, we are told that in anticipation of
the project Gudea spends a night in a (prior) temple of Ningirsu, and
receives a dream or omen directly from the deity authorizing him to
undertake the new building project. One could argue, therefore, that
there is an agency behind that of Gudea: the god without whose
blessing, and indeed vision, the enterprise cannot be carried forward.
The temple is then physically executed by the god’s earthly agent, the
13
For sculpture, see A. Parrot, Tello. Vingt campagnes de fouilles (1877–1933) (Paris,
1948), and F. Johansen, Statues of Gudea—Ancient and Modern (Copenhagen, 1978).
For statuary texts, see H. Steible, Die neusumerischen Bau- und Weihinschriften, Freiburger
altorientalische Studien, Bd. 9 (Stuttgart, 1991), esp. pp. 157–79. For Cylinders A and
B, see D. O. Edzard, Gudea and his Dynasty, Royal Inscriptions of Mesopotamia, Early
Periods, vol. 3/1 (Toronto, 1997), esp. pp. 69–101. For a recent synthesis of text and
image, see C. E. Suter, Gudea’s Temple Building: The Representation of an Early Mesopotamian
Ruler in Text and Image (Groningen, 2000).
ruler. Thus Gudea, given grammatical agency in the narrative, must be

construed as an example of “distributed agency” in Gell’s terms.14
Second, the role of the image of Gudea is elaborately described in
the text written on the statue, which through ritual consecration would
have been empowered to have its own agency when installed in the
temple—yet another sort of distributed agency: that of the referent
through his empowered image.15 And third, the temple itself, once
completed, takes on an agency of its own. This agency is explicitly
marked or claimed in text, and projected as if able to be inferred or
perceived (be “abducted”) by its public.
The first aspect of distributed agency, from the god to Gudea, is
paralleled in a number of cases, even depicted in Sumerian art. One
such instance is seen on official seals and sealings of the Ur III period
(c. 2110–2000 b.c.), which contain lengthy formulaic inscriptions nam-
ing the seal-holder as the servant of the king, along with imagery of
that individual standing before or being presented to the ruler (fig. 4).
The king, like the god for Gudea, is primary agent to the extent that,
both visually and verbally, the legitimacy and power of the official
flows from his empowerment by the state or the ruler. In the imagery,
the ruler occupies the dominant, seated position, while in the text, the
king and his titles are mentioned before the official is named.16 How-
ever, through the transfer of agency (and sometimes even through the
explicit statement that the seal has been gifted by the ruler), the official
functions as the distributed person of the ruler in the exercise of his
office. The seal imagery, therefore, records this doubling of agency, the
visual juxtaposition of the figures paralleled by the verbal juxtaposition
of king and official in the inscription. At the same time, the imagery
allows the seal to signal its mediation of that agency, through cascading
14
Cf. G. Zólyomi, “Directive infix and oblique object in Sumerian: an account of
the history of their relationship,” Orientalia 68 (1999), p. 242, citing Cylinder A, I:20–1:
“In it [the dream], he [Ningirsu] made him [Gudea] see Eninnu [the temple].”
15
See in general L. Marin, Portrait of the King, trans. M. H. Houle (Dexter, MN,
1988); and for the ancient Near East, W. Sallaberger, “Das Erscheinen Marduks als
Vorzeichen: Kultstatue und Neujahrsfest in der Omenserie Šumma ālu,” Zeitschrift für
Assyriologie 90 (2000), pp. 227–62, esp. p. 256.
16
I. J. Winter, “The king and the cup: iconography of the royal presentation scene
on Ur III seals,” in M. Kelly-Bucccellati with P. Matthiae and M. Van Loon, eds., From
Image to Insight: Studies in Honor of Edith Porada (Malibu, 1986), pp. 253–68.
indexical relationships: from ruler to official, from official to his seal,

from seal to its impression.17
Yet another instance may be seen on the Stele of Eannatum of
Lagash of the Early Dynastic Period, often referred to as the “Stele
of the Vultures.” On the obverse of the stele (fig. 5), the god Ningirsu
smites the head of an enemy of Lagash with his mace, at the same
time as he has gathered the enemy in his battle net. Both in the lengthy
text engraved on the stele and visually by virtue of placement on the
obverse, the god is accorded prime agency in the achievement of victory.
At the same time, on the reverse (fig. 6), the ruler of Lagash, Eanna-
tum, is seen leading his troops into battle and ultimately victorious. It
is his physical agency that has effected the outcome of the battle, even
as he declares himself the recipient of the distributed agency of the
deity through selection and empowerment.18 The artwork provides the
instantiation of this flow of agency through its visual message.
Indirectly, one can also argue that the stele itself performs politically,
exerting its agency upon the viewing public which responds to and is
shaped by the rhetoric of the effective ruler as the god’s agent. The
stele would then represent Gell’s “index” of royal agency in its role
as affective monument making visual the verbal account of the ruler’s
ability to act.19
These examples help elucidate the hierarchical chain from principal
agent to distributed agent/index, using illustrations from the Sumerian
corpus. The flow of agency from deity to ruler, or from ruler to official,
can be demonstrated, beginning with Gudea’s account of his receipt of
17
See J. Rappaport and T. Cummins, “Between images and writing: the ritual of
the king’s Quilca,” Colonial Latin American Review 6 (1998), pp. 7–32, and I. J. Winter,
“Introduction: glyptic, history and historiography,” in W. W. Hallo and I. J. Winter,
eds., Seals and Seal Impressions: Proceedings of the XLV e Rencontre Assyriologique Internationale
(Bethesda, MD, 2001), p. 3.
18
The full text is published in Steible, Die altsumerischen Bau- und Weihinschriften, pp.
120–45, especially 128–9. Visual deployment of the deity on the obverse and the ruler
on the reverse has been glossed in terms of a distinction between icon and narrative
(I. J. Winter, “After the battle is over: the ‘Stele of the Vultures’ and the beginning of
historical narrative in the art of the ancient Near East,” in H. L. Kessler and M. S.
Simpson, eds., Pictorial Narrative in Antiquity and the Middle Ages, Studies in the History of
Art 16 (Washington, DC, 1985), pp. 11–32). There is redundancy in text and image
regarding the agency attributed to the deity on the obverse; however, the way the ruler
pulls away to exercise an agency of his own on the reverse of the stele suggests that
text and image must be analyzed separately before social agency is inferred.
19
See further I. J. Winter, “Sex, rhetoric and the public monument: the alluring
body of Naram-Sîn of Agade,” in N. Kampen, ed., Sexuality in Ancient Art (Cambridge,
1996), pp. 11–26.
empowerment from the deity and the actions that follow, as recounted
above. It is the second and third aspects of Gudea’s accounts, how-
ever—those concerning his seated statue and the temple he claims to
have constructed—which fall more directly into Gell’s literal domain
of the agency of the artwork exerting force on its social environment.
An examination of each is useful in the present context.
The Statue
The text inscribed on the seated Gudea “B” (fig. 3) tells the reader
explicitly that the image (Sum. alam) was intended for the ki-a-nağ,
a commemorative, funerary chapel (literally a “place for the pouring
of water,” or libation), dedicated to a ruler and, in this case, located
within the Eninnu temple. Food offerings of grain, beer, and bread to
be regularly presented to the consecrated, hence animate, image, are
enumerated across the upper back of the image.20 Other known images
of Gudea were intended for other shrines, according to the inscriptions
indicating the various deities to whom each was dedicated.21 But what is
also learned from the inscription on the seated Gudea is that once the
image is brought into the temple, the living Gudea, ruler of Lagash,
commands the now consecrated, installed, and presumably activated
statue: “Image, to my lord [the god Ningirsu], speak!”22
By the end of the section it is clear that at least one of the reasons
the statue has been installed is precisely in order to recite Gudea’s own
pious deeds and words to the deity. I have argued elsewhere that one
should distinguish seated statues, which presumably occupied shrines
of their own, from standing statues placed before divine images in the
gods’ shrines.23 As the dyadic partner in this latter interaction, the
cult image of the deity would have been similarly enlivened, able to
both receive and send communications or sensory stimuli. At least, such
agency is attested in text, although we do not have preserved the cult
images themselves, inscribed with explicit attestations to such empow-
erment. Nevertheless, a recent study of omen texts from Nineveh by
20
Steible, Die neusumerischen Bau- und Weihinschriften, SB.1, lines 1–20; discussed in
Winter, “Idols of the King,” pp. 21, 29–30.
21
Winter, “Idols of the King,” pp. 24–6.
22
Steible, Die neusumerischen Bau- und Weihinschriften, SB.7, lines 21–5. These statues
were named and there is likely to have been a correlation between naming and works
intended to exercise agency.
23
Winter, “Idols of the King,” pp. 30, 34.
Walther Sallaberger makes clear that the tradition of according agency

to the image continued in southern Mesoptamia well into the first mil-
lennium b.c. Sallaberger demonstrates how at the New Year’s festival,
the cult image of the Babylonian deity Marduk is expected to play an
active role, mediating through communication acts between heavenly
intention and the earthly community.24
That statues are accorded agency specifically through their ability
to speak is not limited to polytheistic traditions. A primary aspect of
the Christian cult of images, particularly within Roman Catholicism, is
the miraculous intervention of images, especially of the Virgin Mary.
Even in iconoclastic phases, words can be put into the mouth of the
cult image, if only to disavow devotion. Such a case has been studied
by Shirin Fozi, who has discovered a sixteenth-century treatise against
devotion directed to the image of the Virgin of Le Puy in the Auvergne,
in which the image, deemed an idol (idole) in the text, is said to address
the idolaters (idolatres) in direct speech, counseling them not to adore
her through her image:
. . . n’est pas moi que tu dois reverer
Je ne suis rien . . . sinon un . . . malhereux idole
. . . Dieu seul . . . te peut secourir au besoin
. . . it is not I whom you should revere

I am nothing . . . but a . . . miserable idol
God alone . . . can take care of your needs.25
It is ironic that the agency accorded the image in the text should thus
be put to disclaiming agency for the image! But since the text would
only have been written if the populace did accord some degree of
agency to the image, it is a useful addition to Gell’s discussion of images,
where he articulates the line between the non-idolatrous use of images
as aids to piety and the idolatrous empowerment of them as “physical
channels to access divinity.”26 The Le Puy text provides a hint of the
degree to which cults of miraculous images can expose the tensions
within Christianity between the power accorded such images and the
24
Sallaberger, “Das Erscheinen Marduks,” pp. 255–6, 258.
25
Shirin Fozi, “A miracle of darkness and the Madonna of Le Puy” (unpublished).
A number of instances are attested in the New World, where books on cult are written
in the first person as if the narrative were told by the cult image itself (Tom Cummins,
personal communication).
26
Gell, Art and Agency, pp. 133ff., esp. p. 135.
theological anxieties about practice deriving from the Old Testament

commandment to make no graven images and prophetic pronounce-
ments on the subject. Indeed, a number of pioneering studies in this
vein have been undertaken over the past 10 to 15 years, largely by art
historians.27
At the same time, the negative case from the Auvergne, as with the
“Lamentation over Sumer and Ur” discussed in the following section,
provides us with an additional perspective on those instances when an
image is directly accorded power. For, with the true Virgin behind her
cult image, as with the living Gudea instructing his consecrated image,
what we are seeing is indeed a case of the “distributed person”;28 the
agency of the image is only able to be autonomous once the agency
of the referent behind has been (ritually or by belief ) transferred into
the image in a chain linking the originating source or person to the
extended or distributed material person.
What this suggests, therefore, is that one should not lose sight of
the referent, all too easy when the same word, “agency,” is applied to
both Gudea and his statues, or to the Mesopotamian deity and his cult
image. It argues for avoiding the shorthand of referring simply to the
agency of the image in such cases, and for maintaining the modifier
“distributed,” or at least a clear articulation of the indexical relation-
ship graphed by Gell.29 Only with such distinctions can the cline from
empowering source to empowered manifestation be made transparent.
This hierarchy from the inherently agentive to the delegated agent is
important when theorizing the agency of “art”—whether established
through performative sacralization or merely by social accord. If left
undistinguished, Gell’s expressed goal of determining the nexus of
social relations occasioned through the performance of the work in a
specific socio-cultural, historical context is itself undermined.
27
See, for example, K. M. Ashley and P. Sheingorn, “An unsentimental view of
ritual in the Middle Ages or, Sainte Foy was no Snow White,” Journal of Ritual Studies
6(1) (1992), pp. 63–85, and other papers in that issue, devoted to art in ritual context;
also M. Caviness, “Reception of images by medieval viewers,” in C. Rudolph, ed., A
Companion to Medieval Art: Romanesque and Gothic in Northern Europe (London, 2005).
28
Gell, Art and Agency, pp. 96ff.
29
I persist in use of the term “referent,” signaling the associative presence behind the
work (as discussed in Winter, “Le palais imaginaire: scale and meaning in the iconography
of Neo-Assyrian cylinder seals,” in C. Uehlinger, ed., Images as Media: Sources for the
Cultural History of the Near East and the Eastern Mediterranean, Orbis Biblicus et Orientalis
175 (Fribourg, 2000), esp. pp. 75–8), rather than Gell’s term “prototype,” which has
its own ambiguities. For Gell, see Art and Agency, p. 54.
The Temple
At the end of the text known as Cylinder A, in which Gudea recounts
the planning and construction of the Eninnu for Ningirsu, the completed
temple is described with all of its affective properties. We are told that
the great awe-inspiring radiance (ni2-gal ) emanating from the temple
was cast over the land “like the sun(god)” (dutu-gim), that the temple
itself was exuding sumptuousness or allure ( i-li gur3-a) and “stood to
be marveled at” (u6-di-de3 gub-ba).30
That one should wish to translate the final statement in such a way
as to portray the temple as marvelous or wondrous is clear from the
context. The actual phrase, however, includes the Sumerian ideographic
sign u6, which, while it can be rendered as marvelous, actually is com-
prised of component signs for both eye, igi, and house/temple, e2, in
such a way as to suggest the intensified or augmented visual experience
that has induced the sense of the marvelous. A more literal translation,
then, would preserve the underlying visual experience, as in the original
derivation of “admire” from the Latin augmentative prefix ad- joined
to the verb mirari, “to see, look at.” Hence, the temple literally “stood
to be ad+mired.”
This semantic loading is clearly implied in another, negative, case:
the text of the “Lamentation over Sumer and Ur.” In that account,
the temple of the moon god Nanna at Ur, described as once fragrant
with cedar and decorated with gold, silver, and lapis lazuli, has been
destroyed. In the closing summary statement, the author laments:
. . . u6-di-bi ba-gul
(The temple) . . . its admirableness was extinguished.31
We see here the same literary pattern as in the Gudea cylinder’s sum-
mary of the Ningirsu temple: after the attributes contributing to the
building’s original affect have been described, the culminating statement
relates to the visual response the temple (once) evoked. Previously, the
overall affect had been strong and positive; its loss is to be lamented.
Through the negative case, then, the opposite is articulated as well.
The earlier temple had been charged with positive affect, which then
30
Edzard, Gudea and his Dynasty, p. 87: Cyl. A, cols. xxix, 14–xxx, 12.
31
P. Michalowski, The Lamentation over the Destruction of Sumer and Ur, Mesopotamian
Civilizations 1 (Winona Lake, IL, 1989), pp. 62–3.
elicited ad+miration; that is, through the act of looking, a heightened

visual response had been induced.
In later Akkadian texts of the neo-Assyrian period in the early first
millennium b.c., there is an exact lexical equivalent, sometimes even
written with Sumerian logograms, to be read: ana tabrāt ( kiššat) nišê.
Here again, an intensified form of the now Akkadian verb “to see”
(barû) is used, with the whole phrase to be translated: “for the intensified
viewing experience [ hence ad+miration] of (all) the people.”32 In the
rhetorical structure of official texts, these phrases, in both Sumerian
and Akkadian, are generally used to conclude descriptions of major
undertakings, calling attention to the work or project in question by
claiming for it a powerful response.
What is interesting for Gell’s formulation of agency exerted by the
work is the active role the enterprise or product itself can be shown
to take in disseminating its properties. In a Sumerian temple hymn
of the Ur III/Old Babylonian Period (late third to early second mil-
lennium b.c.), we are told that a temple’s interior is “a beaming light
issuing forth.”33 Use of the verb “to go out, issue forth” suggests the
outward movement (extromission) of the properties of the sanctuary
from their source into the perceiving world. It lays the foundations
for our understanding of the active way in which the finished work,
once fully endowed and ritually consecrated, was thought to emit its
qualities affectively.
This brings us to a discussion of early Sumerian as a language dem-
onstrating ergativity: that is, one that distinguishes agent and patient,
rather than subject and object.34 It will be recalled that Gudea’s Cylin-
der A describes the antecedents to and construction of the temple, by
the end of which the building has been physically completed. In the
continuing account preserved in Cylinder B, the temple has been fully
appointed, the deity formally (and ritually) installed. Indeed, once this
32
See on this P. Garelli, “La conception de la beauté en Assyrie,” in T. Abusch,
J. Huehnergard, and P. Steinkeller, eds., Lingering over Words: Studies in Ancient Near Eastern
Literature in Honor of William L. Moran (Atlanta, 1990), pp. 173–7, and I. J. Winter, “The
eyes have it: votive statuary, Gilgamesh’s axe, and cathected viewing in the ancient
Near East,” in R. S. Nelson, ed., Visuality Before and Beyond the Renaissance: Seeing as Others
Saw (Cambridge, 2000), pp. 22–43.
33
Å. Sjöberg and E. Bergmann, The Collections of the Sumerian Temple Hymns, Texts
from Cuneiform Sources 3 (Locust Valley, NY, 1969), pp. 24, 77.
34
Akkadian has related grammatical features, probably due to its proximity to Sume-
rian (Nicholas Postgate, personal communication; H. P. Muller, “Ergative construction
in early Semitic languages,” Journal of Near Eastern Studies 54 (1996), pp. 261ff.).
is accomplished, it is no longer Gudea as builder but the temple that

is described in an active mode:
[e2]-e me-gal-la [s]ag mi-ni-ib2-il2 ni2-me-lim5-ma šu mi-ni-ib2-du7
The temple [+agentive marker -e] . . . lifted its head, unparalleled in radi-
ant splendour35
Significant is the post-position -e attached to the noun e2, “temple/
house,” which marks the fact that the temple itself takes on an agency
directly comparable to that indicated in the Sumerian temple hymn cited
above, in which the temple actively emits its light-bearing properties.
The agency performed by the Ningirsu temple fits well into Gell’s
construct of an active artwork distinguished from an essentially passive
spectator, and in fact helps us to inflect the Sumerian viewer who, as
a result of his or her intensified viewing experience, is ultimately the
recipient of the temple’s affective power. Gell states that the individual
who permits him- or herself to be “attracted to an index, and submits to
its power . . . is a patient, responding to the agency inherent in the index.”36
But from where does he derive his analytical terms? Ironically, he rejects
a semiotic approach to “decoding” the artwork according to linguistic
models,37 while simultaneously appropriating the vocabulary, along with
the “polarity of agent/patient relations,” directly from linguistics.38
Agent::patient relations (ergative-absolutive constructions) are seen
as alternatives to subject::object relations (nominative-accusative). The
literature is extensive, reflecting work done in linguistics especially from
the late 1970s and 1980s, but continuing to the present. Analyses range
from expositions of the nature of ergative systems39 to case-studies of
35
Edzard, Gudea and his Dynasty, p. 97: Cyl. B, col. xvi, 3–4.
36
Gell, Art and Agency, pp. 31 (emphasis mine), 66.
37
Ibid., pp. 6, 25, 66.
38
Equally masked is the important work of Anthony Giddens: for example, Central
Problems in Social Theory: Action, Structure and Contradiction in Social Analysis (Berkeley, CA,
1979), which offers an agency-oriented framework for social analysis; and The Consti-
tution of Society: Outline of the Theory of Structuration (Berkeley, CA, 1984), in which the
agent, agency and power are delineated in terms of the flow of social action. See also
I. Karp, “Agency and social theory: a review of Anthony Giddens,” American Ethnologist
13 (1986), pp. 131–7, esp. p. 134.
39
E.g., B. Comrie, “Ergativity,” in W. P. Lehmann, ed., Syntactic Typology: Studies in the
Phenomenology of Language (Austin, TX, 1978), pp. 329–94; R. M. W. Dixon, “Ergativity,”
Language 55 (1979), pp. 59–138; F. Plank, ed., Ergativity: Toward a Theory of Grammatical
Relations (London, 1980); and A. Cooreman, B. Fox, and T. Giv’on, “The discourse
definition of ergativity,” Studies in Language 8(1) (1984), pp. 1–34. See also, more recently,
R. D. Van Valin, Jr., ed., Advances in Role and Reference Grammar (Amsterdam/Philadelphia,
particular languages,40 including a substantial body of work on early

Sumerian.41 Ergative-absolutive and nominative-accusative construc-
tions can occur within the same language, with a variety of strategies
determining which construct will be deployed in a given usage.
Interesting for our purposes, as discussed by F. R. Palmer, is when
to separate agent::patient roles, which can be communicated in most
grammatical structures independent of language group—often by word
order and/or verbal construction, from specifically ergative systems
in which agency is marked, most often in the nominal case.42 Clearly,
Gell uses the terms agent and patient to indicate roles, independent
of any specific language group and communicable no matter what the
linguistic structure. Nevertheless, it seems important to uncover the
linguistic model, precisely because it does make a difference whether
the role is to be inferred or is actually marked. In the former case, as
1993); R. M. W. Dixon, Ergativity (Cambridge, 1994); C. Manning, Ergativity: Argument,

Structure, and Grammatical Relations (Stanford, CA, 1996); R. D. Van Valin, Jr. & R. J.
LaPolla, Syntax: Structure, meaning and function (Cambridge, 1997); Alana Johns, D. Massam
& J. Ndayiragije, eds., Ergativity: Emerging Issues (Dordrecht, 2006).
40
E.g., W. H. Eilfort, P. D. Kroeber, K. L. Peterson, eds., Papers from the Parases-
sion on Causatives and Agentivity at the twenty-first Regional Meeting, Chicago Linguistic Society,
1985 (Chicago, IL, 1985). Also, M. Silverstein, “Hierarchy of features and ergativity:
grammatical categories in Australian languages,” in R. M. W. Dixon, ed., Grammati-
cal Categories in Australian Languages, AIAS Linguistic Series 22 (Canberra, 1976), pp.
112–71 and “Cognitive Implications of a referential hierarchy,” in M. Hickman, ed.,
Social and Functional Approaches to Language and Thought (Orlando, FL, 1987), pp. 125–64.
S. Chung, Case Marking and Grammatical Relations in Polynesian (Austin, TX, 1978); J. W.
Du Bois, “The discourse basis of ergativity,” Language 63 (1987), 805–55, for ancient
Maya. More recently, M. J. Ezeizaburena and M. P. Carrañaga, “Ergativity in Basque:
a problem for language acquisition?” Linguistics 34 (1996), pp. 955–89, note that “the
acquisition of the ergative/absolutive distinction can be learned through the semantic
feature agentivity” (p. 965).
41
E.g., D. Foxvog, “The Sumerian ergative construct,” Orientalia 44 (1975), pp.
395–425, P. Michalowski, “Sumerian as an ergative language,” Journal of Cuneiform
Studies 32 (1980), pp. 86–103; M. Yoshikawa, “Ergativity and temporal indication in
Sumerian,” in M. Mori, H. Ogawa, M. Yoshikawa, eds., Near Eastern Studies Dedicated
to HIH Prince Takahito Mikasa on the Occasion of his 75th Birthday (Wiesbaden, 1991), pp.
491–507; G. Zólyomi, “Voice and topicalization in Sumerian,” Ph.D. thesis, Eötvös
Lorand University, Budapest, 1993; C. Woods, “Deixis, person, and case in Sume-
rian,” Acta Sumerologica 22 (2000), pp. 303–34; also J. Braun, Sumerian and Tibeto-Burman
(Warsaw, 2001).
42
F. R. Palmer, Grammatical Roles and Relations (Cambridge, 1994), esp. pp. 22–87.
See also B. Byrne and E. Davidson, “On putting the horse before the cart: exploring
conceptual bases of word order via acquisition of a miniature artificial language,”
Journal of Memory and Language 24 (1985), pp. 377–89, which argues that it is “natural”
to refer first to agents, then to patients.
employed by Gell, the agency is abducted43 through a series of signs;

in the latter, it is marked grammatically, and so directly signaled to the
literate, rather than inferred.
In the specific case of Sumerian, ergativity “manifests itself not only
in nominal case marking, but also in pronominal systems and verbal
agreement, as well as in syntax.”44 Early Sumerian, corresponding
chronologically to the third millennium b.c. works of Eannatum and
Gudea, actually manifests a “split ergative” system: ergative in the per-
fective or completed aspect of the verb, nominative-accusative in the
imperfective or ongoing.45 Agentivity and perfectivity are thus conjoined,
and argued to be related to “the notion of pragmatic foregrounding
in the narrative.”46 This fits well with the observation of sumerologist
Thorkild Jacobsen, who noted that the ergative is used especially when
the agent is to be stressed semantically.47
The point is that one sees the use of the ergative marker for the third
person nominal plus a verbal form indicating the perfective or accom-
plished act precisely in rhetorical language such as is used on public
monuments and in official accounts of the third millennium b.c., and
I would argue that this is so because it is the entity manifesting agency
in such cases (the king, the work) that is meant to be foregrounded,
whether in text or image.48
One could object, of course, that an association of ergativity in the
grammatical structure with agency in narrative confuses morphology
with semantics, especially as every language is likely to have ways of
43
Art and Agency, p. 14, re abduction as inference and induction in the service of
explanation.
44
Michalowski, “Sumerian as an ergative language,” pp. 88–9. Michalowski’s point
is that “it is not enough to state that Sumerian is an ergative language because it marks
nouns in a particular manner; ergativity must be investigated throughout the grammar,”
especially as, in some cases, ergative markings can be correlated with specific “semantic
classes of verbs or on the basis of tense or aspect of the verb.”
45
Ibid., pp. 93, 101; also Woods, “Deixis,” esp. pp. 305, 327–8.
46
Ibid., pp. 101–2. Note that the reader’s cue for agency is signaled from word
order and other (semantic) signals in nominative-accusative languages such as English
or Latin, no less than by the ergative marker in early Sumerian. However, I stress the
agentive marker as a particularly powerful and unambiguous sign. In addition, in a
split ergative system such as Sumerian, agency can be indicated more richly precisely
because it has been selected to be marked in the perfective.
47
T. Jacobsen, “The Sumerian verbal core,” Zeitschrift für Assyriologie 78 (1988), pp.
161–220, esp. pp. 204–5 on the “Intrusion of Ergative.”
48
This also fits well with Karp’s discussion of Giddens’s social definition of agency,
in which the agent is one “engaged in the exercise of power in its primary sense of the
‘bringing about of effects’ ” (“Agency and social theory,” pp. 136, 137 n. 1).
identifying agency or causation (as in who built the temple). However,

recent studies have demonstrated that the early acquisition of the
ergative-absolutive distinction is “learned through the semantic feature
agentivity”49—that is, that the two are cognitively intertwined. For our
purposes, in the Sumerian case, I would argue that morphology and
semantics are not discrete categories precisely in those instances where
grammatical marking provides stress that is congruent with narrative
intent. At such moments, grammar can signal meaning (e.g., stress) as
well as structure.
This claim is in fact supported by work in linguistics that discusses
the semantics of case markings. In particular, the work of Suzanne
Kemmerer and Arie Verhagen insists that “the knowledge underlying
grammar is not qualitatively different from other aspects of human
understanding”; they suggest that viewing causative constructions as
cognitive schemata helps to explain why languages with case markings
“to express the causee are rare, if not non-existent, as distinct from
those marking the causer” (i.e., the agent).50
In Sumerian, too, I suggest it is not insignificant that the agent car-
ries the post-position marker -e, while the patient is unmarked. The
agentive -e serves not just as a morphological, but also as a semiotic,
marker of stress. The strategy of choosing ergativity to mark the ruler
as agent in the performance of certain acts, accompanied by the per-
fective aspect of the verb to indicate that the act is completed, directly
complements the imagery that also foregrounds the acts of the ruler
and means that the viewing and reading audiences are brought to an
identical conclusion, one consistent with the goals of the social con-
structor of the work. In short, especially in a split-ergative system such
as early Sumerian, ergativity becomes part of the semantic signaling
system of the language/text/culture/society.
Recognition of this distinction is again important for Gell’s own goal
of probing social relationships through the agency of the artwork, for
what I am arguing is that social relations are no less encoded than
the cultural, linguistic, iconographic “meaning” Gell claims to eschew.
49
E.g., Ezeizaburena and Carrañaga, “Ergativity in Basque,” p. 965.
50
S. Kemmer and A. Verhagen, “The grammar of causatives and the conceptual
structure of events,” Cognitive Linguistics 5 (1994), pp. 115–56, esp. p. 147. See also
S. Wallace, “Figure and ground: the interrelationships of linguistic categories,” in P. J.
Hopper, ed., Tense and Aspect: Between Semantics and Pragmatics (Amsterdam, 1982) and
A. Alexiadou, Functional Structure in Nominals: Nominalization and ergativity (Amsterdam/
Philadelphia, 2001).
That a particular grammatical voice can be chosen—in Samoan, for

example, or in Sacapultec Maya, as in Sumerian—emphasizes the fact
that ergativity is discourse-based, carrying semantic and/or pragmatic
import and marked agency therefore reflects a socially determined as
well as socially deployed empowerment.51
Furthermore, to the extent that “agentive indicates agency and
patientive [indicates] affectedness,”52 the properties or attributes causing
affectedness, which Gell dismisses as mere aesthetics,53 in fact cannot
be so dismissed if one is to be successful in establishing those social
relationships operative within his “art nexus.” Rather, I would say that
only if the values attached to affective properties of works or objects
are understood can one successfully reconstruct the social system in
which the works are actively engaged.54
The Affective Object
That a given artwork exists within a social “art nexus” needs little
argument. By Gell’s definition, “the theoretical study of ‘social rela-
tions in the vicinity of objects mediating social agency’ ” constitutes the
mission of the anthropologist of art.55 One is forced to ask, however,
whether that is the only job of the anthropologist of art. Whose job
then is to show what the affective properties of the work are, how
those properties may operate and change in relation to society, or how
those properties may themselves act to effect change in society, hence
in social relations?
51
Clark, Language 57(1), p. 201; Du Bois, “The discourse basis of ergativity,” p. 806;
Chung, Case marking . . . in Polynesian, ch. 3; Michalowski, “Sumerian as an ergative lan-
guage,” p. 89. It is not insignificant (as noted by Judith Irvine, personal communication)
that ergativity is sometimes related to social distance, hence reinforcing the public as
distinct from personal context of usage. See also F. C. Southworth, “Linguistic marks
for power: some relationships between semantic and social change,” Anthropological
Linguistics 16(5) (1974), pp. 177–91. What I would emphasize, then, is the diacritical
nature of ergative marking, of particular interest for those who would study (inscribed)
public monuments.
52
Palmer, Grammatical Roles, p. 85.
53
Gell, Art and Agency, p. 3 and passim.
54
As a model for such an integrative approach, see J. Tanner, “Social structure, cul-
tural rationalisation and aesthetic judgement in Classical Greece,” in N. K. Rutter and
B. A. Sparkes, eds., Word and Image in Ancient Greece (Edinburgh, 2000), pp. 183–205.
55
Gell, Art and Agency, p. 7.
Gell dismisses attempts by anthropologists and others to pursue indig-

enous aesthetics as misguided efforts that merely “equate the reactions
of the ethnographic Other . . . to our own.”56
Robert Layton, in his review article unpacking Art and Agency,57 sug-
gests that Gell has flattened the discourse on art within anthropology to
date (see also Tanner and Osborne, “Introduction,” this volume). This
is no less the case for art history, and even philosophy and psychology,
where both agency and the affective properties of work have been
under discussion for well over a decade.58 Within anthropology, Howard
Morphy, among others, has made significant contributions to the discus-
sion of affective properties such as light in an indigenous context; and
he has directly engaged the aesthetic as part of anthropology’s mission,
precisely by seeking to uncover socio-cultural conventions and values
within a tradition, rather than by imposing our values upon it.59
That the art historian’s role should, conversely, be limited to aesthet-
ics and formal analysis is equally reductive. The social history of art
has been well theorized at least since the 1950s.60 Underlying Gell’s
reduction of the art historian’s enterprise is an old-fashioned definition
of aesthetics that demands the work be alienated from its originating
context, and an equally outdated notion of formal analysis tied to
attribution studies. “Style,” for example, that important concept and
tool for formal analysis to which Gell devotes an important chapter, can
be shown to have been consciously deployed as an affective property
in its own right, intertwined with iconography toward meaning.61 In
56
Ibid., p. 6.
57
Robert Layton, “Art and Agency: a reassessment,” Journal of the Royal Anthropological
Institute 9 (2003), pp. 447–64.
58
See, for example, Dipert, Artifacts, Art Works and Agency. Certainly, some recent work,
published after the death of Alfred Gell, has risen beyond his thumbnail sketches of
the disciplines: e.g., W. van Damme, “Universality and cultural particularity in visual
aesthetics,” in N. Roughley, ed., Being Human: Anthropological Universality and Particularity
in Transdisciplinary Perspectives (Berlin/New York, 2000), pp. 258–83; N. Carroll, “Art and
the domain of the aesthetic,” British Journal of Aesthetics 40 (2000), pp. 191–208.
59
H. Morphy, “From dull to brilliant: the aesthetics of spiritual power among the
Yolngu,” Man 24 (1989), pp. 21–40. See also Morphy’s comments in T. Ingold, ed.,
“1993 debate: aesthetics is a cross-cultural category,” in Key Debates in Anthropology
(London, 1996), p. 249–94.
60
E.g., A. Hauser, The Social History of Art (New York, 1957). More recent studies
concerned with such issues are M. Baxandall, The Limewood Sculptors of Renaissance
Germany (New Haven, CT, 1980); M. Camille, The Gothic Idol: Ideology and Image-Making
in Medieval Art (Cambridge, 1989).
61
R. Klein, Form and Meaning: Essays on the Renaissance and Modern Art, trans. J. Jay and
L. Wieseltier (New York, 1979), esp. p. 149; also I. J. Winter, “The affective properties of
addition, art historians have themselves been grappling with the issue of
agency for some time: from patronage, to the possibilities for subjective
expression on the part of producing artists, to the power of the work
as socially affective, not merely reflective.62
For anthropology no less than art history, a larger set of dimensions
within which agency is imputed to artworks seems called for: dimensions
that consider not only “social relations in the vicinity of objects,” but
also underlying authority structures, systems of belief, notions of his-
tory, and systems of value—symbolic, material, and aesthetic. If, with
Robert Wicks, part of the response to the artwork, hence its potency,
is “how . . . intention has been realized,”63 then the work cannot be
fully understood in its social dimension unless one can decode what is
affective, what is causing the social force,64 along with the hierarchy of
agentive power discussed above.
This criticism notwithstanding, Gell’s volume has opened the debate
to a more public forum on the nature of the artwork and its role
in society, as understood by anthropologists and art historians alike.
Conceivably, had Gell lived to reflect further on the importance of
his insights, he might have resisted the temptation to represent such
styles: an inquiry into analytical process and the inscription of meaning in art history,”
in C. A. Jones and P. Galison, eds., Picturing Science/Producing Art (New York/London,
1998), pp. 55–77; and M. B. Garrison, “Seals and the elite at Persepolis,” Ars Orientalis
21 (1992), p. 17.
62
Norman Bryson, “The gaze in the expanded field,” in H. Foster, ed., Vision and
Visuality (Seattle, WA, 1988), p. 91, discusses how the work actively erupts into the
visual field of the viewer. A panel on agency and “The affective object in art history”
was mounted at the annual meetings of the College Art Association in 1995, chaired
by medievalist Madeline Caviness, which included papers that dealt with artworks as
active and interactive components in their social environments. More recently, Brian
Just has taken on the issue of agency regarding pre-Columbian sculpture, suggesting
that Gell “downplays the complex multiplicity of social roles art objects can play” (“Ars
longa, vita brevis: beyond agency in ancient Maya sculpture,” paper given at 2003 CAA
session, “Between creation and destruction: the aesthetics of iconoclasm”). Nor is the
issue of the agency of the work particularly new in discourses on art. See, for example,
discussion of an eighteenth-century case in E. Naginski, “The object of contempt,”
Yale French Studies 101 (2002), pp. 32–53, esp. pp. 34–5: “The High Enlightenment
recognized the potential agency of things . . .,” moving beyond “Hobbesian mechanistic
explanations of agency” to the notion that “matter was radically active,” and things
could be “endowed with an intrinsic . . . vitality.”
63
R. Wicks, “Dependent beauty as the appreciation of teleological style,” Journal of
Aesthetics and Art Criticism 55 (1997), pp. 387–400, esp. p. 395, emphasis mine.
64
Cf. M Tamen, Friends of Interpretable Objects (Cambridge, MA, 2001), pp. 2–4, on
works becoming interpretable only in the context of “a community of agreement,”
within which certain objects maybe said to speak.
reductive and essentializing views of either anthropology or art history

in order to make his points—particularly as so many of the issues he
raises are challenging and had already begun to generate vigorous
inquiry in their respective disciplines.
The Sumerian case has been useful as an illustration of the purpose-
ful marking of agency, both in text and in image. This has implica-
tions for how one may understand the power of the artwork within
that society, the work itself serving as a vehicle with agentive force.
Other languages and other visual repertoires will have devised equally
effective signaling systems permitting their speakers or social members
and subsequent analysts to infer agency, causation, and transformative
power. What I am suggesting is that how a tradition and its language(s)
mark or ascribe agency is not a trivial component in any analysis of
the weight to be given to the social power of the agent, including the
agency accorded the artwork.
The Sumerian case also asks us to distinguish text from image in
the articulation of agency, so that represented power or authority may
be interrogated as either consistent or contested before the work, the
period, or the nature of agency itself is assessed. Particularly in the
case of inscribed artworks, things may be said to happen because of
the semiotic relationship between word and image. Furthermore, if
the potentially ambiguous image (as the basket-bearing ruler on the
“Ur-Nanše Plaque”) is in fact anchored by the accompanying text, then
it suggests that one has to have language to conceive or perceive the
image: that, in short, the image is understood by its originating soci-
ety in language, so that when the art historian is talking about a social
history of art or attempting to reconstruct the affective power of the
work, correspondences in language, grammatical structure, and idiom
need to be part of the analysis.
In the end, it is this art historian’s view that what one cannot do is
simply take the unmodified “agency” of the artwork as given. How-
ever, once modified—as with “distributive” or “abducted” agency
(which demands awareness and analysis of the properties of the work
that generate the affect and so permit appropriate inferences)—then
Gell’s concepts as articulated in Art and Agency have much to offer. But
I would also insist that the art historian and the anthropologist working
on aesthetics—indigenous, historically specific systems of value—have
much to contribute to the discussion as well. Once the relational nexus
surrounding and generating the artwork is seen to function as part
of a holistic system, affective properties and meaning intersect with
social agency. That is, instead of having to choose “art as a system

of action intended to change the world rather than encode symbolic
propositions about it,”65 one is enjoined to see art both as a system of
meaning encoding propositions about the world and as a system of
action intended to change the world, precisely because the excitation
generated by the work lies in the interaction between the two.
Acknowledgments
I am particularly grateful to all of the members of a graduate seminar on “The Cult

of Images” in Christianity, taught by my colleagues Thomas Cummins, Ioli Kala-
vrezou, and Hugo van der Velden at Harvard University in the Fall of 2004, which
I had the privilege of auditing, for stimulating discussions on a number of the issues
raised here. I am also grateful to Miguel Civil, Mark Geller, Nicholas Postgate, and
the late Jeremy Black, for cautions and perspective regarding the reading of ergative
construction, which have helped to shape my argument; to Richard Davis, Judith
Irvine, Piotr Michalowski, Erika Naginski, Michael Silverstein, Christopher Woods, and
Kapila Vatsyayan for conversations on related topics; to the editors, Jeremy Tanner
and Robin Osborne, for pertinent comments; and to Robert Hunt, as always, for his
acute reading of an early draft.
65
Gell, Art and Agency, p. 6.
Figure 1. Plaque of Ur-Nanše of Lagash, found at Tello, c. 2600 b.c.;

limestone, ht. 40 cm. Courtesy of the Département des antiquités
orientales, Musée du Louvre.
Figure 2. Detail, registers 2–3 of the Stele of Ur-Namma of Ur, found at

Ur, c. 2110 b.c.; limestone, ht. as restored c. 2.85 m. Shown as originally
restored (disassembled 1988). Courtesy of the University of Pennsylvania
Museum.
Figure 3. Statue “B” of Gudea

of Lagash, found at Tello, c. 2115
b.c.; diorite, ht. 93 cm. Courtesy
of the Département des antiquités
orientales, Musée du Louvre.
Figure 4. Clay Tablet with seal of Ur-Lisi,

ensi of Umma, citing ruler Amar-Su’en of Ur,
Ur III period, c. 2045 b.c. Courtesy of the
Pierpont Morgan Library.
Figure 5. Detail, obverse of the Stele of Eannatum of Lagash, found at

Tello, c. 2460 b.c.; limestone, full ht. as restored 1.88 m. Courtesy of the
Département des antiquités orientales, Musée du Louvre.
Figure 6. Detail, reverse of the Stele of Eannatum of Lagash. Courtesy of

the Département des antiquités orientales, Musée du Louvre.
CHAPTER TWENTY-NINE
“SEAT OF KINGSHIP”/“A WONDER TO BEHOLD”:

THE PALACE AS CONSTRUCT IN THE
ANCIENT NEAR EAST
By folk definition, the palace is where the ruler resides. In the successive
kingdoms of ancient Mesopotamia, however—Sumerian, Akkadian,
Babylonian, Assyrian—the palace was the seat of many activities:
administrative, bureaucratic, industrial, and ceremonial as well as resi-
dential. In brief, it was an “institution,” not just a “residence”; part of
the state apparatus, not merely a container of state apartments.
The word for palace in Sumerian and Akkadian, the languages of
ancient Mesopotamia, is composed of the Sumerian sign for “house”
followed by the adjective “large, great” (é.gal).1 The Akkadian borrowing
is not a literal translation (where Akkadian “house” = bītu and “large”
= rabû), but rather is formed from the Sumerian (ekallu), emphasizing
the composite term as its own cognitive category. At base, the under-
lying adjective denotes scale, but may also be seen to reflect elevated
(enlarged) status and function, such that a more accurate translation
might be “the Great House,” as opposed to merely “the big house.”2
* This article originally appeared as “ ‘Seat of the Kingship’/A ‘Wonder to Behold’:

The Palace as Construct in the Ancient Near East,” Ars Orientalis 23 (1993) issue on
Pre-Modern Islamic Palaces, G. Necipoglu ed., pp. 27–55.
1
For the non-specialist; it may be useful to note that Sumerian-speaking peoples
precede Akkadian speakers in the lower Tigris-Euphrates valley. The two languages
and peoples mixed in the third millennium b.c., after which time we find Sumerian
loan words in Akkadian, and evidence for Akkadian influences and words in Sumerian
as well. Akkadian briefly (ca. 2300 b.c.) and then ultimately (ca. 1900 b.c.) replaced
Sumerian and serves as the language of Babylonia and Assyria. Akkadian scribes and
scholars continued to be taught Sumerian, copying old inscriptions as part of the school
curriculum, and preserving literary and ritual texts through translation and bilingual
word lists. Bilinguals continued into the early first millennium b.c., by which time
Sumerian was no longer a living language. The conventions in modern scholarly texts
are to indicate Sumerian words in normal type and Akkadian words in italics.
2
Note that the Sumerian word for temple, by contrast, consists merely of the
logogram for “house” followed in construct by the name of the deity, as in É.dInanna,
“the house of the god(dess) Inanna.” My translation of “Great House” for the Meso-
potamian palace should not be confused with a structure so indicated and distinct
from the palace/ekallu in Hittite tradition (see H. G. Güterbock, “The Hittite Palace,”
334 chapter twenty-nine
In a Mesopotamian gloss, preserved in the year-name of an Old

Babylonian ruler of the eighteenth century b.c. and written in Sume-
rian, a palace the king has just constructed for his highest administrative
official is referred to as worthy of being “the seat of his kingship” (ki-tuš
nam-lugal-la-na), while a slightly earlier hymn in Sumerian speaks in
the voice of a ruler, who describes how the foundations of his rule were
made strong “in the palace of kingship, in my pure, good seat” (é-gal-
nam-lugal-la ki-tuš-kù-du10-ga-gá).3 These two references imply far more
extended functions for a royal palace than merely the royal residence;
the use of the abstract noun, “kingship,” suggests that the palace is the
center from which rule is exercised and in which the state is run.
I hope to demonstrate here that issues of morphology and decora-
tive program are tied to this extended function. Were the collection of
essays in this volume and the conference it preserves devoted to the
palace in the ancient Near East, contributors would each be taking
a particular region, period, function, or form—as has been done in
the Islamic contributions that follow. Instead, I will attempt to cover
a comparable subfield within a single paper. The service an overview
of the ancient Near East can provide for Islamicists, whose areas of
interest coincide in large measure with the major geographical units
of the ancient Near East, is to offer a relatively broad survey of trends
and a relatively detailed bibliography. This will permit a perspective on
continuity and change across historical, cultural, and religious divides,
and also the possibility of pursuing particular aspects through further
reading. The highlighting of selected examples and aspects of ancient
Near Eastern palaces will lay the foundations for inquiry into whether or
not there are areas of significant overlap in the pre-Islamic and Islamic
Near East, and I hope stimulate others to seek further continuities and
observe meaningful changes as particular interests arise.
Regional divisions of the ancient Near East were no less distinctive
in ethnic, linguistic, and cultural character than those in later, Islamic
periods. The multiple functions of a Mesopotamian palace are thus
in Le palais et la royauté: XIX eme Rencontre Assyriologique Internationale 1971, ed. P. Garelli
[ Paris, 1974], 305–6).
3
J. V. Kinnier-Wilson, The Nimrud Wine Lists: A Study of Men and Administration at
the Assyrian Capital in the 8th Century B.C. (London: British School of Archaeology in
Iraq, 1972), 45: year 34 of Samsuiluna of Babylon; G. Pettinato, review of Römer,
Sumerische ‘Königshymnen’, in Archiv für Orientforschung (1992), 208–10, line 99, re Lipit-
Ishtar of Isin.
“seat of kingship”/“a wonder to behold” 335
not necessarily characteristic of royal residences in the neighboring

city-states of northern Syria and southeast Anatolia, even though they
were also referred to as ekallu.4 These “palaces” were considerably
smaller in scale than their Assyrian and Babylonian counterparts, had
their own characteristic forms, and do seem to have been simply resi-
dences. In an interesting turn-around, when the Assyrian rulers of the
first millennium b.c. adopted and adapted that Western form, they did
not also call it a palace—presumably on the principle that you cannot
call two different things by the same name—and so coined a special
term, bīt-hilāni (on which, see below), in order to distinguish it from
their own larger complexes. Wherever possible, I shall refer to these
regional differences, particularly with respect to Anatolia and Syria-
Palestine; but for purposes of time and space will leave out significant
aspects of ancient Egypt, on the one hand, and ancient Iran, on the
other. Throughout, I shall concentrate mainly on the Mesopotamian
sequence—the most complete and perhaps also the most developed we
have for the ancient Near East.
In selected examples of palaces from the Uruk Period to the Neo-
Babylonian (from ca. 4000 to ca. 500 b.c.), focus will be on the three
aspects referred to above: form and space, including technology; decora-
tive program; and function. At the same time, the reader is referred to
several recent specialized studies that present far more detailed descrip-
tions, analyses, and illustrations than are possible here.5
As is the case for much of the pre-modern Islamic world, the direct
information we have about the ancient Near Eastern palace comes
largely from often incomplete archaeological remains, amplified by tex-
tual reference. Unlike Islam, however, there is no living tradition—hence
no contemporary practitioners or constructions to aid in interpretation
or reconstruction. Our popular impressions of the pre-Islamic palace
come from biblical references, on the one hand (as, for example, the
description of Solomon’s palace in Jerusalem that follows the account
4
And when multiple functions are characteristic of Hittite palaces of the late second
millennium b.c. on the Anatolian plateau (cf. Güterbock, “Hittite Palace”), this is not
to say that they are the same functions. Each historical and cultural case needs to be
examined first in terms of internal evidence for associated activities.
5
E.g., Ernst Heinrich, Die Paläste im alten Mesopotamien, Denkmaler antiker Archi-
tektur 15, Deutsches Archäologisches Institut (Berlin, 1984); Seton Lloyd and H. W.
Muller, Ancient Architecture (New York: Rizzoli, 1986); Jean Margueron, Recherches sur les
palais mésopotamiens de l’âge du bronze, 2 vols. (Paris, 1982); Rudolph Naumann, Architektur
Kleinasiens: von ihren Anfangen bis zum Ende der hethitischen Zeit (Tübingen, 1955).
of his temple, I Kings 7:1–12), and from British watercolor reconstruc-

tions of Assyrian palaces at Nimrud and Nineveh following excavations
in the mid-nineteenth century, on the other.
In the latter, we see splendid, multi-storied structures, elaborately
decorated. Because the ground plans and ground-floor sculptures were
based upon excavated remains, they are generally reliable; however
the rest of the elevation is largely invention: oftentimes a hodge-podge
copied from other known ruins, such as Persepolis in Iran.6 It is also
striking when one compares the restoration drawings of Nimrud and
Nineveh, how very much they resemble the drawings done for the
restoration of the façade of Buckingham Palace in the 1820s! One
is forced to conclude that the draftsmen responsible for the Assyrian
watercolors were themselves “(re-)constructing” palaces according to
their own contemporary desires and imagination—in particular an
imagination that saw the Assyrian “empire” in the mirror of the then-
contemporary British empire.
The account of Solomon’s palace does contain a good deal of useful
information: from his lavish use of cedar in construction (7:3), to the
presence of three rows of windows, suggesting multistory façades (7:4),
the relationship between portico, where the king sits in judgment, and
inner courtyard (7:7), and the separate residence for his wife, pharaoh’s
daughter (7:8). This information has not received much attention, largely
because, as is also the case with Islamic research, scholars have con-
centrated mainly on religious areas and buildings. Yet, not only for the
Solomonic period but throughout the ancient Near Eastern sequence,
to establish a new state, or capital, both a temple to the primary deity
and a palace had to be constructed.
The Uruk Period (ca. 4000–3000 b.c.), in which the early stages of
a complex social hierarchy and large-scale urbanization have been
observed, is a logical starting point for examining the Mesopotamian
palace.7 However, at the type-site of Uruk/Warka on the lower Euphra-
tes, although archaeologists have recovered a large complex of build-
ings identified as temples, with characteristic tripartite plan, bent axis
6
James Fergusson, The Palaces of Nineveh and Persepolis Restored (London: John Murray,
1851), fig. 31 and frontispiece.
7
Hans J. Nissen, The Early History of the Ancient Near East: 9000–2000 B.C. (Chicago,
1988), ch. 3, 39–64.
approach in the cella, altar, and podium,8 nothing clearly recognizable

as a palace has yet been discovered. One anomalous structure has been
excavated in the sacred (Eanna) precinct, levels V–IVa; known as Build-
ing 11 or “Palace E,” it is square in plan, with a large central courtyard
surrounded by banks of rooms (resembling more than anything the
later Islamic four-iwan building type with very shallow iwans; see fig.
1).9 The plan is clearly distinct from that of a temple, so the building
has been suggested as a possible palace. The problem is that we have
no corroborating textual evidence for the building, nor even for the
existence in the early texts of a title that clearly designates a ruler; so
the building could well house “administrative” activities and still be
related to the religious complex.10 How the Mesopotamian state and a
designated ruler emerged is far too complex to discuss here, although
most scholars agree that some sort of hierarchical organization in
governance must have been operative in the Uruk period. At the same
time, since archaeological work at the site of Warka has concentrated
on the sacred quarter, it is certainly possible that a palace exists in unex-
cavated areas. Only further fieldwork can help to determine whether
we are faced with an accidental absence in the archaeological record,
or a meaningful absence in the historical record.11
The first buildings to be clearly identified as palaces date from the
third phase of the Early Dynastic Period (ca. 2600–2430 b.c.), and
coincide with the earliest textual evidence for titles denoting rule: Sume-
rian lugal, “king,” and ensi, “steward” or “governor,” a regional title
equivalent to king in the hierarchy of governance. The best (and earliest)
archaeological evidence comes from Kish, the legendary city where
8
O. Tunça, L’architecture religieuse protodynastique en Mésopotamie, Akkadica, Supplemen-
tum II (Louvain, 1984); Ernst Heinrich, Die Tempel und Heiligtümer im alten Mesopotamien
(Berlin, 1982).
9
See plan in M. S. B. Damerji, The Development of the Architecture of Doors and Gates in
Ancient Mesopotamia, trans. T. Takaso and Y. Okada (Tokyo: Institute for Cultural Studies
of Ancient Iraq, 1987), fig. 147; also published in Heinrich, Die Tempel, fig. 118a, and
Nicholas Postgate, Early Mesopotamia: Society and Economy at the Dawn of History (London:
Routledge, 1992), fig. 4.2.
10
See discussion in Postgate, Early Mesopotamia, 137ff., and also reference to the issue
of the connections between the origins of royalty and the palace in early Mesopotamia
in O. Tunça, review of Margueron, Recherches sur les palais, in Zeitschrift für Assyriologie
74 (1984): 318.
11
See on this period, Heinrich, Die Paläste, 9–13, where some possible candidates
for palaces are noted, based on morphological comparisons with later palaces, but
without definitive evidence.
kingship as an institution is said to have “descended from heaven.”12

Although there seems to be a great deal of variation in overall plan, the
two buildings designated as palaces contain a large number of rooms of
differing size and shape, suggesting many functions, and all seem to share
what is later a defining characteristic of palaces in Mesopotamia—a
large central courtyard.13 Since this is a period in which autonomous
city-states were distributed across the Mesopotamian alluvium, it is to
be anticipated that each central city would have had its own palace,
however incomplete the present evidence.
In the succeeding, Akkadian Period (ca. 2334–2154 b.c.), political
development within the period marks a significant change toward a
centralized nation-state, incorporating formerly autonomous city-states
within a single polity under the hegemony of Agade. Unfortunately,
the capital of Agade has not been definitively identified or excavated.
A number of other sites have produced large buildings dated to the
Akkadian period and identified as palaces (e.g., Khafaje, Tell Asmar,
Tell al-Wilayah, possibly Assur), all of which have features in common:
at least one central court, perimeter walls with a primary entrance,
evidence of residential use together with other activities.14 The most
complete palace plan preserved is that at Tell Brak, a site in the
Habur region of northern Mesopotamia.15 The building is identified
through bricks stamped with the name of Naram-Sîn, king of Agade
(2254–2218 b.c.). As reconstructed, it is essentially square, ca. 80
meters a side, but incomplete to the south and southwest, and known
only from foundations. Nevertheless, features preserved are common
to other palaces: massive perimeter walls; single, monumental entrance
on axis with large courtyard, surrounded by rooms; and at least three
additional, smaller courtyards also flanked by banks of rooms (fig. 2).16
12
See on Kish, E. Mackay, A Sumerian Palace and the “A” Cemetery at Kish, pt. 2
(Chicago, 1929); P. R. S. Moorey, “The ‘Plano-Convex Building’ at Kish and Early
Mesopotamian Palaces,” Iraq 28 (1964): 83–98; Heinrich, Die Paläste, 14–28; Postgate,
Early Mesopotamia, 137.
13
For a discussion of the spatial properties of that central court in Palace A at
Kish, see J. Margueron, “Remarques sur l’organisations de l’éspace architectural en
Mésopotamie,” in L’archéologie de l’Iraq du début de l’époque néolithique à 333 avant notre ère,
Colloques internationaux du Centre National de la Recherche Scientifique, no. 580 (Paris:
Editions du CNRS, 1980), 165–68.
14
See Heinrich, Die Paläste, 43.
15
See David Oates, “Excavations at Tell Brak, 1978–81,” Iraq 44 (1982): 187–204,
which also includes bibliography for the early excavations of the building.
16
Heinrich, Die Paläste, fig. 22.
However, since Brak is a site at the very northern periphery of Akka-

dian political influence, this building is probably more a fortress-cum-
governor’s-palace/provincial administrative center than a royal seat.
If we expand our definition of palace to include not only residences
associated with the exercise of power by the highest absolute authority,
but also by the highest local authority in any given political structure,
then provincial governors and dependent local princes can certainly also
occupy “palaces.” Indeed, the Old Babylonian year-name cited above,
which referred to the residence of a high-ranking official as a palace
[é.gal], suggests that the extended administrative functions performed
in such a building may be the most operative variable in defining the
term, which should then be applicable to more than the royal seat, even
within the capital.17 In any case, while the Brak building may well be
called a palace, it is conceivable that the Akkadian kings’ palace(s) back
in Agade would have been larger, perhaps less regular, the exterior walls
perhaps less massive, and might well have contained a greater variety
of room types, correlated with a wider range of activities. In short, it
is not clear that one can generalize from this plan to the capital, or
for the period as a whole.
The recently excavated site of Tell Mardikh (ancient Ebla) in North
Syria provides us with further evidence that palaces of this period
(roughly the late Early Dynastic/early Akkadian Period) were not merely
residences, great houses of local hierarchical rulers, but were also cen-
ters of political and administrative activity. In the royal palace of level
IIB1 were found hundreds of cuneiform tablets the contents of which
range from treaties with foreign rulers to daily economic records, all
carefully stored on shelves and in baskets within specially designated
archives.18 While no extant southern Mesopotamian palace has produced
such archives, this is likely to be a result of the palaces having been
cleaned out and often razed to their foundations to permit subsequent
building. The demonstrated epigraphical relationship between the
Ebla tablets and texts found in non-palatial contexts in Mesopotamia,
17
This would certainly seem to be true at the Assyrian site of Khorsabad, where
the royal palace situated on the upper citadel was replicated in smaller scale in several
buildings on the lower terrace. One of the lower buildings has been identified as the
residence of the king’s brother and chief vizier; the other possibly belonged to his son
as crown prince (Gordon Loud, “An Architectural Formula for Assyrian Planning, Based
on the Results of Excavations at Khorsabad,” Revue d’Assyriologie 33 [1936]: 153).
18
See on Ebla the many publications of P. Matthaie, for example, I tesori di Ebla
(Rome and Bari: Editori Laterza, 1985), 25–54 and figs. 14–36, 41–47.
in conjunction with known political and military events that link the
two regions, allows us to posit the existence of similar palace archives
in contemporary and even earlier Sumerian and Akkadian palaces as
well.19 That the practice continues into the early second millennium is
evident from the accumulated tablets and sealings found in the palace at
Tell Leilan in the Habur region of North Syria, the contents of which
attest to a broad network of communications between related polities
of the Old Babylonian/Old Assyrian Period.20
Happily, a relatively well-preserved palace has been excavated at Mari
on the middle Euphrates, which was apparently in use over a number
of reigns from the late Ur III/Isin-Larsa to the early Old Babylonian
Period (ca. 2000–1758 b.c.). It is in this palace, consisting of some three
hundred rooms and courts, that all of the spatial configurations plus
decorative schemes and administrative functions that characterize later
Assyrian and Babylonian palaces can be observed as part of a coher-
ent complex of features.21 Clearly recognizable is the primary entrance
into a large paved courtyard, with subsidiary rooms ranged around that
space and an additional, smaller inner court (see plan, fig. 3). While
19
It is indeed unfortunate, given the number of administrative texts and literary
texts preserved from the Ur III period (ca. 2112–2004 b.c.) in southern Mesopotamia,
that we have not recovered any of the major palaces of Ur III rulers (cf. Postgate, Early
Mesopotamia, 141). It is therefore difficult to estimate how characteristic of the period
was the provincial palace built by an Ur III governor of Eshnunna in the Diyala River
region to the northeast (Henri Frankfort, Seton Lloyd, and Thorkild Jacobsen, The
Gimil-Sin Temple and the Palace of the Rulers at Tell-Asmar, Oriental Institute Publications
43 (Chicago, 1940).
20
Dominique Parayre and Harvey Weiss, “Cinq campagnes de fouilles à Tell Leilan
dans la haute Jezireh (1979–1987): Bilan et perspectives,” Journal des Savants ( Jan.–June
1991), 3–26. Unfortunately, however, not enough of the plan of the palace has been
excavated to permit comparison with other buildings.
21
See on Mari the initial excavation reports of Andre Parrot, Mission Archéologique
de Mari, vol. 2, pt. 1: Le Palais: architecture; pt. 2: Le Palais: peintures murales; and pt. 3:
Le Palais: documents et monuments (Paris, 1958–59); and more recently, Heinrich, Die Paläste,
68–81; Margueron, Recherches sur les palais, 209–380 (the most extensive discussion to
date); Marie-Henriette Gates, “The Palace of Zimri Lim at Mari,” Biblical Archaeologist
( June 1984), 70–87. Interpretive studies include Y. al-Khalesi, The Court of the Palms:
A Functional Interpretation of the Mari Palace, Bibliotheca Mesopotamica 8 (Malibu, 1978);
Jean-Marie Durand, “L’organisation de l’éspace dans le palais de Mari: le témoinage
des textes,” in Le système palatial en Orient, en Grèce et à Rome, Université des Sciences
Humaines de Strasbourg, Travaux du Centre de Recherche sur le Proche-Orient et
la Grèce antiques 9 (Leiden, 1987); Jean Margueron et al., “Les appartements royaux
du premier étage dans le palais de Zimri-Lim,” M.A.R.I. 6 (1990): 433–62; Béatrice
Pierre, “Décor peint à Mari et au Proche-Orient,” M.A.R.I. 3 (1984): 223–54, and
Béatrice Pierre-Muller, “Une grande peinture des appartements royaux du palais de
Mari (Salles 219–220),” M.A.R.I. 6 (1990): 463–558.
not all scholars agree on the functions attributed to specific rooms or

areas of the palace,22 there is no doubt that ovens and food-storage
features indicate the residential nature of the building. Indeed, a small
group of cuneiform tablets found in the northwest wing of the palace
attest to the delivery of delicacies for visiting dignitaries.23 When that
evidence is seen in conjunction with the later Assyrian administrative
texts known as the Nimrud Wine Lists, it is clear that at any given
moment the palace household included large numbers of individu-
als—members of the royal family, high court officials, eunuchs, guards,
workers, and visitors—all of whom were being fed and provisioned from
palace stores.24 In addition, a vast collection of administrative texts has
been preserved in rooms around the outer court. The range of subjects
covered by these texts makes it clear that the palace was engaged in
administering the king’s own estates and production industries, as well
as affairs of state.25
The Mari palace also preserves for us the first attestation of a particu-
lar constellation of formal reception suites well known from palaces in
later periods. At Mari, the reception suite is set parallel to the northern
end of the inner court (see fig. 3, court 106 and rooms 64 and 65).
A central doorway connects the first room to the courtyard. There is
evidence of a podium on the south wall opposite that central door. The
podium was plastered and whitewashed, giving it special prominence,
and could have been used either as a base for a statue, or, more likely,
as a platform for the throne of the ruler himself, for those occasions
that called for him to be in view, and, reciprocally, with a full view
of the courtyard. The inner room included a second podium on the
short, west wall, which then faced down the length of the room toward
an elevated niche that, it has been suggested, may have contained an
image of the local goddess, Ishtar. Identical suites, with the innermost
room being the formal throne room and a throne base preserved along
a short wall, are also to be found in Assyrian palaces of the first mil-
lennium (see below). What is more, evidence of a developed program
of decoration in wall paintings is preserved at Mari that also echoes
the decorative programs of later Assyrian palaces.
22
See chart in Gates, Biblical Archaeologist ( June 1984) for divergent opinions.
23
M. Burke, “Une reception royale au palais de Mari,” Revue d’Assyriologie 53 (1959):
139–46.
24
Kinnier-Wilson, Nimrud Wine Lists, 32ff.
25
Postgate, Early Mesopotamia, 149.
In the Mari palace, two sets of wall paintings were found fallen from
the northern wall of court 106 that gives access to the throne-room
suite.26 One set preserves what is likely to be a royal figure attendant
upon a sacrifice; the other what seems to be a scene of investiture of a
ruler, identifiable by his headgear and garment, by the goddess Ishtar,
within an elaborate setting of trees and plants (see detail, fig. 4). This
last scene was placed on the façade just to the right of the doorway
entering the throne-room suite, and al-Khalesi has suggested that the
space depicted in the painting in fact replicates the physical space of
the inner throne room. Especially if this is so, but even if the iconog-
raphy merely asserts the special selection of the ruler by the goddess in
general terms, the presence of the “investiture” painting suggests that
the façade of the throne-room complex serves as an important convey-
ance for statements of royal rhetoric and state ideology—a pattern we
will see in both Assyrian and later Babylonian palaces.
A third set of paintings comes from room 132, a small chamber
opposite the main entrance off the large court 131. The floor level of
this chamber is raised slightly and the entrance emphasized by con-
centric semicircular steps that jut out into the courtyard. This special
focus, in combination with the fact that the imagery includes the figure
of a ruler pouring libations before a seated deity, has led to suggestions
that the chamber is a small chapel. I find such a suggestion persuasive,
particularly as both textual and other archaeological evidence attest to
the presence of ritual spaces in other palaces.27 In Hittite Anatolia as
well, the king and queen were expected to perform certain ritual acts
daily, and Güterbock has adduced the likelihood of a sanctuary as a
regular feature of the palace.28 A ritual function has also been attrib-
uted to certain suites within the Assyrian palace of Assurnasirpal at
Nimrud;29 and in historical times, the association of a royal chapel with
26
See on this Anton Moortgat, “Wandgemalde im Palaste zu Mari und ihre histo-
rische Einordnung,” Baghdader Mitteilungen 3 (1964): 68–74, al-Khalesi, The Court of the
Palms; and Pierre, “Décor peint à Mari.”
27
See especially, Frankfort et al., The Gimil-Sin Temple and the Palace of the Rulers,
where the palace of the provincial governor of Eshnunna when the city was under
the hegemony of Ur included two shrines, one presumably to the local deity, the other
to the deified ruler of Ur.
28
Güterbock, “Hittite Palace,” 310.
29
Marc Brandes, “La Salle dite ‘G’ du palais d’Assurnasirpal II à Kalakh, lieu de
ceremonie rituelle,” Actes de la 17ème Rencontre Assyriologique Internationale (Gembloux, 1970),
147–54. Note, however, that Julian Reade is not persuaded by Brandes’s argument
that the suite around room G was used for ritual lustrations and suggests banquets as
the palace is certainly well known. The Mari palace thus adds evidence
for an important religious component to supplement the various aspects
of the Mesopotamian complex.
Most recently, Margueron et al. and Pierre-Muller have published an
additional series of paintings, unfortunately fragmentary, that appar-
ently decorated a reception suite in a second story or storey in the
southeast wing of the same palace (above areas E, F, and room 120
on the plan, fig. 3).30 Its fragmentary nature and the difficulty of recon-
struction notwithstanding, this group of paintings is extraordinary for
containing in its repertory a number of motifs—ruler in combat with
a lion, ruler victorious over a fallen enemy, and ruler receiving some
delegation of approaching individuals—that find direct counterparts
in the more complete decorative programs of later Assyrian palace
reliefs. This implies an iconography of rule associated with palace
decoration in use over at least a thousand-year period. Furthermore,
the presence of a second story that includes a reception space fits well
with both the Solomonic description cited above and later textual and
pictorial evidence. In particular, in the eleventh-century b.c. Egyptian
text of Wenamun, the Egyptian envoy who visits Byblos on the Medi-
terranean coast is received by the local prince in the upper chambers
of his palace, from the windows of which one could see the sea.31
Both the texts and this new evidence from Mari suggest that we may
be missing quite essential parts of Near Eastern palaces known only
from ground floors or foundations, and we should therefore not try to
distribute all palatial functions across the ground-floor plan as if that
were the entirety of the building.
Recent analyses of wall construction at Mari suggest several rebuild-
ings and a long period of use before Mari was conquered by Hammu-
rabi of Babylon in his thirty-fourth year (ca. 1758 b.c.); Moortgat had
an alternative (“Assyrian Architectural Decoration: Techniques and Subject Matter,”

Baghdader Mitteilungen 10 [1979]: 85). The argument for both is based upon the imagery
of the seated king holding a phiale-like bowl on the reliefs, and whether he is lustrating
or drinking is not certain. However, on scenes where individuals are clearly banqueting,
as in the reliefs of Sargon II at Khorsabad, their drinking vessels are obviously being
brought to the lips, which is not the case here, so I tend to be more persuaded by the
symbolic than the literal in the present scene.
30
Cited above, n. 21.
31
Cited in I. J. Winter, “Art as Evidence for Interaction: Relations between the Assyr-
ian Empire and North Syria,” in Mesopotamien und seine Nachbarn: politische und kulturelle
Wechselbeziehungen im Alten Vorderasien, ed. Hartmut Kühne, Hans J. Nissen, and Johannes
Renger, Berliner Beiträge zum Vorderen Orient, vol. 1, pt. 1 (Berlin, 1982), 363.
earlier argued that the wall paintings in the palace reflected different
phases as well.32 This raises an additional caution for the scholar, for,
while it is important to emphasize the degree of continuity that exists
from the second into the early first millennium b.c. with respect to
some aspects of spatial configuration, decoration, and function, it is
also the case that we cannot assume each palace represents a single,
coherent program. It is only in the Neo-Assyrian period that we have
a sufficient number of examples and degree of preservation to attempt
fuller readings of attitudes toward rule and to assess experiences of
authority—what I have called elsewhere “royal rhetoric”—as articulated
in palace construction and decoration.
During the early second millennium b.c., the region around the upper
Tigris, near modern Mosul, had established its political independence
from the south. In the early first millennium, this area constituted the
heartland of Assyria, from which, in a series of military maneuvers over
a period of some three hundred years, the state expanded its territory
until it reached from the Zagros in the east to the Mediterranean in the
west, and from the Taurus in the north to Babylon and Egypt in the
south and southwest. Over this period, virtually every successive ruler
initiated the construction of a new palace, as the capital shifted from
Assur to Nimrud, to Khorsabad, and finally to Nineveh. Although there
are no explanations for these shifts in the several preserved Assyrian royal
inscriptions, they have been understood as a function of statecraft.33 As
with Solomon in Jerusalem, a new ruler established the authority of his
reign in part through palace and other building campaigns.
The complete circuit of city walls has been traced at Nimrud and at
Khorsabad, with enough preserved at Nineveh to suggest that a similar
pattern prevailed (fig. 5). Essentially, rectilinear enclosure walls pierced
by gates in all directions surrounded large areas. Set into and sometimes
breaking the line of the exterior wall were two types of construction,
often at different ends of the city: a raised citadel containing royal
32
Moortgat, “Wandgemalde im Palaste zu Mari”; also discussed in Moortgat, The
Art of Ancient Mesopotamia: The Classical Art of the Near East (London, 1969), 82–84.
33
Khorsabad is a particularly good example since, like Samarra and Fatehpur Sikri,
it was built and occupied by a single ruler, Sargon II of Assyria. The phenomenon
of new palace construction by successive rulers was noted by Oppenheim, Ancient
Mesopotamia, and has been discussed in part by Sylvie Lackenbacher, Le roi bâtisseur:
Les récits de construction assyriens des origines à Teglatphalasar 3 (Paris: Editions Recherche
sur les civilisations, 1982): 76, as a way of demonstrating that the ruler surpassed all
of his predecessors.
palace(s) and temples, and a building known in Akkadian/Assyrian as

an ekal-māšarti, a fortified palace, or arsenal, based on the general plan
of a residential palace, but often larger in scale and with a simpler
distribution of rooms around each courtyard.34
The consistency of the Assyrian pattern highlights the difference
from neighboring first-millennium citadels. In the Assyrian case, the
royal palace and citadel are set into the rectangular perimeter wall,
often overlooking a distinctive natural feature in the landscape, like
a river; in the capital cities of the principalities of North Syria and
southeast Anatolia to the west, as in the kingdom of Sam al at Zincirli,
the pattern is rather to contain the citadel and palace at the center, a
round perimeter wall more or less equidistant at all points from the
citadel enclosing the lower town.35 In all cases, the citadels are raised,
and access is limited via controlled routes and gates, in a way similar
to later Islamic practice in the Near East. The royal palaces of the
ancient Near East are also themselves frequently set on raised platforms,
so that a continuous sequence of physical elevations may be read as
progressive elevations in status.36
In the case of building techniques and materials, a combination
of environmental factors plus wealth, labor force, and extension of
trade network conspires to dictate materials and methods, which pres-
ent significant regional variation. Throughout Mesopotamian history,
the primary building material was mud brick—making use of the
most abundant natural resource in a region of virtually no stone or
construction-size wood. At Zincirli, set in the foothills of the amply
wooded Amanus mountains, stone foundations are overlaid with walls
that combine wooden beams and brick. On the Anatolian plateau to
the northwest, stone construction was common. In Assyria, brick was
used for bearing structural loads; however, proximity to sources of stone
and wood allowed rulers under the influence of the west and northwest
34
Fort Shalmaneser, excavated at Nimrud between 1949 and 1963, is a good example
of the type—on which, see M. E. L. Mallowan, Nimrud and Its Remains (Aberdeen, 1966),
vol. 3, plan VII. For the city walls, with citadel and ekal-māšarti set into the perimeter,
see plan in Moortgat, Art of Ancient Mesopotamia, fig. 102.
35
Naumann, Architektur Kleinasiens, fig. 259.
36
Note also the point made by Postgate, Early Mesopotamia, 137–40, that in the ear-
lier periods in Mesopotamia, the location of the palace varied significantly from one
city-state to another and from one period to another: sometimes located close to the
traditional temple/sacred precinct, sometimes in newer sectors of town, away from
the older traditional areas, and this was closely tied to the relationship of a particular
ruler to the state apparatus and the religious authority.
to introduce stone revetments and orthostat slabs as decorative skins on

the walls of their public buildings, and to employ a variety of precious
woods as well. From the Middle Assyrian Period (second half of the
second millennium b.c.) through the Neo-Assyrian Period (first half of
the first millennium b.c.), kings describe in display texts the lavish con-
struction materials assembled for their palaces and, in the later phase,
actually depict on palace reliefs the cutting of wood and quarrying of
stone blocks for sculpture.37
Limited by the preservation of buildings, we are reduced to recon-
structing façades, lighting, and roofing methods from the occasional
ancient representations of architecture, in monumental scale on palace
reliefs or in miniature on cylinder seals.38 From these images it would
seem that massive exterior walls with niched façades and crenellated tops
were prevalent at least from the mid-second millennium. And, although
post-and-lintel construction was likely to have been the principal way of
spanning space and bearing weight, there is again evidence for vault-
ing in palaces from the same period.39 At Khorsabad, both complete
barrel vaults and intact arches over major entries were well preserved,
as was arched wall construction in the so-called Governor’s Palace at
Nimrud.40 All of this suggests that while the antecedents of Islamic
construction may be found most immediately in the great arches of the
Sasanian Period at sites like Ctesiphon, the beginnings of that tradi-
tion may have reached considerably farther back into antiquity than
is generally acknowledged. Indeed, I should not be at all surprised if
37
Cf., for example, A. Kirk Grayson, Royal Inscriptions of Mesopotamia 2 (Toronto,
1991): 38–45, for the palace of Tiglath Pileser I at Assur and 227–28 for the Northwest
Palace of Assurnasirpal at Nimrud; John M. Russell, “Bulls for the Palace and Order
in the Empire: The Sculptural Program of Sennacherib’s Court VI at Nineveh,” Art
Bulletin 69 (1987): 520–39, and Sennacherib’s Palace without a Rival at Nineveh (Chicago:
University of Chicago Press, 1991). Although the Neo-Assyrian palaces are far better
preserved, it would seem from the Middle Assyrian texts that kings employed many of
the same building techniques as are archaeologically attested later—for example, the
lining of the walls with basalt and alabaster slabs, and the installation of large gateway
figures, noted in the Tiglat Pileser I inscription cited above.
38
See especially Edith Porada, “Battlements in the Military Architecture and in the
Symbolism of the Ancient Near East,” in Studies in the History of Architecture Presented to
Rudolf Wittkower (New York: Phaidon, 1967), 1–12; and Margueron, Recherches sur les
palais, pt. 2, ch. 2, on open and closed spaces in the Mesopotamian palace.
39
Heinrich, Die Paläste, 89–91 and fig. 46; Margueron, Recherches sur les palais, 521–22;
H. Gasche and W. Birschmeier, “Contribution à l’étude de la voûte en briques crues,”
Akkadica 24 (1981): 1–16; and Damerji, Doors and Gates, 116 ff., for doors with arched
construction.
40
Heinrich, Die Paläste, fig. 76.
one day we find evidence of simple dome construction in the ancient

Near East as well!
Great attention was paid in the Assyrian palaces to the scale and deco-
ration of major gateways and entrances, including threshold inscriptions
and the colossal human-headed bulls and lions that flanked principal
doors (fig. 6).41 It is not clear whether these great stone colossi actually
carried the weight of doorway arches, or, like the orthostats, simply
lined the walls; but their iconography is one of menace and protection.
The placement of monumental stone sculpture at doorways seems to
have been borrowed by the Assyrians from the West—where gateway
lions and sphinxes are known from second-millennium Hittite sites on
the Anatolian plateau, and then later, from first-millennium Neo-Hittite
citadels, like Zincirli, with which the Assyrians came into contact during
the ninth-century military campaigns in the area.42
The larger lesson to be learned from interactions of this sort, which
must surely be relevant for subsequent periods as well, is that in some
aspects of architectural practice, like the shapes of perimeter walls and
placement of elite citadels, neighboring states may, despite contact,
remain distinct; however, in other aspects, like building techniques,
materials, or decorative schemes, they may change once contact is
established. In the case of Assyria and the West, Assyrian palace con-
struction owes a good deal to foreign contact. In the case of Assyria
and the East, by contrast—as seen at the site of Hasanlu in northwest
Iran, exposed to Assyrian contact around the same time—we see the
converse: Assyrian elements were adopted, as illustrated by the addi-
tion of glazed plaques and new porticoes to embellish local building
façades.43 In this latter case, it would seem that there was a desire to
emulate practices associated with the major political force in the region.
In the case of the Assyrian adaptation of Western elements, there may
also have been some positive charge associated with the incorporation
41
On attention to doors and gates in general, their construction and plan, see
Damerji, Doors and Gates.
42
On this, see Winter, in Mesopotamien und seine Nachbarn, esp. 356. Note, however,
that Reade rightly points out Assyrian colossi have Mesopotamian prototypes as well in
smaller-scale free-standing figures of glazed terracotta or stone (“Assyrian Architectural
Decoration,” 18).
43
It should be noted, however, that interior spatial configurations did not change.
See on this I. J. Winter, “Perspective on the ‘Local Style’ of Hasanlu IVB: A Study in
Receptivity,” in Mountains and Lowlands: Essays in the Archaeology of Greater Mesopotamia,
ed. L. D. Levine and T. C. Young, Jr. (Malibu: Undena Press, 1977), 371–86.
(appropriation?) of a highly developed tradition just as the Assyrian

polity was expanding.
In later historical periods, it is possible to document the spread of
new techniques and modes of construction as part of the general dis-
semination of architectural knowledge and practice. While this is not
possible for the ancient Near East, I do wish to underscore the impor-
tance of seeking to distinguish between transmission of knowledge as
part of practice and culturally charged borrowings that carry with them
coordinates of reference and meaning. Furthermore, I would stress the
fact that regionally distinct traditions in morphology and decoration are
not fixed, but rather, within the constraints of resources and cultural
practice, can respond to historical/political exigencies.
The earliest of the relatively well-preserved Neo-Assyrian palaces,
the Northwest Palace of Assurnasirpal II (883–859 b.c.) at Nimrud, sits
on the western edge of the citadel, overlooking the Tigris. Although
many rooms of the western sector have been eroded away, the basic
configuration of the ground plan can be read (see fig. 7). Typical of
most Neo-Assyrian palaces, it conforms to the basic type established
by the palace at Mari, in which space is divided into two main sectors
organized around an outer and an inner courtyard.44
Dividing the two courts on an east-west axis is the throne-room suite
of two long rooms, one with a throne base preserved in situ (cf. Room
B on plan). Primary access to the throne room is via the large outer
courtyard, where monumental pylons with flanking door guardians mark
the entrance. This entrance is on the long, north wall, necessitating a
90-degree turn to face the throne on the short, east wall. The pattern
of access and layout is one seen already in the palace at Mari, and is
repeated in the reception/throne-room suites of all of the major Assyrian
44
The literature on this palace is vast, as it is one of the best preserved in the
ancient sequence. For a recent study of the architectural basis for reconstruction,
see Richard Sobolewski, “Beitrag zur theoretischen Rekonstruktion der Architektur
des Nordwest-Palästes in Nimrud (Kalhu),” in Palast und Hütte: Beiträge zum Bauen und
Wohnen im Altertum (Mainz: von Zabern, 1982), 237–50. For this particular two-court
plan, see Moortgat, Art of Ancient Mesopotamia, 127, following Loud, “An Architectural
Formula,” 156, in which the outer court is referred to as the “gate-court” or bābānu
in Akkadian, the inner court as the “house” or “residential court,” Akk. bitānu. My
objections to this terminology would take too long to argue here, but in summary can
be related to the likelihood that the rooms flanking the inner court were also devoted
to ceremonial or public functions, while the thickness of the walls (some 5 meters)
argues again for a second story, with the residential quarters more likely to have been
located on the upper floor(s).
palaces.45 Oppenheim suggested that this represents a conscious mod-

eling of the royal audience chamber upon the bent-axis plan of the
early Sumerian sanctuary, in effect sanctifying the ruler without formally
deifying him.46 In any case, the ruler is not on axis with, or visible from,
the outer court, as he would have been in the anteroom at Mari, and
as was traditionally the case in the straight-axis throne rooms of the
later Neo-Babylonian and Persian Periods, where the king was seated
opposite the main door (for example, in the Southern Palace of Nebu-
chadnezzar II in Babylon, fig. 8).47 These traditions of visibility and
sight-lines, both of and by the ruler, can become significant indicators
of cultural and national attitudes toward authority and the person of
the ruler, and analysis of patterns thereby goes beyond description
toward the reconstruction of experience within the built environment,
as Islamicist Eric Schroeder called for nearly forty years ago, and as
Gülru Necipoğlu pursues in her paper from the present conference.48
Subsequent Assyrian rulers also built palaces on the citadel at
Nimrud, but the remains are too fragmentary to read the complete
plans. When Sargon II (721–705 b.c.) decided to shift the capital to
Khorsabad, he constructed the citadel in such away that his palace
and attached temples were the only buildings at the highest level, with
subsidiary palaces and administrative buildings in a separate, lower
enclosure (see reconstruction of the citadel, fig. 9, and plan of the
palace, fig. 10). Sargon preserved the organizational principle of two
main courtyards; however, his throne room (room VII on plan) is no
longer between the two courts, but rather is set longitudinally along the
southwestern wall of the inner court. Nevertheless, the configuration of
45
Geoffrey Turner, “The State Apartments of Later Assyrian Palaces,” Iraq 32
(1970): 177–213, and discussion in I. J. Winter, “Reading Concepts of Space from
Ancient Mesopotamian Monuments,” in Concepts of Space Ancient and Modern, ed. Kapila
Vatsyayan (New Delhi: Abhinav, 1991), 57–73.
46
Ancient Mesopotamia, 328, as cited in Winter, “Reading Concepts of Space,” 63.
47
See on this Marian Feldman, “The Presentation of Kingship: The Neo-Baby-
lonian Throneroom Façade of the Southern Palace at Babylon,” unpublished paper,
1992, as well as S. Polony and G. Winkler, “Statische Untersuchung des Thronsaals
der Südburg in Babylon,” Mitteilungen der Deutschen Orient-Gesellschaft 99 (1968): 55–58.
The Book of Esther of the Hebrew Bible, 5:1–2, records this straight axis in use on
into the Achaemenid period of the second half of the first millennium as well, in its
description of the Persian king seated on “his royal throne in the royal house opposite
the gate,” and from there seeing Esther standing in the courtyard.
48
“Scientific Description of Art,” a review of Donald Wilber, Architecture of Islamic Iran,
in Journal of Near Eastern Studies 15 (1956): 93–102. G. Necipoğlu, Ars Orientalis 23.
the throne-room reception suite remains constant, as it did throughout

the Neo-Assyrian period.49
As noted at the beginning, a very different sort of royal palace is
found in contemporary Neo-Hittite and Aramaean sites to the west of
Assyria. Here, small self-contained structures are marked by columned-
portico entrances into banks of lateral rooms, often with service rooms
at one or both ends (for example, Hilani III at Zincirli, fig. 11). They
either stand independently or are grouped around enclosed courts (as in
the “Upper Hilani” complex at Zincirli).50 The columned portico and
limited size are characteristic of royal buildings in Syria and Palestine
from at least the mid-second millennium onwards. While the polities
they represent are considerably smaller than the large urban states of
Babylonia and Assyria, the reduced size of the Syro-Palestinian palaces
is not merely proportional to their population or territory. Even consid-
ering that they may well have stood several stories high, on the model
of Mari and the Wenamun text cited above,51 the limited number and
type of rooms and spaces suggest that they could not have served as
many and diverse functions or constituents as an Assyrian palace.
It is presumably this smaller type of royal palace to which the Assyr-
ian ruler, Tiglath Pileser III (744–727 b.c.) referred when he declared
that he had constructed a palace in the western manner, which he
called a bīt-hilāni. Sargon II also claims to have constructed such a
building at Khorsabad. The literature on this building type in Assyria
is long,52 and I shall refer to it only briefly here; but I believe it may
have ramifications for the later development of the iwan in the Islamic
architectural tradition, as well as serving as an important historical case
of cultural borrowing.
49
Turner, “State Apartments of Later Assyrian Palaces,” cited above, n. 45.
50
See the lengthy discussion in Naumann, Architektur Kleinasiens, 354–78, for the sites
of Zincirli, Carchemish, and Tell Halaf, among others. More recently on Tell Halaf,
see Jeanny V. Canby, “Guzana (Tell Halaf ),” in Ebla to Damascus: Art and Archaeology of
Ancient Syria, ed. Harvey Weiss (Washington, D.C., 1985), 332–38. Indeed, in the earlier,
Hittite palace of the second millennium as well, although it was apparently a large
complex, Güterbock has suggested that residence and administrative quarters may have
been in separate structures (cf. Güterbock, “Hittite Palace,” esp. 308).
51
Cf. reconstruction in Naumann, Architektur Kleinasiens, fig. 455.
52
See Winter, “Art as Evidence for Interaction.” Note also that J. Börker-Klähn (“Der
bit-hilāni im bit-šahūri des Assur-Tempels,” Zeitschrift für Assyriologie 70 [1980]: 258–72) has
recently documented an instance in which a bīt-hilāni was associated with a temple of
the Neo-Assyrian period in Assur. She reconstructs the building (Börker-Klähn, “Der
bīt-hilāni,” fig. 4) with portal sculpture flanking a large entry, and an upper colonnaded
balcony above the entrance.
Scholars have debated just where Sargon’s bīt-hilāni might have been
located and how its principal features might be recognized. The most
salient feature of the plan of known western palaces, the columned
portico, has been the marker sought by most scholars, largely because it
is observable on the ground. It is this feature that connects the building
type to the iwan: and when as at Zincirli, several buildings are grouped
around one court, we may in fact see antecedents for the three- or
four-iwan building.53 However, it is possible that the term hilāni could
be related to modern Hebrew and ancient Ugaritic ln, “window,”
and may therefore actually be identified less by its columns than by a
multistoried façade with windows, such as is described in I Kings 7 for
the palace of Solomon and as has been reconstructed at Mari, with
perhaps the second-story overhang supported by a columned portico.
In such a case, the bīt-hilāni may not always reflect a separate building,
but rather a suite or complex incorporated in the main palace.
One possibility for the Assyrian bīt-hilāni is that the building was
not in the city or on the citadel at all, but rather was located in some
landscaped area outside the walls, as is depicted in a hunting park on
one of Sargon’s reliefs (fig. 12) and on a relief of Assurbanipal from
room H of his North Palace at Nineveh, where small pavilions with
columned porticoes stand amid trees and watercourses.54 A second can-
didate for the bīt-hilāni at Khorsabad is a small, free-standing structure
(often labeled a temple, but on no solid evidence) that is set on the
western corner of the citadel (see plan, fig. 10). Yet another possibility
is that the attached suite of rooms at the northwestern end of the royal
palace, which extends out beyond the line of the city wall, constituted
a specially designated wing (rooms 1–8 on plan, fig. 10).
A clue to the character of the structure may be contained in Tiglath
Pileser’s description of the building as built “for his pleasure;” that is,
despite the formulaic nature of this phrase, which is used by several
kings, one is led to think of the structure, free-standing or attached, as
distinct from the official apartments and reception areas. This would
apply to all of the three possibilities noted above. Along with the sepa-
rate building on the citadel and the park pavilion, the attached suite of
53
I have even wondered whether there might not be some etymological relationship
between the Aramaean/Akkadian hilāni and Arabic (1) iwan, but such speculation goes
well beyond my own competence.
54
Richard D. Barnett, Sculptures from the North Palace of Ashurbanipal at Nineveh (London:
British Museum, 1976), pl. XXIII (room H).
rooms at Khorsabad would lend itself well to repose. The northwest

edge of the citadel looks out over the course of the river Khosr, thereby
providing both view and fresh air. Sargon tells us that he laid out a
landscaped park at Khorsabad, the siting of which would most appro-
priately be beyond the city to the northwest. In addition, the orientation
of room 7 is such that its doorway is aligned with that of room 4, to
look out to the northwest; and it is precisely in the reliefs of room 7
that we see banquet scenes and an elaborate frieze of the king’s hunt-
ers in a park. Moreover, the trees and river that are represented in the
hunting park conform to Sargon’s description of the park he created,
which he tells us explicitly was modeled on a western landscape. How
better to enjoy the park than in a western-style structure?
However the bīt-hilāni in its original or borrowed form may be identi-
fied in future, its association with leisure and park land introduces the
connection of gardens and purposeful landscaping to Assyrian palaces.
This is attested by the Middle Assyrian period, when Tiglath Pileser
I records taking both hardwood and fruit trees “from the lands over
which I had gained dominion,” and filling the orchards of Assyria with
them, while in a second text he records planting a royal garden for
his “lordly pleasure, in the midst of which he built a palace.”55 The
tradition is perpetuated throughout the Neo-Assyrian period, as seen
from the Banquet Stele text of Assurnasirpal II, in which the king,
describing the founding of the new capital at Nimrud, enumerates the
various trees and plants gathered in his travels and incorporated with
abundant water canals into luxuriant gardens at home.56 Scholars of the
ancient Near East are just beginning to look for archaeological evidence
of such landscaping, particularly associated with palaces;57 but the line
from Assyrian to Babylonian to Achaemenid to Islamic palace gardens
and orchards can at least be affirmed. Terms utilized in describing
55
See Grayson, Royal Inscriptions, 27, 55, and, for a similar account by Assur-
bel-kala, 105.
56
Cf. Donald J. Wiseman, “A New Stele of Assurnasir-pal II,” Iraq 14 (1952):
24–39, and more recent translation in A. K. Grayson, Assyrian Royal Inscriptions, pt.
2 (Wiesbaden: Harrassowitz, 1976), para. 678, as well as discussion in Ann Shafer,
“A Monument to the Center of Empire: Ashurnasir-pal II’s ‘Banquet Stela’ in Context,”
unpublished paper, 1992.
57
See the early study by A. Leo Oppenheim, “On Royal Gardens in Mesopotamia,”
Journal of Cuneiform Studies 24 (1965): 328–33; also Donald J. Wiseman, “Palace and
Temple Gardens in the Ancient Near East,” in Monarchies and Socio-Religious Traditions
in the Ancient Near East, ed. H. I. H. Prince Takahito Mikasa (Wiesbaden: Harrassowitz,
1984), 37–43; and discussion in Lackenbacher, Le roi bâtisseur, 124–29.
these early gardens all denote pleasure and joy. It should come as no
surprise that in these early periods no less than in later times, wealth
and power would be associated with management of the landscape
for purposes of delectation, not just mere sustenance—especially in an
environment where gardens were expensive and difficult to establish and
maintain; but one can also go a step further in suggesting that such a
display could be part of a public statement of wealth, power, and even
territorial appropriation through reference to the lands of origin of the
various trees and plants collected.
As noted above, Assurnasirpal II also refers to the varied types of
wood he employed in his palace at Nimrud, and like the gardens, the
building materials would have conveyed wealth and power indepen-
dent of narrative content. Far more explicit statements of wealth and
power, not to mention political ideology, are possible and attested in
the decoration of the actual palace buildings, through the addition of
applied verbal and visual messages.
It is in the incorporation of inscribed texts into the scheme of palace
“iconography” that Assyrian practice may come closest to later Islamic
practice. Although calligraphic script was never developed in the ancient
Near East to the extent that it was in later Islamic periods, one does
see a distinctly “lapidary style” employed for the palace texts, which
needs further study as part of the overall visual effect in the decorative
program as a whole (visible, for example, on the block surrounding the
doorway colossus, fig. 6). Russell has done the most complete study to
date of the role played by various sorts of inscriptions in a single palace,
where the ruler makes use of each type of text for different rhetorical
purposes.58 For regions brought into the Assyrian polity that retained
58
For the palace of Sennacherib at Nineveh, see Russell, Sennacherib’s Palace without
a Rival, ch. 2, 7–33; see also Pamela Gerardi, “Epigraphs and Assyrian Palace Reliefs:
The Development of the Epigraphic Texts,” Journal of Cuneiform Studies 40 (1988): 1–35,
for a special case of epigraphs applied to the narrative reliefs; and Julian E. Reade,
“Sargon’s Campaigns of 720, 716 and 715 b.c.: Evidence from the Sculptures,” Journal
of Near Eastern Studies 35 (1976): 95–104, for the coupling of distinct types of texts with
equally distinct types of imagery in the palace of Sargon II at Khorsabad. In I. J.
Winter, “Royal Rhetoric and the Development of Historical Narrative in Neo-Assyrian
Reliefs,” Studies in Visual Communication 7, 2 (1981): 2–38, I have attempted to demon-
strate that in the palace of Assurnasirpal II at Nimrud, even the structure of the text
finds parallels in representational practice. In addition, shifts in grammatical structure
when referring to the ruler (from third person singular, as part of narrative accounts,
to first person singular as self-presentation) are indicative, and not without parallels to
the account by Sheila Blair (see article in this volume) of Ilkhanid texts concerning the
ruler (that shift from third-person narratives to second-person exhortatives).
local rulers, we find bilingual inscriptions on palaces that juxtapose

Assyrian Akkadian to the local language. This is not unlike the situation
described by Catherine Asher (see Ars Orientalis 23 (1993)) for Mughal
India, where a local Hindu ruler could include inscriptions in both
Persian and Sanskrit. The ancient Near Eastern and South Asian cases
show intriguing similarities in that both evince significant differences
in nomenclature in the local versus the official court language. One
ruler of Guzana to the west of Assyria, for example, is referred to as
“king” in the local Aramaean, but only as “governor” in the Akkadian,
appropriate to his subordinate status vis-à-vis Assyria.59
The extent to which the Assyrians developed the application of
orthostat stone relief carvings to exterior and interior palace walls was
unprecedented and, as noted above, seems to be derived from contact
with North Syria and Anatolia. The Assyrians employed limestone and
alabaster in their carvings; the Syrian and Anatolian sites often used
basalt as well as limestone, and it is interesting that the alternation of
black and white stone for decorative purposes, as on the Long Wall
and Herald’s Wall at Carchemish,60 as well as the use of lions as door
or gateway figures, can still be attested in the same general region well
into the Islamic period (on which, see the Soucek, Redford, and Tabbaa
articles in Ars Orientalis 23).
The Assyrian orthostats stand some two meters high, and are carved
in relatively high relief, often incorporating inscriptions over or as part of
visual representations. Traces of color suggest that they were originally
painted; and Layard’s account of the throne room of Assurnasirpal II
in the Northwest Palace includes references to fragments of plastered
wall paintings along with the reliefs—presumably from the upper
parts of the wall surfaces and the ceiling (see reconstruction, fig. 13;
although note that the human figures are too small in scale). In addition,
Assurnasirpal’s Banquet Stele mentions the decoration of his palace with
glazed brick and with bronze door bands, examples of which have been
found elsewhere,61 and Postgate, following the speculations of Reade,
59
Ali Abou-Assaf et al., La statue de Tell Fekherye et son inscription bilingue assyro-araméenne
(Paris: Editions Recherche sur les Civilisations, 1982), 64, re line 19 in the Akkadian
text and line 13 in the Aramaic.
60
Winfried Orthmann, Untersuchungen zur späthetitischen Kunst (Bonn: Habelt, 1971),
501–2, 503–5, and pls. 24, 26–28.
61
Glazed bricks as part of palace decoration are attested in the Middle Assyrian
period by Tiglath Pileser I, who continues a palace begun by his father, adding a
façade of glazed brick (the color) of obsidian (black), lapis lazuli (blue), pappardilû-
argues for the probability of textiles, no longer extant, as having been

another important medium of palace decoration.62
Numerous studies in recent years have investigated the sorts of politi-
cal and cultural messages articulated in Assyrian decorative programs.
These messages are conveyed by placement, as well as by content. Com-
posite, protective creatures stand at doorways and at corners. Within
rooms, the king’s figure is often given prominence opposite doorways,
or in the center of a wall, regardless of the subject of the scene.63 On
one such example, a scene of the Assyrian king Sennacherib receiving
prisoners in the field after the siege of Lachish, we see the enthroned
ruler positioned exactly in the middle of the northwest wall (fig. 14).
In the field above the king’s face is a rectangle containing four lines of
explanatory text. The introduction of textual labels into visual narra-
tives, first attested on reliefs in the eighth century, served to emphasize
the image or the narrative moment with which they are associated,
thereby complementing or augmenting the visual program.64
When we look at the sorts of motifs commonly represented in pal-
ace decoration, many commonalities occur across the entire range of
preserved evidence. Foremost among these is the presence of the palm
tree, either as an independent element or in association with the image
of the king himself. In the Babylonian palace of Nebuchadnezzar II
(604–562 b.c.), as at Mari, the palm occurs in a repeating frieze on
the outer façade of the throne room, flanking the central door through
which the ruler could be visible at selected times.65 In the Northwest
Palace of Assurnasirpal, the same tree is depicted throughout the palace,
most frequently flanked by symmetrical genii.66 In the throne room,
which constitutes a special case, the king himself is shown duplicated
stone and parutu-alabaster (which must be yellow and white, respectively, for that is the
color palette found on later exterior glazed façades and walls). For the Assurnasirpal
text, see Grayson, Assyrian Royal Inscriptions, 2:para. 677; indeed, the Banquet Stele on
which the reference occurs could itself be considered part of the palace program,
as it stood in niche EA of the great courtyard D of the palace, just adjacent to the
throne room (see Shafer, “Monument to the Center of Empire”). The only problem
with this is uncertainty, given the later occupations of the palace, that this was the
stele’s original location.
62
For discussion of the throne-room textiles, see Postgate, Early Mesopotamia, 143.
63
Julian E. Reade, “The Architectural Context of Assyrian Sculpture,” Baghdader
Mitteilungen 11 (1980): 86.
64
Gerardi, “Epigraphs and Assyrian Palace Reliefs.”
65
Feldman, “Presentation of Kingship.”
66
See on this motif, now, Barbara N. Porter, “Sacred Trees, Date Palms and the
Royal Persona of Ashurnasirpal II,” Journal of Near Eastern Studies 52 (1993): 129–39.
on either side of the tree, presumably participating in its ritual care

(fig. 15). If Castriota is correct in suggesting that on some occasions the
tree stands metonymically for the ruler, precisely because maintenance
of the fertility of the land through proper ritual performance is a major
function of kingship,67 then the repetition of the motif in Babylon and
throughout the Northwest Palace is not merely for purposes of decora-
tion, but also conveys the powerful message that rule is grounded in
nature, i.e., in cosmology.
The motif of the king and tree is accorded pride of place in the
throne room of Assurnasirpal; it appears both directly behind the
king on his throne on the eastern wall and directly opposite the major
doorway of the north wall. Throne rooms, as relatively public ceremo-
nial and political loci, are especially likely to be highly invested with
charged imagery—in the best preserved cases, incorporating a number
of motifs that in total reflect the full panoply of royal activities and
attributes.68 In the throne room of Assurnasirpal II at Nimrud, scenes
of hunt and battle are distributed along the long, south wall and parts
of the north wall, with the king himself depicted at the far west end. I
have argued elsewhere that the assemblage of images can be read as a
unified program, recapitulating in both content and structure the king’s
“Standard Inscription” that is written over every slab (see, for example,
on throne-room slab B.23, fig. 15), and signifying all of the major
attributes of the ruler appropriate to his stewardship of the state: ritual
67
See David Castriota, “Divinity, Kingship and Metonymy: The Imagery of the
Sacred Tree in the Art of Syria and Mesopotamia in the Second Millennium b.c.,”
unpublished paper.
68
Unfortunately, however, not all throne rooms in Assyrian palaces have been com-
pletely recovered. The reliefs of the throne room of Sargon at Khorsabad (room VII
on plan) were lost in an accident after their removal, those of Assurbanipal at Nineveh
are poorly preserved, those of Tiglath Pileser III were dismantled in antiquity to be
reused in a later palace. It is also likely that no matter how complete the archaeological
remains of a given building may be, we will never fully recover all of the elements that
contributed to the decorative program. The loss of textiles as a contributing factor has
been noted. In addition, the actual thrones on which the ruler sat could well have been
decorated with motifs appropriate to the ideology of rule. This is certainly the case with
the two decorated stone throne bases that have been preserved: one of Shalmaneser III
from Fort Shalmaneser (see P. Hulin, “The Inscriptions on the Carved Throne-Base of
Shalmaneser III,” Iraq 25 [1963]: 48–69, and the recent analysis by Michelle I. Marcus,
“Geography as an Organizing Principle in the Imperial Art of Shalmaneser III,” Iraq
49 [1987]: 77–90), the other of Sargon II from Khorsabad (discussed in Winter, “Royal
Rhetoric,” 19 and fig. 17). Neither of the throne bases adds a rhetorical element not
also preserved in wall decoration, although the emphasis on state diplomacy on the
Shalmaneser III base offers a different nuance by virtue of its primacy.
performance, virile strength, military victory, and statecraft.69 Russell

has recently demonstrated that in the later reign of Sennacherib, this
lexicon was expanded to include new themes related to civic construc-
tion, which convey more explicit messages pertaining to maintenance
of the “center,” i.e., the capital, in contrast to earlier formulations that
emphasized the maintenance of the state through territorial acquisition
and the establishment of boundaries.70
These Sennacherib reliefs are situated in the large court giving onto
his principal throne room. As noted for Mari, the courtyard wall that
doubles as throne-room façade is especially adapted to proclamations of
rule, and in the Neo-Assyrian palaces of Assurnasirpal II and Sargon,
at least (e.g., fig. 16), we find particular attention given on that wall
to processions of foreign delegations bearing tribute to the ruler—a
topos conveying the ruler’s ability to command both wealth and stately
attention.71
69
Winter, “Royal Rhetoric” (text in Grayson, Assyrian Royal Inscriptions, 2:paras.
650–53). (For a different reading of this throne room, see Luc Bachelot, “La fonction
politique des reliefs néo-assyriens,” in Marchands, diplomates et empereurs: Etudes en honneur
de Paul Garelli, ed. D. Charpin and F. Joannès [ Paris 1991], 109–28.)
70
Russell, “Bulls for the Palace,” 520–39.
71
In the palace of Assurnasirpal at Nimrud, there is clear indication by dress that
the tribute bearers are foreigners from the West. While just one national contingent is
represented, at least in what is preserved of the Court D façade, Assurnasirpal mentions
that envoys from twelve countries were invited to his inauguration festivities in the text
of his banquet stele, which was set up in that same courtyard. It is to be expected that
all would have brought gifts, and I see the throne-room façade as a kind of presag-
ing of the delegations anticipated at the inauguration. Sargon II, who restored the
Northwest Palace at Nimrud, similarly represents a tribute-bearing foreign delegation
on the façade of his throne room (cf. Heinrich, Die Paläste, fig. 93 = the west wall of
court VIII, leading into the throne room, room VII, our fig. 16). An indication of the
sort of high-level gifts between rulers appropriate to the opening of a new palace is
preserved in an exchange of letters between the king of Babylon and the king of Egypt
in the later second millennium, in which the Babylonian king declares that he has just
built a new “house,” is planning a “house-opening,” and invites the Egyptian king
to attend. The Egyptian king responds by sending luxurious furnishings for the new
palace, including ebony furniture overlaid with ivory and gold (cf. William L. Moran,
The Amarna Letters [Baltimore and London: The Johns Hopkins University Press, 1992],
EA 3 and EA 5), very much on a par with the chandelier sent by Queen Victoria to
the Ottoman court on the completion of the Dolmabahçe Palace in the nineteenth
century (for which reference I am indebted to Jülide Aker). That these tribute scenes
have a broader valence than merely diplomatic gift exchange has been suggested in a
most interesting paper by Jack Cheng, in which he argues that the representation of
tribute is a means of proclaiming the stability of the economic base of the state through
the effective accumulation of wealth, and as such constitutes an important trope in the
iconography of the successful ruler (cf. “Tribute Scenes in the Program of Legitimation
by Sargon II of Assyria at Khorsabad,” unpublished paper, 1992).
Through verisimilitude in landscape elements and dress, military

narratives are made to be more than generic victories; they refer to
actual campaigns of the king’s reign. The representation of at least
a half-dozen separate campaigns in the reliefs of Assurnasirpal II’s
throne room confirms the king’s account in his Banquet Stele of how
he depicted on his palace walls the “glory of my heroism across high-
lands, plains and seas.” In later Assyrian reigns (as, for example, the
Lachish siege of Sennacherib, fig. 14), these military scenes proliferated
throughout the entire palace. By concentrating them in the throne room
during the reign of Assurnasirpal and by placing the throne room itself
at the center of the palace, the ninth-century king conveyed the fun-
damental message that, as the throne room is the heart of the palace,
so the palace is the heart of the state.
The use of extended decorative programs as vehicles for the articula-
tion of ideology is not unusual in the history of royal palaces, and many
art historical studies have attempted to reconstruct those programs,
along with their ideological underpinnings.72 In the palaces of the
ancient Near East, the “official” public statements about the ruler and
the state as they appear in the decorative program serve to underscore
the institutional nature of the palace as part of the larger state appara-
tus.73 To the extent that “the palace” can serve as metonym for the ruler
(as “the White House” does for the American president), and thereby
for the state, the palace is the source of ideology; and to the extent that
the palace is the physical manifestation of a program of royal rhetoric,
it is also the vehicle for that ideology. Thus, we return to issues of func-
tion, and the role of the palace-qua-institution, with which we began.
Clearly, the ruler and a large extended household resided in the pal-
ace and had to be accommodated. On the practical level, this required
private apartments, cooking facilities, and stores. Evidence exists that the
king’s mess included large parties of his sons and officials and that alloca-
tions were made for the rest of the palace on a regular basis.74 In addi-
tion, periodic festivities must have been organized, which would have
necessitated the banqueting of very large numbers of individuals, as on
72
For example, Jonathan Brown and W. H. Elliott, on the palace of Philip IV of
Spain, A Palace for a King: The ‘Buen Retiro’ and the Court of Philip IV (New Haven and
London: Yale University Press, 1986).
73
As defined and discussed by Louis Althusser, “Ideology and Ideological State
Apparatuses,” in Lenin and Philosophy and Other Essays, trans. B. Brewster (London: New
Left Books, 1971), 131–49 (originally published in French, 1966).
74
Kinnier-Wilson, Nimrud Wine Lists, 32 ff.; Postgate, Early Mesopotamia, 145.
the occasion of the inauguration of Assurnasirpal II’s new capital at

Nimrud, when the king records he fed some 70,000 people for ten days.
The size of the Assyrian throne room (ca. 10 × 45 meters for both
Assurnasirpal II and Sargon II) and its decoration argue for its nature as
a public reception suite in which the ruler would give audiences, although
whether on a regular basis or occasionally is uncertain. Visual evidence
that the ruler at least received selected members of the court and highly
placed officials is preserved in scenes carved on cylinder seals, which
show individuals presented before the seated ruler—a tradition that must
have been not unlike the Mughal darbār illustrated in several miniatures
of the seventeenth-century court of Jahangir and Shah Jahan.75 Scenes
on reliefs showing Assurnasirpal II with cup or bowl in hand may attest
to the king’s judicial function, in keeping with a long-standing iconog-
raphy of the ruler rendering just decisions in Mesopotamia.76 Whether
the ruler exercised legal office inside the palace (perhaps in the throne
room) or outside is not certain for this period, however. On the basis of
the Solomonic reference (see above, with regard to I Kings 7:7) and a
Sumerian literary text regarding the legendary king Lugalbanda, who
“takes [or exercises] office in the outer courtyard, in ‘The Gate that
Brings in Myriads’,”77 one is encouraged to seek possible evidence for
similar use of the palace courtyard and gate in the Assyrian period.
It must also be considered that the throne room could have functioned
as a venue for legal hearings.
Tribute scenes on courtyard façades leading into the throne rooms
at Nimrud and Khorsabad (e.g., fig. 16) both illustrate and assert state
reception of foreign delegations. That the palaces served as the reposi-
tories of such gifts, along with the booty seized in foreign campaigns,
we know from Assurnasirpal’s repeated assertion that he brought pre-
cious metals and other rich booty to the palace, and also from Sargon’s
statement that he restored the Northwest Palace in order to place in
75
See discussion in Irene J. Winter, ‘The King and the Cup: Iconography of the
Royal Presentation Scene on Ur III Seals,” in From Image to Insight: Studies in Honor of
Edith Porada, ed. M. Kelly-Buccellati (Malibu: Undena Press, 1986), 263, and Vishakha
N. Desai, Life at Court: Art for India’s Rulers, 16th–19th Centuries (Boston: Museum of Fine
Arts, 1985), figs. 12 and 13, for the Mughal paintings.
76
Winter, “The King and the Cup,” 253–68; Postgate, Early Mesopotamia, 150.
77
Thorkild Jacobsen, “Lugalbanda and Ninsun,” Journal of Cunieform Studies 41
(1989): 69–86, esp. lines 58–59, pp. 71 and 73.
it booty from his victory over Carchemish in 717.78 Not unlike later
palaces, from Fatehpur Sikri to Versailles, the display of valuable goods
and elaborate appointments served as signifiers of the success of the
ruler, and hence of the state.
Tribute scenes, overall decorative program, and display all attest not
only to specific events, but also to the very fact that the palace was the
site where statecraft was conducted. Texts from Nimrud further docu-
ment that the extended household supported in the palaces included
ministers, administrative officials, and scribes, whose job it was to run
not only the palace but the state.79 Assurbanipal’s famous “library” at
Nineveh may reflect the special case of an unusually literate ruler;
however, the archives of other palaces, such as Mari, strongly argue
for the palace as repository of central state records as well.
The iconography of room 132 at Mari and the suite around room
G at Nimrud, as was mentioned earlier, may suggest that some rooms
in the palace were devoted to ritual activities, and I suspect that if any
new palaces were to be investigated with modern excavation methods,
we would find considerably more evidence to support such a contention.
The importance of ritual performance and court ceremonial in other
times and places argues strongly for the presence of such spaces within
the palace.80 If we were to include procession as part of ceremonial
display, then the fact that the processional route from the Ishtar Gate
to the temple of Marduk at Babylon passes along the east wall of
Nebuchadnezzar’s palace (see fig. 8) could imply an active role for the
palace and/or the king in the procession.81
Finally, I would argue that a significant component of function is
“affect”: the impact of the building upon subjects of the state and upon
foreigners. Lackenbacher has studied Assyrian narratives of royal build-
ing activities, with particular focus on palace construction.82 In a number
of instances, rulers take credit for innovations in technique (such as
78
Grayson, Assyrian Royal Inscriptions, 2:para. 653 (= Assurnasirpal II, Standard
Inscription); D. D. Luckenbill, Ancient Records of Assyria and Babylonia 2 (Chicago, 1927):
para. 138 (= Sargon II).
79
Kinnier-Wilson, Nimrud Wine Lists, 95ff.; Postgate, Early Mesopotamia, 141.
80
Cf. Gülru Necipoğlu, Architecture, Ceremonial and Power: The Topkapi Palace in the
Fifteenth and Sixteenth Centuries (Cambridge, Mass.: MIT Press, 1991), 120–22, 140–41.
Ars Orientalis 23 (1993).
81
On the processional route and recent archaeological work in Babylon, see H. Trenk-
walder-Piesl, “The Procession-Street of Marduk in Babylon,” Sumer 41 (1985): 36–40.
82
Le roi bâtisseur, esp. 73–81; see also Lackenbacher, Le Palais sans Rival: Les récits de
construction en Assyrie (Paris: Editions la Découverte, 1990), passim.
Sennacherib’s reference to bronze casting for column bases) or include

statements about how skillfully the palaces have been constructed.83 In
addition, rulers express personal gratification concerning their palaces
(e.g., “palace of my joy” and “my royal residence that I love”).84
Perhaps most important of all, we find references to intended impact.
Assurnasirpal refers to his new palace as “fitting and splendid,” “palace
of all the wisdom of Kalhu” (Nimrud).85 The king is clearly celebratory;
but at the same time he proclaims the palace as concentrating within
itself all that is of value in the capital. Nearly two hundred years later,
Sennacherib calls his new residence “Palace without a Rival.”86 He says
of the limestone reliefs, “I made them objects of astonishment”; of his
colossi, “I made them a wonder to behold”; and of the palace as a
whole, “To the astonishment of all peoples I raised aloft its head.”87
The importance of this phrase, “to [or for] the astonishment of all
peoples” (ana tabrāt kiššat nišē ) cannot be too strongly emphasized. It is
an exact translation of a Sumerian formula of reference to impressive
building, largely applied to temples in the earlier periods (u6-di un šar).
In Neo-Assyrian usage, both temples and palaces are so described,
but it is especially characteristic of texts referring to new palace con-
structions.88 It would be interesting to survey extant attestations to see
83
See, for example, the Chicago Assyrian Dictionary, vol. ‘N’, p. 187, entry under nakliš:
Sennacherib refers to an earlier palace, whose construction had not been artistic/skill-
ful enough, ša ekalli . . . epištaš la naklatma; his son Esarhaddon declares that he “had [a
palace] built skillfully as his royal seat and for the pleasure of his lordship,” ana mūšab
šarrūtiya u multa ūti bēlūtija nakliš ušēpišma. This is clearly a continuation of Middle Assyrian
rhetorical practice, as when Tiglath Pileser I states that he built his palace at Assur with
understanding and skill, decorating it in a splendid fashion, and making it fitting as a
royal residence (cf. Grayson, Royal Inscriptions of Mesopotamia, 2:TP I.4.65–66, 77–89).
In some of the royal correspondence from the later reigns of the Neo-Assyrian period,
there is also evidence that rulers were not only kept informed but engaged through
correspondence in decisions regarding construction and decoration. See, for example,
Simo Parpola, State Archives of Assyria 1 (Helsinki: Helsinki University Press, 1987): nos.
60, 61, 110, 133, and State Archives of Assyria 5 (Helsinki: Helsinki University Press, 1990):
nos. 15, 56, 282, 293, concerning Sargon II and the work on Khorsabad.
84
Cited in Lackenbacher, Le roi bâtisseur, 74.
85
Grayson, Assyrian Royal Inscriptions, 2:paras. 680 and 653, translates it as “in splen-
did fashion,” and para. 682, the “palace full of wisdom.” I find more persuasive the
translations of Ann Shafer in “Monuments to the Center of Empire: Ashurnasirpal
II’s ‘Banquet Stela’ in Context,” unpublished paper, 1992, in the first instance because
it is more grammatically accurate and in the second because it implies an “epithet”
applied to the palace, as opposed to mere description.
86
Literally, “Palace for which there is no second [or, no equal].”
87
Luckenbill, Ancient Records of Assyria and Babylonia, 2:paras. 367, 389, 394.
88
See, for example, Grayson, Royal Inscriptions of Mesopotamia, 2:296, where Assur-
nasirpal uses a comparable phrase for his Sharrat-Nip i temple at Nimrud.
whether it is possible to determine a time when “astonishment” was

accorded to palaces as well as temples, and whether this correlates
with any significant developments in the Mesopotamian state and in
the institution of kingship. Garelli has discussed some of the attributes
of royal palaces intended to astonish. He noted that standard words
for “beautiful” or “well-built” were not used for royal buildings; rather,
one finds a vocabulary focusing on qualities also applied to the person
of the king. By Garelli’s account, because the palace was the work of
the king, it was possessed of the same “splendor” and “majesty” as the
king himself.89 However, it should be remembered that it is the ruler’s
own voice in the texts that articulates the qualities of the palace, so that
the king actually imbues the building—or asserts that his buildings are
imbued—with those very qualities which he would also have ascribed
to himself. The palace is thus set up as a mirror of the king. It is a
physical manifestation of the ruler’s power and ability to build; and
at the same time, by having built so impressively, the ruler has further
demonstrated his power and ability to command resources, induce
astonishment, and create a fitting seat of government—in short, to
rule. The rhetorical function of the palace, as exemplified through its
affect, is, I would argue, as essential as its residential, administrative,
productive, and ceremonial functions.
Throughout the preceding survey, I have tried to demonstrate that

morphology, decorative program, and function are not independent
variables. Rather, room type, organization of space, individual decora-
tive motives, and overall decorative scheme are fundamentally linked
to function. Given the nature of the archaeological and textual record,
in any scholarly study of the palace we are limited to the expression
of the royal voice, which privileges rhetoric and intention over actual
practice. Obviously, as is all too well known in modern times, buildings
can be poorly designed for anticipated functions. Equally, when buildings
are secondarily occupied, or when historical events precipitate change,
then their original form and decoration can either constrain function or
have little relationship to new usage. Nevertheless, recovery of the ideal
89
Paul Garelli, “La conception de la beauté en Assyrie,” in Lingering over Words:
Studies in Ancient Near Eastern Literature in Honor of William L. Moran, Tzvi Abusch, John
Huehnergard, and Piotr Steinkeller, eds. (Atlanta, Ga.: The Scholar’s Press, 1990), 176
and passim, with respect to terms like tašri tu, “splendor, majesty, sumptuousness,” from
the verb šarā u, “to render majestic, sumptuous, imposing.”
schema and the associated originating rhetoric is a necessary first step

toward any critique of the fit between intention and actual practice.
In the Mesopotamian schema, and apparently in the Hittite one as
well,90 the palace was conceived as incorporating a bundle of activities
and functions: residential, political, administrative, industrial, ritual,
ceremonial, and affective. Storage and display of surplus and luxury
goods served as extensions of elaborate decorative programs that articu-
lated state ideology, and spaces were designed to meet the functional
needs of the palace as an institution.91 Limited comparisons with other
palace types have suggested the importance of regional diversity; but at
the same time, it can be demonstrated that regionally specific building
forms and decorative practices could be transmitted across regional
boundaries under certain political or cultural conditions. There are
many ancient Near Eastern palaces not mentioned in this brief survey,
and many palaces that have been mentioned have received less than
adequate description or analysis. Because I have been sketching with a
broad brush, there has been a tendency to emphasize similarities in the
Mesopotamian sequence across some two thousand years. In many ways
this is not unjustified. From early royal hymns in Sumerian to later royal
inscriptions in Assyrian Akkadian, indications are that the palace was
construed as the seat of kingship, not merely as the residence of the king.
Nevertheless, micro and macro shifts in form and decoration need to
be studied more closely with respect to the many political changes in
state development over this long period. What I have tried to stress
throughout is the role of the palace within the context of the state and
the rhetorical function of the palace as embodiment of the state.
If there turn out to be significant continuities in building materials
and techniques, decorative programs (especially for non-figural motifs
of symbolic value), and ceremonial/administrative functions from the
pre-Islamic to Islamic periods in the Near East, beyond the few I have
90
Güterbock, “Hittite Palace.”
91
Clearly, more work needs to be done in systematic study of the shifting role of
the palace as institution over time in the ancient Near East, as seen through changes in
decorative program, room, and spatial distribution, sight lines, and avenues of approach.
What is interesting is that the evidence provided by Mari and the Neo-Assyrian palaces
suggests the palace decorative scheme consisted not merely of a single, generalized
message appropriate to the “seat of kingship,” but of a series of accumulated messages,
communicated in individual rooms and areas, that were specifically tailored to the
function of associated spaces. One must therefore “read” the sum of those messages
to get closer to the overall rhetorical program of the palace.
noted here, it will not surprise me at all; nor will I be surprised if distinct
regional traditions within Islam actually reflect recognizable subdivi-
sions in earlier periods as well. At the same time, one will want to take
care to distinguish features apparently similar over time that are merely
the consequence of relatively limited ways of representing/organizing
authority (what evolutionary biologists call spurious homologies) from
features that truly represent continuity in underlying concepts and
traditions. When continuity cannot be demonstrated, it then becomes
necessary to account for the differences as artifacts of differing historical
practice. I confess that a significant part of my mission in the forego-
ing survey has been to convince historians of Islamic architecture (and
culture) that the pre-Sasanian pre-Islamic world should be included in
their scholarly purview. Yet it has also become apparent that, however
much we acknowledge the significant divide between before and after
the introduction of Islam, students of the ancient Near East have
much to learn from considering the more complete historical record
of Islamic practice—by which I mean not only building practice, but
also cultural and political practice.
In the end, what is so clear as to be obvious, but still needs to be
stated, is that any study of the palace, whatever the historical period,
is fundamentally linked to the study of concepts of authority and rule.
To understand the palace, one must see it as the locus of a particular
practice of governance. Furthermore, when continuities of morphol-
ogy and/or decoration occur across spatio-temporal boundaries, one
cannot immediately assume continuity in meaning; whenever possible,
it is necessary to establish associative significance independently. Con-
versely, it is possible that differences in morphology and/or decoration
nevertheless represent quite similar social and political systems.
For the ancient Near East, the play in the subtitle of this survey was
a conscious one: the palace is both a physical and a mental construct,
both built and construed. It is at once the concentrated center of rule,
“the seat of kingship,” and also the concrete expression of rule, “worthy
of being” the seat of kingship, “for the astonishment of all peoples,”
“a wonder to behold.”
Acknowledgments
My sincere thanks to Jülide Aker, Jack Cheng, Marian Feldman, and Ann Shafer for
permission to cite unpublished works produced as seminar papers at Harvard University,
and to Jülide Aker and Margaret Ševčenko for tough readings of an earlier draft of
this paper. My thanks also to Barbara N. Porter for her generosity in providing a copy
of her paper on the date palm for use by me and my students prior to its publication.
Conversations with many colleagues will have found their way into a number of aspects
of this broad survey. I am grateful to all.
Figure 1. Uruk/Warka. Plan of Eanna Complex, level IV, ca. 3200 b.c.,
including “Palace” E (Building 11). From Postgate, Early Mesopotamia,
fig. 6:4.
Figure 2. Tell Braq. Plan of Naram-Sin Palace, ca. 2300 b.c. From Heinrich,
Die Paläste, fig. 22.
Figure 3. Mari. Plan of Palace, early 2nd millennium b.c. After Margueron
et al., M.A.R.I. 6, fig. 1.
Figure 4. Mari. “Investiture” painting, Palace, court 106, detail. From

Parrot, Mission archéologique de Mari II, pl. XI.
Figure 5. Khorsabad. Plan of city walls with citadel, built by Sargon II, 8th
century b.c. From Frankfort, Art and Architecture, fig. 165.
Figure 6. Nimrud. Colossal gateway

figure from entry ‘a’ into throne room B
of Northwest Palace of Assurnasirpal II.
Photo: courtesy Trustees of the British
Museum.
“seat
of kingship”/“a wonder to behold”
Figure 7. Nimrud. Plan of Northwest Palace of Assurnasirpal II, 9th century b.c. Composite from several
sources, including Mallowan, Nimrud and Its Remains, 165, and Heinrich, Die Paläste, fig. 55.
369
370
chapter twenty-nine
Figure 8. Babylon. Plan of Southern Palace of Nabopolassar and Nebuchadnezzar II, 6th century b.c.
From Heinrich, Die Paläste, fig. 122.
“seat
of kingship”/“a wonder to behold”
Figure 9. Khorsabad. Isometric reconstruction of citadel and Palace of Sargon II, 8th century b.c. From Levine,
Bulletin of the Society for Mesopotamian Studies, fig. 2, drawing by Rob Mason.
371
Figure 10. Khorsabad. Plan of Palace of Sargon II, 8th century b.c. From
Loud, Khorsabad II, after Place.
Figure 11. Zincirli. Plan of Hilani III, 8th century b.c. After Naumann,
Architektur Kleinasiens, fig. 448.
Figure 12. Khorsabad. Relief showing pavilion in wooded area, from room 7,
Palace of Sargon II. Photo: courtesy Oriental Institute, University of
Chicago.
Figure 13. Nimrud. Watercolor reconstruction of the throne room of

Assurasirpal II, Room B, Northwest Palace. From Layard, Monuments of
Nineveh, 1849, pl. 2.
Figure 14. Nineveh. Relief showing ruler and epigraphic text, in aftermath of
the siege of Lachish, from room 36 of the Southwest Palace of Sennacherib,
early 7th century b.c. Photo: courtesy Trustees of the British Museum.
Figure 15. Nimrud. Relief of king flanking palmette tree, from Throneroom
B, Northwest Palace of Assurnasirpal II. Photo: courtesy Trustees of the
British Museum.
Figure 16. Khorsabad. Reconstruction of the throne room façade, court VIII,
Palace of Sargon II. From Heinrich, Die Paläste, fig. 93.
CHAPTER THIRTY
OPENING THE EYES AND OPENING THE MOUTH:

THE UTILITY OF COMPARING IMAGES IN WORSHIP IN
INDIA AND THE ANCIENT NEAR EAST
For me as a scholar of the ancient Near East, comparison of archaeo-

logical and textual evidence from Mesopotamia with the living prac-
tice of India—particularly with respect to the rituals and devotional
practices related to images in worship in Hindu temples—has been
both illuminating and enabling. The particular venue of the University
of Pennsylvania Center for the Advanced Study of India’s thematic
seminar on “Pilgrimage, Art, and Ritual” occasioned a more reflective
articulation of the utility of comparing two cultural traditions not only
2000 miles apart but also separated by some 4000 years; however it
should be noted at the outset that such an endeavor is not without its
skeptics. I therefore welcome this opportunity to bring the issue before
a primarily Indological audience.
The enterprise began through an interest in the sculpture of Gudea,
ruler of the city-state of Lagash toward the end of the third millen-
nium b.c.e. Some 16 free-standing images, both seated and standing,
were discovered in excavation at the satellite town of Girsu (modern
Tello) early in the 20th century (Parrot 1948), and are presently in the
Louvre.1 From discovery to the present, various museum installations
have grouped all of the statues in a single gallery, and as an assemblage
they create a powerful impression in the modern viewer. Yet inscriptions
on many of the images tell us that each was dedicated to an individual
deity in the Mesopotamian pantheon, and would have been placed in
the temple or shrine of that deity (Steible 1988; Winter 1992; Edzard
* This article originally appeared as “Opening the Eyes and Opening the Mouth:
The Utility of Comparing Images in Worship in India and the Ancient Near East,” in
Ethnography and Personhood: Notes from the Field, M. W. Meister, ed., Jaipur and New Delhi:
Rawat Publications, 2000, pp. 129–162.
1
A number of unprovenienced pieces subsequently have become known, presumably
the result of clandestine excavation at Tello. Most of these are in other public collec-
tions, such as the British Museum and the Metropolitan Museum in New York.
378 chapter thirty
1997). Therefore it is important to keep in mind when viewing these

images that the museum is not the natural home of the sculptures; nor
would they ever have been seen as a group in their original context.
Even when a curator makes a major effort to evoke the “original
experience” of an image through exhibition techniques like spacing,
color, and lighting—as at the Philadelphia Museum of Art for the
exhibition Manifestations of Shiva (Kramrisch 1981, and review, Meister
1981)—it is still a far cry from the actual context of experience and
use. What the processional bronzes of the Shiva exhibition and the
various Gudeas have in common is that both classes of image have
been conceived to be in worship, and were treated as living entities in
their temples: fed, dressed and ornamented; subject to ritual attention
throughout the diurnal/festival cycle.
Western inquiry into such images, Indian and Mesopotamian alike,
has, until the present generation of scholarship, been largely based upon
formal analysis: technique of casting or carving, physical properties,
style, attribution, followed by the consequences of the above, i.e., dating,
sequence, historical patronage. In my field, at least, this has been due
in part to the fact that scholarship has tended to be divided between
those who worked on material culture and those whose research was
primarily philological, with art historians and archaeologists receiving
very little or no training in ancient languages. Once one puts texts
together with the images, both those inscribed directly upon the objects
and those written on independent clay tablets or cylinders, there is
no doubt of the enlivened status and ritual attention offered to the
Mesopotamian images.2
The familiar analogy between Gudea with his clasped hands and
European Medieval or Renaissance donor portraits, which has been
easy to suggest due to the formal similarity in pious posture and earnest
gaze (cf. Braun-Holzinger 1977), must therefore be considered as only
partially accurate. In fact, although some of the Gudea images were
indeed placed before a deity, others were installed in their own chapels,
and all were themselves objects of devotional practice; thus, a better
analogy in western practice would be to Roman Catholic statues of
2
In one of the more complete accounts, Esarhaddon, ruler of Assyria 680–669
b.c.e., records the installation of (the image of ) the god Assur in his temple, the place-
ment of other (images of ) deities and various animal attributes around him, and the
establishment of regular cult offerings (Borger 1967: Assur A, ¶2 vi 28–vii 19, pp. 5–6).
See also Winter 1992; Walker and Dick 1998.
opening the eyes and opening the mouth 379
saints and the Virgin that can similarly be represented in reverence to

Christ and/or be set for veneration in their own chapels.
While there is much one could do in pursuing this equation, the
theological position that argues against what is called “idolatry” in the
Hebrew Bible (e.g., Leviticus 19:4, “Turn ye not unto idols, nor make
yourselves molten gods”; see, Dohmen 1987) precludes our taking the
analogy one step further to an examination of the image as manifesta-
tion—as presentation, rather than mere representation [re-presentation] of the
referent. The ascription of living manifestation makes the comparison
of Mesopotamian images with those of Hindu India far more apt
and likely to be more rewarding in terms of understanding the Meso-
potamian case. And yet, in the present era of relativist claims for the
uniqueness of each cultural system and practice, to argue for the utility
of such comparison is not simple.
Over the course of the past ten years, since I have been studying
the attention given to and the theological grounding for Hindu images
in worship, it has proven necessary to develop an explicit argument in
support of that utility. The argument has its origins in the practice of
cross-cultural comparison within anthropology (on which, see Eggan
1954; Northrup and Livingston 1964; Ember 1991; Mace and Pagel
1994; Hunt, in preparation; plus many others), and has been signifi-
cantly developed in the sub-field of “ethnoarchaeology” as practiced
by anthropologically-trained archaeologists of the late 1960’s to the
present. The term itself was introduced by K. C. Chang (1967) as he
sought to elaborate on how a present/living tradition could be used
to amplify data from excavation, and even explain the archaeological
record in the absence of living informants. The potential has been
advanced by a number of archaeologists, many of whom engaged in
specific fieldwork in contemporary cultures, setting their research designs
with a view toward filling gaps in our knowledge of the past by analogy
with the present (e.g., for the Near East, Matson 1974; Ochsenschlager
1974, 1995; Watson 1978; Kramer 1979, 1982; Stone 1981; for India,
Jayaswal 1989, 1990; Kamp 1993; Vidale, Kenoyer, and Bhan 1993).
Virtually all of the archaeologists who have done such ethnoar-
chaeological studies have confined their investigations to the same geo-
graphical area as the ancient societies they are studying. This strategy
is essentially based upon the premise that consistency in (at least some)
environmental and ecological factors, along with (occasionally some)
historical and linguistic continuities, would increase the likelihood of
continuity in practice and meaning. Such consistencies notwithstanding,
380 chapter thirty
the dangers implicit in even this limited reach have been stressed in
recent critical literature, and a few counter-cases have been adduced as
cautions (Heider 1967; Wylie 1982, 1985; Galloway 1992; Lamberg-
Karlovsky 1989). How much more dangerous, therefore, to work across
both time and space, and across very different ecological, linguistic,
historical, and cultural boundaries!
Support for such an enterprise cannot be gathered from any premise
of continuity, which would require the demonstration of either histori-
cal contact or some type of contextual similarity (e.g., environmental).
Instead, one must have recourse to an argument based upon the
demonstration of systemic parallels between the two members of the
comparison. Only following upon such a demonstration can one then
proceed to an analysis, not assumption, of the possibility that particu-
lar aspects of the societies under discussion may be indeed analogous
and/or that known aspects of practice/behavior/belief in one tradi-
tion are such as to suggest that, in the absence of data from the other,
the former may serve by analogy to explain or amplify understanding
in the latter. This approach has been discussed by Bloch (1970), who
argued for comparison between entities that do not belong to the same
historical stream, and can be distant in time and place, in which case, a
“typological approach” is advocated. Bloch’s argument is related to that
of Hunt (in preparation), who speaks to the validity of cross-cultural
comparison through the use of clearly defined dimensions and variables.
Similarly, Wylie insists that “analogical inference is not radically faulty
or categorically misleading” (1985; emphasis mine). She suggests that
arguments by analogy are as subject to systematic testing as any other
arguments, and asserts that such explorations are needed, in order to
attempt to bring the “unfamiliar and otherwise inaccessible aspects of
the past into view” (ibid.: 107).
In the specific instance, I suggest that the sculptural practices and
temple procedures I wish to study should yield well to amplification
through analogy because they were developed and used by societies
that demonstrate basic systemic similarity at fundamental levels. The
systemic parallels I would cite to support the Mesopotamia-Hindu India
comparison are as follows:
1. both developed within states,

2. manifesting a stratified social hierarchy,
3. organized within an agrarian-based economy,
4. characterized by large urban centers,
5. in which there has been a significant investment in large-scale religious

architecture,
6. devoted to cultic practice,
7. related to a polytheistic religious system,
8. for the management and maintenance of which there resides a per-
manent, specialist priesthood [often characterized by hereditary office],
and
9. which in turn presides over and directly engages with images in
worship.
Over the past two decades, there has been a growing literature on
historical connections between the Indian subcontinent and the ancient
Near East from early times through the Roman period (see Begley
and De Puma 1991; Boucharlat 1987–8; Inizan 1993; Pingree 1987,
1989, 1992; Reade 1979, 1996; Sluszkiewicz 1980; Smith 1989; Vickers
1994; Finnish Oriental Society 1993). These studies, however, stress the
material (archaeological) evidence for contact, whereas the question of
systemic parallels between ancient Mesopotamia and historical India,
particularly based upon the dimensions noted above, has not been
explored;3 and it is precisely the demonstrable systemic similarities
between ancient Mesopotamian temples and contemporary Hindu
practice in India that, I would argue, provide the basis for fruitful
cross-cultural comparison.
By this distinction, Hindu India presents a far better case for com-
parison if one is seeking to understand the “lost” practices of a dead
civilization by analogy with a living tradition than would a non-urban-
ized, non-stratified, non-polytheistic society, whether or not a specific
element, such as ritual attention to three-dimensional images, could be
demonstrated to be held in common. I would go even further in sug-
gesting that religious practices and attitudes, behaviors and beliefs, are
more likely to serve as analogues for one another within two polytheistic
societies than if the members of the comparison start from very different
3
I suggest that, despite recent industrialization in India, the temple-household
system currently in place still reflects its origins in an agrarian-based economy, one
which seems to have developed within the early states of Medieval India (9th–14th
century a.d.). From preserved texts on cultic practice in India it is also clear that
many temple practices are marked by considerable theological conservatism, hence
continuity with the past.
382 chapter thirty
theological attitudes toward deity, not to mention different theological

attitudes toward images.
Scholarly interest in aspects of Hindu ritual practice related to tem-
ples and their images has gained ground in recent years (for example,
Blurton 1993; Bühnemann 1988; Dagens 1985; Goudrian 1970; Hein
1978; Waghorne 1992 and 1985; Welbon 1984; Younger 1986), and
is considerably more developed than the even more recent inquiry
into comparable Mesopotamian practice (on which, Oppenheim 1964;
Jacobsen 1987; Hallo 1988; Winter 1992; Matsushima 1993; Berlejung
1997; Walker and Dick 1998). In seeking to amplify aspects of devo-
tional practice and attitudes characteristic of polytheistic Mesopotamia,
an informed investigation of Hindu practice is rewarding on more
than one level. First, I would suggest that it is possible to give life to
highly telescoped references or equally schematized images through
comparison with the cultic treatment of and relationship to images.
Second, I argue that it has been possible to provide at least a tentative
interpretation of some Mesopotamian practices by direct application of
analogy. And third, it is possible to stimulate the search for evidence of
practice in hitherto unexamined data within the Mesopotamian record
upon observation of Hindu practice.
The animate status of the Mesopotamian image—god or enshrined
ruler—is attested by textual evidence for rituals of the “opening of the
mouth” and the consequent opening of the eyes plus the rest of the
sensorium that actually brought the image to life (Smith 1925; Walker
and Dick 1998). These texts refer to divine images, rather than royal,
and are late copies, dated to the first millennium b.c.e.; however they
are bilingual in both Sumerian and Akkadian, suggesting a tradition
that goes back into the third millennium, despite the fact that early
divine statues, most likely of wood overlain with precious metals, have
not been found. That the royal images of Gudea with which we began
were also subject to such animation is clearly implied by the fact that
every inscribed statue of Gudea uses the verb for “birthing” (Sumerian
tud), rather than the normal verb for the “making” of things (Sum.
dim2), in the assertion of the bringing into being of the image (see on
this, Winter 1992).
The Mesopotamian tradition of an initial “mouth opening ceremony”
is also documented in ancient Egypt—for royal statuary, and presum-
ably also for divine images (Otto 1960; Goyon 1972; Assmann 1992;
Milde 1994). It finds its closest parallel in Hindu and Buddhist practice
in the “opening of the eyes” of a sacred image (Sanskrit nētrōnmīlana),
well attested in text (cf. Tripathi 1978a; Gombrich 1966; Bentor 1992,
1996) and still in practice today at the consecration of a new image.
The fact that the ancient Near East and Egypt privilege the opening
of the mouth while South Asia has privileged the opening of the eyes
reminds us that one cannot expect to find exact correspondences in
every domain. However, that each tradition developed the ceremonial
means by which inert physical matter could be transformed into ani-
mate, sentient matter through the ritual infusion of “life” remains of
utmost importance.
The ritual animation of the Gudea figures is necessary so that they
may then play an active role, whether in the god’s shrine or in their
own chapels. The inscription contained on the well-known seated Statue
‘B’ contains the revealing passage in which the statue is commanded
to speak directly to the local deity, Ningirsu, on behalf of the living
ruler: “Image, to my lord [the god Ningirsu], speak!” (Edzard 1997:35,
col. vii, lines 24–25).
I have suggested elsewhere that the standing statues of Gudea
may be distinguished from the seated figures in that the former were
intended to be placed before the image of a presiding deity within the
cult shrine, while the latter were intended for their own chapels (Winter
1992:25–26). A key to this is that the seated position was the dominant
in Mesopotamia, so that subordinate figures stand before a seated one,
as on Gudea’s own seal (fig. 1), where the ruler is depicted being led
by an interceding deity into the presence of a seated figure of the god
of wisdom and sweet waters, Ea/Enki—recognized by the flowing
streams that issue from a vase held in his hand and from his shoulders
to encircle the throne. This spatial convention is consistent with that
used in earlier periods as well, as on a seal of the Early Dynastic period
(fig. 9), ca. 2600 b.c.e., where cult personnel approach a seated deity
(probably the moon-god Nanna/Su’en). Another indicator is that the
great majority of inscribed standing Gudea statues carry their inscrip-
tions on the back, to be read/seen from behind when the visual and
spatial bond—as in Hindu temples, not to be interrupted—would have
been between (the image of ) the devotee/ruler and (the image of ) the
deity within the shrine.
When one attempts to translate this rendering of figures in profile
along a single ground line into three-dimensional space, the ruler would
stand on axis with the divine image, as is captured in many Hindu devo-
tional paintings, or by actual photographs of worshipers (fig. 2). Such
cross-cultural juxtapositions, while they do not tell us much more than
384 chapter thirty
could be deduced by projecting the “presentation scene” into real space,

nevertheless provide a window into actual experience in space. They thus
serve to augment the otherwise largely sterile reconstructions of sacred
architectural space drawn from archaeological plans—as, for example,
the reconstruction of the cella of the Middle Assyrian Ishtar Temple at
Assur of ca. 1350 b.c.e., in which physical space is indicated as devoid
of furnishings, with a barely identifiable cult image at the far end of
the room, and a single individual entering at the right margin. The
austerity of such reconstructed spaces matches in no way the sensory
stimuli of olfactory, visual, and auditory receptors characteristic of a
place of active worship, whether Roman Catholic, Eastern Orthodox,
Buddhist, or Hindu.
Before the image has been introduced into the sanctuary, however, its
own sensorium is awakened, following upon which, the image is bathed
in sacred water as if a new-born, and then ritually introduced into its
new home, the temple (Smith 1925; Jacobsen 1987; Winter 1992; Walker
and Dick 1998). It is here that one can begin to report upon the ritual
activities observed and recorded in Hindu temple practice.
I have been accorded privileged access to consecration ceremonies
of the awakening and installation of images on the occasion of a new
temple built for the Indian community in Ashland, Massachusetts, in
1989; and also to festival and daily worship over the course of several
years in the Rādhārama a Temple to Srī Krishna as Rādhārama a-jī,
“he who dallies with Rādhā,” in Vrindavan—the town on the Yamuna
river that marks the place where the god as cowherd sported with the
gopis on the Gangetic plains. The Laxmi temple in Ashland was built
according to traditional plans and procedures, overseen at all stages by
priests/artisans (sthapatis) brought from Mamallapuram in South India,
just as recorded by Parker (1992) for the Srirangam temple construction
in Tamil Nadu. The Krishna temple in Vrindavan belongs to the Sri
Caitanya Sampradaya, established under Mogul mandate in the 16th
century. The deity is manifest as a svayam-praka a svarūpa, a “self-revealed
form” in stone (śālagrāma) that appeared to a disciple of Caitanya,
Gopāla Bhatta, and since 1542 has been enshrined in the temple.
In Mesopotamia, the act of “bringing in” to the temple is marked
by particular ceremonial, the image then “seated” or enthroned in the
shrine. In the Indian case, once life and breath has been infused into
the image ( prā aprati hā), the image is properly installed as well, and
then provided with an auspicious name (nāmakara a). In the Mesopo-
tamian case, each inscribed statue bears on it its given “name”—for
the images we have been discussing, not “Gudea” per se, but a noun-
phrase-cum-prayer that constitutes both the identity of the statue and
the message the statue is to convey to the deity (e.g., on Gudea Statue
‘D’: “May Gudea, the one who built the temple, have a long life”;
Edzard 1997:39–40). That there be no ambiguity, the “name” is clearly
marked by the formulaic phrase: “for its name, it was given. . . .” (Sum.
mu-še3 mu-na-sa4).
Specific Mesopotamian rituals of installation include sacrifice/offering
and placement, as well as name-giving. From that point onward, the
image, now manifestation of god or individual, is considered resident in
his or her temple, a necessary component of which is the placement of
the image on a dais, enthroned, and often with footstool (as in figs. 1,
9). Clearly, much of the ritual process in Mesopotamia was considered
esoteric knowledge, not for the uninitiated, and so not consigned to
text. The (art) historian attempting to reconstruct such processes must
then rely upon inference adduced from the referential in text, the read-
ing of images, and the distribution of archaeological artifacts. While
important components of ritual knowledge within Hindu tradition are
similarly constrained, the textual record does preserve much lacking
from Mesopotamia (e.g., Dagens 1985; Goudrian 1970; Mittal 1966a &
b; Welbon 1984). For the Rādhārama a temple in particular, a printed
handbook for a parallel sampradāya, or tradition of Krishna-worship,
the Śrī Vallabhapu tiprakāśa (Ragunāthaji ivaji, 1986; henceforth, VPP)
provides a schematic plan of a related temple-form, and also records
a great deal of the daily and festival ritual calendar, as well as illustra-
tions of the temple inventory described in the text.4 The plan indicates
a central porticoed space for devotees, an intermediary and elevated
level on which priests caring for the deity can operate for certain ritual
procedures (the jagmohan), and the inner shrine (garbhag ha), residence
of the god, on axis with the open central space. This configuration
can be favorably compared with the standard Mesopotamian “straight-
axis” temple—for example, the temple to the deified ruler Šu-Sin of
Ur at Tell Asmar of ca. 2050 b.c.e., or to the goddess Ishtar Kititum
at Ischali of the later third/early second millennium b.c.e., where,
similarly, the deity would have been enshrined in a niched chamber
4
The ritual compendium for the Śrī Rādhārama a temple is the Haribhaktivilasa of
Śrī Gopala Bhatta—see S. K. De 1961 and Joshi 1959.
386 chapter thirty
on axis with the antechamber and buttressed entry from a large court
( Jacobsen 1989:79–81).
The inventory plates of the VPP illustrate “head ornaments” and
jewels that are directly parallel both to actual ornaments worn by
Krishna as Rādhārama a-jī in Vrindavan and to the textual invento-
ries preserved from Mesopotamian temples (for example, of Ishtar of
Lagaba—Leemans 1952). While we have known for a long time that
Mesopotamian deities accumulated large numbers of ornaments, gar-
ments and attribute elements such as weapons, symbols and standards
in their temples, to see the pictorial documentation in the VPP along
with the actual deployment of these elements in living practice (fig. 8)
has been enormously instructive with respect to the attention given
the living presence of the god on a day-to-day basis. And just as the
svayampraka a image of Śrī Rādhārama a-jī is adorned at particular
festivals with jewels paralleled by those on the VPP plates, so also the
ornaments worn by Mesopotamian deities depicted on clay plaques
of the Old Babylonian period, for example, are likely to have cor-
responded to actual ornaments in their respective temples—whether
Ishtar of Lagaba, or Ishtar Kittitum from Ischali (Leemans 1952:12).
Both the VPP plates and my own photographs of the deity at different
moments in the liturgical year chart the changes of dress and ornament
as Krishna is adorned daily, the changes keyed to season and event that
have associations with particular colors and types of garment. The
inventory also includes thrones or pedestals upon which the deity is
seated/placed, which coincides well with the Mesopotamian record of
thrones and podia of precious materials for various deities, presented
as votive offerings by a variety of rulers.
The annual festivals particular to the mythology and history of
each god are mentioned occasionally in the Mesopotamian record,
but without description of the specific ritual procedures followed.
One of the most instructive aspects of my time in the Rādhārama a
Temple has been as witness and recorder of the “bathing ritual” (Skt.
abhi ēka), known as an integral part of Hindu, Jain and Buddhist wor-
ship (see Klimberg-Salter 1982), and here held bi-annually: just after
the monsoon season at Janmā amī, the birth anniversary of Lord
Krishna, and again just before the next monsoon season at the full
moon known as Bodhapūr imā (figs. 2, 4, 6). This bathing ceremony
of the divine form may be seen as the Hindu analogue to the Meso-
potamian “mouth-washing” ceremony, which is attested for both royal
and divine images on a regular ritual calendar cycle.5 Both traditions,

therefore, share the periodic renewal and cleansing of the devotional
image (thought to have been recorded visually on the reverse of the
great Stele of Ur-Namma of the Third Dynasty of Ur, ca. 2110 b.c.e.
(fig. 3; Börker-Klähn 1975): refreshing and pleasurable, and also neces-
sary to remove any pollution the image may have accumulated during
the previous calendrical cycle, thereby re-establishing it as a pure and
sacred vehicle of material manifestation.
In the course of the Hindu ritual in Vrindavan, the image is brought
forward from the inner sanctuary to the jagmohan and placed upon a
special silver pedestal supported by four lion-paw legs, with lion-heads
at each corner (fig. 4). The god wears only a white linen wrap, and
over the course of the day, milk and other materials are poured in
abundance over the image by members of the priestly community and
some (male) members of the lay community as well.
Hundreds of gallons of milk are poured throughout the day, along
with other substances—clarified butter, honey, curds, saffron water—
offered in lesser quantities, mixed with special herbs and spices. The pri-
mary substances—clarified butter, curds, milk, honey—are also attested
as divine offerings in Mesopotamia. To pursue the analogy related to
temple offerings any further, one would have to do a systematic study
of the distinction between daily (food) offerings and festival provisions
intended for the god’s consumption; then compare these to “ritually
spent” materials, as in the bathing festival. For the present, two aspects
of the abhi ēka are worthy of mention. First, before participating in the
living ritual, my own mental image of the Mesopotamian offerings had
been one of neat dishes, bowls or containers, discretely disposed about
the image. While this does tend to be the case in India when food is
being given to the god for consumption, and may well have been the
case in ancient Mesopotamia, I was quite unprepared for the display of
abundance in dispensed liquids in the “bathing” ritual. On reflection,
this makes good sense if at least part of the theological rationale is to
cleanse absolutely as well as to treat the image luxuriantly, and it serves
to remind one that regularity and quantity are stressed for a reason
in Mesopotamian prescriptive texts. Second, this abundance seems
5
Mouth washing (Skt. ācamana) is also a prominent part of the daily ritual of bathing
the deity and of the twice-yearly abhi ēka at the Rādhārama a temple (Śrī Shrivatsa
Goswami, personal communication).
388 chapter thirty
particularly apt when one recognizes that the first of the bi-annual
bathing rituals marks the festival of the birth of Krishna ( Janamā amī)
within the monsoon season, and the second occurs at the full moon
preceding the anticipated beginning of the monsoon season in India,
once the harvest has been taken (Bodhapūr imā). The flow of milk in
the first, synonymous with nourishment in the early stages of “life” itself,
replicates and parallels the desired fertility of the agricultural year to
come through the production of the cows that constituted the primary
resource of the Gangetic plain, generously and lovingly lavished upon
a beloved deity; the second closes that agricultural cycle, just before
the rains that will determine the production of the coming agricultural
year. And, just as Gudea tells us in his Cylinder inscriptions that the god
Ningirsu promises (and then has delivered) a good harvest and agricul-
tural abundance (Sum. he2-gal2) as a consequence of the ruler’s pious
patronage of the temple (Cyl. A, col. xi:7–17; Cyl. B, col. xix:12–15,
Edzard 1997:76, 98), so also as one is lavish in provisioning Krishna
and demonstrating devotion to the deity, then comparable abundance
may be hoped for. In short, each of these traditions articulates and
performs its devotion within the context and the calendar of the local
mode of agrarian production.
Now, one of the primary criticisms to which this comparative work
in India has been subjected by scholars of both India and Mesopotamia
is that nothing has really been discovered that could not have been
known from looking at the Mesopotamian materials alone. But here I
would argue that this is only apparently the case because subsequent to
observation of some aspect of ritual practice in India it has been pos-
sible to go back to the Mesopotamian record and find corroboration
that had not been previously noticed, and/or because what had been
previously suggested quite tentatively can now by analogy be consider-
ably buttressed as interpretation. Three brief examples with respect to
Mesopotamian works make this abundantly clear:
1. tall spouted vessels and a series of conch shells + imitations in other

materials from the Royal Cemetery of Ur of the Early Dynastic
IIIA period (ca. 2600 b.c.e.);
2. an Early Dynastic period carved plaque from the lower levels of the
giparu, or residence of the EN-priestesses, at Ur; and
3. the calcite “Disk of Enheduanna,” of the Akkadian period from Ur
(ca. 2250 b.c.e.).
1. Spouted vessels have been found in funerary contexts in Early

Dynastic Mesopotamia, particularly in the graves of the Royal Cemetery
of Ur, and in the temple to the moon-god, Sin, at Khafajeh—both in
silver and in bronze. They are also depicted in a number of narrative
scenes on works such as a cylinder seal that shows a libation poured
before a seated deity (fig. 9; also fig. 5), and a shell plaque from the
Royal Cemetery showing a priest pouring a libation before a divine
standard (see on this, Winter, 1999, where typology and functions are
worked out). The use of similar spouted vessels as containers for liquid
can be documented in the abhi ēka of the Rādhārama a Temple (fig. 4)
and in the inventories of the VPP. They serve a multiplicity of uses in
the course of the ritual performance, from conveyance of the liquid
substance for libation, to hand-washing, to direct pouring of sanctified
liquid into devotees’ cups at the end of the ritual.
The utility of a spouted vessel for pouring is obvious, and based on
the contemporary images preserved from Mesopotamia, the excavator
of the Royal Cemetery had initially identified this vessel with ritual
libation. What the Indian case offered was a methodological caution
that the range of depiction might not represent the entire range of usage;
and indeed, when one returns to the Mesopotamian record, there turns
out to be hitherto unexamined evidence that such vessels also had a
range of uses in ancient times, including handwashing, and the filling
of cups at ritual banquets.
At the same time, a rather large group of conch shells and shallow
vessels of metal, alabaster, and shell found in the Royal Cemetery had
been identified by the excavator as “lamps”, although they showed no
signs of soot or residue from burning. The group consists of some 72
objects from 61 graves, of which 38 are of shell, 23 of copper, seven
of silver, three of alabaster and one of gold. The shell containers are
characterized by an open “bowl”, often cut open with some of the
whorls removed in order to enlarge the natural orifice, with an open
channel at one end. These cut shells are thought to have constituted the
original form from which the similarly-shaped vessels in other materi-
als had been copied. Their designation as lamps was presumably due
to a similarity in shape to later Roman lamps, with the open channel
serving as bedding for the wick, although the excavator recognized that
the absence of any sign of burning in the trough argued “against the
identification of the metal vessels as lamps” (Woolley 1934:283). He
noted that the shape could have been used “for some kind of libation,
390 chapter thirty
or for pouring unguents over the deceased”; however in the end the
objects were designated as lamps. Questions about this designation were
raised independently by Postgate (1977), Müller-Karpe (1993), and the
present author (in press), but it was important as validation of their
alternative pouring capacity to find such conch-shells deeply inscribed in
Vaishnavite ritual practice in India. In the context of the annual abhi ēka
in Vrindavan, conches serve as conveyors of the liquids poured directly
over the deity in the course of the ritual. Whether milk, curds, honey,
clarified butter, or saffron water, the liquid is first poured from a large
container into a spouted pot of the sort discussed above, then via the
spout into the conch, from which the liquid flows over the god (fig. 4).
It is said that the flow, when spiraling through the right-hand turnings
of the interior of the conch, conveys auspiciousness to the deity. It is
also the case that the quantity of flow from the limited container of
the conch is perhaps more controllable than would be the case were
the liquid poured directly from the larger spouted vessel.
Thus, while doubts about conches as lamps may have predated cross-
cultural observation, Hindu practice both reinforced such doubts and
also stimulated further inquiry into the Mesopotamian visual record.
And, indeed, it turns out that strong corroboration was present within
the assemblage of the Royal Cemetery itself for establishing such vessels
as conveyors of liquid furnished from a larger vessel. The evidence is
found on a shell inlay that forms the frontal plaque of one of the large
lyres found in tomb PG/789 of the Cemetery (Woolley 1934:105). In
the second register from the bottom, a jackal-like animal walking on
his hind legs is depicted carrying an offering table piled with portions
of meat, while a lion follows behind, holding a large jar suspended in
reed netting in his left hand; a flat open container lies across his right
palm. In both shape and size, the container is closely reminiscent of
the silver and gold “imitations” of the Royal Cemetery conches. The
relation between the two vessels carried by the anthropomorphized lion
is clearly functional, the smaller intended to serve either as a ladle or
conveyor of the liquid held by the larger jar. This imagery, once noted,
certainly strengthens the view that the conches and related vessels were
not lamps, but intended for pouring; however, it was the Indian use of
such vessels that stimulated the search for alternative usage.
2. A square plaque carved in low relief found in the giparu at Ur
(Moortgat 1969: Fig. 116) shows libations being poured from a spouted
vessel, comparable to scenes on contemporary seals (e.g., fig. 9): in the
bottom register, the libation is before an architectural (temple) facade,
and in the upper register, before a seated deity, likely to be the moon-
god of Ur, Nanna/Su’en. In a study of the office of en-priestess as
represented in Mesopotamian art, it was observed that while libations
were poured in both registers, before the exterior of the shrine and
before the deity, only in the lower register was a figure in procession
behind the libating priest carrying a small kid or lamb, presumably for
sacrifice. It was therefore suggested that the temporal sequence went
from bottom to top, as the lower scene represented ritual activity before
and outside the shrine, which would have been a more likely space for
blood sacrifice in any case; while in the upper register, the same set of
figures was then shown in ritual activity inside the shrine, that is, before
the image of the resident deity (Winter 1987:195).6
Parallels in Hindu practice strengthen this hypothesis. In temples
dedicated to the goddess Kālī, for example—one of the rare sects that
preserves animal sacrifice to this day—animals to be sacrificed are
tethered and slaughtered directly in front of the temple entrance. In
temples to other deities, vegetal substitutes such as gourds or melons
with red flesh are instead smashed upon the threshold of the temple on
certain occasions in lieu of animal sacrifice. This I was able to observe
during the inaugural ceremonies of the Laxmi temple in Ashland.
While, once again, the hypothesis regarding the giparu plaque existed
prior to observation, the parallel in Hindu tradition serves a dual
role: first, by demonstrating an analogous practice, it strengthens the
likelihood of the Mesopotamian interpretation; and second, as such,
it brings the Mesopotamian tradition alive, encouraging us to look not
only at instances of sacrifice outside the shrine, but also alerting us to
the importance of the threshold—known abstractly as liminal space,
marked in later Assyrian palaces by important inscriptions, and cited in
magico-medical texts as a dangerous place where evil spirits can lurk,
but unexamined to date as a locus for ritual activity.
3. The calcite disk associated with the en-priestess Enheduanna,
daughter of the first Akkadian ruler, Sargon of Agade (ca. 2340 b.c.e.),
and sister of two subsequent rulers, bears her inscription on one side
(see bibliography, Winter 1987). On the other side is a horizontal
band of relief showing a libation poured into a palm plant by a priest
6
This sequence is conflated on the Early Dynastic seal (fig. 9) where the divine
image (also probably the moon-god of Ur) is shown seated before the entrance to
his shrine.
392 chapter thirty
followed by three figures, the leading one a woman and assumed to be

Enheduanna herself (fig. 5). Her role as chief priestess in the cult of
the moon-god, Sin, at Ur, has long been known, although no records
preserve her role in ritual activity by virtue of the office.
In the relief, Enheduanna stands behind the priest actually pouring
the libation, with one hand raised in a gesture of greeting; she does
not pour the libation herself. While this had previously been noted
in describing the imagery, it was not remarked upon as an indica-
tor of the distinction between chief-priestly status and functionary in
actual cult performance. However, in the Vrindavan abhi ēka as well,
the primary ritual specialist and chief priest of the temple does not
himself perform the ritual libations or fire offerings to the deity after
the bathing festival, but rather directs the activity of other priests and
selected members of the lay community officiating in the offertory rota
(fig. 6). The analogous situation in this case serves to open an avenue
of investigation into the roles of priestly officials and lay participants,
suggesting a hierarchy of status and a direct correlation with certain
types of performance roles.
Thus, while in the three cases noted above one could have come
to similar conclusions by an analysis of the objects and their imagery,
augmented by archaeological and textual data, it was the observa-
tion of actual practice that either buttressed or helped to eliminate
previous interpretations, and further stimulated the search for internal
corroboration.
I should add, however, that observation does not automatically
translate into analogy. One has to be appropriately familiar with and
informed about both originating and parallel tradition, if one is to avoid
naive assumptions and/or gross error. The particular analogy of the
conch as a vessel for pouring liquids has a wry association for me, since,
when first noticed, I returned triumphantly to announce categorically in
a public talk that “conches were used for libation in Hindu ritual. . . . .” It
was then pointed out to me by an Indologist colleague that they were
used in ritual practice of the Vaishnavite sect, within which falls the
worship of Krishna, but not in all Hindu tradition, and that I must be
much more careful how I represent the tradition! This has been quite
salutary, for I had clearly fallen into the trap of the amateur, in gen-
eralizing when I had only a specific instance before me. It is a caution
for all those who attempt to move into an area of scholarship in which
one has not been primarily trained, and underscores the importance
of becoming an “informed” consumer, as well as of checking regularly
with experts. In short, it is critical to move beyond the browsing stage

of the “neophyte-cum-tourist” to one who controls the materials being
consumed; otherwise, the results are likely to be quite simplistic and
superficial, if not downright inaccurate.
A second caution is the reminder that depth of knowledge of and
familiarity with a tradition does not come all at once; nor is any tradition
likely to reveal itself except sequentially. Without multiple visits and the
privilege of recording the abhi ēka over a number of years, between 1987
and 1998, I would not have been able to make the connections noted
here. It is my sense that in the course of those visits my seriousness
was tested; and just as my own questions became more complex and
informed with each visit, so also more was revealed or made available
to me on successive occasions, while much remains unrevealed. Thus,
it was not until my sixth visit that I was invited to observe some of the
preparations for the ritual, including the cleaning of the silver throne
on which the deity would be installed after the abhi ēka (figs. 7, 8).7 And
even then, those ritual objects that were being readied in the temple
kitchen remained off-limits, as access to that part of the temple was
limited to members of the priestly household.
Finally, the obvious should be articulated: that engagement in a living
tradition (the “participant-observer” role so well-known in anthropologi-
cal fieldwork) does not permit the same cool distance as does the study
of texts and excavated objects. One makes friends; and sometimes loses
them. One walks the grounds and smells the flowers, and drinks water
one probably shouldn’t because one is hot and thirsty, and feels divinely
protected when no harm results. One begins to recognize roles and
players in the ritual performance, and the range of individual perfor-
mances across time. One rejoices in the god’s renewal along with the
devotees and ritual participants, and grumbles at the monkeys who steal
one’s temple sweets. In short, one weaves oneself in and is woven into
the ritual process; and then one must face the problem of adequately
representing the tradition/experience in one’s own domain.
In the end, just as no tradition will ever fully be accessible from
the outside, or fully committed to externalization through text, so also
7
Note here that when enthroned, Śrī Rādhārama a-jī is accompanied by cere-
monial vessels and pairs of attribute animals—in particular the cattle of the plains,
just as the seated moon-god Nanna-Su’en (fig. 9) is flanked by his attribute animals, a
pair of bull-calves, on the Early Dynastic seal.
394 chapter thirty
no two traditions separated by time and space will ever be entirely

analogous, or perhaps analogous in form but not in meaning. And
there are likely to be fundamental areas of difference between ancient
Mesopotamian and modern Hindu practice. One such area may be the
role of the devotee in public congregations of observance during ritual
performance. In Vrindavan, for example, the sanctuary is opened at
regular intervals on ordinary days, so that devotees may have a recip-
rocated viewing of the god (darśana), and on festivals such as the great
abhi ēka, devotees gather in the central space of the temple for the entire
day, from the awakening of the god in the morning, through his bathing
on the podium ( jagmohan; fig. 4), to his presentation, fully clothed and
ornamented, on the throne, set up on the jagmohan for evening worship
(fig. 8). The other extreme is represented by the Hebrew tradition of the
utter isolation of the (unimaged) deity in his sanctuary of the temple in
Jerusalem, which the High Priest entered only once a year, and that only
after great mental and spiritual preparation (Harlow 1978: 601–609).
In the absence of evidence from Mesopotamian sources, one cannot
project public engagement in the sanctuary or even direct sight-lines
of the revealed deity through the sanctuary door. All we know is that
the royal statue was placed within the shrine; and representations of
“audience” scenes with the gods suggest that the ruler and the priest-
hood at least did have direct access to the divine image (figs. 1, 9). There
is parallel textual evidence suggesting that the ruler sought audience
with the deity at particular times; and there is further evidence that
the priestly community related to the cult of a particular deity would
have had access to the shrine in the course of the care and feeding of
the god. On these grounds, the parallels with India seem compelling;
but at the same time, it is important to resist the temptation to assume
similarity—the “you are there!” rush, as articulated by Patty Jo Watson
(1979)—rather than feel impelled to demonstrate it.
These cautions notwithstanding, I can only say that the experience
in India in general, and in the Rādhārama a temple in Vrindavan in
particular, has been extraordinarily enriching—not only for myself
personally, but for the possibility to connect to the Mesopotamian
implements their use in a once-living tradition, and not merely to see
them as inert archaeological artifacts.
One may therefore return to the issue of the utility of analogy dis-
cussed at the beginning, and make three points: first, in the absence of
information from the ancient culture, but given sufficient grounding in
both the initial tradition and the tradition brought forward for com-
parison, one is able by analogy to suggest function for some artifacts,
and equally, to propose some interpretations not incompatible with
the broader context of cultural practice within the original tradition;
second, in the presence of weak internal evidence leading to relatively
tentative hypotheses concerning the initial tradition, it is possible, by
analogy, to strengthen existing hypotheses by reference to the compared
tradition; and third, in the presence of parallel praxes, it is possible,
by analogy, to heighten understanding through direct experience of a liv-
ing tradition.
Given the systemic parallels adumbrated above, it is likely that
many more enlightening analogies could be drawn between ancient
Mesopotamia and Hindu practice—not least of which could be in the
organization of the temple economy: landholding, priestly office, and
prebends (as per Bongenaar 1997); or attitudes toward light, radiance,
and “ornament” as multilayered means of expressing divinity, fertility,
and/or other positive cultural values (see, for example, Parpola 1993).
In addition, the possible historical connections of these two traditions,
posited in isolated studies (Dalley 1997; Golzio 1983; Pingree 1987,
1989, 1992; H. R. H. Prince Peter of Greece 1951), although not
essential for the present undertaking, may yet bear fruit. What my
own fieldwork in India and with the Hindu community in the United
States has demonstrated is that for the Mesopotamian sculptural images
under discussion, while one can certainly take such once-consecrated
works out of their current museum-context and re-place them into their
respective temples as part of a mental exercise, the actual experience of
the Hindu image in worship provides a far greater sense of an active
cult: complete with garments, ornaments, ritual vessels, incense (which
we know to have been part of Mesopotamian cult activity), music, and
flowers—the providers of visual, olfactory, auditory, tactile, even gusta-
tory sensation. It is my point that, pursued with rigor, such analogical
situations can serve the past, offering what Carl Nylander feared would
always elude the archaeologist: the music to the flute, the grief to the
grave (1969:3). In the present case, what the cross-cultural comparison
affords is a window into the sensory surround, at both artifactual and
experiential levels, that would have animated Mesopotamian ritual
performance as components of a living tradition.
396 chapter thirty
Acknowledgments
In mind are conversations with and profound debts to Cathy Asher, Pramod Chandra,
Shrivatsa Goswami, Robert C. Hunt, Asim Krishna Das, Michael W. Meister,
R. Nagaswamy, Kapila Vatsyayan, and Guy Welbon, without all of whose generosity I
could not have pursued the fieldwork and subsequent study necessary to write on this
topic.
References
Assmann, Jan. 1992. “Semiosis and Interpretation in Ancient Egyptian Ritual,” in

Interpretation in Religion, Vol. 2, S. Biderman & B. Scharfstein, (eds.), 87–109. Leiden:
Brill.
Begley, Vimala and Richard Daniel De Puma (eds.), 1991. Rome and India: The Ancient
Sea Trade, Madison: The University of Wisconsin Press.
Bentor, Yael. 1992. “Sūtra-style Consecration in Tibet and its Importance for Under-
standing the Historical Development of the Indo-Tibetan Consecration Ritual for
Stūpas and Images,” Tibetan Studies: Proceedings of the 5th Seminar of the International
Association of Tibetan Studies, Narita 1989. Narita.
——. 1996. “Literature on Consecration (Rab gnas),” in J. I. Cabezon and R. R. Jackson
(eds.), Tibetan Literature: Studies in Genre, 290–311. Ithaca: Snow Lion.
Berlejung, A. 1997. “Washing the Mouth: The Consecration of Divine Images in
Mesopotamia,” in K. van der Toorn (ed.), The Image of the Book: Iconic Cults, Aniconism,
and the Rise of Book Religion in Israel and the Ancient Near East, Leuven: Peeters.
Bloch, M. 1970. “Two Strategies of Comparison,” in A. Etzioni and F. L. Dubow (eds.),
39–41. Comparative Perspectives: Theories and Methods. Boston: Little, Brown.
Blurton, Richard. 1993. Hindu Art. Cambridge, MA: Harvard University Press.
Bongenaar, A. C. V. M. 1997. The Neo-Babylonian Ebabbar Temple at Sippar: Its Admini-
stration and its Prosopography. Istanbul: Nederlands Historisch-Archaeologisch Instituut
te Istambul.
Borger, R. 1956. Die Inschriften Asarhaddons Königs von Assyrien [AfO Beiheft 9]. Graz:
Self-published.
Börker-Klähn, J. 1975. “Šulgi badet,” ZA 64: 235–40.
Braun-Holzinger, E. A. 1977. Frühdynastische Beterstatuetten [Abhandlung der Deutschen
Orient-Gesellschaft, 19]. Berlin: Gebr. Mann.
Bühnemann, Gudrun. 1988. Pūjā: A Study in Smārta Ritual, Vienna: Gerold & Co.
Chang, K. C. 1967. “Major Aspects of the Interrelationship of Archaeology and
Ethnology,” Current Anthropology 8: 227–43.
de Coppet, D. 1992. “Comparison, a universal for anthropology: From ‘representation’
to the comparison of hierarchies of values,” in A. Kuper (ed.), Conceptualizing Society,
59–74. London and New York: Routledge.
Dagens, Bruno, trans. 1985. Mayamata: An Indian Treatise on Housing, Architecture and
Iconography, New Delhi: Sitaram Bhartia Trust, esp. Ch. 12: “The Foundation
Deposit,” 50–62.
Dalley, Stephanie. 1997. “Babylonian Influence in the Far East through the Buddha
and Mani,” Journal of Ancient Civilizations 12: 25–35.
De, S. K. 1961. The Early History of Vaisnava Faith and Movements in Bengal. Calcutta:
Firma KLM.
Dohmen, C. 1987. Das Bilderverbot. Seine Entstehung und seine Entwicklung im Alten Testament,
2nd ed. Frankfurt am Main: Athenaeum.
Edzard, D. O. 1997. Gudea und his Dynasty [Royal Inscriptions of Mesopotamia, Early
Periods, Vol. 3/1]. Toronto: University of Toronto Press.
Eggan, Fred. 1954. “Social Anthropology and the Method of Controlled Comparison,”
American Anthropologist 56: 743–63.
Ember, Melvin. 1991. “The Logic of Comparative Research,” Behavior Science Research
25: 143–53.
Finnish Oriental Society. 1993. Ancient Near East and India: Intercultural Religious Parallels.
The Franco-Finnish Symposium, 10–11 November 1990, The Finnish Institute, Paris [Studia
Orientalia 70]. Helsinki: Puna Musta.
398 chapter thirty
Galloway, Patricia. 1992. “The Unexamined Habitus: Direct Historic Analogy and
the Archaeology of the Text,” in J. C. Gardin & C. S. Peebles (eds.). Representations
in Archaeology, 178–95. Bloomington: Indiana University Press.
Gombrich, Richard. 1966. “The Consecration of a Buddhist Image,” Journal of Asian
Studies 26.1:23–36.
Goudrian, T. 1970. “Vaikhānasa Daily Worship according to the Handbooks of Atri,
Bh gu, Kāsyapa and Marīci,” Indo-Iranian Journal XII:161–215.
Goyon, Jean-Claude. 1972. Rituels funéraires de l’ancienne Egypte. Paris: Editions du
Cerf.
Hallo, W. W. 1988. “Texts, Statues and the Cult of the Divine King,” Supplement, Vetus
Testamentum 40:54–66.
Harlow, Rabbi Jules. 1978. Mahzor for Rosh Hashanah and Yom Kippur, 2nd edition. New
York: The Rabbinical Assembly.
Heider, Karl G. 1967. “Archaeological Assumptions and Ethnographical Facts: A Cau-
tionary Tale from New Guinea,” Southwest Journal of Anthropology 23:52–64.
Hein, E. 1978. “Temple, Town and Hinterland: The Present Network of ‘Religious
Economy’ in Puri,” in A. Eschmann, H. Kulke & G. C. Tripathi (eds.), The Cult of
Jagannath and the Regional Traditions of Orissa [South Asia Institute, S. A. Studies VIII],
439–68. New Delhi: Manohar.
H. R. H. Prince Peter of Greece. 1951. “Possible Sumerian Survivals in Toda Ritual,”
Bulletin of the Madras Government Museum N.S. VI:3–24.
Hunt, Robert C. (In preparation). Apples and Oranges: Comparability in Cultural Anthropo-
logy.
Inizan, Marie-Louise. 1993. “At the Dawn of Trade, Cornelian from India to Mesopo-
tamia in the Third Millennium: The Example of Tello,” South Asian Archaeology 1991:
Proceedings of the 11th International Conference of the Association of South Asian Archaeologists,
121–34. Stuttgart: Franz Steiner Verlag.
Jacobsen, Thorkild. 1987. “The Graven Image,” in P. D. Miller, Jr., et al. (eds.), Ancient
Israelite Religion: Essays in Honor of Frank Moore Cross, 15–32. Philadelphia: Fortress
Press.
——. 1989. “The Mesopotamian Temple Plan and the Kititum Temple,” Eretz Israel
20:79–81.
Jayaswal, Vidula. 1989. “Socio-Ritual Significance of Ancient Terracottas in the Gan-
getic Plains: The Ethnoarchaeological and Literary Evidence,” in J. M. Kenoyer (ed.),
Old Problems and New Perspectives in the Archaeology of South Asia [Wisconsin Archaeological
Reports vol. 2], 253–62. Madison: Univ. of Wisconsin Press.
——. 1990. “Ethnoarchaeological Study in Northern India Terracottas: A Case Study,”
in A. Ray and S. Mukherjee (eds.), Historical Archaeology of India: A Dialogue between
Archaeologists & Historians, 227–37. New Delhi: Books & Books.
Joshi, Rasik Vihari. 1959. Le rituel de la devotion K snaite. Pondicherry: Institut Français
d’Indologie.
Kamp, Kathryn A. 1993. “Towards an Archaeology of Architecture: Clues from a
Modern Syrian Village,” Journal of Anthropological Research 49:293–18.
Kramer, Carol. 1979. “An Archaeological View of a Contemporary Kurdish Village:
Domestic Architecture, Household Size, and Wealth,” in Kramer (ed.), Ethnoarchae-
ology: Implications of Ethnography for Archaeology, pp. 139–63. New York: Columbia
University Press.
——. 1982. Village Ethnoarchaeology: Rural Iran in Archaeological Perspective, Studies in
Archaeology. New York: Academic Press.
Kramrisch, Stella. 1981. Manifestations of Shiva. Philadelphia: Phila. Museum of Art.
Lamberg-Karlovsky, C. C. 1989. “Ethnoarchaeology: Legend, Observations and Critical
Theory,” in L. de Meyer & E. Haerinck (eds.), Archaeologia Iranica et Orientalis: Miscel-
lanea in Honorem Louis Vanden Berghe, 953–75. Ghent: Peeters Press.
Mace, Ruth and Mark Page. 1994. “The Comparative Method in Anthropology,”
Current Anthropology 35:549–64.
Matson, Frederick R. 1974. “The Archaeological Present: Near Eastern Village Potters
at Work,” American Journal of Archaeology 78:345–47.
Matsushima, Eiko. 1993. “Divine Statues in Ancient Mesopotamia: Their Fashioning
and Clothing and their Interaction with the Society,” in E. Matsushima (ed.), Official
Cult and Popular Religion, 209–19. Heidelberg: Universitätsverlag C. Winter.
Meister, M. W. 1981. “Display as Structure and Revelation: On Seeing the Shiva
Exhibition,” Studies in Visual Culture 7:84–89.
Milde, H. 1994. “ ‘Going out into the Day’: Ancient Egyptian Beliefs and Practices
Concerning Death,” in Jan Maarten Bremer, et al. (eds.), 15–34. Hidden Futures: Death
and Immortality in Ancient Egypt, Anatolia, the Classical, Biblical and Arabic-Islamic World,
15–34. Amsterdam: University Press.
Mital, P. D. 1966a. Braj kā sāmskrtik itihās. Mathura: Agraval Press.
——. 1966b. Braj ke utsav, tyauhār aur mele. Mathura: Sahitya Samsthan.
Moortgat, A. 1969. The Art of Ancient Mesopotamia. London and New York: Phaidon.
Müller-Karpe, M. 1993. Prähistorische Bronzefunde, Abteilung II, 14. Band: Metalgefässe im
Iraq, I. Stuttgart: Steiner Verlag.
Northrup, F. S. C. and Helen H. Livingston (eds.), 1964. Cross-cultural Understanding:
Epistemology in Anthropology. New York: Harper and Row.
Nylander, Carl. 1969. The Deep Well. London: Allen & Unwin.
Ochsenschlager, Edward. 1974. “Modern Potters at al-Hiba, with Some Reflections on
the Excavated Early Dynastic Pottery,” in Christopher B. Donnan and C. William
Clewlow, Jr. (eds.), Ethnoarchaeology [Monograph 4], 149–57. Los Angeles: University
of California, Institute of Archaeology.
——. 1995. “Ethnographic Evidence for Waterbuffalo,” Bulletin on Sumerian Agriculture
VII:1–9.
Oppenheim, A. Leo. 1964. “The Golden Garments of the Gods,” in Ancient Mesopotamia,
183–98. Chicago: University of Chicago Press.
Otto, E. 1960. Das ägyptische Mundöffnungsritual, Wiesbaden: Harassowitz.
Parker, Samuel K. 1992. “Contemporary Temple Construction in South India,” Res,
Anthropology and Aesthetics 21:110–13.
Parpola, Asko. 1990. “Bangles, Sacred Trees & Fertility,” South Asian Archaeology 1987:
Proceedings of the 9th International Conference of the Association of South Asian Archaeologists,
263–84. Rome: ISMEO.
Parrot, André. 1948. Tello. Vingt campagnes de fouilles (1877–1933). Paris: Editions
Michel.
Pingree, David. 1992. “Mesopotamian Omens in Sanskrit,” in La circulation des biens,
des personnes et des idées dans le Proche-Orient ancien [XXXVIIIe RAI], 375–79. Paris:
Editions Recherche sur les Civilisations.
——. 1989. “MUL.APIN and Vedic Astronomy,” in H. Behrens et al. (eds.), Dumu-
E2-dub-ba-a: Studies in Honor of Åke W. Sjöberg, 439–45. Philadelphia: The University
Museum.
——. 1987. “Venus Omens in India and Babylon,” in F. Rochberg-Halton (ed.), Lan-
guage, Literature, and History: Philological and historical studies presented to Erica Reiner [AOS
67], 293–315. New Haven: American Oriental Society.
Postgate, J. N. 1977. “Excavations at Abu Salabikh, 1976,” Iraq 39:269–99.
Ragunāthaji Śivaji, ed. 1986. Śrī Vallabhapu tiprakāśa. Bombay: Khemrāja Śrīkr -
adāsa.
Reade, J. E. 1979. Early Etched Beads and the Indus-Mesopotamian Trade [BM Occasional
Papers, 2]. London, The British Museum.
——. ed. 1996. The Indian Ocean in Antiquity. London: Kegan Paul.
Sluszkiewicz, Eugeniusz. 1980. “Pāli Bāveru ‘Babylon,’ ” Rocznik Orientalistyczny XLI:
107–17.
400 chapter thirty
Smith, Morton R. 1989. “India and Mesopotamia: Gods, Temples, and Why,” CSMS
Bulletin 17:29–35.
Smith, Sidney. 1925. “The Babylonian Ritual for the Consecration and Induction of
a Divine Statue,” Journal of the Royal Asiatic Society. 37–60.
Steible, H. 1988. “An Attempt at Establishing a Chronological Order for the Statues
of Gudea of Lagash. . . .” Unpublished lecture, University of Pennsylvania.
Stone, Eliz C. 1981. “Texts, Architecture and Ethnographic Analogy: Patterns of
Residence in Old Babylonian Nippur,” Iraq 43:19–33.
Tripathi, G. C. 1978a. “Navakalevara: The Unique Ceremony of ‘Birth’ and ‘Death’
of the Lord of the World,” in A. Eschmann, H. Kulke & G. C. Tripathi (eds.), The
Cult of Jagannath and the Regional Traditions of Orissa [South Asia Institute, S. A. Studies
VIII], 223–64. New Delhi: Manohar.
Vickers, Michael. 1994. “Nabataea, India, Gaul and Carthage: Reflections on Hellenistic
and Roman Gold Vessels and Red-Gloss Pottery,” AJA 98:231–48.
Vidale, M., Jonathan Mark Kenoyer and Kuldeep Kumar Bhan. 1993. Ethnoarchaeo-
logical Excavations of the Bead Making Workshops of Khambhat: A View from
beneath the floors,” South Asian Archaeology 1991: Proceedings of the 11th International
Conference of the Association of South Asian Archaeologists Verlag, 273–88. Stuttgart: Franz
Steiner.
Waghorne, Joanne P. 1992. “Dressing the Body of God: South Indian Bronze Sculpture
in its Temple Setting,” Asian Art V:9–33.
Waghorne, Joanne P. and Norman Cutler (eds.), 1985. Gods of Flesh, Gods of Stone: The
Embodiment of Divinity in India, Chambersburg, PA: Anima.
Walker, C. B. F. and M. B. Dick. 1998. “The Induction of the Cult Image in Ancient
Mesopotamia: The Mīs Pî Ritual,” in M. B. Dick, ed. Born in Heaven, Made on Earth:
The Making of the Cult Image in the Ancient Near East, Winona Lake, IN: Eisenbrauns.
Watson, Patty Jo. 1979. “The Idea of Ethnoarchaeology: Notes and Comments,” in
Carol Kramer, ed. Ethnoarchaeology: Implications of Ethnography for Archaeology, 227–88.
New York: Columbia University Press.
——. 1978. Archaeological Ethnography in Western Iran [Viking Fund Publications in
Anthropology, No. 57], Tucson: University of Arizona Press.
Welbon, Guy R. 1984. “Mahāsamproksana, 1981: Agama and Actuality in a Con-
temporary Temple Renovation,” in Agama and Silpa: Proceedings of the Seminar held in
December, 1981, 69–102. Bombay: Ananthacharya Indological Research Institute.
Winter, Irene J. 1987. “Women in Public: The Disk of Enheduanna, the Beginning
of the Office of EN-priestess, and the Weight of Visual Evidence,” in J-M. Durand
(ed.) La Femme dans le Proche-Orient Antique, 189–201. Paris: Editions Recherche sur
les Civilisations.
——. 1992. “ ‘Idols of the King’: Royal Images as Recipients of Ritual Action in
Ancient Mesopotamia,” Journal of Ritual Studies 6:13–42.
——. 1999. “Reading Ritual in the Archaeological Record: Deposition Pattern and
Function of the Two Artifact-Types from the Royal Cemetary of Ur,” in H. Kühne,
et al. (eds.). Fluchtpunkt Uruk, Archäologische Einheit aus methodischer Vielfalt, Aufsätze für
Hans J. Nissen, 229–256. Rahden: M. Leidorf.
Woolley, C. Leonard. 1934. Ur Excavations II: The Royal Cemetery [Publications of the
Joint Expedition of the British Museum and the Museum of the University of
Pennsylvania to Mesopotamia]. London: The British Museum.
Wylie, Alison. 1982. “An analogy by any other name is just as analogical: A commentary
on the Gould-Watson dialogue,” Journal of Anthropological Archaeology 1:382–401.
——. 1985. “The Reaction against Analogy,” Advances in Archaeological Method and Theory
8:63–111.
Younger, Paul. 1986. “The Citamparam Temple Complex and its Evolution,” East
and West 36:205–26.
Figure 1. Seal of Gudea, from Tello (after Moortgat 1969: pl. N1).
Figure 2. Rādhārama a temple, Vrindavan, abhi ēka (May 1991).

402 chapter thirty
Figure 3. Stele of Ur-Namma of Ur, fragment of reverse, ca. 2110 b.c.e.

(The University Museum, Philadelphia: CBS 16676).
Figure 4. Rādhārama a temple, abhi ēka, milk pouring from spouted vessel
into conch shell (May 1991).
Figure 5. Disk of Enheduanna, from Ur, Akkadian period,

ca. 2300 b.c.e. (The University Museum, Philadelphia: CBS 16665).
Figure 6. Rādhārama a temple, abhi ēka, chief priest and officiating

figure before deity (May 1991).
404 chapter thirty
Figure 7. Rādhārama a temple, abhi ēka, preparation of throne

before ritual (May 1991).
Figure 8. Rādhārama a temple, abhi ēka, Śrījī enthroned (May 1991).
Figure 9. Seal, Early Dynastic period, provenience unknown (perhaps

Umma?), ca. 2600 b.c.e. (Staatliche Museum, Berlin: VA 3878).
CHAPTER THIRTY-ONE
THE AFFECTIVE PROPERTIES OF STYLES:

AN INQUIRY INTO ANALYTICAL PROCESS AND THE
INSCRIPTION OF MEANING IN ART HISTORY
In the wider arena of the way(s) in which the arts and the sciences
generate appropriate terms and concepts to be used as instruments
in analytical operations, the term/concept “style” occupies a rather
special place: applicable both to the ways in which the operations are
undertaken1 and to describable characteristics of the objects of analysis.
In the present chapter, I wish to pursue, on the one hand, the lack of
discreet boundaries between “style” as it is manifest in a work and sub-
ject matter—hence, content and meaning—and, on the other hand, the
hermeneutic problems raised by attempts to correlate style and meaning
through “stylistic analysis” as operationalized in art history.
I have chosen my terms carefully to mirror the language used for
certain mathematical operations, as I believe the analogy holds well,
and in the hope it will raise questions of methodology common to both
the sciences and the humanities. To the extent that style is initially an
artifact of the hand of a maker, it is applicable, no less than a plus sign,
to the entire domain of (all possible) figures; but for a specific task, from
that broad domain, individual items are selected (i.e., certain figures);
then, with the addition of “style” to those figures, that is, the hand of the
maker, there results from the range of possibilities the specific image.
Much of what one can do with style depends upon how one defines
style; and I would assert from the beginning that there is no absolute
definition of style, but rather, a range of operative definitions varying
with user and analytical task to be performed. Art historians generally
* This article originally appeared as “The Affective Properties of Styles: An Inquiry

into Analytical Process and the Inscription of Meaning in Art History,” in Picturing
Science, Producing Art, C. A. Jones & P. Gallison, eds., New York & London: Routledge,
1998, pp. 55–77.
1
See for example, A. C. Crombie, Styles of Scientific Thinking in the European Tradition,
3 vols. (London, 1994).
406 chapter thirty-one
revert to Meyer Schapiro’s basic definition of 1953, as “the constant

form—and sometimes the constant elements, qualities, and expres-
sion—in the art of an individual or group.”2 This definition implies
that a given style is characterized by a particular attribute or attributes
observable in a work or group of works, which in turn permits the
construction of “sets” and “boundaries” on the basis of the presence or
absence of defining variables. What Schapiro did not fully account for,
however, was the element of agency in the manipulation and organi-
zation of form; nor did he engage the issue of the necessity of style in
the materialization of content.3 What is more, neither Schapiro’s basic
definition nor his extended discussion takes on the question of whether
style in fact inheres in a work, or rather is made to adhere to the work
as a product of description, comparison, and classification undertaken
by an external analyst.
For my purposes, I would take the position that once there is anything
in the work we can call form, then there is also style, but that it is also
important to keep what is intrinsic to the work distinct from what is
extrinsic to it, by consistently referring to post-hoc determinations as
the products of “stylistic analysis.” In that way, style is a function of a
period, place, workshop, or hand; it is inherent in the work, and it is
thus what is apparent to the perceiver. Stylistic analysis then introduces
the conscious observation, selection, and articulation of manifest proper-
ties to the act of perception. That distinction having been made, what
I wish to bring to discussion is the fiction that the so-called style of a
given work is merely a passive by-product, an artifact of “making” as
divorced from “meaning.”
To pursue this issue, it is crucial to see art history within the larger
picture of European intellectual history. In that larger picture, it becomes
possible to see why art historians at a particular moment in the history
2
M. Schapiro, “Style,” in Anthropology Today, ed. A. L. Kroeber (Chicago, 1953),
pp. 287–312. The subsequent literature is enormous; only a limited selection will
be referred to below; however, many of the discussions themselves contain extensive
bibliography.
3
In gross terms, this leads to the question of whether there can be a work without
style, as it is sometimes said of individuals, who “have no style.” To certain architectural
historians, as well—Viollet-le-Duc, for example—a building not executed within the
terms of a certain “order” can indeed be said to be without style. I would argue that
this is exactly a case in point of an “operative” (and limiting) definition of the term.
the affective properties of styles 407
of the discipline needed to separate the act(s) and sign(s) of making

from the range of cultural meanings attached to the finished work
(and equally, why this is no longer either necessary or desirable).4 And
it is also possible to pursue the degree to which the how of representa-
tion enters into the domain of choice—whether consciously as a tool
deployed by individuals and/or cultures, or subconsciously as generated
from/by a body of ideas and attitudes. This is surely no less true of
verbal or musical art forms than of the visual. And, as Joseph Koerner
has demonstrated for the writing of German philosopher Hans Blumen-
berg,5 it should be subject to analysis in all forms where construction
is integrally intertwined with content.
Art-historical analysis to date has tended to privilege subject matter
as the vehicle by which meaning is conveyed to an audience through
the work of art, and hence “iconography” as the analytical procedure
by which to arrive at an understanding of meaning. In two cases from
the ancient Near East, however—one of Phoenician and North Syrian
ivory carving of the early first millennium b.c., the other of Mesopota-
mian royal sculpture of the third and first millennia—it is possible to
explore two distinct ways in which style may be said to enter into the
arena of meaning. In the first case, I shall argue that questions of style
intrude as less conscious expressions of underlying cultural attitudes and
patterns; whereas in the second, style actually functions as a consciously
deployed strategic instrument with specific rhetorical ends. In both cases,
one may match meaning from subject matter in a particular cultural
context with expressive content in style. The resultant correlations
imply that style in fact plays an important role in complementing or
even activating the more overt message(s) provided by content, and as
such, (1) cannot be divorced from meaning in any study of the affective
properties of the work, and (2) should be considered in any historical
analysis of meaning.
4
On this, see W. Sauerlander, “From Stilus to Style: Reflections on the Fate of
a Notion,” Art History 6 (1983): 254, where style is referred to as a “highly condi-
tioned . . . hermeneutical ‘construct,’ worked out at a distinct moment in social and
intellectual history.”
5
J. L. Koerner, “Ideas about the thing, not the Thing Itself: Hans Blumenberg’s
style,” History of the Human Sciences 6, 4 (1993): 1–10.
Case I: Style and Cultural Meaning
The first issue, that of style as an expression of underlying attitudes and

patterns, arises from a comparison of two groups of ivory carvings of
roughly the eighth century b.c., found at the Assyrian capital of Nimrud
and a number of other sites in the ancient Near East.6 Neither group
is native to Assyria, and the original objects of which the remaining
plaques were clearly components must have been part of the impres-
sive booty and tribute in ivory attested in Assyrian text and depicted
on Assyrian reliefs.7
One group of ivories has been identified as the product of Phoeni-
cian work, congruent with the modern Levantine coast, while the other
group has been located within a region centering around northern
Syria and southeastern Anatolia—both areas in which the Assyrian
army mounted massive military campaigns.8 When one juxtaposes a
“Phoenician”-style furniture plaque depicting a winged female sphinx
with a “Syrian” example of the same motif, one is immediately struck
by the lack of “Egyptian” features in the Syrian work (figs. 1 & 2).
None of the decorative details—crown, head-cloth, and pectoral or
uraeus bib—known from Egyptian representations and quite accurately
reproduced in the Phoenician work, is included on the Syrian plaque.
Very different also is the extremely round face, puffy cheeks, and broad
nose of the Syrian sphinx as opposed to the more oval face and delicate
features of the Phoenician; or the elongated slender and long-legged
body of the Phoenician sphinx as compared with the heavy proportions
and short legs of the Syrian. In addition, on the Phoenician plaque,
6
See, among others, the primary publications of R. D. Barnett, A Catalogue of the
Nimrud Ivories (London, 1957); M. Mallowan and G. Herrmann, Furniture from SW 7
Fort Shalmaneser [ Ivories from Nimrud, fasc. III] (Aberdeen, 1974); and G. Hermann,
Ivories from Room SW 37 Fort Shalmaneser [Ivories from Nimrud, fasc. IV] (London, 1986),
where extensive bibliography is provided through the date of publication.
7
Narrative reliefs include both the removal of furniture as booty from captured
citadels and the subsequent use of such identical furniture, as in the well-known scene
of Assurbanipal and his queen in a garden (A. Paterson, Assyrian Sculptures: Palace of
Sinacherib [ The Hague, 1915], passim; R. D. Barnett, Sculptures from the North Palace of
Ashurbanipal at Nineveh [ London, 1976], Pl. LXIV).
8
A full analysis is presented in I. J. Winter, “Phoenician and North Syrian Ivory
Carving in Historical Context: Questions of Style and Distribution,” in Iraq XXXVIII
(1976), pp. 1–26, including a history of these distinctions, along with the social context
in which the schools of carving functioned.
the sphinx’s wings curve up in a delicate arc over the back, creating
a counter-rhythm to the horizontal body and vertical head, with the
embarrassing juncture of wing and shoulder hidden by the pectoral;
on the Syrian plaque, the wings jut up at an awkward angle, and are
folded back parallel with the body, so that one’s eye does not move
when looking at the piece, but rather is fixed on the massive block of
the head and body. The Syrian carver, then, seems to have chosen to
emphasize the sense of massive power in the image of the sphinx, at
the expense of elegance, grace, and detail of design.
This overall impression is reinforced further by the surrounding space.
On the Phoenician plaque, an interplay of filled and empty space has
been achieved; the animal is well planned into and comfortably con-
tained within its borders, and the plant elements are spaced to fill the
voids between legs and between the head and wing. The Syrian sphinx,
by contrast, presses up against the limits of its plaque, as if it were
simply too large to be contained within. Very little extra space exists
within the rectangle, and even the curving tendrils of a tree stump at
the right press up against the animal’s body. In short, when one looks
at the Phoenician piece, one is struck by the balance, elegance, and
careful design of the plaque as a whole. We may infer from allusions
in textual sources and from the context of usage in representation that
this composite mythological creature had symbolic significance beyond
its decorative function;9 but for the Phoenician sphinx, by virtue of the
attention given to its design, and the more remote profile view, one is a
step further removed from the impact of the motif than with the Syrian
representation, where everything conspires to confront the viewer with
the power vested in the sphinx as a semidivine being.
Many of these same distinctions can be made in comparing yet
another motif: that of a male figure slaying a griffin—a theme linked
with myths of the youthful hero-god, Ba’al, promoting fertility and life
through his victory over the destructive powers of the sun. Once again,
Phoenician and Syrian examples separate themselves quite readily
9
See, for example, the reference in the Hebrew Bible to a Phoenician master crafts-
man being employed in the construction of the temple of Solomon plus the decoration
of the cella with sphinxes (Hebr. cherubim), (I Kings 6:23–30) or the presence of pairs
of sphinxes along with griffins and bulls as heraldic pairs in a ritual scene painted on
the courtyard wall of a second millennium palace at Mari on the middle Euphrates
(A. Parrot, Mission Archéologique de Mari, II (2). Le palais: peintures murales [Paris, 1958],
pl. XI).
(figs. 3 & 4). The Phoenician griffin-slayer is shown with an accurately

rendered Egyptian-style wig. Both hero and griffin are winged, and
the spacing of the two sets of wings is complementary; their elegant
upward curve creates a rhythm that would have contrasted with the
downward diagonal of the spear, were the plaque complete. All four
of the griffin’s feet are planted on the ground, and only his head twists
back to receive the spearpoint. The horizontal form of his body bal-
ances the diagonals created by the forward-leaning stride of the hero.
No individual detail is allowed to predominate, and the delicate rhythm
of opposing elements, as with the Phoenician sphinx, creates a sense
of harmony and elegance. By contrast, the Syrian griffin-slayer wears
headgear borrowed from—but sadly misunderstanding—an Egyptian
royal crown. Only the griffin is winged here; yet at first glance, it is
difficult to determine to whom or where the wings are attached. The
hind legs of the griffin are thrown up against the sides of the plaque,
and this diagonal body, plus the bent knee of the hero, expresses the
force necessary to subdue such an adversary. This force is further
emphasized by the strong diagonal thrust of the sword, which seems
to dominate the whole piece. Man and animal are packed into the
rectangular space of the plaque, and seem to fight against the very
borders, as if ready to burst out of the frame.
These distinctions, whether applied to ivories or to other media,
presumably allowed the discerning ancient to identify place of origin,
just as today they allow the archaeologist and art historian to divide
the finds from Nimrud and elsewhere into two coherent stylistic groups,
Phoenician and Syrian, even in cases where motifs are common to
the two groups. Their attribution to place of origin is then fixed by
comparison with large-scale fixed stone monuments in the two adja-
cent geographical regions. The consistency of these elements brings to
mind Heinrich Wölfflin’s well-known pairs of attributes that permitted
a distinction between Renaissance and Baroque—his a distinction in
time, mine a distinction in space.10 In fact, Wölfflin’s terms, “calm,
complacent, graceful, still, in a state of being,” for the Renaissance,
versus “restless, overwhelming, pathological, in a state of becoming,”
10
H. Wölfflin, Principles of Art History: The Problem of the Development of Style in Later
Art, trans. M. D. Hottinger (New York, 1956, orig. in German, 1915); Renaissance und
Barock (Munich, 1888).
for the Baroque, sound significantly like a description of the differences

between Phoenician and North Syrian.
In my initial study of these ivories, the foregoing distinctions made
it possible to define distribution patterns of the two groups, and so
to speak of distinct regions of production and economic spheres of
interaction for what was in its time the most important luxury-good-
cum-artwork in the ancient Near East. In that regard, I was using
stylistic analysis, according to Meyer Schapiro’s 1953 definition, as an
archaeologist: one for whom style represents a diagnostic feature or a
series of symptomatic traits permitting one to locate the work spatially
or temporally.11 The operation was based upon a series of premises:
(1) a material work possesses (visual) properties; (2) those properties are
observable and describable; (3) they are then applicable to other works,
according to which, groups or clusters can be established; and (4) they
therefore permit of generalization. Furthermore, since the observable
properties are inherent in the object, one can come to them without
specific insider knowledge. In short, one does not have to control local
rules in order to create meaningful clusters.
This approach has not been limited to archaeologists; many of
Wölfflin’s contemporaries and successors have used style in this way
within the discipline of art history. As perspective on such usage, two
glosses on Wölfflin’s contribution are important here. First, however
reductively individual pair-bonds of his descriptive terms have been
employed by subsequent practitioners of art history—e.g., separating out
single elements, such as “painterly” to be opposed to “linear,” “open”
as distinct from “closed” forms—Wölfflin himself was explicit that these
were not independent variables, but rather existed as sets consisting of
co-varying elements—the consideration of all of which was necessary in
order properly to assign membership to one group or another. Indeed,
the use of sets of attributes comes very close to scientific methodology in
multivariate analysis, frequently employed in the study of archaeological
materials; and there, too, it has been shown that techniques that take
11
Schapiro, “Style,” p. 287. “Archaeology” has changed appreciably since Schapiro
wrote, however, and archaeologists have come to use style as a measure and expres-
sion of a variety of sociocultural factors, no longer merely for purposes of plotting
distribution—see, for example, D. D. Davis, “Investigating the Diffusion of Stylistic
Innovations,” in Advances in Archaeological Method and Theory, vol. 6, ed. M. B. Schiffer
(New York, 1983), pp. 53–89, and the essays included in M. W. Conkey, Style in
Archaeology (New York, 1989).
account of more than one variable, and particularly the associations

between variables, give far stronger (i.e., more informative) results.12
Second, the era that generated Wölfflin’s distinctions was entirely
consistent with their being used “archaeologically.” At the time, the
world of scholarly inquiry was sufficiently cognizant of the important
contributions of Linnaeus and other natural scientists that classification
was a primary intellectual endeavor, both scientific and historical.
The period also followed closely upon late-eighteenth-century analy-
ses of commodity and capital, a significant consequence of which had
been the reification, if not fetishization, of property—which was quickly
translated into the art market, where value, especially in painting,
was tied to recognition of a particular master’s hand. Style—or, more
properly, “stylistic analysis”—was thus used largely as a diagnostic tool,
à la Giovanni Morelli,13 in order to establish period, place, workshop,
and/or artist; in short, it was concerned less with meaning than with
attribution.
In the present climate of scholarly inquiry, however, other questions
of, hence other considerations of and possibilities for, the concept of
style and stylistic analysis intervene. Important for our particular case is
that Syria and Phoenicia were two contemporary cultural entities, with
closely related languages, a common pantheon and a shared mythologi-
cal tradition. How then to get beyond classification, to account for the
differences in what Schapiro called “the meaningful expression” carried
by differing styles used to render like motifs?
This is the point, of course, at which “outsider” observations must be
augmented by independent, localized evidence that will keep in check
projections grounded in mere non-disconfirmation, and instead sustain
a hermeneutic supported by data. In the specific instance of the ivories,
evidence abounds to demonstrate that the cities of Phoenicia were
linked historically to Egypt in ways that the cities of Syria were not;
they also looked out geographically on the Mediterranean seaways and
socially onto a more multicultural world in ways that the inland cities
of Syria did not. When these data are put together with the descrip-
tive properties of the ivories, it becomes possible to contextualize the
affective properties of the stylistic characterizations made above. If the
12
S. J. Shennan, Quantifying Archaeology (Edinburgh, 1988), cited in M. Roaf, “Pottery
and p-values: ‘Seafaring Merchants of Ur?’ Re-examined,” Antiquity 68 (1994): 776.
13
G. Morelli, “Principles and Methods,” in Italian Painters: Critical Studies of Their
Works, vol. I, trans. from the German by C. Ffoulkes (London, 1900), pp. 1–63.
Phoenician works are at once more elegant and more removed, with
greater balance between plein et vide, while the Syrian works are more
intense, with greater dynamic impact, then it might be hypothesized
that some bearers of Phoenician culture—at least, those producers and
users of elite objects—embodied more sophisticated and at the same
time less engagé attitudes and emotional states than their counterparts
in Syria.
Such claims, precisely because they are relatively consistent with our
own cultural distinctions, were easier to make in the days before the
past was identified as a foreign country and the act of “essentializing”
whole cultures or stereotyping subsets had itself been essentialized as a
tool of hegemonic discourse. Indeed, ever ahead of his times, Schapiro
had already noted the problems attendant upon attempting to read
“racial” and cultural worldviews from styles.14 Not surprisingly, there-
fore, the scientifically based processual archaeologists of the 1960s and
1970s insisted that we were ill equipped to be “paleo-psychologists,”15
and argued that such interpretive exercises should not be part of the
ancient historian’s purview. And yet, there are patterns of culture, just as
there are consistencies, if not correlations, in cultural/historical styles, as
Schapiro himself acknowledged, and as Winckelmann had observed as
early as the mid-eighteenth century.16 The tricky part is the attachment
of cultural meaning to those patterns by adducing non-retrojective,
non-anecdotal sources of evidence in support of such assignments, and
the location of those meanings in the appropriate societal band (whole
culture, elite, identifiable subculture, etc.).17
14
Schapiro, “Style,” pp. 306–10.
15
L. R. Binford, “Archaeological Systematics and the Study of Cultural Process,”
American Antiquity 31 (1965): 203–10.
16
J. Winckelmann, Histoire de l’art chez les anciens, vol. I (Amsterdam, 1766), pp. 12,
63, who first described the principal characteristics of Egyptian art, and then insisted
upon the degree to which those characteristics were to be found in all Egyptian works,
regardless of date.
17
Michael Baxandall’s “period eye,” as developed in his study, The Limewood Sculptors
of Renaissance Germany (New Haven and London, 1980), esp. pp. 143–63, was an attempt
at this, although since criticized as too “normative” and monolithic, but at least histori-
cally grounded. Similar interpretive forays are encouraged by the recent school of post-
processual archaeologists, for example, Ian Hodder, Reading the Past: Current Approaches
to Interpretation in Archaeology (Cambridge, 1986). Unfortunately, however, as applied to
date, in many cases the interpretations put forward are undersupported by evidence. But
see now the exploratory piece by Margaret W. Conkey, “Making Things Meaningful:
Approaches to the Interpretation of the Ice Age Imagery of Europe,” in Meaning in the
Visual Arts: Views from the Outside, ed. I. Lavin (Princeton, 1995), pp. 49–64.
For our case, interesting parallels have been suggested by William

Rathje in his study of ports of trade and contemporary inland sites of
the pre-Columbian New World. According to Rathje’s analysis, seacoast
towns in general are characterized by more cosmopolitan culture, are
less single-minded or committed to any particular religious or social
pattern, and, as a reflection of their domination by trading interests,
are open to many eclectic influences; inland areas, by contrast, tend
to be both more religious and more intensely committed to definite
cultural patterns.18 However one hesitates to characterize for fear of
moving from description to caricature, it was certainly possible to find
these same traits in the respective cultural identities of modern, pre-
1967 Lebanon and Syria, home to the ancient Phoenician and Syria
carving centers; and I believe there is sufficient evidence from the
literary and historical record to suggest that similar patterns of social
interaction and cultural adaptation pertained in the area during the
early first millennium b.c. as well.
The implications of the Phoenician/North Syrian case for the history
of art are not trivial, for the case suggests first, that attributes may be
identified that convey through style the “meaningful expression” of a
work or group of works, and second, that a “meaningful correlation”
may be established between a given style and the broader cultural
outlook of a region or social group. It then forces upon us the dual
methodological problems of how to identify significant attributes—i.e.,
what unites the hands of Michelangelo and Raphael as manifesta-
tions of the Italian Renaissance, rather than what distinguishes them
as individual artists19—and then, how to read style effectively as a
barometer of underlying cultural, regional, group, or personal attitudes
appropriate to particular moments of time or states of mind, without
engaging in massive retrojections of value and meaning existing only
in the observer?
18
W. Rathje, “The Last Tango in Mayapan: A Tentative Trajectory of Produc-
tion Distribution Systems,” in Ancient Civilization and Trade, ed. J. A. Sabloff and C. C.
Lamberg-Karlovsky (Albuquerque, 1975), pp. 409–45.
19
Another case of the relationship between style and meaning—one outside the
concerns of the present paper—would be when the style in fact reveals the hand of
a particular artist (as per Sauerlander’s distinction of stilus, Art History 6, cited above),
and the “meaning” of the work therefore takes on special significance as the work of
that artist and no other (on which, see J. L. Koerner, The Moment of Self-Portraiture in
German Renaissance Art [Chicago, 1993]). In such cases, the “sign” of the artist—her/his
“hand”—assumes an indexical value, independent of subject matter.
With this latter question, one evokes not only Schapiro’s definition
of style as “inner content,” but also Riegl’s Kunstwollen, or “artistic
volition,” that underlying drive toward a particular style in a particular
time and place.20 Panofsky, in his analysis of the concept of Kunstwollen,
articulated it as, in part, the “psychology” of a given period, in which
the collective will becomes manifest in the artistic creation21—and
indeed, it is possible to see early art history, especially in Germany, as
an extension of nineteenth-century studies of social psychology, influ-
enced strongly by Dilthey and others.22 Although neither Riegl nor
Panofsky makes it explicit, what Riegl was reaching for was a concept
of “culture” with a small “c,” operative in the social sciences as a system
with certain definitions and certain boundaries, within which both the
artist and his work were to be situated, however strong the mark of
the individual; and not “Culture” with a capital “C,” as the products
of (certain elements in) a society. It is this that I believe also underlies
Wölfflin’s observation (Wölfflin, whom subsequent art historians have
reified for the articulation of variables that have led to the decontextu-
alization of formal analysis!) that “not everything is possible at all times
in the visual arts.” By this I understand him to be saying that there
is a degree of historical, if not cultural determinism in any given period;
that, in his words, we can determine the “feelings” (the Lebensgefühl )
of a period from its style; and that “a new Zeitgeist [or, period spirit]
demands a new form” (i.e., style).23 This I would amend for the case
20
A. Riegl, Stilfragen (Munich, 1977, orig. publ. 1893, 2nd ed., 1923). And see on this,
H. Zerner, “Alois Riegl: Art, Value and Historicism,” Daedalus 105 (1976): 177–88.
21
E. Panofsky, “Das Problem des Stils in der bildenden Kunst,” republished in Aufsatz
zu Grundfragen der Kunstwissenschaft (Berlin, 1964, orig. publ., 1915); also, “Der Begriff des
Kunstwollens,” Zeitschrift für Aesthetik und allgemeine Kunstwissenschaft XIV (1920), recently
translated by K. J. Northcott and J. M. Snyder, “The Concept of Artistic Volition,”
Critical Inquiry VIII (1981): 17–33.
22
Discussion in L. Dittmann, Stil, Symbol, Struktur: Studien zur Kategorien der Kunstgeschichte
(Munich 1967), esp. chapter 1. In part, Wölfflin’s work was in answer to more deter-
ministic notions of style deriving from material and/or technique—although Heather
Lechtman has revived the issue of interdependence between imagery and production
techniques (“Style in Technology: Some Early Thoughts,” in Material Culture: Styles,
Organization and Dynamics of Technology, eds. H. Lechtman and R. Merrill [St. Paul,
Minn., 1977], pp. 3–20).
23
See, for example, Wölfflin, Principles, p. 11. Michael Podro (The Critical Historians of
Art [New Haven, 1982], pp. xxiv–xxv) provides the background for this in his reference
to Wölfflin’s first work Renaissance und Barock, as an attempt to “explain changes in style
as changes in attitude,” and Wolfflin’s later development, based upon the work of Adolf
von Hildebrand, of the “reciprocal adaptation of subject matter and the material of
visual representation” (i.e., style).
of the contemporary Phoenicians and Syrians to: “a different Kulturgeist

[or, cultural spirit] demands a different form.”
This having been said, it is nonetheless true that attaching historically
accurate “feelings” to a given visual manifestation is a major problem
for the historian, as is the recovery of historically meaningful interpre-
tations of those feelings within a cultural sphere. And obviously, these
tasks become increasingly difficult the further one is in time and place
from the making, coding, and culture of the original. An early caution
came from Dante Gabriel Rosetti, in his “The Burden of Nineveh,”
where, occasioned by his confrontation with ancient Assyrian art,
he warned against unwarranted interpretation of the character of a
civilization. Legend has it that Rosetti was emerging from the British
Museum in 1851, just as a great winged gateway lion from Nimrud was
being hoisted up the front steps (see Illustrated London News, February 28,
1852), whereupon he was prompted to meditate upon the constraints
placed on understanding a distant culture known only through isolated
remnants (although when we see an individual standing alongside one
of these great colossi, 4 to 5 meters in height, it is hard to escape a
response to its scale, whatever one’s culture, or not to ascribe size as
one of its meaningful attributes in antiquity).
Nevertheless, there is risk involved in any attempt to correlate visual
attribute with sociocultural interpretation. Not very fruitful attempts
were made by several anthropologists in the 1950s, who used then
recently developed “achievement-motivation” tests for individuals
emerging from our own culture as measures of whole cultural tradi-
tions (for example, the ancient Greeks and the Maya). The goal was to
determine when the phase of greatest drive toward achievement had
occurred—largely on the basis of the frequency of certain extremely
simplistic diagnostic traits, such as diagonal lines, in the artifactual
assemblages of the particular tradition.24 These studies led to a dead
24
For example, A. F. C. Wallace, “A Possible Technique for Recognizing Psychologi-
cal Characteristics of the Ancient Maya from an Analysis of their Art,” American Imago
7 (1950): 239–58; E. Aronson, “The Need for Achievement as Measured by Graphic
Expression,” in Motives in Fantasy, Action and Society, ed. J. W. Atkinson (Princeton,
1958), pp. 249–65. Nor were anthropologists the only proponents of such readings.
J. N. Hough, as early as 1948 (“Art and Society in Rome,” in Transactions of the American
Philological Association 79 [Philadelphia], p. 341), posited correlations of artistic techniques
with Roman social structure: i.e., between rigidity of design and the rigidity of society,
between relief sculpture on a single plane and autocracy, and between spatial illusion-
ism and social participation in government. The work of E. Davies (“This is the way
Crete went—not with a bang but a simper,” Psychology Today 3 [1969]: 43–47), is a
end because their categories were far too gross, and the underlying
assumption that the meanings attached to particular diagnostic cat-
egories were universal and therefore could be universally applied was
never tested, hence never confirmed. The same criticism of assumed
universality could be applied to the cross-cultural search for binary
opposition in the structuralism of the 1960s;25 but what structuralism
contributed was the formulation that underlying cultural patterns were
manifest and could be discerned in the material products (namely, “art”)
of a given cultural universe.
This brings us back to what Henri Zerner has referred to as Riegl’s
“radical historicism”—his total rejection of normative aesthetics toward
the culturally and historically specific.26 It is certainly true that, follow-
ing upon post-structuralist and postmodernist critiques, we now under-
stand cultures (and historical moments) to be less discretely bounded
and homogeneous than initially perceived, and selected voices—as in
cultural and political elites—differentially recorded.27 But unless we
as art historians wish to concede the impossibility of any historically
grounded knowledge, the result of these critiques must be to raise the
standards of argument and evidence, rather than to relinquish any hope
of explanation. A major problem lies in confirmation. It is not possible
to know for certain whether the emotional and/or experiential values
attributed to the coherent variables that constitute the Phoenician as
opposed to the North Syrian style are valid for the first millennium
b.c. in the ancient Near East. One can certainly seek, and even think
to find, corroborating contextual information and/or historical analogy
that can be brought to buttress interpretations of the work. But even to
do so, one must be operating under an initial premise that there be a
meaningful correlation between the manifest elements of style and the
experiential nexus from which they derive—a premise that has not to
date been subject to hypothesis formation and testing.
by-product of that era, in which the psychological variables needed for achievement
were concretized into the number of diagonal lines, s-shapes, and unattached forms
manifest in early Greek pottery.
25
C. Lévi-Strauss, Anthropologie structurale (Paris, 1958); La pensée sauvage (Paris, 1962),
etc., and the many analytical studies of structuralism, e.g., Dan Sperber, Le structuralisme
en anthropologie (Paris, 1968).
26
Zerner, Daedalus 105:185.
27
E.g., James Clifford, The Predicament of Culture (Cambridge, 1987).
Case II: Style, Agency, and Agenda
The Phoenician/North Syrian case suggests that certain visual attributes

derive from the special geographical and/or historical situation of the
producing culture, and that they represent not-necessarily-conscious
reflections of worldview and experience held by at least some mem-
bers of that culture. My second case, that of certain elements of style
employed in Mesopotamian royal sculpture, requires briefer discussion,
but raises the important issue of the degree to which a style may be
consciously deployed as an active agent in constructing a worldview, for
particular rhetorical-cum-ideological ends.
Panofsky left this possibility open in his 1920 study of Riegl’s Kunst-
wollen, when he referred to the collective will of a period, manifest in
artistic creation and apprehended either consciously or unconsciously.28
There is certainly little disagreement in the field that the decorative
programs devised for public spaces can embody very conscious con-
structs. The palaces of Neo-Assyrian rulers from the ninth through the
seventh centuries b.c. prove no exception. In the Northwest Palace
of Assurnasirpal II (883–858 b.c.), for example, the four epithets of
ideal royal attributes found in the king’s Standard Inscription, carved
on every slab of his palace reliefs, find their exact counterparts in the
four ways in which the king is represented: “attentive prince” shows
him seated in an ancient posture related to rendering good judgment;
“keeper of the gods” shows him attendant upon the sacred tree, under
the aegis of the god Assur; “fierce predator” shows him in battle with
wild bulls and lions; and “hero in battle” shows him victorious over
enemy citadels.29 But the power invoked in his epithets is also manifest
in the size, proportion, and musculature of the human body, as seen
not only in the king’s own figure, but also in those of his protective
genii that flank the doorways and mark the vulnerable corner spaces
of the palace (fig. 5).30
In many respects, the way of rendering the ruler in early Neo-Assyr-
ian art shows continuity with the preceding Middle Assyrian period of
28
Panofsky, in Critical Inquiry VIII:20, emphasis mine.
29
I. J. Winter, “Royal Rhetoric and the Development of Historical Narrative in
Neo-Assyrian Reliefs,” Studies in Visual Communication 7 (1981): 2–38.
30
J. Curtis and J. E. Reade, eds., Art and Empire: Treasures from Assyria in the British
Museum (New York, 1995), p. 43; J. B. Stearns, Reliefs from the Palace of Ashurnasirpal II
(Graz, 1961), pls. 44–51.
the second millennium b.c. For example, on a carved altar of Tukulti-

Ninurta I (1244–1208 b.c.), the earlier king carries a similar mace
and wears a similar wrapped garment, while in other Middle Assyrian
representations, the royal tiara is similar.31 What has shifted by the early
first millennium is precisely the heavier proportions of the royal figure,
and, most clearly manifest on the kilted genii, the emphasis on massive
musculature that is characteristic of early Neo-Assyrian “style.”32
I would argue that this is not a random shift. It corresponds to an
equal intensification in contemporary texts of references to might and
power, coequal with the extraordinary military expansion of the Assyr-
ians in the period. In short, the power invoked in subject matter is the
power manifest in style is the power at issue in the state.
The well-known statues of Gudea of Lagash (ca. 2110 b.c.) constitute
a parallel situation some fifteen hundred years earlier (see fig. 6). They
are recognizable by the ruler’s characteristic headgear, cylindrical body,
clasped hands, and enlarged, staring eyes. While these are all diagnostic
features of Gudea statues, and thus serve to date unexcavated works
and associate all with the Neo-Sumerian period, it is only when the
statues are seen in the context of the inscriptions accompanying virtually
every one of the over twenty extant works that we understand we are
in the presence of a true confluence of style and meaning. For one of
Gudea’s chief epithets in Sumerian, indeed, written directly upon one
of his statues in a lengthy dedicatory text, is a2 sum-ma dNin-dar-a-ke4
(“arm strengthened [lit., given] by [the god] Nindara”), and I would
argue that this specific quality appropriate to rulership, rather than
the random development of a more “realistic” or plastic mode of ren-
dering, is what best accounts for the massive musculature of Gudea’s
uncovered right arm. We are further told that he possesses wisdom,
which in Sumerian translates literally as one “of wide ear” (geštu-dagal),
and is regarded with a “legitimizing gaze” by his god, upon whom he
is to concentrate in return, thus accounting for the enlarged ears and
eyes.33 In short, as with the Assyrian reliefs, form (style) has been used
to convey intended meaning: the power, authority, and appropriate
31
Reproduced in A. Moortgat, The Art of Ancient Mesopotamia (London and New
York, 1969), fig. 246.
32
Compare, for example, the rendering of such a genius figure on Middle Assyrian
cylinder seals, e.g., ibid., pl. K: 6 and 7, with the Neo-Assyrian palace figures.
33
See extended discussion of this in I. J. Winter, “The Body of the Able Ruler:
Toward an Understanding of the Statues of Gudea,” in Dumu-É-dub-ba-a: Studies in
Honor of A. W. Sjöberg, ed. H. Behrens et al. (Philadelphia, 1989), pp. 573–83.
attributes necessary to rule, possessed by virtue of representation by

one claiming the right to rule!34
The covariance apparent in the Gudea and Assurnasirpal statuary
between intended message/content and style of rendering strengthens
the generalizable relationship between form and meaning suggested
above, by adding conscious choice to the construct. This relationship
has been cognized more readily in literary studies, and is increasingly
evident in the field of literary as well as art history in recent years,
from Gary Saul Morson’s study of social realism in the Soviet novel
to David Summers’s study of contrapposto, among many others.35 The
capacity of style to carry value, and to be purposely deployed in order
to represent specific values, has even been recognized by art historians
whose approach has been largely “formalist”—as, for example, Sidney
Freedberg in his study of the stylistic revolution occurring in Italian
painting around 1600, in which he noted that the “manner of employing
basic elements of a style may be altered . . . to accord with the artist’s
sense of the nature of his subject.”36 Such a situation of conscious choice
served as the core of the important dissertation by Leslie Brown Kessler
with respect to the work of Domenichino and Lanfranco, where she
showed that differing aims on the part of contemporary artists could
call forth not only different subjects, but also distinct styles considered
appropriate to those subjects.37 Although the textual (and cultural)
record of the ancient Near East neither identifies individual artists nor
34
One could continue making this case. For example, as one moves to the end of
the Neo-Assyrian sequence, to the seventh-century reliefs of later kings such as Assur-
banipal, the apparent stylistic shifts covary with developments in narrative complexity:
the proliferation of information and the expansion of the visual field reduce emphasis
on the king himself, whose figure then becomes both smaller and less massively pro-
portioned—see discussion in Winter, Studies in Visual Communication 7, pp. 27, 30.
35
G. S. Morson, “The Socialist Realist Novel and Literary Theory” (unpubl. ms.),
p. 15; D. Summers, “Contrapposto: Style and Meaning in Renaissance Art,” Art Bulletin
59 (1977): 336–61. For the general issue of the impossibility of a neat division between
form and content in poetics, see A. C. Bradley’s “Oxford Lectures on Poetry,” in The
Problem of Aesthetics: A Book of Readings, ed. E. Vivas and M. Krieger (New York, 1960),
p. 569. See also Jeffrey Ruda, “Flemish Painting and the Early Renaissance in Flor-
ence: Questions of Influence,” Zeitschrift für Kunstgeschichte 2 (1984): 210–36, where the
argument is made explicitly that it is essential to look at the relationship between style
and iconography (p. 236).
36
S. Freedberg, Ca. 1600: A Revolution of Style in Italian Painting (Cambridge Mass.,
1983), p. 8.
37
L. B. Kessler, “Lanfranco and Domenichino: The Concept of Style in the Early
Development of Baroque Painting in Rome” (Ph.D. dissertation, University of Penn-
sylvania, 1992).
includes conscious exegeses on art-making, we can nonetheless observe

(and highlight) those related instances—largely court art executed
within a domain of political ideology—in which style may be seen to
carry value and therefore convey meaning, as well as instances when,
as in the later Neo-Assyrian period, styles have been altered in order
to accord better with rhetorical ends.38
While these observations will come as a surprise to no one, I do
believe it is important to put them into the context of current issues in
the practice of art history. The initial isolation of “style” from “icono-
graphy” as two discrete tools of analysis—the one related to form, the
other to content and meaning—served the field well, up to a point. The
analysis of style came to serve as the means by which authentication or
attribution could be attached to a given work, and was privileged by
some practitioners of the history of art, thereby leaving iconography
to another set of practitioners, with each subgroup subject to intellec-
tual fashion.39 Yet the unit, and the unity, is ultimately the work as a
whole, of which “style” and “iconography” are analytical subsets. In
any individual undertaking, therefore, isolated stylistic or iconographi-
cal analyses can only be partial. Too often, one or the other has been
taken for the whole: the whole of the work, or the whole of the art
historical endeavor.
At a moment more than one hundred years since these analytical
tools were developed for use in art history, it is important to keep in
mind that they have been constructed by us, to serve for particular
procedures. Often, as is the case with advances in technology, such
tools are discovered to possess properties that permit other analytical
operations not thought of when the tools themselves were invented. If
38
Additional cases within the art of the ancient Near East could be cited to buttress
those presented here. For example, the so-called “Court Style” in Achaemenid Persian
seal carvings of the sixth to fourth centuries b.c. has been suggested to represent a
conscious response to demands made upon carvers to construct a “visual language of
control and empire” (M. B. Garrison, “Seals and the Elite at Persepolis: Some Observa-
tions on Early Achaemenid Persian Art,” Ars Orientalis 21 [1992]: esp. p. 17).
39
The intellectual climate of the ascendancy of iconographical analysis, following
upon the work of Panofsky, was the stimulus, for example, for George Kubler’s The
Shape of Time: Remarks on the History of Things (New Haven, 1962), as he attempted to
resurrect the validity of a notion of style in understanding art and cultures. Henri
Focillon’s The Life of Forms in Art (New York, 1943, orig. La Vie des formes), which
articulated the premise that certain forms give rise to other forms as such without the
necessary mediation of culture, helped to keep alive a sharp divide in the field between
adherents of one approach or another.
the division of style and iconography as discrete analytical tools in art

history initially became equated with a comparable division between
form and meaning, suggesting that meaning was to be revealed through
the iconographic enterprise and not through an analysis of style, it is
now time to reconsider that division. The degree to which it is no lon-
ger sufficient is the degree to which we insist more on the many ways
in which an artwork can “mean,” along with a better understanding
of the various contextualizations of the work today—plural in the face
of a postmodern awareness of positionality and polyvalence, but still
allowing for more than the fact of the work’s production in assessing
the cultural and historical climate of its production and subsequent
reception.
In this respect, we have come to assume Riegl’s negative attitude
toward any theory that severed art from history.40 As concerns the
historical divide between form and meaning, it is also apparent that
some creative and analytical art historians had pointed out the theo-
retical limitations of this division quite early. In particular, I would
cite Robert Klein, whose work has been too little considered since his
untimely death in 1967.41 Indeed, as noted above, it has been repeat-
edly demonstrated for individual cases that both style and iconography
in fact carry meaning; that often the meaning they carry is either
identical or complementary; and that when it is not, we must account
for the discrepancy—purposeful subversion, incongruence—by further
analysis of meaning. Therefore, I emphatically underscore once again
the importance of the challenge to the exclusion of style from investiga-
tions into the domain of meaning—not just in particular cases in the
art-historical literature, but as a general principle.
This has been perhaps best understood in studies of clothing styles,
from A. L. Kroeber and Roland Barthes to Dick Hebdige and Kennedy
40
Zerner, Daedalus 105, p. 179; and see also, Reigl, “Das holländische Gruppenpor-
trait,” Jahrbuch der Kunsthistorischen Sammlung des Allerhochsten Kaiserhauses 23 (1902), pp.
71–278, esp. p. 73; where he states that “the task of art history is to . . . decipher . . . the
essential character” of the period.
41
R. Klein, Form and Meaning: Essays on the Renaissance and Modern Art, trans. M. Jay
and L. Wieseltier, with a foreward by H. Zerner (New York, 1979, orig. La Forme et
l’intelligible [Paris, 1970]). See esp. “Thoughts on Iconography,” p. 149, where Klein
refers to conscious quotations of antique style in nineteenth-century painting (e.g.,
Gros’s Napoleon at Eylau) that provide “nonexplicit, wordless meaning”—i.e., one not
carried overtly through subject matter.
Fraser,42 and is apparent today in both clothing store windows and

advertising layouts, where a whole universe of value is subsumed within
the category of “taste.” Display in advertising and in shop windows—
lighting, color, accessories, posture, and grouping of models—serves to
set up emotional linkages to merchandise that itself manifests particular
properties of style and is embedded in a vast nexus of signification.43
The acculturated individual who then chooses to dress in a certain style
has elected to signal the attendant meanings and values conveyed by
the signs upon her/his body.44 Potential consumers who react to fashion
store windows and/or viewers who react to an individual’s dress style
also represent insiders who know and operate under understood sets
of coded references. They should in principle be directly analogous to
contemporary audiences for Phoenician or North Syrian ivories or the
reliefs of an Assyrian throne room, whose responses would equally be
determined by their familiarity with and sensitivity to the full range of
associations afforded by the visual stimuli.
The key to “style-as-meaning” lies, I would argue, in cultural con-
text and in the emotional response invoked/provoked by the work.
Here I would build upon an essay by James Ackerman, in which the
42
A. L. Kroeber, Style and Civilizations (Ithaca, 1957); R. Barthes, The Fashion System,
trans. J. Ward and R. Howard (New York, 1983); D. Hebdige, Subculture: The Meaning of
Style (London, 1979); Kennedy Fraser, “On and Off the Avenue: Feminine Fashions,”
a series of articles for the New Yorker from November 1970 into the early 1980s. (See,
in particular, the beautiful piece on Bianca Jagger in the issue of February 24, 1973,
in which “style” itself becomes an attribute, going beyond description to the ascription
of positive “affect.”)
43
B. Means, “Clothing Store Windows: Communication through Style,” Studies in
Visual Communication 7 (1981): 64–71.
44
The emotional component, and the individual identity constructed in association
therewith, distinguishes this view of (choice in and associations with) clothing style from
many anthropological studies that emphasize instead an emotionally de-cathected set
of attributes permiting the conveyance of information regarding social group member-
ship and maintenance and/or serving as signs in the construction of social boundaries.
See, for example, M. Wobst, “Stylistic Behavior and Information Exchange,” in For
the Director: Research Essays in Honor of James B. Griffin (Anthropology Papers, Museum
of Anthropology, University of Michigan, no. 61), ed. C. E. Cleland (Ann Arbor,
1977), pp. 317–42, with specific reference to clothing styles. In addition, see M. W.
Conkey, “Style and Information in Cultural Evolution: Toward a Predictive Model
for the Paleolithic,” in Social Archaeology: Beyond Subsistence and Dating, ed. C. L. Redman
et al. (New York, 1978), pp. 61–85 and R. N. Zeitlin, “A Sociocultural Perspective
on the Spatial Analysis of Commodity and Stylistic Distributions,” in The Human Uses
of Flint and Chert, ed. H. de G. Sieveking and M. H. Newcomer (Cambridge, 1986),
pp. 173–81, where the notion of style as marker of social and political affiliation is
extended to other types of goods/works.
impact of a work is seen as the result of a combination of intellectual

knowledge plus sensory perception.45 It is style, I would argue, that sets
up the parameters for and the emotional linkages of affective experience,
via the culturally conditioned sensory motors of visual perception. And
in that respect, issues of style engage both properties of the work and
functions of response. In short, style both inheres in a work and lives
in the eye of the beholder.
With this, we may return to the aims of the present volume. For at
the level of sensory perception, the observation and experience of style
as a manifest cluster of attributes links the humanist to the scientist,
the historian of art to the historian of science.46 And furthermore, as
an analytical tool, stylistic analysis functions like any scientific attribute
analysis, requiring description, classification, and systemic contextualiza-
tion—goals of the scientist no less than of the humanist.47
I have argued for a further component in understanding style, how-
ever—one that requires moving from description and classification to
experience. It is therefore implied that the analyst of style in any given
historical manifestation not only replicates certain scientific procedures
in the course of analysis, but also functions as a social scientist in the
attempt to capture historicized experience, just as the contemporary
experiencer of stylistic properties can only do so as a social being.
45
J. S. Ackerman, “Interpretation, Response: Suggestions for a Theory of Art
Criticism,” in Theories of Criticism: Essays in Literature and Art [Occasional Papers of the
Council of Scholars of the Library of Congress, no. 2] (Washington D.C., 1984), pp.
33–53. This is the sort of problem pursued more by perceptual psychologists than art
historians: see, for example, Salek Minc, “Significant Form and Physiological Stimuli
in Art Perception,” Australian UNESCO Seminar: Criticism in the Arts University of Sydney,
May 1968 (Canberra, 1970), pp. 68–80, who based his distinction on the work of Clive
Bell [Art (New York, 1957)], which by that time was largely out of date. However,
what he (Minc) was after, the relationship of visual stimulus to the human autonomic
response system, not merely the cultural system, is the problem that Ackerman was
also raising—one that has yet to be systematically explored.
46
Herein lies the answer to the problem raised by David Topper (“The Parallel
Fallacy: On Comparing Art and Science,” British Journal of Aesthetics 30 [1990]: 311–18):
the illogical propositions that create a false analogy between “art” and “science” become
logical once one compares “art history” and “science” as parallel analytical traditions.
47
The related issues of the assumption of discrete boundaries in descriptive sets
and the authority assumed by the analyst who, in the very act of determining what
variables will be privileged in the making of those boundaries, inscribes exclusion
into the analytical process, have not been pursued here, but should be acknowledged
as problematic. I would argue that it is important to take them into consideration in
specific analytical undertakings.
I have further suggested that it is only in the unity of “form-plus-con-

tent” that a given work of visual art realizes its ontological identity—
whether for its own original time and place, or for the viewer/analyst
at a distance. Since subject matter must be given physical form in
order to convey itself visually, the very act of making produces a way
of making; and if one accepts that that way of making is manifest as
style, then it is style that not only gives form but also “affective agency”
in the psychological sense to the meaning of the subject matter. Or,
put another way, style itself then becomes a sign existing between the
maker and the world, to be processed no less than subject matter. If it
is easier to describe the physical properties of a style than it is to assess
their affective value, that is not a license to ignore the latter, or to avoid
developing methodologies that will permit access to them.
Schapiro closed his 1953 article on style with the statement that,
“a theory of style adequate to psychological and historical problems
has still to be created.” That statement remains true more than fifty
years later, although we can be said to have made some progress.
On the basis of issues raised here, I would propose that at the very
least a theory of style must (1) consider the proposition that there is
a nonrandom relationship at the macrolevel between a style and the
culture/period within which it is produced, before one ever gets to the
relationship between a given style and the psyche of a specific indi-
vidual (i.e., artist) making “art”; (2) acknowledge that style is closely
allied with the psychological stimulus known as “affect,” and as such
is an integral component in the communication of meaning, hence in
the response that the work elicits; (3) take account of the fact that the
potential use and value of style as a concept depends entirely upon
the nature of the analytical operation(s) in which it is employed; and
(4) move toward methodologically sound ways to test the hypotheses
generated to explain style and/or to explain the relationship between
style and other aspects of culture.
Throughout all of the above, it is essential to keep in mind that the
concept of style gave rise to its use as an analytical tool, and therefore
to place both the concept and its subsequent deployment squarely
within the broader history of ideas. To the extent that all analytical
concepts can—indeed, must—be scrutinized both as products of a
particular moment or moments in history and within the context of a
particular set of tasks to be accomplished, the concept of style for the
art historian then takes its place with comparable analytical concepts
in the history of science.
Acknowledgments
The general issues dealt with in this paper were presented in a College Art Association
panel in 1987. Although the case studies used have not changed since then, I am most
grateful for this opportunity to reformulate the problem. I would also like to thank a
number of graduate students, now colleagues, who over the years have put their good
minds to nuanced definitions of style; many will see echoes of themselves in what is
presented here. In particular, I would cite Jülide Aker, Jack Cheng, Harry Cooper,
Marian Feldman, Elizabeth Herrmann, David Joselit, Brandon Joseph, Leslie Brown
Kessler, Michelle Marcus, Steven Nelson, Scott Redford, John Russell, Ann Shafer,
and Yuejin Wang. In addition, my thanks to Garth Isaak for help with respect to
the mathematical metaphor I was seeking, and to Robert Hunt for acute and critical
comments on an early draft.
Figure 1. Sphinx, North Syrian style. Ivory plaque, Fort Shalmaneser, Nimrud.
Iraq Museum, Baghdad.
Figure 2. Sphinx, Phoenician style. Ivory plaque, Fort Shalmaneser, Nimrud.

Iraq Museum, Baghdad.
Figure 3. Griffin-slayer, North Syrian style. Ivory plaque, Fort Shalmaneser,

Nimrud. Museum of Fine Arts, Boston.
Figure 4. Griffin-slayer, Phoenician style. Ivory plaque, Fort Shalmaneser,

Nimrud. Iraq Museum, Baghdad.
Figure 5. Protective Genius. Limestone relief from the Northwest Palace of

Assurnasirpal II, Nimrud. British Museum, London.
Figure 6. Gudea of Lagash, Diorite statue, found at Tello. Musée du Louvre,

Paris.
VIEWING (IN) THE PAST AND THE PRESENT
CHAPTER THIRTY-TWO
THE EYES HAVE IT:

VOTIVE STATUARY, GILGAMESH’S AXE, AND
CATHECTED VIEWING IN THE ANCIENT NEAR EAST
For H.B.:
zi-ša3-gal2-la šu-dagal-du11-ga
The enlarged, staring eyes of Mesopotamian votive statues (figs. 1–3)
have often been remarked upon as a characteristic stylistic feature, only
occasionally with the proposition that their function must have been
to denote attentiveness toward the presumed object of their gaze.1 Yet
it is now known that these statues were placed in shrines either seated
within their own chapels or standing in direct visual contact with the
resident deity.2 The effect is frequently enhanced by inlay: shell for the
white of the eye, lapis lazuli or bitumen for the pupil. The panoply of
assembled statues from the Square Temple at Tell Asmar in the third
millennium b.c.e., as photographed for the University of Chicago
Diyala Expedition in the 1930s, gives an almost eerie sense of absolute
and focused attention (fig. 3).3 In later periods, inscribed banded agate
* This article originally appeared as “The Eyes Have It: Votive Statuary, Gilgamesh’s
Axe, and Cathected Viewing in the Ancient Near East,” in Visuality Before and Beyond
the Renaissance, R. Nelson, ed., Cambridge: Cambridge University Press, 2000, pp.
22–44.
1
See, for example, Helen Gardner, Art Through the Ages, 6th ed. (New York: Harcourt,
Brace, Jovanovich, 1975), 51–2 and Figs. 2–10 to 2–12 and 2–18: “One notices at
once . . . how disproportionately large the eyes are. . . . The reason for this is convention.”
“The purpose of these votive figures was to offer constant prayers to the gods in behalf
of their donors, and thus their open-eyed stare may symbolize the eternal wakefulness
necessary if they are to fulfill their duty.”
2
I. J. Winter, “ ‘Idols of the King’: Consecrated images of rulers in ancient Meso-
potamia, Journal of Ritual Studies 6 (1992): 13–42.
3
The two largest images in the Diyala group, one male holding a cup, the other
female, have been considered by some scholars to be deities—due to their scale, and an
argument that the divine emblem on the base of the male figure precludes his having
been an ordinary votary—although my own opinion is that the rest of their attributes
and properties place them in the votive category (see partial discussion in A. Spycket,
La Statuaire du Proche-Orient ancien [Handbuch der Orientalistik, Abt. 7, Bd. I]) (Leiden:
E. J. Brill, 1981), 54, and in I. J. Winter, “Review of Spycket, La Statuaire,” Journal of
Cuneiform Studies 36 (1984): 106.
434 chapter thirty-two
“eye-stones” actually bear the name of, and occasionally messages

of dedication by, the donor that are literally “beamed” at the deity.4
All of this suggests a tradition in which visual attention is stressed, a
complement to the verbal attention that is contained in letter-prayers,
and even in instructions given to the statues themselves to “speak” to
the deity on behalf of the devotee.5
Less studied has been the power of the visual experience articulated
in ancient Mesopotamian texts, where the references are not merely
implicitly suggested but explictly acknowledged. Indeed, exhortations
to view an extraordinary phenomenon often precede description, as
if the intended audience were expected to see and experience the object
of the gaze as distinct from merely processing descriptive informa-
tion intellectually. Thus, the standard Neo-Assyrian version of the
Gilgamesh epic opens with instructions concerning the sacred Eanna
Temple precinct in Uruk: “Look at its wall that gleams like copper(?);
inspect its inner wall, the likes of which no one can equal!” Then the
poem further instructs: “Go up on the (city-)wall of Uruk and walk
around; examine its foundation, inspect its brickwork!”—the implicit
assumption being that the ensuing inspection will be sufficient to assure
for the city-state of the young ruler its rightful place as the preeminent
cultural center of its time.6
The direct relationship between exhortation and ekphrasis immediately
calls to mind classical Greek constructions, as in Euripides’ Ion, where
the chorus, comprised of citizens of Athens, first views, then describes
the temple of Apollo at Delphi. A series of exhortations to look alternate
4
For the “eye-stones,” see Stephen Langdon, “The Eyes of Ningal,” Revue Assyrio-
logique 20 (1923): 9–11; Wilfred G. Lambert, “An Eye-stone of Esarhaddon’s Queen
and Other Similar Gems,” Revue d’Assyriologie 63 (1969): 65–71; idem., “An Eyestone of
Ibbi-Sin,” Iraq 41 (1979): 44; B. K. Ismail, “Onyx Bead with Sumerian Votive Inscrip-
tion of Adad-apla-iddina,” Sumer 37 (1981): 114–15; and recent discussion in F. Tallon,
ed., Les pierres précieuses de l’Orient ancien (Paris: Réunion des Musées Nationaux, 1995),
72–3 re object No. 105. There are nearly 30 such eye-stones known, from the late
Neo-Sumerian period in the end of the third millennium b.c.e. to the Neo-Babylo-
nian period in the middle of the first. The brownish color of the agate with its white
center closely resembles a human eye. One of the most complete dedicatory inscrip-
tions reads: “To Baba, her mistress, for the life of Ibbi-Sin [the last ruler of the Ur
III period], Aman-ili, lady (wife) of Ir-nanna, governor of Lagash, dedicated (this
object)” (Lambert, Iraq 41: 44).
5
Discussed in Winter, Journal of Ritual Studies 6: 21.
6
The Epic of Gilgamesh, trans. Maureen G. Kovacs (Stanford, Calif.: Stanford Uni-
versity Press, 1985), 3, Tablet I, l. 12.
the eyes have it 435
with descriptions of specific aspects of the sculptural program suppos-

edly before them, as the iconographic program is revealed.7 Nor does
this event stand alone; rather, as demonstrated by Froma Zeitlin, it is
part of an identifiable genre in which sightseers come to a temple pre-
cinct or another public place and feast their eyes upon an architectural
or artifactual phenomenon.8
Vision and visuality have been extensively studied in connection with
Greek sources, ranging from specific cases, such as the erotic transfer
from beloved to representation in the Alcestis,9 to the broader “phenom-
enology of sight”.10 That this emphasis, both cultural and scholarly,
should be evident in the classical tradition will surprise no one; that
parallel concepts should have been richly developed in ancient Meso-
potamia, from Sumerian through Neo-Babylonian times, is, if not less
expected, certainly less well known, and to date has not been explicitly
studied. Yet the literary instances found in the Gilgamesh epic, no less
than those of the Ion, serve as rhetorical invitations to share experience
through a common exhortation to view. They beg the question of vari-
able responses, presuming the mutuality of sensory response in the guise
of a request to the audience to see what the author has determined
should be seen. In royal inscriptions, which constitute the largest class
of relevant documents in the Mesopotamian corpus, this presumption
takes on the role of a constitutive statement. When a palace is built,
for example, and stated to be “without equal,” it becomes so precisely
because it has been proclaimed by royal authority to be so.
7
Euripides, Ion, trans. W. S. Di Piero, with Introduction by Peter Burian [The
Greek Tragedy in New Translations] (New York and Oxford: Oxford University Press,
1996), ll. 164–225: Chorus: “Look! Images of the gods housed here. . . . Look at this!
Herakles, son of Zeus.” See also on this the study by Froma I. Zeitlin, “The Artful
Eye: Vision, Ecphrasis and Spectacle in Euripidean Theatre,” in Simon Goldhill and
Robin Osborne, eds., Art and Text in Ancient Greek Culture [Cambridge Studies in New
Art History and Criticism] (Cambridge: Cambridge University Press, 1994), 138–96,
in particular the discussion, 145, of ‘hyperviewing’: “those moments . . . [that] join word
and image in pictorial language.”
8
Zeitlin, op. cit., 148. See also, her discussion, 157 and n. 50, of Iphigenia in Aulis,
1. 190–1, 203–5, and 232–4, where, like the Ion, the act of looking carries with it an
emphasis on the emotive and aesthetic aspects of the experience: the desire to see
(190–1), the joy taken in sating one’s eyes with a visual spectacle (232–4), the particular
aesthetic stimuli that arouse an astonished gaze (203, 205).
9
See, for example, Charles Segal, “Cold Delight: Art, Death and the Transgres-
sion of Genre,” in Euripides and the Poetics of Sorrow (Durham, N.C., and London: Duke
University Press, 1993), 37–50.
10
Raymond A. Prier, Thauma Idesthai: The Phenomenology of Sight and Appearance in
Archaic Greek (Tallahassee, Fla.: The Florida State University Press, 1989).
More rarely, particularly in poetic literature, we are given a glimpse

of visual experience through direct speech. In an early second-millen-
nium b.c.e. fragment of the Gilgamesh epic dated to the Old Baby-
lonian period, for example, we are able to step outside of the official
corpus of often-formulaic royal inscriptions. Using the literary device
of direct speech, Gilgamesh relates a dream to his mother, describing
his vision of an artifact, an axe of unusual appearance. He goes on to
report: āmuršuma a tadu anāku arāmšuma kīma aššatim, “I saw it, and I felt
(such) joy; I loved it as (one would) a woman.”11 The hero then falls to
caressing the axe, declaring it his brother (i.e., his trusted companion).
Now, for purposes of the epic narrative, the dream is interpreted as
a foreshadowing of the coming of Enkidu, the semi-wild man of the
steppe who will become a companion to him; but what is important
for our purposes is the coupling of the visual experience with such a
strong emotional cathexion to the object.
It would be a truism to insist that a given aesthetic object can only
be experienced and its impact appreciated by the acts of looking and
seeing, were it not for the fact that Mesopotamian tradition reinforces
the act, not merely the object.12 That there was a well-developed
vocabulary for seeing and looking in both Sumerian and Akkadian,
the languages of ancient Mesopotamia, needs no argument. A plethora
of verbs (Sumerian igi . . . du8, igi . . . la2, igi . . . bar, igi . . . gal2; Akkadian
amāru, barû, dagālu, hâ u, na ālu, palāsu [naplusu]) cover as many nuanced
aspects as one could find in modern English: to see, behold (by eye
hold!); to regard, look at, observe, inspect; to survey, explore, examine;
11
Benjamin Foster, “Gilgamesh: Sex, Love and the Ascent of Knowledge,” in John H.
Marks and Robert M. Good, eds., Love and Death in the Ancient Near East (Guilford, Conn.:
Four Quarters Publishing Co., 1987), 27 (= Gilg. OB Pa i 31–6, esp. l. 32–3).
12
Recognition that sight is vested in the human species—indeed, in all animate spe-
cies—is, of course, self-evident. The extent to which responses to particular visual stimuli
similarly reside in the neurophysiology of the species have yet to be determined. Some
recent studies have begun to investigate these questions, dealing with both hard-wiring
responses to certain patterns, forms, and compositions, and the emotional responses (like
awe) that ensue (see, for example, Jeremy Dronfield, “Entering Alternative Realities:
Cognition, Art and Architecture in Irish Passage-Tombs,” Cambridge Archaeological Journal
6:1 [1996]: 37–72). Nevertheless, precisely because sight is common to the species, it is
the individual cultural adaptations and attendant valuations that are interesting. As with
many cultural patterns, it is important to gather cross-cultural data before one can begin
to discuss whether similarities are the product of historical interaction, independent
invention within a range of possibilities, or “universals” explainable through reference
to the neurophysiology of the species.
to stare.13 Parallel to this developed vocabulary is the number of occa-

sions in Mesopotamian literary and historical texts in which the act of
looking is described, or visual inspection is invited. These occurrences
range from love poetry and heroic epics, in which the regard of the
beloved or the display of the hero are woven into evaluative passages,
to building accounts in which the finished product is presented to an
intended audience.
The most frequent responses to an envisioned object occur in
association with human love or affection, the object then being the
beloved—sexual partner or companion. One bilingual text describes a
well-endowed young man as “one who is fitting (i.e., satisfying) to look
at” (Sum. igi.bar e2.du7 = Akk. ša ana naplusi asmu).14
In the Neo-Assyrian version of the Gilgamesh epic, when Gilgamesh
is described to Enkidu before the two actually meet, Enkidu first is
invited to “see him, look into his face” (amur šâšu u ul pānišu).15 Only then
does the passage go on to describe the hero. In the following passage,
Gilgamesh is called a “well-formed youth, bearing vitality; adorned with
allure (is) his whole body” (el ūta bani balta iši/zu’una kuzba kalu zumrišu).16
The poetics of alliteration call for axial symmetry of each line based
upon initial word sounds (vowel:b:b:vowel in the first line, z:k:k:z in
the second), which delivers a sort of litany of attributes when spoken
aloud; but even in English, one gets a strong mental image once we
have been instructed through Enkidu to see the heroic figure.
Indeed, this verbal description conforms most closely with the
rendering of the well-formed body of the Akkadian ruler Naram-Sin
(ca. 2250 b.c.e.) on his stele found at Susa (fig. 5), such that it has
13
See the bilingual tablet from Nineveh, B. Landsberger and O. R. Gurney, “igi-
du -a = tāmartu, short version,” Archiv für Orientforschung 18 (1957): 81–6, for Sumerian
and Akkadian equivalents. The Akkadian terms are discussed in the relevant entries
in the Chicago Assyrian Dictionary, henceforth CAD, and in W. von Soden, Akkadisches
Handwörterbuch (Wiesbaden: Harrassowitz, 1972), henceforth AHw. For the Sumerian,
to date, most publication series have had to include their own glossaries, but see also
brief definitions in M. L. Thomsen, Sumerian Language: An Introduction to Its History and
Grammatical Structure. (Copehagen: Akademisk Forlag, 1984), as well as the volumes of
the fledgling Philadelphia Sumerian Dictionary, A 1 and 2 and B. (Note that Sumerian
words are generally rendered in normal font; Akkadian in italics or underlined.)
14
CAD N1: 306, naplasu, “to look, glance.”
15
Gilg. Nineveh I v 15. Kovacs, Gilgamesh, 10 renders the same line: “Look at him,
gaze at his face.”
16
Nineveh I v 16–18. Kovacs, Gilgamesh, 10: “He is a handsome youth, with fresh-
ness (?), his entire body exudes voluptuousness.”
been suggested that the relationship is not accidental.17 Nor are the
descriptions of Gilgamesh and his axe the only instances of picturing
vision in the epic. In the Old Babylonian version cited above in the
episode with the axe, once Gilgamesh has found his companion, he is
made to say in direct speech: ana alka Enkidu kīma ilim tabašši, “I look
at you, Enkidu, (and) you are like a god.”18 In other words, in both
cases, where description could have sufficed (Gilgamesh is well formed;
Enkidu is like a god), the literary devices of exhortation or firsthand
report are employed in such a way as to focus upon the act of viewing
that underlies the visual assessment.
In addition to the positive view of the beloved, the lover’s regard is
itself described positively, as in an example of Sumerian love poetry
in which the lover says to his sister/beloved: igi-za igi du8-ru-na-bi
ma-dug3, “the gaze of your eye/your regard is pleasing to me.”19 The
positive regard of gods or kings, described in terms of glowing or shin-
ing (not unlike the English “beaming” at someone), is also seen as a
mark of favor: ina būnīšunu namrūtim l[ū i]ppalsūnim, “with their shining
faces, truly they looked at me”; and rulers are flattered by courtiers
who speak of their desire to enter before the king “in order to see
his gracious face,” ana dagāli panišu damqūta.20 Verbs of seeing can also
be stressed, so that individuals are invited to stare/look hard at, not
merely see, particular phenomena, as in one twelfth-century bilingual
Babylonian text concerning the image of a god: ibtarrâ niš māti lānšu
elâ, “[ T ]he people of the land stared at his [i.e. the deity’s] tall/lofty
figure.”21 Thus, the relationship between being on view and appreciative
17
I. J. Winter, “Sex, Rhetoric, and the Public Monument: The Alluring Body of
Naram-Sîn of Agade,” in Natalie B. Kampen, ed., Sexuality in Ancient Art (Cambridge:
Cambridge University Press, 1996), 11–26, esp. 18–19. Note also that on the whole
monument, all eyes are on the ruler—both the soldiers that follow him over ascend-
ing terrain and the enemy whose necks bend at increasing angles the farther they are
down the right-hand margin, in order to keep the victorious Naram-Sîn in their line
of vision.
18
Gilg. P. ii 11, cited in CAD N2: 122, na ālu; see also, discussion in Foster, “Gil-
gamesh,” 21–42.
19
Bendt Alster, “Sumerian Love Songs,” Revue d’Assyriologie 79 (1985): 143–4 (=
SRT 31:4).
20
See Amarna Letter, EA 151: 18, cited CAD D: 21, dagālu, and an Old Babylonian
bilingual of the reign of Samsuiluna (ca. 1749–1712 b.c.e.), the Sumerian version
reading: igi.zalag.ga.ne.ne.a u.mu.ši.in.bar.re.eš—cited CAD B: 320, būnu A.
21
Bilingual text of Nebuchadnezzar I of Babylon (1124–1103 b.c.e.), referring to
an image of the god Marduk, cited CAD B: 115, barû, and published most recently
by Grant Frame, Rulers of Babylonia from the Second Dynasty of Isin to the End of Assyrian
Domination [Royal Inscriptions of Mesopotamia: Babylonian Periods, vol. 2 (henceforth, RIMB 2)]
visual assessment is well established with respect to persons and imputed

to deities in the literature, and indeed, the conclusion of the line of
the Babylonian text just cited is that, the people once having seen the
god’s lofty (and fitting, and majestic, and bright) image, they all paid
attention to him—that is, they focused upon the god himself, as well as
cathecting to the image. The aesthetic aspects of that concentration
of emotional energy are demonstrated through the qualities of form
and stature selected for and described in conjunction with favorable
viewing; the importance of visual experience is indicated through being
acknowledged as the necessary pathway to engagement.
This emphasis on vision and the subsequent process of assessment
through focused viewing can also be demonstrated with respect to works
such as temples, palaces, and precious objects. The Old Babylonian
Gilgamesh fragment cited clearly illustrates the degree to which visual
affect was acknowledged in conjunction with inanimate objects, when
the hero’s dream-axe is described as having elicited such a powerful
response. Since the majority of other textual references come from
the official corpus of royal documents, with their own propagandis-
tic agendas, the visual attributes recorded for royal undertakings are
generally given positive value, and in such cases, the response to the
specific visual stimulus is expected to be admiring. Nevertheless, even
through these carefully managed texts, it is possible to glean a sense of
the importance of seeing, and thereby the importance of audience, for
works we would call “art,” as well as for animate objects of desire.
Often the gods are the selected audience, particularly in royal inscrip-
tions where the building of temples is recorded. When the Middle Assyr-
ian ruler Shalmaneser I (1274–1245 b.c.e.) rebuilds the Assur temple
after the destruction of an earlier one by fire and the god is brought
in to take up residence in the new temple, the ruler describes his hope
that his work will bring satisfaction in terms of the god’s first seeing
and then being pleased with his new abode: epšēti nimurti E2 šâtu līmurma
li dâ, “[M]ay he (the god Assur) see the bright/shining works of that
(Toronto: The University of Toronto Press, 1995), 30 (= Neb I. B.2.4.9, lines 15–16).
The Sumerian un.ma.da igi.kar2.kar2.ra.ab e2.gar8.bi sukud.da duplicates the verb, as
igi.kar2.kar2, but this could be either a function of intensification or of response to the
plurality implied in the people of the land, un.ma.da. The Akkadian is clear, however:
The verb barû is conjugated in a form that provides the same intensification of the
simple verb for seeing as “to smash” would be for the simple verb “to break.”
temple and rejoice.”22 The Neo-Assyrian ruler Esarhaddon (680–669

b.c.e.) also consistently concludes his temple-building accounts with
the hope that the deity involved may “look joyously” upon the king’s
work.23 In one of the more complete accounts, Esarhaddon first records
the installation of the god Assur in his temple, the placement of other
deities and various animal attributes around him, and the establishment
of regular cult offerings. Then the god’s reaction is provided, not as
an aspiration but as a matter of record: epšētia damqāte kēniš ippalisma
ēli libbašu, “When he (Assur) looked steadfastly upon my good works,
his heart rejoiced.”24
Deities are not the only audience whose gaze is invoked. On the
colossal gateway lion from the Sharrat Nip i temple in Nimrud, Assur-
na irpal II (883–859 b.c.e.) records that he built the temple “for the
eternal regard of rulers and princes,” ana nan[mar m]alkī.MEŠ u3 NUN.
me ša dārâte ēpuš, and the same construction is used on blocks from
the king’s Northwest Palace at Nimrud.25 It may well be the case that
the public location of the lion inscription and the palace orthostats
accounts for the particular audience addressed—those rulers who will
enter the temple or palace in later years as equals and be in a position
to restore them, as expressed explicitly toward the end of the Sharrat
Nip i inscription.26 However, in the Northwest Palace inscription, no
specific instructions for repairs and maintenance in the future follow;
rather, after being invited to gaze, the future ruler is simply told that
22
A. Kirk Grayson, Royal Inscriptions of Mesopotamia: Assyrian Period, vol. 1 [henceforth,
RIMA 1] (Toronto: University of Toronto Press, 1987), 185 (= Shalm. I.1.148–51).
23
For example, R. Borger, Die Inschriften Asarhaddons, Königs von Assyrien [henceforth
Es.] [AfO Beiheft 9] (Osnabrück: Biblio Verlag, 1967), ¶20, Borsippa A 20, regarding
the temple of the goddess Gula in Borsippa, also now published in Frame, RIMB 2,
74 (= Es. B.6.31.10, l. 20). The text relates that the king restored its ruins, then closes:
“[M]ay [Gula] look upon this work joyously” (i.e., with pleasure), [šip]ru(?) šuāti adīš
lippa[lisma]. Indeed, this formulation is typical of all of Esarhaddon’s temple building
and restoration in Babylon: See, for example. Frame, op. cit., 177, temple for the god-
dess Queen-of-Nippur; 179, for the god Enlil; 184, for the goddess Ishtar. This was
apparently formulaic, as it is used in Babylon by local rulers as well as by Assyrians:
e.g., Marduk-apla-iddina II (Merodach-Baladan II, 721–710 b.c.e.), who requests of
the goddess Ishtar that when she “looks joyously upon” the work he has done, she may
bestow a long life upon the king, in Frame, RIMB 2, 138 (= B.6.21.a, l. 30–1).
24
Borger, Es. 5–6 = ¶2, Assur A vii 18–20.
25
A. Kirk Grayson, The Royal Inscriptions of Mesopotamia: Assyrian Period, vol. 2 [hence-
forth, RIMA 2] (Toronto: University of Toronto Press, 1991), 296 = ANP II.32.10–11;
Northwest Palace = RIMA 2, 301–2 (= ANP II.35.8).
26
RIMA 2, 297 (= ANP II.32.15–18).
the king made the palace fitting and splendid “for eternity,” ana dārâte,
installing colossal beasts of limestone and alabaster in its doorways.27
What these last citations provide is both the expansion of the audience
beyond the gods to future generations of rulers, and also the juxtaposi-
tion of spectacular works to the assertion that the palace has been built
to be seen for eternity.
A text of the Neo-Babylonian ruler Nebuchadnezzar II (604–562
b.c.e.) goes even further, linking the viewing of luxurious appointments
with admiration and stating its intended audience to be all peoples,
not merely kings and princes: bīta šâtim ana tabrâti ušēpišma ana dagālim
kiššat nišē lulê ušmallīša, “I had that temple built for admiration, (and)
for the viewing/regard of all the people I filled it with delight(s).”28 We
shall return to the particular goal of provoking admiration, but for the
present, the linkage of viewing, spectacle (in this case, the temple), and
response is clearly established, as is the range of audience, from god,
to future rulers, to “the people.”
The exhortation to view works that by our definition would be
included in the category of “art,” and/or the aspiration to have such
works viewed (and admired), extends to goods as well as buildings, and
one can find a similar range of audiences invoked. In a Neo-Babylonian
text of Nabonidus (555–539 b.c.e.), for example, the god Shamash is
enjoined to “look joyously upon the precious products of the king’s
hands, his auspicious works,” lipitti qātīa šūquru epšētūa damqāta . . . adīš
naplisamma, and amongst these, the “royal image” is specifically men-
tioned.29 In other words, the gods are specifically invited to look upon
works in which there has been artistic investment and respond joyfully,
that is, favorably.
In a text of Assurnasirpal I (1049–1030 b.c.e.), an ornate bed made
for the inner chamber of the temple of the goddess Ishtar is elaborately
described. The king tells us that the bed has been made of taskarinnu-
wood (boxwood?) and gold, adorned with precious stones, and as a
result, it shines “like the rays of the sun(god).” He then closes with the
assessment: “for viewing it is suited,” an[a n]a āli asmat, a construction
not unlike that used for the well-formed youth ša ana naplusi asmu, one
27
RIMA 2, 302 (= ANP II.33.9–10): ussimši ušarri ši umami KUR-e u A.AB.BA.MEŠ
DU3-šunu ša NA4 puli BABBAR-e u NA4 parūte DU3 ina KA2.MEŠ-ša ušziz.
28
Cited CAD D: 22, dagālu (= VAB 4 118 ii 53).
29
Cited CAD B: 320, būnu (= VAB 4 258 ii 21).
who is “suited to being looked at,” that is, worth seeing.30 In a Neo-
Babylonian text some 500 years later, Nebuchadnezzar II also speaks of
having deposited within a temple “everything that is (worthy of ) being
looked at”: mimma ša inna alu.31 In neither of these texts is the audi-
ence who will do the looking specified. Rather, it is the fact of being
see-worthy that is emphasized, as if this in itself were a sufficient state-
ment of the work’s impact and value. The Assurnasirpal I text, being
the more complete, actually provides us with the entire sequence: from
construction through completion to response. First, the act of making
the bed and its decoration is given attention; then, once completed,
the bed shines like the sun, as a positive attribute; and finally, as such,
it is worthy of regard.
Similarly, in a text of Shalmaneser III of Assyria (858–824 b.c.e.),
the king records a statue ša epšētušu ana dagāli lullâ šūturū bunnanûšu,
“whose workmanship is pleasurable to behold, whose appearance is
extraordinary.”32 Here, too, issues of facture (its workmanship) and
representation (its extraordinary appearance) are articulated, and both
are linked to viewing. But perhaps even more important: if through a
combination of workmanship and visual attributes value is achieved,
it is also the case that through seeing, value is perceived.
The Shalmaneser text further adds the emotional response of pleasure
that comes as a result of seeing something satisfying. Surely, here, we
come close to developed Western and South Asian notions of aesthet-
ics as engaging sensory delight, not merely intellection. This is further
attested in instances where the Assyrian kings project divine response
to their acts of pious patronage. For the gods in question, it is quite
explicit that seeing is the means by which satisfaction is to be ascertained.
In one text, for example, Esarhaddon projects a goddess’s response
upon seeing his restoration work in Babylonia, when he describes the
goddess Nanâ as “joyous in her regard (of ) that work,” šipir šâšu dNanâ
hadīš ina naplusiša.33 The same mechanism is articulated as Esarhaddon
describes his rebuilding of the temple of the goddess Gula; however,
there he expresses the hope that “she may then look joyfully upon this
30
W. von Soden, “Zwei Königsgebete an Ištar aus Assyrien,” Archiv für Orientforschung
25 (1974–7): 37–49; cited also CAD N2: 124, na alu.
31
Cited CAD N2: 128, na alu (= PBS 15 79 i 40 and duplicate CT 37 8 i 39 [ Nbk]).
32
Cited CAD B: 318, bunnanû, from the Kurba’il text of Shalmaneser III, originally
published J. V. Kinnier-Wilson, “The Kurba’il Statue of Shalmaneser III,” Iraq 24
(1962): 94, 1.37–8.
33
Borger, Es., 77 (= ¶50, Uruk D 14–18).
work.”34 Similarly, his son, Assurbanipal (668–627 b.c.e.) requests the

goddess Ninma to “look joyfully upon my auspicious works,” epšētia
damqāti hadīš lippalis.35
This formulation goes back at least into the second millennium b.c.e.,
as is evident from the statement of Shalmaneser I that concludes his
description of the rebuilding of the Assur temple, earlier destroyed by
fire. The full passage reads: “When Assur, the lord, enters that temple
and joyfully takes his place on the lofty dais, may he see the shining
workmanship of that temple and rejoice,” enūma dAššur EN ana E2 šâtu
ibâ’uma BARA2-šu īra adīš irammû epšēti nimruti E2 šâtu līmurma li dâ.36
As this study represents an initial attempt to lay out the sources for
the Mesopotamian visual experience, I have selected from a broad
range of texts, both chronological and typological. Future research may
demonstrate different attitudes toward the nature of vision and viewing,
depending upon whether sources tend to be so-called historical (i.e.,
royal) texts or literary texts, such as epics and love poetry. One could
argue that the appreciative response desired in royal texts is largely rhe-
torical. Indeed, there are instances when the intended quid pro quo is
made quite explicit, as when Esarhaddon expresses the hope that when
the goddess Nanâ is “joyous in her viewing” of the restoration work
he has done on her cult sanctuary, adīš ina naplusiša, she might put in
a good word for him before the god Nabû.37 On another occasion, he
addresses the same goddess directly, saying: “if/when you are joyous in
your dwelling,” adiš ina ašābiki, and then proceeds to list what benefits
might come to him—that she might speak well of him, establish a good
fate for him, and so forth.38 Nevertheless, the status accorded the gaze
of the divine beneficiary, as the necessary act of perception prior to
affirmation and benevolent patronage, reinforces the power of visual
cathexion in Mesopotamian tradition.
Yet another type of visual response is consistently recorded: that
of “the people” to a major work. Whereas the reaction of “joy” is
ascribed to gods and rulers, who are the direct beneficiaries of these
works, which by our definition we call “art,” lower-echelon viewers are
34
Borger, Es., 32 and Frame, RIMB 2, 177 (= ¶20, Borsippa A 20), cited in n. 21.
35
Cited CAD H: 23, adīš (= Streck ABP.240.15).
36
Grayson, RIMA 1, 185 (= Shalmaneser I.1.148–51).
37
Borger, Es., 75–6 (= ¶48, Uruk B) and Frame, RIMB 2, 189, l. 18–20, concerning
the restoration of the E iliana, the cella of the goddess Nonaia at Uruk.
38
Borger, Es., 76 and Frame, RIMB 2, 187, l. 16–19.
said, rather, to respond with “admiration” and, on occasion, with “awe”

(Sum. u6-di and ni2-me-gar; Akk. tabrītu and qalu). In a Sumerian text
recording the dedication of a cultic vessel on behalf of Rim-Sin of
Larsa (ca. 1822–1763 b.c.e.), for example, the sacred gate of the great
courtyard of a temple to Inanna is called the “place of ad+miration
of the land,” ki-u6-di-kalam-ma-ka.39
I have broken the word admiration apart in the last translation in
order to conform to its Latin prefix-plus-root, because it preserves
exactly the relationship of viewing/seeing (mirāri ) plus enhancement
(ad-) which, when joined together, signify the same intensified sensor
activity as does the Sumerian u6-di and its Akkadian equivalent, tabrītu
(to be discussed)—that is, augmented viewing leading to positive response. On
its own, the cuneiform sign for u6 includes the ideographic sign for eye,
igi. As a word, the Sumerian u6 has lexical equivalents to the Akkadian
verbs of seeing, barû, na ālu and naplusu.40 Indeed, Thorkild Jacobsen has
shown that the base meaning of u6 is the act of looking, and the verbal
construction u6-e or u6-di is precisely the imperfective (i.e., ongoing) act
of looking; so that the actual denotation of the u6-di-response, that of
“ad+miration,” is predicated precisely upon continued looking.41
The Sumerian term u6-di has often been translated by modern
scholars as “wonder,” “marvel,” “astonishment,” thereby correspond-
ing well to the Greek thauma idesthai, that “amazed stupor produced by
seeing,” or “looking in wonder and astonishment,” so well documented
in classical sources.42 I believe the Sumerian does suggest all of these
things in the lexicon of experience, and for a poetic translation, a
single word like “wonder” would render the Sumerian (or Akkadian)
well. But I would argue that for purposes of cultural analysis, only the
more literal “ad+miration” captures the relationship between the visual
spectacle and the spectator’s response that is primary in the Sumerian and
subsequent Akkadian/Neo-Assyrian usage.
39
Douglas Frayne, Royal Inscriptions of Mesopotamia, Early Periods, vol. 4 [henceforth,
RIME 4] (Toronto: University of Toronto Press, 1991), 303 (= Rim-Sin I.23.27–9).
40
See Hermann Behrens and Horst Steible, Glossar zu dem altsumerischen Bau- und
Weihinschriften [ Freiburger altorientalische Studien 6] (Wiesbaden: Steiner Verlag, 1983),
349, citing MSL XIII 168, 14’–16’.
41
Thorkild Jacobsen, “Lugalbanda and Ninsuna,” JCS 41 (1989): 78 re u6-dug4 =
hâtu, and its imperfective, u6-e/di. Jacobsen reads Enmerkar 63: “may the people look
at me admiringly,” [nam]-lu2-lu6 u6 du10-ge-eš u-mu-un-e, and the bilingual text 4R
28 2:7, “you looked not at your house,” e2-zu u6 li-bi2-du3 = bītka ul tamur, where the
Akkadian, from the verb amāru, “to see,” is very clear.
42
Prier, Thauma Idesthai, especially 81–97.
It seems not irrelevant, therefore, that the components making up the

Sumerian sign u6 are precisely those for eye, igi (as noted) and house,
e2. Although one greets most psychophilology with justified scepticism,
in the present case the resultant ideogram, eye+house, seems to picture
exactly the spectatorial role here being described, and physically articu-
late the visual underpinning of the reaction: ad+miration.
Deities and rulers can on occasion also evoke this response. The
goddess Inanna, in a late third millennium hymn, is called nin u6-di-
kalam-ma, “lady, [object of] admiration of the land.”43 Shulgi of Ur
(2094–2047 b.c.e.) is described in a royal hymn as su6-za-gin3 gaba-
ku3-ga u6-di, “(king with a) dark lustrous beard, a pure/shining breast,
a sight (to behold),” and implicitly, therefore, admirable, a wonder, a
marvel.44 And later on in the same hymn, the king is addressed directly:
mes-zi kurun3-na-gun3-a-gim u6-di-du10-ga-me-en3, “like a true mes-tree,
sparkling with irridescent fruits, you are a sweet sight.”45
The description of Shulgi is not unlike that given for Dumuzi, consort
of the goddess Inanna, in one of the Sumerian temple hymns, prob-
ably of the same period: once again, “he of the pure/shining breast,
a sight [to behold],” u6-di.46 Sjöberg, in his edition of the hymns, has
translated the phrase “marvelous to behold,” and indeed in context, this
translation serves the sense quite well. In others of the temple hymns,
the temples themselves, not their deities, are referred to in the same
terms. The temple of Inanna in Kullaba is said to be u6-di i-li gur3-
ru, literally, “admirable, full of allure.”47 And the temple of Numušda
in Kazallu is presented as u6-e gub-ba, “set up (or established) for
[unending] admiration.”48 Finally, in the hymn recording the building
of the Ekur in Nippur by Shulgi’s father, Ur-Namma, respondents are
introduced. We are told specifically that the king established it (the
temple) “for the admiration of the multitudes,” u6-di-bi-še3 un-sar2-
sar2-ra-ba ši-im-ma-gub.49
43
Royal hymn of the Isin period, cited by Françoise Bruschweiler, Inannna: La déesse
triomphante et vaincue dans la cosmologie sumérienne (Geneva: Université de Genève, 1987),
33, 117, 120.
44
Jacob Klein, Three Šulgi Hymns (Ramat Gan: Bar Ilan University Press, 1981), 72
(= Shulgi D.7).
45
Ibid., 72 (= Shulgi D.33).
46
Åke W. Sjöberg, Sumerian Temple Hymns (Locust Valley, N.Y.: Augustin, 1969), 30
(= TH 17:217), on the temple of Dumuzi in Badtibira.
47
Ibid., 29 (= TH 16: 199), translated by Sjöberg as “marvelous.”
48
Ibid., 40 (= TH 31:397).
49
Jacob Klein, “Building and Dedication Hymns in Sumerian Literature,” Acta
Sumerologica II (1989): 48, 51 (= Ur-N B.21). Klein renders the same line: “He displayed
it for wonder among the multitudes of the people.”
Nowhere is the sense of visual effect clearer than in the several refer-
ences by Gudea of Lagash (ca. 2127–2110 b.c.e.) to his Eninnu temple,
built for the god Ningirsu in the city of Girsu. Gudea’s text represents
the most eloquent account of a Sumerian temple thus far preserved.
Statements of the temple’s admirable qualities are repeated throughout
Gudea’s lengthy account of its building and consecration. In the first
part, related to the building of the temple and known as Cylinder A, it
is stated on more than one occasion that the temple was established to
be admired, u6-di-de3 ba-gub.50 In the final summary description-cum-
hymn to the temple at the end of that cylinder, we are told:
29: 14 e2-a ni2-gal-bi The temple, its awesome radiance
15 kalam-ma mu-ri was cast over the land;
16 ka-tar-ra-bi its praise-song
17 kur-re ba-ti . . . reached the mountains . . .
30: 11 i-li gur3-a [the temple], bearing allure,
12 u6-di-de3 gub-ba is established to be admired.
It is of particular interest that the closing point in the passage, which
stands for the affect of the temple as a whole, is its nature as an object
worthy of what we have been calling “ad+miration.” Thus, the implica-
tion in modern language, which follows clearly from the context, is that
once established, the result is nothing short of astonishing, marvelous,
wondrous.
This is echoed further in Gudea’s continuing account of his temple,
known as Cylinder B, which describes the temple’s consecration. After
another laudatory description, we are told that not only the people but
even the generic Anunna-gods “made their way to view (i.e., admire)
it”: da-nun-na u6-di-de3 im-ma-šu4-šu4-ge-eš2.51 That these undifferenti-
ated, lower-level deities have journeyed to the temple and then respond
positively urges us to understand that even if the underlying sense of
the term u6-di is “to see,” the implication of that seeing is a positive
loading of the experience, hence “ad+miration.”
50
Gudea Cyl. A 24: 17 and 24: 25; see the text edition in D. O. Edzard, Gudea and
his Dynasty [Royal Inscriptions of Mesopotamia, Early Periods, vol. 3/1] (Toronto: University
of Toronto Press, 1997, 84). A translation of the complete cylinder also appears in
Thorkild Jacobsen, The Harps that Once: Sumerian Poetry in Translation (New Haven, Conn.,
& London: Yale University Press, 1987), 386–425 (this line = 419).
51
Cyl. B 1: 11; Edzard, RIME 3/1, 89 and Jacobsen, Harps, 426 for this line.
The sense of semantic loading is also clearly implied in the one

negative case in the literature, the text known as the “Lamentation over
Sumer and Ur,” in which the temple of Nanna at Ur, the god’s holy
dwelling, described as once fragrant with cedar and decorated with
gold, silver, and lapis lazuli, has been destroyed. Using the construction
we have been considering, the text then states: “[ T ]he house/temple,
the admiration of which was [once] so good, that admiration is (now)
destroyed.” e2 u6-di-bi i3-du10-ga-ri u6-di-bi ba-gul.52 Once again we
see the same pattern of composition: After the affective attributes are
presented, the summary statement relates to the visual response the
temple had evoked. The context makes clear that the previous visually
based admiration had been charged with the positive (du10, “goodness,”
“sweetness”), and the loss is to be lamented. In the negative, we are
able to see its opposite even more clearly—just how much the visual
experience had previously been charged with positive affect.
Individual parts of temples, along with cult objects in which great
care was invested, are also discussed in terms of their power to elicit
admiration. The main courtyard of the Shamash temple in Larsa, the
Ebabbar, in a text of the ruler Siniddinam (1849–1843 b.c.e.), was called
the “place of admiration of the land,” k[isa]l-ma eš3 e2-babbar-ra/
ki-u6-di-kalam-ma-ka.53 This is essentially the same phrase as that used
for the courtyard of the Eme’urur temple in Larsa when a precious cult
vessel for Inanna was set up there in honor of the ruler Rim-Sin.54
As for objects, a cultic šakan-vessel made by Ibbi-Su en of Ur
(2028–2004 b.c.e.), artfully formed and decorated with wild bulls and
snakes, is celebrated as a work “the admiration of whose adornment
is without end,” še-er-ga-an-du11-ga-bi u6-di nu-til-le-dam.55 A chariot
for the god Enlil is further eulogized in direct speech by the ruler Išme-
Dagan of Isin (1953–1935 b.c.e.): “[Y]ou are admirable [i.e., a thing
of wonder] to behold, your appointments are outstanding,” nig2-u6-di
52
Cited Pennsylvania Sumerian Dictionary (henceforth PSD) AI: 163, a-sal-bar
(= Lamentation over Sumer and Ur 426). See also P. Michalowski, The Lamentation
over the Destruction of Sumer and Ur (Winona Lake, Ind.: Eisenbraun’s, 1989), 62–3, who
renders the line rather differently, but retains the sense of the visual through his use of
“admired” and “admiration” in English: “The admired temple that used (to receive)
first class oil, its admiration was extinguished.”
53
Frayne, RIME 4, 158 (= Sin-iddinam 1.36–7).
54
Ibid., 303 (= Rim Sin 23: 23–9).
55
H. Steible, Die neusumerischen Bau- und Weihinschriften, Freiburger altorientalische
Studien 9 (Stuttgart: Steiner Verlag, 1991), Ibbi Su’en 9: 17–28.
gal2-me-en giš.šu-kar2-zu dalla ma im-e3.56 These last citations make

clear that not only buildings but also precious objects can elicit responses
of admiration. In addition, the Išme-Dagan text makes a substantive
out of the combination of nig2 (thing) and u6-di, such that we have
before us a “thing of admiration [or wonder].”
The Akkadian version of the same phenomenon is an exact lexi-
cal equivalent of the Sumerian, not infrequently actually written with
Sumerian logograms.57 Thus, for example, the Sumerian u6-di un-sar2-ra
is the equivalent of Akkadian tabrât kiššat nišē, for the “admiration of
all the people,” with the Akkadian word kiššatu actually written with
the Sumerian sign sar2 (and nišē often with the sign for un, “people”).
What is more, the word tabrītu is a nominal formation from the verb
barû, “to see,” comparable to the visual component incorporated into
the Sumerian logogram for u6. Usage follows a similar pattern as well.
Both Sennacherib (704–681 b.c.e.) and Esarhaddon of Assyria make
frequent use of this formulation for their palace inscriptions—Sennach-
erib also singling out his great gateway colossi as objects of admiration,
if not astonishment.58 Esarhaddon, in creating the palace of the crown
prince for his son Assurbanipal, further amplifies the phrase by add-
ing that it was ana tabrât kiššat nišē MES lulê ušmalli, “for the ad+miration
[i.e., the admiring gaze] of all the people, I filled it with delight(s).”59
In a text cited earlier with respect to public viewing, Nebuchadnezzar
of Babylon makes it even more explicit. By doubling the verbs related
to sight, he emphasizes the visual experience even further: bīta šâtim ana
tabrâti ušēpišma ana dagālum kiššat nišē lulê ušmalliša, “I had that temple
built for ad+miration; for the regard of all the people, I filled it with
delight(s).”60 In other words, to provoke admiration is a motive in his
doing, and to that end, he filled the temple with delight(s) so that it
would be (perpetually) gazed at by all the people.
56
Klein, Acta Sumerologica 11, 37–41 (= Išmedagan I.9–10).
57
See von Soden, AHw, 1299: tabrītu.
58
See discussion in P. Garelli. “La Conception de la beauté en Assyrie,” in Tzvi
Abusch et al., eds., Lingering Over Words: Studies in Ancient Near Eastern Literature in Honor
of William L. Moran (Atlanta, Ga.: The Scholar’s Press, 1990), 175.
59
Borger, Es., 72 (= ¶43, Tarbisu A 43 30). Esarhaddon adds a variation as well,
in his account of the stele he set up in Sam’al, a province to the west. On the stele
itself he declares that it was set up ana tabrât kiššat nakirī, “for the wondering gaze of
all enemies!”
60
See n. 28.
Nabopolassar (625–605 b.c.e.) adds another piece in the argument

of intense admiration-stimulated-by-viewing, when, in speaking of his
work on the temple of Marduk, he also includes the sense of “fitting”
as a positive quality of the finished building: ana tabrâtim lū ušāsimšu, “for
(the) admiration (of the onlookers) I truly caused it to be suitable/fit-
ting.”61 Both the quality of being “fitting” and the admiration that that
provokes are part of the summary statement of the temple. As with
the accounts of Sumerian temples already cited, the statement that a
work is for, or provokes, admiration is generally the closing affirmation
of the work in the texts.
Paul Garelli has suggested that those palaces and other works that
provoke admiration do so through the great effort expended and the
precious materials utilised.62 From a survey of all instances where the
construction is used, his criteria can be extended to include a wide
variety of properties related to material, skill, and scale, plus attributes
of form and artful manufacture and the presence of what to us seems
ineffable, like radiance.63 But I would argue that the sum of all of the
properties is the ability to evoke the u6-di/tabrītu response of intense
“ad+miration.”
As noted, when the texts specify those who respond to works in this
way, they are neither the high gods nor rulers. The case of the Anunna-
gods in Gudea’s Cylinder B can be understood as hyperbole that reaches
beyond the temple’s earthly spectators; but since the Anunna-gods are
a class of undifferentiated lower divinities, not the named gods of the
pantheon, the descriptive terminology is still restricted to the general
audience, as distinct from direct beneficiaries of the work. The only
instance I have found in which a high god is so engaged is a reference
in one of the Shulgi hymns of the Third Dynasty of Ur (late third
millennium B.C.E.) to the goddess Inanna, who looks in ad+miration as
the king enters her shrine.64 This latter instance can best be understood
if Shulgi’s entrance into the shrine is viewed in conjunction with his
performance of the “Sacred Marriage” rite, in which the bride Inanna
receives her bridegroom; it is then specific to the context of the royal
61
Cited CAD A2: 329, asāmu in the causative Š-stem, šūsumu (= VAB 4 64 iii 30),
“to make fitting.”
62
Garelli, “La Conception de la beauté,” in Abusch et al., eds., Lingering Over Words,
175–6.
63
This will be developed further in a study under preparation.
64
See Klein, Three Shulgi Hymns, 136 (= Šulgi X.11).
hymn, where the lover (the goddess) may be said to be overwhelmed

by the sight of the beloved (the ruler). What is certain is that to date,
no responses of the high deities to cultic objects or temples made/built
on their behalf have been preserved describing their experience of the
works in terms of “wonder,” or “ad+miration,” u6-di or tabrītu; rather,
theirs is a response of “joy,” or “delight.”
I believe the identification of respondents and their differential
responses to be a not-trivial distinction. Gods, or kings, constitute the
primary recipients of building/work projects. It is with them in mind
that the projects have been undertaken; they are in effect the intended
beneficiaries of the undertaking, and thus, active recipients of the work.
Their response is therefore described in terms of direct gratification:
joy, pleasure, delight.65 By contrast, when “the people,” or the minor
deities like the Annunaki reported by Gudea, stand before a major
work, they constitute a more passive audience: not the project’s direct
beneficiaries, but its onlookers. They react to the qualities and overall
affect of the particular enterprise as spectators; and as such, they are
neither proprietary nor directly gratified; they are instead impressed, if
not overwhelmed. Both responses—joy/delight and ad+miration—are
grounded in the act of seeing; however, for the gods, seeing is the point
of departure, the trigger for an emotional response in anticipation of
active engagement, whereas for the populace, the end experience is
the act of seeing itself, intensified as the powerful emotional reaction
of admiration.
A less frequently used formula actually articulates a response com-
prised not only of admiration but also of awe so powerful that it reduces
respondents to a state of combined jubilation-cum-speechlessness (Sum.
nig2-me-gar; Akk. qâlu).66 In this way do the people respond to the
goddess Inanna;67 but even more interesting is the combination u6-di-
65
The one exception to this with which I am familiar is a royal statement that
Esarhaddon’s workmen respond to the completion of their task in building the royal
palace with joy and jubilation. Here, however, it could be said that the rhetorical nature
of the text has governed the report as a self-reflexive statement of celebration.
66
CAD Q: 72, qâlu A, “to become silent, to stay quiet; to pay attention (with the
sense of a focus of attention).” The bilingual equivalent is found in Irving L. Finkel,
The Series SIG7 . ALAN = Nabnītu [Materials for the Sumerian Lexicon 16] (Rome: Pontifi-
cium Institutum Biblicum, 1982), Tablet III.168–73 and IV.232: nig2-me-gar = qâlum,
rīšātum, resp.
67
See Bruschweiler, Inanna, 117, citing the text “The Exaltation of Inanna,”
nin-me-šar2-ra, l. 21–2. This text has been reedited by Annette Zgoll, Der Rechtsfall
nig2-me-gar-sag-gi6-ga, “the admiration and awe of the black-headed

people (the Sumerians),” in response to the god Enlil.68
The two are combined again in a description of the cultic ma-gur8-
boat of Shulgi approaching the quay, “a thing of awe, its admira-
tion does not end,” nig2-me-gar-am3 u6-di-bi nu-til3-le.69 Sjöberg has
described nig2-me-gar as an unsounded intake of breath. It is an
immediate and physiological response to the experience of power,
as is clear in the hymn to the Ulmash temple of Inanna, where it is
the temple’s affect that makes awed silence fall upon (?) enemy lands,
nig2-me-gar ki-bala-a du7-du7.70 Sjöberg has also commented upon
the close connection of nig2-me-gar with u6-di in the sense of a joined
awe or anticipation plus ad+miration.71 Together they heighten the
intensity of the viewer/reactor’s response. The entity, god or work, pro-
voking awe and admiration, is then something sufficiently powerful and/
or impressive that it is experienced with extremely strong emotion.72
I find myself tending to think of the ad+miration/awe responses
in terms of a sort of WOW! effect on the people, parallel to, but dif-
ferent from, the joy response of the gods. It is as if the impact of the
work when viewed by the people is perceived as impressive, if not
overwhelming, whereas when viewed by the gods, their own inherent
power matches that of the work, and so their response is seen not in
der Enhedu-Ana im Lied nin-me-šara [Alter Orient und Altes Testament 246) (Münster:
Ugarit-Verlag, 1997).
68
From the text known as “Enlil in the Ekur,” l. 129—the text edition made available
to me by the editors from the files of the PSD, gratefully acknowledged.
69
From the text known as “Inanna & Bilulu,” 4 rev. I: 14—see S. Langdon, Historical
and Religious texts . . . Nippur [Babylonian Expedition of the University of Pennsylvania, Series A,
vol. 31] (Munich, 1914).
70
Sjöberg, Sumerian Temple Hymns, 47 (= TH 40: 510).
71
Ibid., 144, discussion to Temple Hymn 40.
72
I have, in fact, been present at a moment of such collective response, in the
early-morning opening (darśan) of the shrine of the ruby image of Siva Natarāja in
the temple of Cidambaram in South India. As is well known in Hindu tradition, the
reciprocal viewing of deity and devotee constitutes an important aspect of worship—on
which, see Diana Eck, Darśan: Seeing the Divine Image in India, 2d ed. (Chambersburg,
Pa: Anima, 1985). At Cidambaram, in the period before the god was to be on view,
an increasingly anticipatory silence pervaded the assembled devotees, and then, at the
moment when the shrine was opened, literally everyone simultaneously drew in a deep
breath as he/she experienced the powerful visual impact of the god’s presence. That something
similar was intended in the Mesopotamian context is strongly suggested by a Babylonian
text that refers to “seeing the god’s face” as a euphemism for the cultic manifestation
of the deity through his image: aššum muppalsāta ātamar pānika, on which, see W. Mayer,
Untersuchungen zur Formensprache der babylonischen ‘Gebetsbeschwörungen’ [Studia Pohl, Series
Maior 5] (Rome, 1976), 478–81, re l. 15–18.
terms of being overwhelmed but rather well met. What to the gods
produces delight, to those on a lower level evokes admiration and awe;
and therefore, vision must be understood not as an absolute but as a
social phenomenon.
When seen in this light, it is further possible to understand the use
of the terms u6-di/tabrītu in so many royal texts as part of the ideologi-
cal apparatus of the state. It is a way, after all, of calling attention to
the investment in and the merit to be obtained from a given work or
project, by claiming for it a powerful visual affect and/or response.
I would suggest that this is why so many official texts conclude their
descriptions of important projects with the summary statement of the
work’s ability to elicit admiration, or “wonder”—the ultimate affirma-
tion of an enterprise. Thus, the closing passage of Gudea’s Cylinder A
on the building of the Eninnu moves from description of the affective
properties of the temple to an assessment of its impact. And in the
same way, the Assyrian building accounts also initially lay the ground-
work for closure of the project by declaring the undertaking to have
been done to perfection. Only then is the work in question declared
to be perceived as intended, with admiration, by future princes or by
the people.
In generalizing for all of Mesopotamian culture over a complex 2000-
year period, during which many historical shifts and changes in domi-
nant language occur, I have chosen to focus upon continuities, rather
than differences. This seems justified both by demonstrable continuities
in cultural patterns and by the fact that the present study constitutes an
initial investigation of the phenomenon of visual response. It should be
stated, however, that there is not always a perfect correspondence across
periods or languages where activities or terms related to intensified visual
experience are in evidence. For one thing, it will have been noticed
that exhortations to view have been cited from the Akkadian, not the
Sumerian, literary corpus. For another, later Neo-Assyrian rulers seem
to balance references to temple construction with accounts of palace
building, both of which elicit strong visual responses, whereas to date I
have not found references to Sumerian palaces that claim to affect their
viewing public in the ways described above. What this difference may
signify, if indeed it turns out to be consistent and meaningful, is that
the class of affecting works has been expanded over time and/or across
the Sumerian-to-Akkadian/Assyrian historical divide. Nevertheless, a
sufficiently large number of Sumerian logograms are preserved in the
written versions of Akkadian terms, and enough connections can be
made with individual elements—such as the temples and cult objects

eliciting strong responses, as well as the distinction between divine and
popular responses to the works celebrated—that I would not want to
posit any less importance accorded to vision in Sumerian, as opposed to
Akkadian, tradition. For architecture and objects alike, the association
of special properties with enhanced visual response forms an essential
part of the account of a work, while parallels in writing and mean-
ing of the primary terms suggest that the completed work was no less
“affective” in earlier periods than it was later. Finally, in both languages,
Sumerian and Akkadian, we find quite comparable rhetorical strategies
on the part of Mesopotamian rulers: in particular, the attribution of
the people’s positive response of ad+miration and awe as the closing
statement and ultimate affirmation of their enterprises.73
Let us now return to the third millennium Sumerian sculptures with
which we began (figs. 1–3). I would submit that the key to their enlarged
eyes does indeed lie in the votive nature of the images. But I would add
that in addition to the focus of attention, what we are seeing in those
eyes is literally the physical manifestation of individuals struck by the
u6-di response to the deity before him/her, a deity whose awe-inspiring
nature is then reflected in the wide-eyed stare.
In this way, the enlarged eyes of the statues serve both as the expres-
sion of devotion and the reflection of the ad+miring response to the
awe-inspiring divinity whose image would presumably have been
installed in the shrine where the votive statues were placed. Here again,
as in the case of the powerfully muscled right arm in the statues of
Gudea of Lagash that articulates the divine empowerment of the ruler,
form is directly invested with meaning.74
That the votive statue, which after all represents the pious devotee,
should manifest this visible marker of affect makes sense when it is
recalled that deities as well as works are documented as eliciting the
identical response resulting from intense visual experience. In fact, if
73
Anticipated for the future would be chronological studies of the variations in terms
and usage over time (and studies of variation across space!), along with a careful and
systematic study of these terms in direct relation to the actual works preserved to us,
to see precisely how the attributes claimed in textual sources may have been manifest
visually as “style” in any given period.
74
On this issue, as manifest in Mesopotamian art, see I. J. Winter, “The Affective
Properties of Styles: An Inquiry into Analytical Process and the Inscription of Meaning
in Art History,” in C. A. Jones and P. Galison, eds., Picturing Science, Producing Art (New
York & London: Routledge, 1998), 55–77.
what “the people” are responding to in the works, or at least what is

claimed for the works, is the same power as is inherently possessed by
the gods, then it should come as no surprise that the responses should
be identical. Or, inversely, if for Mesopotamia a primary role played by
affective sensory (i.e., aesthetic) experience was to touch and experience
(the awesome power inherent in) the divine, then faced directly with the
divinity him/herself, one should expect the homologous response. Put in
the proper hierarchical order, the people respond to the gods with an
awe born of the impact derived from direct and intense visual experi-
ence of the sacred; and what they respond to in the work—temple,
object, or even royal abode—is precisely those properties that link it
to the divine.75
Nor should this preclude finding enlarged eyes associated with divine
images in the Mesopotamian repertoire, despite the notable scarcity
of divine images in the archaeological record.76 Not only are the gods
consistently described as “seeing” or “viewing” the objects of their
pleasure, as documented in this chapter, but they are also described
as viewing the pious in interactions. Thus, in a text of Assurbanipal,
the king records that the gods Assur and Ninlil “lift their auspicious
eyes,” and “look joyfully upon the king,” ina nīš ī[nī]šunu damqāti . . . adīš
[it]taplasu,77 just as, in Christian worship, the Virgin is invoked by the
pious worshiper in the “Salve Regina”: “turn thy compassionate eyes
toward us,” illos tuos misericordes occulos ad nos converte.
75
An interesting aspect of this point that cannot be developed in the present context
is the distinction between the role of vision in Mesopotamian tradition and the role
of “hearing” as the means of experiencing the divine in the Hebrew Bible. Stephen
A. Geller has pursued the history of the two senses as vehicles for the transmission of
divine will in “Fiery Wisdom: Logos and Lexis in Deuteronomy 4,” Prooftexts 14 (1994):
103–39, where he sees a definite development from an originally privileged role for sight,
which then over time gave way to an ultimate privileging of the sense of hearing God’s
“word”—see esp. 122–4, 134, and 139, n. 24. The contrast is extremely instructive
for apprehending just what was entailed in the Mesopotamian visual experience. The
issue is also discussed by David Chidester, Word and Light: Seeing, Hearing, and Religious
Discourse (Urbana, Ill.: University of Illinois Press, 1992)—for which reference I am
indebted to Robert Nelson.
76
Spycket, La statuaire, 77, 144, 185, 203. Only a sample of divine statues will permit
us to assess whether gods’ eyes are as enlarged as those of votive statues, however.
77
Cited in CAD B: 320, būnu A (= Asb). Note that the verb form derives from
the verbs palāsu, naplusu, the same verbs that appear frequently in association with
cathected viewing. That this phraseology has a long history in Babylonian texts can be
demonstrated: See, for example, the text of Nebuchadnezzer I of the twelfth century
b.c.e., in Frame, RIMB 2, 32 (= Neb I B.2.4.10, l. 12), noting that the sungod “looked
joyously (upon him)” and gave him the kingship over all people.
In the visual record, this relationship is best preserved in narrative

relief, where we see kings standing in ritual service before a deity (e.g.,
the Stele of Ur-Namma of Ur), or in direct audience with the god
(as on the Stele of Hammurabi, fig. 5). In both cases, one Sumerian,
the other Old Babylonian, the reciprocity of visual exchange is clearly
marked. That the god’s “viewing” has implications beyond mere sight
toward positive acknowledgment and benevolence is clear from an Old
Babylonian hymn to the goddess Ishtar, where she is exalted: naplasuša
bani bu āru, “prosperity is created by her gaze.”78 In the end, it is this
prosperity (bu āru) in Akkadian, abundance ( e2.gal2) in Sumerian, that
is the ultimate gift of the gods, and thus, to have represented Hammu-
rabi in the same period literally eye to eye with the sun-god Shamash
is tantamount to proclaiming to viewers of the stele that the king is
truly in the god’s favor, that his authority to propound the laws also
contained on the stele derives directly from the deity, who, if order is
properly maintained, will look well (sic) upon not only the king, but
the kingdom.79
The consequence of putting the textual material together with the
sculptures has been to make clear how much stress was placed upon
vision and visuality within Mesopotamian tradition, and how well vision
was understood as a cognitive and symbolic act. In the sculpture, eyes
are emphasized, even disproportionally in some periods, and this would
seem to signal not only the intensity of the visual bond between, say,
devotee and deity but also the augmented visual experience that results.
In the texts, gods and rulers see not only each other but also works—tem-
ples, cult appurtenances, royal images, devotees—and they are pleased:
their hearts rejoice, their faces shine; and the hoped-for result is that
they will respond benevolently to the patron/ruler responsible for the
work in question, or to the pious supplicant. The generic “people” and
the lower-level gods also see works—temples, palaces, statuary, luxury
goods—and are declared to respond appropriately: with the admiration
and awe that results from a powerful visual experience, the bowled-over
spectator before the spectacular. In addition, individuals cathect visually to
individual objects: the lover to the beloved, Gilgamesh to his dream-axe.
78
Cited CAD N1: 306, naplusu (= RA 22 170: 15).
79
See now on the relationship between god and king on the Law stele of Ham-
murabi, I. J. Winter, “Art in Empire: The Royal Image and the Visual Dimensions
of Assyrian Ideology,” in S. Parpola and R. M. Whiting, eds., Assyria 1995 (Helsinki:
University of Helsinki Press, 1997), 366–7.
In turn, objects and public works are declared worthy of being seen, as a
way to underscore their value. In short, as has been argued for classical
Greece and the antique in general,80 so also for ancient Mesopotamia:
the act of viewing arouses the spectator’s responses. Visuality is coded
as affective and powerful, and vision is explicitly acknowledged as the
primary path to both religious and aesthetic experience.
Acknowledgments
I should like to thank the editor, Robert Nelson, for the opportunity to participate in
the present volume, since the issues with which I have been concerned mesh so well
with those that governed his original symposium in Los Angeles. I would also like to
thank Benjamin Studevent-Hickman and Jülide Aker for close readings of an earlier
version of the manuscript; and Froma Zeitlin for a series of wonderfully stimulating
conversations, plus a number of sources otherwise not known to me. I am deeply
indebted to the work of Thorkild Jacobsen, with whom many of these questions were
discussed before his death. Finally, several of the observations contained in the pres-
ent study, especially those related to Sumerian terms for sight and their logographic
writing, owe their present form to rich exchanges with Hermann Behrens. That he
did not live to receive the fruit of those exchanges has been, and will remain, a source
of great sadness.
80
For Greece, Prier, Thauma Idesthai, passim, and Zeitlin, “The Artful Eye. . .”, 141–2.
Zeitlin also cites R. Padel, “Making Space Speak,” in R. Winkler and F. I. Zeitlin, eds.,
Nothing to Do with Dionysos? Athenian Drama in its Social Context (Princeton, N.J.: Princeton
University Press, 1990) and D. Seale, Vision and Stagecraft in Sophocles (Chicago: Univer-
sity of Chicago Press, 1982). For the antique in general, J. Elsner, “Image and Ritual:
Reflections on the Religious Appreciation of Classical Art,” The Classical Quarterly 46,
no. 2 (1996): esp. 522; and K. Pomian, “Vision and Cognition,” in C. A. Jones and
P. Galison, eds., Picturing Science, Producing Art, 211–13.
Figure 1. Detail, Statue of Ibi -il, found at Mari. Early Dynastic III Period,
ca. 2500 b.c.e. (Photo M. Chuzeville, courtesy Département des antiquités
orientales, Musée du Louvre, Paris).
Figure 2. Detail, Statue of Gudea, ruler of Lagash, found at Tello. Neo-Sumerian

Period, ca. 2110 b.c.e. (Photo by Jill Casid, reproduced courtesy Département
des antiquités orientales, Musée du Louvre, Paris).
Figure 3. Votive statues from the Square Temple, Tell Asmar. Early Dynastic
Periods II and III, ca. 2750–2400 b.c.e. (Photo courtesy The Oriental Institute,
University of Chicago).
Figure 4. Detail, Stele of Naram-Sin, found at Susa Akkadian Period, ca.

2250 b.c.e. (Photo courtesy Département des antiquités orientales, Musée
du Louvre, Paris).
Figure 5. Stele bearing the Laws of Hammurabi, found at Susa. Old Baby-
lonian Period, 1792–1750 b.c.e. (Photo M. Chuzeville, courtesy Département
des antiquités orientales. Musée du Louvre, Paris).
CHAPTER THIRTY-THREE
BABYLONIAN ARCHAEOLOGISTS OF THE(IR)

MESOPOTAMIAN PAST
That ancient Mesopotamians had a clear sense of historical tradition,

and were wont to use that tradition for their own purposes, will come
as no surprise to anyone who has followed the scholarly literature of
the field. Studies of the reception of the third-millennium b.c. ruler
Sargon of Agade, his sons, and his grandson Naram-Sîn (e.g., Glassner
1986), for example, show how the dynasty was at times reviled, at times
reified as a golden age of the past. The Babylonians of the mid-first
millennium b.c.e. have actually been called “antiquarians”, particu-
larly engaged with their past (Clay 1912; Unger 1931: 227); but the
archaeological component of their activities has not been investigated
systematically. In the present paper, I would like to move beyond issues
of the availability of sources and an awareness of the past, to review
the actual practice of excavation as both a technique and a strategy
for recovery of the past in ancient times. The evidence is largely tex-
tual, and has generally been the concern of text-based historians, as
distinct from archaeologists. Yet when this evidence is looked at from
the perspective of ancient claims to knowledge it reinforces the prem-
ise that in the first millennium b.c.e. at least, under the guise of royal
patronage and purpose, the Babylonian past was actively sought in the
field. The resultant finds then served a variety of purposes that bear a
rather striking resemblance to our understanding of the “uses of the
past in the present” today.
What is demonstrable is that they, like us, mounted campaigns to
actively recover ancient remains; and that they declared themselves as
having dug in order to reveal works attributed to the ancients. Finally,
they also, like at least some of us, proclaimed these finds to be the
results of a (divinely directed) research design geared to an empirical
* This article originally appeared as “Babylonian Archaeologists of the(ir) Mesopo-

tamian Past,” in Proceedings of the First International Congress of the Archaeology of the Ancient
Near East, P. Matthaie et al., eds., Rome, 2000, pp. 1785–1798.
462 chapter thirty-three
and positivist recovery of true “traces” of the past—that is, a decidedly

processual as distinct from post-processual set of assumptions!
For late 20th century archaeologists of our era, excavations are
expected to yield evidence of ancient systems of cognition through
patterns of behavior manifest by material culture: architecture, artifacts
and texts. So too the Babylonians. While they may not have subjected
their finds to modern chemical, osteological, or paleobotanical analysis,
they did very much claim to have discovered both ancient materials
and evidence of ancient cognition, and to have studied them accord-
ingly. In what follows, I shall cite a number of cases to demonstrate:
I. The mounting of Field Campaigns, II. The exposure of Architectural
Remains, III. The discovery of Texts and Artifacts, which, once found,
were subject to analysis, and IV. The subsequent Display of a selected
sample of finds.
I. Field Campaigns
The Babylonian king Nabonidus (556–539 b.c.e.) is the ruler perhaps

best known for his field activities. Interpreting the motives for this
engagement has proved to be a bit of a Rorschach test for the intel-
lectual concerns and historical moments of interpreters—from “purely
antiquarian interests” (e.g., Hommel 1885 to Albright 1946: 241–244),
to an ascription of religious piety (Goossens 1948), to a more recent
assessment of profane inclinations and/or overtly political agendas
(Beaulieu 1989; Powell 1991). Given Beaulieu’s demonstration of the
careful selection manifest in Nabonidus’ citation of previous rulers
and attention to particular sites (1989: 138–142), I am inclined to see
the ruler’s engagement as a reflection of a joint political-cum-religious
strategy embedded within a system of beliefs that included the expo-
sure of signs of the past as a means of serving divine intention. What
is important for our purposes is that Nabonidus went about exposing
those signs by excavation.
That these antiquarian/archaeological pursuits were not limited to
Nabonidus, but rather were practiced by Neo-Babylonian rulers as a
whole, particularly Nabopolassar (625–605 b.c.e.) and Nebuchadnezzar
II (604–562 b.c.e.), has been demonstrated by Goossens (1948: 149).
Exactly where to excavate was often revealed by divinely-inspired dream
or divination; a combination of exploration and subsequent building
carried out in Babylon, Uruk, Nippur, Larsa, Sippar, Ur, Agade and
babylonian archaeologists of the(ir) mesopotamian past 463
Kish. The success rate claimed in recovering ancient sites and building
foundations suggests theirs was not a significantly inferior method to
those employed in more recent times. And, as in modern field excava-
tion, targets were occasionally revealed by accidents of exposure due
to natural phenomena.
It is recorded, for example, that a sandstorm revealed to Nebuchad-
nezzar the location of the early foundation walls of the shrine of
Shamash, the Ebabbar, at Larsa, built by the Kassite king Burnaburiash
(Goossens 1948: 154). Not one to be left behind, Nabonidus claimed
that another storm exposed the ruins of a platform and ziggurat of
Hammurabi, some 700 years earlier than Burnaburiash, which he,
Nabonidus, then restored (ibid.: 155). It is not my aim to argue whether
these claims were merely rhetorical, actual, or a combination of
both—indeed there is evidence that both the Ur III ziggurat and the
giparu, residence of the en-priestess of Nanna, in Ur had been explored
and added to in Neo-Babylonian times (Woolley 1925: 15; Wiseman
1985: pl. IV; Weadock 1975: 112–113). What I would suggest is that
embedded within the rhetoric lay an understanding of multiple ways
in which earlier ruins could be revealed.
Large-scale exploratory works were undertaken in Sippar and Babylon
as well as at Ur, Larsa and Agade. The Neo-Babylonians apparently
had no trouble identifying the latter site—an endeavor that has eluded
or divided some of our own archaeologists—although if Tell Muham-
mad is indeed Agade, then the temple revealed there in recent excava-
tions, which shows work by both Nebuchadnezzar II and Nabonidus
(Wiseman 1985: pl. VIIIa; Nasir 1979), may well attest to the verac-
ity of the Neo-Babylonian claims! Both kings record work there, but
only Nabonidus claimed to have been successful in finding the actual
foundations of the Eulmash, the temple to Ishtar of Agade (Goossens
1948: 154; Frame 1993: 23). Nabonidus’ account describes following
in trenches begun by Nebuchadnezzar, at first with little success. Then,
just as the excavators were about to abandon their search, a fierce rain
storm (alternative to the sandstorm at Larsa) is said to have revealed
the foundations of the original Naram-Sîn temple.
The various accounts of Nabonidus provide myriad details of the
archaeological undertakings: duration of campaigns, trenches exca-
vated, work personnel, and even oversight by individuals who would
correspond to our “field directors” today (Goossens 1948: 153–154).
The texts concerning the search for the foundations of the Ishtar
temple in Agade are perhaps the most informative, of which accounts
are preserved from Ur, Sippar and Babylon. In the Babylon version,
the king’s purpose is made explicit: I sought to (re-)build this temple; and
in order to do so, he says: I opened up the ground inside Agade and looked for
the foundation (Frame 1993: 30). In the lengthier Sippar version whether
fictively to make his own success the more salient, or as historical record,
Nabonidus provides the historical background (BM104738): Kurigalzu
king of Babylon who preceded me, looked for the foundation of Eulmash in Agade,
which had not been known since the time of Sargon, king of Babylonia and his
son (sic!) Naram-Sîn . . . but he did not find it. He wrote and set up an inscription
which said ‘I looked ceaselessly for the foundations of the Eulmash but did not find.
Nabonidus then credits Esarhaddon of Assyria, his son, Assurbanipal,
and Nebuchadnezzar with also having looked for the temple. As for
Nebuchadnezzar, he called up his numerous workmen and looked ceaselessly . . . he
dug deep trenches but did not find the foundation. Nabonidus then continues:
For THREE years I excavated in the trench of Nebuchadnezzar (emphasis
ours) . . . I looked to the right and left . . . to the front and rear of the trench . . . (But
then) a downpour occured and made a gully . . . I said . . . to them “Dig a trench in
this gully”. They excavated in that gully and found the foundation of Eulmash.
One could think it unlikely that previous kings would have left
inscriptions stating that they had failed in their exploration attempts;
however, as G. Frame has documented, in some of the Babylonian
inscriptions of Esarhaddon, the king actually does state that he had
sought in vain for the original emplacement of a temple in the Eanna
in Uruk (Borger 1967: 74 § 47:32), and Assurbanipal twice noted that
he sought the foundations of a temple before rebuilding—in Sippar
and in Uruk (Frame 1993: 35). Thus, the Assyrians in Babylonia can
also be attested to have engaged in archaeological reconnaissances, and
did in fact record failure as well as success.
Finally, following upon the rebuilding of the Eulmash, subsequent
excavations were apparently undertaken in Agade in the old bit-akitu and
the palace of Naram-Sîn (Beaulieu 1989: 141–142; Joannès 1988). A
scribal expert in early scripts, one Nabû-zer-lišir, is credited with several
finds as if he were serving as the “field director” of the operations. The
reconstruction of the scribe’s career by Joannès helps us to view his
archaeological operations as serious professional undertakings; indeed,
Weiss had earlier referred to the individual as an “epigrapher-archae-
ologist” (1975: 447). As Beaulieu has noted, the sequence of discover-
ies at Agade follows an identical pattern to that at Sippar: excavation,
architectural restoration and “finds”—all of which suggests that there
was a clear conception of procedure.
II. Architectural Remains
In the vast majority of the Neo-Babylonian archaeological enterprises,

the main object of the search was stated to be the foundations of a
previous temple, long since fallen into ruin; and the ultimate strategy
of recovery included first rebuilding, and then restoration of the old
cult. Divine intervention was said to have been involved not only in
locating the site, but often also in controlling the conditions of discov-
ery—as Adad is credited with having sent the rainstorm that revealed
the ancient foundations in Agade (Frame 1993: 32).
Recovery of the original site must have been important because of
the history of religious practice at a sacred and sanctioned place—and
perhaps, even, as in the selection of Hindu temple sites in India, because
the site was initially determined by a series of divinatory and cultic
practices to ensure that that location and no other had the requisite
auspicious properties. In any case, the fact of Nebuchadnezzar having
built a temple to Ishtar on a site other than the original at Sippar was
given as the reason for the building’s having fallen rapidly into decay
(Goossens 1948: 155). By contrast, Nabonidus’ discovery of the origi-
nal foundation, or temennu, is presented as evidence for the goddess’s
favor and the happy future for the restoration project. This was clearly
important enough that, when rebuilding the temple to Ishtar at Agade,
he declared that he constructed his own brickwork directly “above that
(original) foundation . . . not (allowing the foundations) to protrude by one
finger’s breadth nor (allowing them) to recede by one finger-breadth . . .”
(Lambert 1968–69: rev.iv.18–26; Frame 1993: 26, ii.76–iii.2).
Similarly, Nabopolassar, concerning the wall IMGUR-ENLIL in Babylon,
“looked for its ancient foundation platform and found it”, then laid
new bricks in the original place (al-Rawi 1985: 6, ii.33–41). In the text
describing this event, Nabopolassar goes on to identify himself by the
epithets: he “who searches for the ancient foundation platforms . . ., who
discovers bricks of the past, who rebuilds . . . on the original platform”
(ibid.: iii.4–5).
One of the most interesting aspects of these discoveries is that
the Neo-Babylonian rulers refer not only to the wall foundations but
also to the royal foundation deposits of these earlier constructions—as
Nabopolassar seems to have recorded for the wall in Babylon (ibid.: iii.
16–22) and as Nabonidus recorded that Nebuchadnezzar had found
at Sippar belonging to the Middle Babylonian ruler Burnaburiash
(Tadmor 1965: 362).
The phenomenon of the foundation deposit has been studied by

Ellis (1968), and needs no introduction to modem archaeologists. The
texts—either on independent tablets or inscribed directly onto anthro-
pomorphoid pegs—were said to have been read in antiquity, allowing
subsequent rulers to know which kings’ efforts had preceded them.
Thus Nabonidus records for Agade: I saw/looked upon the old foundation of
Naram-Sîn, an earlier king, and I read the tablets of gold, lapis and carnelian about
the building of the Ebabbar (Tadmor 1965: 359; Powell 1991: 24).1 Such
foundation tablets, as recorded by Nabopolassar, when encountered,
were then carefully replaced in the original foundations by the later
ruler, or included with the new foundation deposit of the discoverer (al-
Rawi 1985: 6).2 In Nabonidus’ account of the foundation inscription of
Naram-Sîn found in the Ebabbar, the king is said to have looked upon
it, and not changing its position, he put with it his own inscription (Lambert
1968–69: 7 = rev. iv. 26–28). He further found a partly-effaced and
damaged image of Sargon of Agade, which Nabonidus arranged to be
properly restored by his master craftsmen, and again, not altering its
original location, had it replaced in the Ebabbar, to the accompaniment
of ritual oblation (ibid., rev. iv. 29–36).3
Apart from the obvious value of an ancient description of a practice
confirmed by modern archaeological discovery, what is important for
us as modem excavators is the care with which Nabonidus and Nabo-
1
Beaulieu (1989: 141) records a clay impression of an inscription of the son of
Naram-Sîn, Šar-kali-šarri with a Nabonidus inscription on the back, saying the impres-
sion was made from a diorite slab in the palace. This offers further confirmation that
the excavation was indeed undertaken.
2
Note that al-Rawi refers to a “statue” of an early king discovered in the wall, and
implicit is that the king then replaced the “statue” along with his own great foundation
deposit “together with my image” (al-Rawi 1985: 6 = iii.16–21). The word translated as
“statue”, however, Akk. almu, means “image” and does not make clear what form—two
dimensional or three—the image would have taken (see on this, Winter 1997: 365), so
it is not clear whether Nabopolassar intends us to understand a foundation figurine
or something other.
3
The text seems to describe this image of “Sargon” as a foundation figurine, as it
is not only found when the foundations of the new Ebabbar were being laid, but, we
are told, it was found within that foundation—Lambert 1968–69: 7, obv. iii. 29–30:
a-lam mšarru-u2kin . . . ina qi2-rib te-me-en-na ša2-a-šu2 ip-pa-al-li-is-ma . . ., “an image of
Sargon . . . in the midst of that foundation he gazed upon . . .” Beaulieu, however, treats
it as a full-standing “statue” of Sargon, since at least 8 texts from the archive of the
Ebabbar of Sippar record on-going offerings to an image of Sargon (1989: 135–136,
citing Kennedy 1969: 79). Unless the foundation discovery and the image receiving
cultic attention represent two different works, the offerings do suggest a three-dimen-
sional royal image distinct from the usual foundation figurine; but the question must
remain open for the present.
polassar claimed to have replaced the original foundation deposits and

performed the necessary cultic observations for them. This, when taken
together with the pronouncements commonly found in the curses at
the ends of foundation tablets—curses that abjure future rulers to treat
their tablets with appropriate respect and ceremony—may give us even
further insight into what we find in the ground today. New construction,
especially in temple areas (e.g., the Sin Temple sequence at Khafaje),
generally leaves several courses of brick of the preceding building phase,
when it would have been just as easy to clear it all off as to level the
wall stubs. This may well have been at least in part a matter of piety-
cum-proper-action regarding the prior construction, precisely because a
previous builder had marked it with his own, divinely sanctioned, record of
foundation. Leaving a few courses of brick preserved both the memory
of the building and the pious foundation document—thereby neither
negating what the gods had previously approved, nor provoking the
wrath threatened in the texts’ curses.4
III. Texts and Artifacts
Like many of our colleagues, the Babylonian kings seemed especially

pleased when they could claim to have discovered an ancient “text”.
Bricks of the earlier period found in Agade are carefully husbanded in
the reign of Nabonidus (Beaulieu 1989: 141–142); and that king also
claimed to have been able to read (ancient) tablets brought to him that
no one else could understand (Machinist—Tadmor 1993: 149, who
provide an alternate reading to that of Lambert 1968–69: 4).
The unusual case of an inscribed stone known as the Cruciform
Monument, found by Rassam at Sippar in 1881 along with two
Nabonidus cylinders (Walker—Collon 1980: 103), underscores the
importance accorded these “traces” of the past. Purporting to be an Old
Akkadian text of Manishtusu, son of Sargon of Agade, it declares the
renewal of rites and privileges of the cult of Shamash at Sippar. The
text was shown by Sollberger (1967–68) to be not earlier than the Old
4
If this is correct, then the architectural sequence in the archaeological record,
which we so value for stratigraphic and chronological purposes, is in itself an artifact
of the Mesopotamian belief system.
Babylonian Period.5 For our purposes, if, as argued by Powell (1991),

the Cruciform Monument is in fact a Neo-Babylonian forgery, its text
constitutes a clear awareness in the later period of what it would take
to re-construct a temple according to the past, and underscores the
weight placed upon the ancient textual record.
Nabonidus’ scribe, Nabû-zer-lišir, clearly had been trained in deci-
phering and copying early texts discovered in “excavation”. In one docu-
ment found in the bīt-akītu of Agade, we are told that the original of the
text had been “on a brick which Nabû-zer-lišir, the scribe, . . . found and
excerpted in Agade” (emphasis ours; Beaulieu 1989: 142 = BM 22457).
In another case, that same scribe took an impression in clay of the
inscription on an Akkadian stone slab found in the palace of Naram-
Sîn; then, in good excavation recording tradition, added his own gloss
on the back, noting the find-spot (ibid.: 141). There are other cases as
well, particularly with regard to the “Sun Tablet” found in the Shamash
Temple at Sippar, where moulds, corresponding to modern “squeezes”,
have actually been found with the original (figs. 1, 2; Walker-Collon
1980: 102–103). These practices suggest that the modern activities of
copying, summarizing, reproducing and cataloguing texts were not
unknown to the scholars of antiquity.
Although the Neo-Babylonian rulers refrain from any mention of
pottery in their excavation trenches, the discovery of ancient objects
seems to have been welcomed by them along with texts. As noted above,
foundation inscriptions were often said to have been accompanied by
ancient “images”, presumably foundation figurines. Nabonidus also
claims to have found a jasper cylinder seal dedicated by Assurbanipal
to the god Sîn, which he (Nabonidus) then placed upon the neck of
his own cult image of Sîn at Harran (Lee 1993: 134).6 Clearly, the seal
5
More recently, Powell has argued that the name of the ruler should rather be
restored as Naram-Sîn, and suggested that the whole thing is an ancient (Neo-Babylo-
nian) forgery in an early genre, made expressly to deceive the sign-hungry Nabonidus
(1991: 26–27)—perhaps by the very priesthood of Shamash at Sippar itself, to buttress
the king’s support for restoring the Ebabbar. The reading of Manishtusu has been
recently re-confirmed by al-Rawi and George (1994: 146), however. Whoever the
Akkadian ruler cited, the text remains a pious fraud; what needs to be determined is
whether or not the fraudulent act is Neo-Babylonian in origin.
6
A related phenomenon may be seen in the 9th century seal for Marduk,
dedicated by Marduk-zakir-šumi, which consisted of an image of Marduk stand-
ing upon his attribute animal, the mušhuššu, and according to the seal’s inscription
was intended to be hung around the neck of the god’s cult image in the Esagil
in Babylon. A further parallel to the various situations described here is to be
was valued for its material and its age; but it is also explicitly stated
that the seal contained on it a likeness of the god, so the imagery was
considered to be especially apt.
The discovery of such prior images is often given as the basis for a
later king’s ability to re-construct a lost cult image. In the Nabonidus
inscription regarding the seal, we are told that it depicted the god
“whose features had been revealed from ancient days” (ibid.: 135, ll.
44–45)—thereby providing an “authentic” model for his own cult
image. In a similar vein, Nabonidus recorded the discovery at Ur of
an “old narû”, or stele, with a representation of an en-priestess [nin.
dingir.ra] of the cult of Nanna (Dhorme 1914; Lambert 1968–69: 4;
Beaulieu 1989: 23–24; Powell 1991: 30), including a (presumably tex-
tual) description of her attributes, attire and jewellery. This discovery
rather conveniently permitted the king to re-establish the institution
as part of his refurbishing of the cult of the Moon god, and to install
his daughter in the post—as Sargon of Agade and Rim Sin of Larsa
had done before him.
So also, the text of the “Sun-God Tablet” of Nabû-apla-iddina (fig. 1)
describes the discovery “on the other (that is, the Western) bank of
the Euphrates” of a baked clay image preserving the appearance and
attributes of the Sun god, Shamash (King 1912: 120–127; Brinkman
1976; Woods 1998).7 This discovery, too, was not without “ interest”.
In the manner of the Ur stele, it enabled the 9th century ruler to
reconstruct not merely a cultic office, but the very cult image of the
deity in the traditional manner.
The “Sun-God Tablet” is of interest in the pursuit of ancient archae-
ology on more than one level. First, on its own terms, the text attests
found in the fact that this seal, too, was discovered out of its original context—in
the house of a bead-maker of the Parthian Period (on which, see Black—Green
1992: fig. 105).
7
= BM91000 (BBSt. No. 36) iii. 19–30–iv. 2—for the transliteration of which I am
indebted to C. Woods and K. Slanski.
u2- ur-ti al-mi-šu2 A drawing of his image,
ser-pu ša2 a-as-bi . . a . . . of fired clay . . .,
ina e-bir-ti on the other bank
id
pu-ra-ti of the Euphrates . . .
in-na-mir-ma was found, and
PN . . . PN . . .,
É.MAŠ uruSip-par . . . šangu-priest of Sippar . . .
giš
UR al-mi šu-a-tu4 . . . the plan of that image
. . . u2-kal-lim-ma . . . presented (to the king).
to the use and value of “found objects” from the past as models in the
present. And indeed, the veracity of the claims by Nabû-apla-iddina
that he used an ancient image as the model for his new cult image is
corroborated by a comparison of the image of the Sun god on the
tablet with very similar Old Babylonian or Akkadian representations
of the god on the one hand, and the clear stylistic contrast between
the archaizing figure of the deity with the contemporary Babylonian
ruler and priest of the tablet on the other hand (see now Woods 1998).
The rendering of shoulders and general proportions of the latter are
characteristic of their own stylistic period—as seen when juxtaposed to
other 9th century images, such as Nabonidus’ stelae (fig. 3)—thereby
emphasizing the archaizing nature of the cult image. But then, second,
the “Sun-God Tablet” itself must have been still visible, or at least re-
discovered by the late 7th century Neo-Babylonian ruler Nabopolas-
sar more than 200 years later, for at this time not only were the two
moulds of the tablet made, but also a 19-line inscription of the later
king was added on the reverse of one of them, outlining various cultic
offerings to Shamash (= BM91002, Walker—Collon 1980: 102–103;
Woods 1998: 4).
I have suggested in passing that the various finds of the Babylonian
“archaeologists” are generally recorded not for their own sake, but in
conjunction with specific royal “interest”. In that, the Babylonians’ use
of objects may be said to have differed from our contemporaries’ search
for “pure knowledge”; although the degree to which their “interests”
were not unrelated to proving or disproving their own belief systems
and/or to their personal ambitions in legitimizing their practices and
predilections might actually on closer examination prove familiar (on
which, see Silberman 1982).8
Once again, then, with the “finds” of the Neo-Babylonians we have a
series of phenomena that parallel the modern archaeological situation.
First, the discovery of ancient texts and a range of objects—cylinder
seals, clay plaques, stone stelae, statues and/or figurines—attested from
the Akkadian, Old Babylonian, Kassite and Assyrian periods. Second,
8
This “interestedness” in and enthusiasm for what was considered diagnostic works
might also account for why a forgery like the Cruciform Monument could have been
so enticing to Nabonidus. N. Wasserman has also recently argued that a purported
Old Babylonian clay amulet is most likely a Neo-Babylonian product, its inscription
executed in a purposefully archaizing script (Wasserman 1994, citing other examples
as well, as noted in Berger 1973).
the valuation of these objects for a variety of complex reasons as

“traces” of a purportedly true past. And third, parallel to these dis-
coveries, an industry that apparently specialized in the passing off of
recently-manufactured works as “authentic”—although perhaps less
from a commercial desire to gull the unsuspecting or greedy institu-
tion/individual than as a politically- or culturally-motivated attempt to
demonstrate a connection with the past.
IV. Display
While the excavated foundations of ancient buildings often served as

the basis for architectural reconstruction, material remnants of earlier
periods—texts, inscribed bricks, objects—were evidently gathered,
labeled, and even put on display in a variety of ways. A number of
early objects were found together by Rassam in the late Neo-Baby-
lonian Shamash temple at Sippar—including an inscribed Akkadian
mace-head, the statue of a king of Mari, the Cruciform Monument,
and the “Sun Tablet” group, along with stone bowls of the Late
Uruk and Early Dynastic Periods that had been repaired in antiquity
(Walker—Collon 1980). This assemblage argues for the temple having
functioned as a local treasury, if not actual museum (Walker—Collon
1980: 111; also Wilcke 1982), much in the way of Western cathedral
treasuries; and as in Western cases, here, too, some objects appear to
have been local to Sippar while others probably came from outside as
booty.9 Nor was the Ebabbar in Sippar likely to have been unique in
accumulating early works. A letter from the Eanna archive in Uruk
(YOS III.86, cited Beaulieu 1989: 232) suggests that at least some of
the statuary in Babylon, possibly in the Esagil of Marduk, had been
taken by Nabonidus from the temple of Nanaya in Uruk.10
9
Specifically, the ED stone bowl, originally from Lagash. Of course, there is a long
history of objects moving as a result of military action in the region, notable amongst
which is the presentation of objects in the sacred precinct at Susa following campaigns
in Mesopotamia by the Elamite ruler, Shutruk-nahunte (cf. discussion Klengel-Brandt
1990: 45; Braun-Holzinger 1997).
10
It will be remembered that Nabonidus also did a good bit of rebuilding in the
Eanna precinct at Uruk, and reinstituted cultic offerings he claimed to have been long
neglected (Frame 1991: 54). The reciprocal relationship between investment in the site
and display of objects from the site is important to underscore.
An area on the western flank of the Northern Palace at Babylon

has been understood to represent a purposeful museum, with some of
the objects said to have been found in the assemblage introduced in
the reign of Nabopolassar, most gathered by Nebuchadnezzer II, and
others likely added by Nabonidus (Wiseman 1985: 65, pl. Ib, citing
Unger 1931: 224–228). This has recently been disputed by Klengel-
Brandt (1990), who has demonstrated from unpublished excavation
records that a number of works attributed to the “museum” in fact
were found elsewhere on the site.
While it was enticing to see the situation in Babylon as a mirror-
image of our own museum practices, to dismantle the Northern Palace
“museum” is not to negate either the antiquarian mentality or the
practice of display in the Neo-Babylonian period. Bracketed between
the third- and second-millennium practices noted above and the situ-
ation found at Hellenistic Tello later, when Adad-nadin-aªªe gathered
statues of Gudea of some 2000 years earlier to stand in his palace
courtyard (Parrot 1948: 152–156),11 the accumulated instances within
the Neo-Babylonian Period suggest an interest not only in recovery but
in demonstration of that recovery—whether as trophies, models, or re-
introduced into cult practice as liturgical objects. Evidence of repair in
antiquity (as Lambert 1968–69: obv. iii, 32–35, where the broken face
of an image is restored), continued usage, and often re-inscription of
works goes along with conspicuous placement to suggest a long-standing
Mesopotamian interest in both collection and “display” that if anything
was heightened in this period.12
Indeed, physical repositories of the material traces of the past must
be understood as part of the same cultural pattern as is evident in the
11
It is probably the case that the solid black lines on the plan of the palace (Parrot
1948: fig. 33b) represent the foundations of Gudea’s own temple to the god Ningirsu,
and the colossal statue of Gudea now in the Louvre was probably originally placed
there, too large to move when the later ruler re-occupied the site. The other statues,
with dedications to different deities of Girsu, however, were most likely gathered from
different loci, particularly the sites of those gods’ temples, elsewhere in the city. This
response in Tello is to be seen not as a unique and idiosyncratic phenomenon, but as
part of the larger picture of an antiquarian theology recognizable throughout Hellenistic
Mesopotamia, as has been noted for Seleucid Uruk as well (Beaulieu 1992).
12
An interesting iconographic co-efficient of such an antiquarian interest is reported
by Ehrenberg (1998), where archaisms observable in 6th century seals from Uruk can
be tied to Nebuchadnezzar’s reinstallation of the “true” Ishtar to her sanctuary in
the Eanna. The (re-) introduction of earlier visual attributes would then represent a
conscious attempt gesture, not unlike the use of an early model for the cult image of
the Sun-god at Sippar.
excavations and the use of earlier objects as models. S. Stewart, in

her book On Longing, speaks of the “collection” in general as offering
exempla with the status of metaphor (1993: 151). She notes that the
collection—whether of the modern museum filled with archaeological
objects or the ancient—represents a form of “ahistorical enclosure”
in which “all time is made simultaneous or synchronous within the
collection’s world”; and through it, the past is put “at the service of
the collection”. In this, Stewart articulates the necessary interestedness
of the collector—Babylonian or otherwise—in serving his/her own
time’s ends.
It is surely no surprise, then, that this “interestedness” was recog-
nized in antiquity. Many of the cult images gathered into Babylon by
Nabonidus, for example, were apparently taken against the wishes of
their local populations, and were then returned to their respective cities
by Cyrus soon after the fall of Babylon (Beaulieu 1989: 232). This could
be seen in light of the cultural heritage protection laws that have come
into force in recent years, calling for the return of objects taken without
the consent of their original populations—although it would not fully
capture the political ideology of protection for the capital city and the
state that may have been at stake in the initial appropriations.
Just how one understands the archaeological proclivities of the
Neo-Babylonian period depends a great deal upon whether one sees
Nabonidus and the others as antiquarians, religious reformers, or
politically-motivated arrivistes who required buttressing by a variety of
legitimating strategies in order to establish themselves within a historical
tradition.13 But there is also a less-manipulative reading of the events
and attitudes charted above: one which foregrounds Babylonian notions
that revealed knowledge was given by antediluvian sages (Lambert
1957: 9), and so, whatever comes from the past is that much closer
to original “truth”. If, in the conceptual structure of the Babylonian
world, poetry was deemed capable of offering “pictures of unsurpass-
able vividness” (Landsberger 1926/1976: 14), so also an ancient image
or object could well have been considered an “unsurpassably vivid”
13
Biggs (1998: 75, citing Beaulieu 1989) has stressed Nabonidus’ attempt to forge links
with the Akkadian dynasty through the revival of the institution of entu of the moon-
god at Ur, emphasizing his religious zeal and that of his mother. It is also interesting to
see Nabonidus’ interests in restoring cultic practices related to the worship of the astral
deities Shamash and Sîn as a possible attempt to break the power of the priesthood
of Marduk, and the antiquarian ethos of Hellenistic Uruk as equally challenging the
dominant Marduk-Nabû theology of Babylon (Beaulieu 1992: 68).
model; and Nabonidus’ (or Nabu-apla-iddina’s) strategy for associating

his reign with appropriate ethics and an auspicious future would have
required just such archaic or archaizing models in the construction and
reconstruction of the tradition at the center, that is, in Babylon.
Such a cultural strategy of employing archaisms as well as archaic
monuments and texts for contemporary ends can be demonstrated
from many traditions at particular times in their history, and is in no
way unique to Mesopotamia, much less to the Neo-Babylonian period.
What may be unique to Mesopotamia is the insistence upon acquiring
knowledge of the past through excavation and the actual recovery of
ancient works. As demonstrated, the Neo-Babylonian rulers claimed
not only to have excavated, but to have recovered ancient building
foundations and their foundation deposits, inscriptions and artifacts.
And they also recorded putting them on display, as well as using them
as models in the production of contemporary works.
In all of these instances, whether actually practiced or merely
ascribed, what is reflected is a clear sense of the methodologies and
tactics we associate today with field archaeology. The lack of scientific
rigor might associate Babylonian field activities more with a 19th century
than a late 20th century archaeological practice; but it is nonetheless
clear that they were already doing two and a half millennia ago what
we have only been doing systematically for the past two and a half
centuries!
It is certainly hoped that current excavations at sites mentioned by
Neo-Babylonian rulers may turn up evidence of their interventions (see
Huot 1985, for Larsa; de Meyer 1980, for Sippar). In closing, I should
simply like to review the two observations that have consequences for
a reading of the archaeological record today. First, a caution: that
objects seemingly out of chronological context—even when not in as
obvious a secondary locus as the Gudeas in the Hellenistic palace at
Tello—may well NOT have been moved by serendipitous agents (e.g.,
rodents or the digging of late pits), but rather have served a cultural
function within the later period, as purposefully gathered representatives
of an imagined past; and second, a reminder: that those walls which
have been preserved only as a few courses of brick in multi-period
sites—particularly temples and palaces—may NOT merely reflect a
practical leveling of the ground, but rather, a cultural response to a
system of belief.
This we can learn from the Babylonian archaeologists of the(ir)
Mesopotamian past!
Acknowledgments
This paper could not have been written without the help of my colleague, P. Machinist,
to whom I am greatly indebted for his many insights and for the many bibliographical
sources he provided. 1 am also grateful to E. Braun-Holzinger and J. Cheng, and to
K. Slanski and C. Woods for permission to cite their unpublished editions of the text
on the “Sun-God Tablet” of Nabû-apla-iddina. Readers will note that the issues raised
in this paper converge with those discussed in the contribution of M. Roaf to the
present volume; it is our hope that this convergence will provide added dimensions
to the issues.
Bibliography
Albright, W. F. 1946. From the Stone Age to Christianity, Baltimore.

al-Rawi, F. 1985. Nabopolassar’s restoration work on the wall Imgur-Enlil at Babylon,
Iraq 47: 1–13.
al-Rawi, F. and George, A. 1994. Tablets from the Sippar library III. Two royal coun-
terfeits, Iraq 56: 135–148.
Beaulieu, P.-A. 1989. The Reign of Nabonidus, King of Babylon 556–539 B.C. (= YNER
10), New Haven-London.
——. 1992. Antiquarian theology in Seleucid Uruk, Acta Sumerologica 14: 47–75.
Berger, P. R. 1973. Die neubabylonischen Königsinschriften (= AOAT 4), Neukirchen-Vluyn.
Biggs, R. D. 1998. Review of G. Jonker, The Topography of Remembrance: The Dead, Tradi-
tion, and Collective Memory in Mesopotamia, Leiden 1995, JNES 57: 74–75.
Black, J. and Green, A. 1992. Gods, Demons and Symbols of Ancient Mesopotamia: An Illus-
trated Dictionary, Austin.
Borger, R. 1967. (1956) Die Inschriften Asarhaddons, Königs von Assyrien (= AfO Beiheft 9),
Osnabrück.
Braun-Holzinger, E. 1997. Verschleppte Bau- und Weihinschriften der Herrscher von
Lagaš, Acta Sumerologica 19: 1–18.
Brinkman, A. J. 1976. A note on the Shamash cult at Sippar in the eleventh century
B.C., RA 70: 183–184.
Clay, A. T. 1912. An ancient antiquary, MJ 3: 23–25.
Dhorme, E. 1914. La fille de Nabonide, RA 11: 105–117.
Ehrenberg, E. 1998. Sixth-century seal impressions at Yale, Paper delivered at Yale
University, as part of the 45th Rencontre Assyriologique Internationale.
Ellis, R. S. 1968. Foundation Deposits in Ancient Mesopotamia (= YNER 2), New Haven-
London.
Frame, G. 1991. Nabonidus, Nabû-šarra-uÉur, and the Eanna temple, ZA 81: 37–86.
——. 1993. Nabonidus and the history of the Eulmaš temple at Akkad, Mesopotamia
28: 21–50.
Glassner, J.-J. 1986. La Chute d’Akkadé (= BBVO 5), Berlin.
Goossens, G. 1948. Les recherches historiques a l’époque néo-babylonienne, RA 42:
149–159.
Hommel, F. 1885. Geschichte Babyloniens und Assyriens, Berlin.
Huot, J.-L. 1985. L’Ebabbar de Larsa durant le 1e millénaire (travaux de 1983), Akka-
dica 44: 14–20.
Joannès, F. 1988. Une lettre néo-babylonien, N.A.B.U. 1988: 5.
Kennedy, D. 1969. Realia, RA 63: 79–82.
King, L. W. 1912. Babylonian Boundary Stones and Memorial Tablets in the British Museum,
London.
Klengel-Brandt, E. 1990. Gab es ein Museum in der Hauptburg Nebukadnezars II.
in Babylon?, Forschungen und Berichte 28: 41–46.
Koldewey, R. 1913. Das wieder-erstehende Babylon, Leipzig.
Kuhrt, A. 1990. Nabonidus and the Babylonian priesthood, in M. Beard – J. North
(eds.), Pagan Priests. Religion and Power in the Ancient World, London: 119–155.
Lambert, W. G. 1957. Ancestors, authors and canonicity, JCS 11: 1–14.
——. 1968–69. A new source for the reign of Nabonidus, AfO. 22: 1–8.
Landsberger, B. 1976 (1926). The conceptual autonomy of the Babylonian world, transl.
T. Jacobsen et al., Sources & Monographs (= MANE l. 4), Malibu: 1–15.
Lee, T. G. 1993. The jasper cylinder seal of Assurbanipal and Nabonidus’ making of
Sin’s statue, RA 87: 131–136.
Machinist, P. and Tadmor, H. 1993. Heavenly wisdom, in M. E. Cohen et al. (eds.),

The Tablet and the Scroll: Near Eastern Studies in Honor of Wm. W. Hallo, Bethesda MD:
146–151.
de Meyer, L. (ed.) 1980. Tell ed-Der III: Sounding at Abu Habbah (Sippar), Leuven.
Nasir, M. 1979. The temple of Ishtar of Agad, Sumer 35: 78–81.
Parrot, A. 1948. Tello. Vingt campagnes de fouilles (1877–1933), Paris.
Powell, M. A. 1991. Naram-Sîn, son of Sargon: Ancient history, famous names, and
a famous Babylonian forgery, ZA 81: 20–30.
Silberman, N. A. 1982. Digging for God and Country: Exploration in the Holy Land, 1799–1917,
New York.
Sollberger, E. 1967–68. The cruciform monument, JEOL 20: 50–70.
Stewart, S. 1993. On Longing: Narratives of the Miniature, the Gigantic, the Souvenir, the Col-
lection, Durham-London.
Tadmor, H. 1965. The inscriptions of Nabunaid: Historical arrangement, in Studies in
Honor of Benno Landsberger (= AS 16), Chicago: 315–364.
Unger, E. 1931. Babylon. Die heilige Stadt, Berlin.
Walker, C. and Collon, D. 1980. Hormuzd Rassam’s excavations for the British Museum
at Sippar in 1881–82, in L. de Meyer (ed.), 1980: 93–114.
Wasserman, N. 1994. BM 78613. A Neo-Babylonian imposture of an Old Babylonian
amulet?, RA 88: 49–57.
Weadock, P. 1975. The giparu at Ur, Iraq 37: 101–128.
Weiss, H. 1975. Kish, Akkad and Agade, JAOS 95: 434–453.
Wilcke, C. 1982. Zum Geschichtsbewusstsein im Alten Mesopotamien, in H. Müller-
Karpe (ed.), Archäologie und Geschichtsbewusstsein, Munich: 31–52.
Winter, I. J. 1997. Art in empire: The royal image and the visual dimensions of Assyrian
ideology, in S. Parpola & R. Whiting (eds.), Assyria 1995, Helsinki: 359–381.
Wiseman, D. J. 1985. Nebuchadrezzar and Babylon, Oxford.
Woods, C. 1998. The Sun-God tablet of Nabû-apla-iddina revisited: Archaic iconog-
raphy, the search for legitimacy, and the transcendence of divinity, Unpubl. paper,
Harvard University.
Woolley, C. L. 1925. Excavations at Ur, 1924–25, Antiquaries Journal 5: 347–402.
Figure 1. Sun-God Tablet of Nabû-apla-iddina, found Sippar. (= WAA 91000;

photo courtesy The Trustees, The British Museum).
Figure 2. Sun-God Tablet in inscribed box, with Nabopolassar copies. (Photo

courtesy The Trustees, The British Museum).
Figure 3. Nabonidus stele (= WAA 90837; photo courtesy The Trustees, The
British Museum).
CHAPTER THIRTY-FOUR
EXHIBIT/INHIBIT: ARCHAEOLOGY, VALUE,

HISTORY IN THE WORK OF FRED WILSON
If I am an archaeologist, Fred Wilson is also an archaeologist. Not,

perhaps, by the standard definition of “the study of the past”—but
rather, by a contemporary definition meaning “the study of a particu-
lar culture’s view of a past.” To that end, Wilson excavates mercilessly,
placing before us artifacts from ancient Egypt through the American
colonial heritage to the 1930s and 40s. And we are faced with not only
the “history” we have seen fit to create out of that past, but also its
vulnerability to alternative readings.
The hot debates in the field of archaeology during the 1960s cen-
tered around the work of Lewis Binford, and whether archaeological
hypotheses could be tested and subjected to the rigors of science—a
kind of formal minimalism of environmental and economic adapta-
tion. And, whether the armchair study of the past counted as “true”
archaeology, or whether one had to be out in the field, generating data.
In the 1970s, the debate centered on the degree to which modern
ethnography could illuminate, through the practices of living cultures,
the activities and systems of meaning of the past, plus long debates
about whether archaeology was truly a discipline, or merely a method-
ology for the recovery of data. In the 1980s and 90s, the challenge to
recover not only material culture and praxis, but cognition, appeared
simultaneously with the more profound challenge of determining
whether one could ever really “know” the past, anyway. Suddenly, the
archaeology being practiced was Foucault’s. In contrast to the work of
Binford or Winckelmann, whose mid-eighteenth century archeology
consisted of fetishizing the “trace” of the past and inscribing it in his
own image, Foucault’s was a theoretical alternative to mere “history,”
because it (archaeology), by virtue of its methodology, had to remain
conscious of the embeddedness of discourses and the processes by
* This article originally appeared as “Exhibit/Inhibit: Archaeology, Value, History in

the Work of Fred Wilson,” in New Histories, L. Gangitano and S. Nelson, eds., Boston:
Institute of Contemporary Art, 1996, pp. 182–91.
482 chapter thirty-four
which knowledge accumulated.1 The past was a foreign country. . . . A

culture cannot/can only be “known” by a non-native. . . . There were
“other ways of knowing”. . . .
If this trajectory seems to map quite comfortably over the contem-
porary art world’s sequence from materialist minimalism through neo-
expressionism and biography to autobiography and the postmodernist
crisis of nontransparency across cultural, subcultural, and historical
boundaries—in short, to “new histories” in the visual arts—it is no
accident. Not, at least, if one was raised a structuralist, grasping for
ways not to throw the baby out with the bathwater: hoping that the
entirely justified critique of a too-rigid linguistics-based structuralism
can give rise not to post- but to neo-; to better-calibrated, more mul-
tivalent ways of describing structures without abdicating the need to
account for the positionality of the analyst or the social embeddedness
of a given discourse.2 What is to be redeemed thereby is not only the
legitimate seeking of structures, but also the theoretical possibility that
an historical moment can be marked by patterns to be found in (some)
artistic and in (some) intellectual practices.
At present, the doing of archaeology can be understood not so much
as the means by which one knows the past, but as a way of generating
the past.3 And the worker in the field, the archaeologist, then, provides
the materials from which that past is generated, from which history is
written. To become “knowledge,” archaeological materials need more
than merely to exist. They must be ordered, classified, analyzed, and
attached to a narrative. Premise, therefore: romanticism and rituals by
which the inanimate is declared animate aside, it is not the materials
that “speak,” but the writer(s) of the narrative.
1
See, on this, the development from Michel Foucault’s The Order of Things; An Archaeology
of the Human Sciences (New York: Random House, 1970) [orig. publ. as Les Mots et les
choses: Une Archéologie des sciences humaines (Paris: Gallimard, 1966)] to The Archaeology of
Knowledge, trans, by A. M. Sheridan (New York: Pantheon, 1971) [orig. published as
L’Archéologie du savoir (Paris: Gallimard, 1969)], where the “archaeology” of the
subtitle moves into the title, and becomes in effect a metaphor for that consciousness
that the practice of “history” is said to lack.
2
Manfred Frank, What is Neostructuralism? [Theory and History of Literature, 45]
(Minneapolis: University of Minnesota Press, 1989).
3
Among the many discussions on this topic within the field of archaeology, see Proc-
essual and Postprocessual Archaeologies: Multiple Ways of Knowing the Past, ed. by Robert Preucel
[Center for Archaeological Investigations, Occasional Paper No. 10] (Carbondale, IL:
Southern Illinois University, 1991), for an attempt to reconcile earlier methodologies
with more current perspectives.
exhibit/inhibit 483
The work of Fred Wilson interrogates some of those prior narratives,

and the history they have produced, through a process of re-examination
and re-inscription. His presentation of almost-but-not-quite-entirely
Egyptian artifacts in a series of vitrine installations at the Whitney
Biennial of 1993 played with our heads in obvious ways, like the
rendering of a black rather than an idealized white Nefertiti, or the
presentation of an unfamiliar series of shawabti figures to accompany
the dead; but in more profound ways, the change of venue from an
archaeological/natural history museum to one of contemporary art
and the play of label content raised the specter of Western civilization’s
most self-serving manipulations (for example, fig. 1).
Wilson is not alone in turning to the archaeological record as a vehicle
for probing the ironies of the social sphere. Sam Wiener, in collabora-
tion with his alter-ego, Evangeline Tabasco, presented a large-scale
installation at the O. K. Harris Gallery in 1980 and in Philadelphia
in 1981, entitled Splendors of the Sohites, produced by the Metropolitan
Container of Art (sic!), that purported to be “an exhibition of master-
pieces excavated in SoHo.”4 However, ancient Egypt was the immediate
visual referent: the interior of the standard, grafittied New York City
trash container was outfitted with a series of study cases holding the
artifacts which were displayed to recall the Metropolitan Museum’s
blockbuster exhibition of the treasures of Tutankhamen. The installa-
tion participated in a known genre of archaeological spoof and fake
historical reconstruction, through the invention of a pseudo-archaeo-
logical assemblage and its excavation history, and then the inversion of
archaeological interpretation such that the “not dis-confirmed” ascrip-
tion of meaning borders on the ridiculous. Here, the subtext was less
a challenge to ethnocentricity than a send-up of over-interpretation
and the fetishization of cultural icons—as best known, perhaps, from
literary works of the same era, such as David Macaulay’s 1979 book
Motel of the Mysteries.
Over the past two decades, the work of Susan Hiller, American-born
artist working in London, has used archaeological materials to create
discomforting juxtapositions between language and image that have
powerful social referents. In one piece, entitled A’shiwi/Native (fig. 2),
she combines a poem called “Indian Children” with a dozen pottery
4
Sam Wiener and Evangeline Tabasco, Splendors of the Sohites (New York: The Met-
ropolitan Container of Art, 1980).
sherds from the American Southwest of the pre-contact period.5 The

poem is reproduced as a page of a 1920s primary school reader, along
with a typical silhouette illustration of the period, showing two clearly
“native children” (fringed clothing, feathered throwing stick), one male,
one female, in a natural setting of mountains and pine trees; it is set
in the lid of a carefully crafted box, while the sherds are equally care-
fully arranged inside the box in plastic envelopes, pinned in rows like
entomological specimens. The juxtaposition creates words between the
words of the poem and the artifacts, even if one does not immediately
recognize the sherds as representative of a Native American culture:
“Where we walk to school each day/Indian children used to play. . . .,”
says the beginning of the saccharine verse, to which the sherds bear
material witness. But the sherds are in fact evidence of an adult’s, not
a child’s, cultural production; it is we who have gathered and classified
the artifacts; and the final lines of the poem make the cultural transi-
tion to dominance clear: “What a different place today/Where we live
and work and play!” (emphasis mine).
While the naive presumption of eminent domain in the poem makes
one cringe today, the placement of the sherds in proximity to the primer
page serves as a reminder of other histories—Wounded Knee, small-
pox in blankets—and the more recent challenges to ethnocentricity, as
well as to the “rights” of one culture to do archaeology upon another
culture in the name of a purported universalist drive for knowledge.
This last issue is the subject of much debate in the archaeological
community as well, as traditional cultures claim their own “rights” to
sacred artifacts and practices and, above all, to the acquisition and
distribution of knowledge.6
Wilson’s Egyptian work is also about ethnocentricity—the ethno-
centricity of archaeological interpretation, itself subject to a historical
5
Susan Hiller, After the Freud Museum (London: Book Works, 1995), No. 010, boxed
1993. (My thanks to Yve-Alain Bois for this reference and the initial loan of his copy
of the volume!)
6
See, for example, Who Owns the Past? ed. by Isabel McBryde, Papers from the
Annual Symposium of the Australian Academy of the Humanities (Oxford and New
York: Oxford University Press 1985); Phyillis M. Messenger, The Ethics of Collecting Cul-
tural Properties: Whose Culture? Whose Property? (Albuquerque: University of New Mexico
Press, 1989); Timothy Murray, “Communication and the Importance of Disciplinary
Communities: Who Owns the Past?” in Archaeological Theory: Who Sets the Agenda? ed.
by Norman Yoffee and Andrew Sherratt (Cambridge: Cambridge University Press,
1993), 105–116; Frederick A. Winter, “Who Owns the Past? Ethical Dilemmas in
Contemporary Archaeology,” in Explorers Journal 69 (1991): 128–133.
exhibit/inhibit 485
particularity with its own cultural text.7 However, his is not a revisionist
reading that claims a different history as truth, à la Cheikh Anta Diop’s
The African Origin of Civilization: Myth or Reality, Civilization or Barbarism—a
reading that inscribes Egyptian civilization in Africa, and gives rise to
a powerful countervalent force often labeled “Afrocentrism.”8 Rather,
neatly avoiding the determination of criteria by which validity may be
established, Wilson destabilizes the authority of the Eurocentric reading,
and thereby any monolithic, exclusionary claim on history.
This process of destabilization is applied not only to Egypt and “other
cultures,” but also to art works in the Western canon. In an invited
intervention in the Parrish Art Museum, Southampton, New York, in
1992, Wilson tore out pages from H. W. Janson’s classic text, History of
Art, scattered them on the floor, stuck them on the wall and tucked them
under frames of religious works by Renaissance artists (fig. 3).9 The
challenge he posed was that the intellectualized description, attribution
studies, and the chronological sequence of the survey text might not be
the only, or even the best, way to gain access to the work.
What I find most unsettling is that Wilson’s installations and interven-
tions not only destabilize the “history” traditionally attached to specific
works and cultures, they also destabilize me! By posing a challenge to
governing paradigms and juggling with classificatory systems, he puts
the viewer in an odd position vis-à-vis her own “history,” departing from
the logic of Linnaeus and moving toward the surrealism of Borges’ odd
clusters.10 I am forced to conclude that an African-American viewer
could have a very different response to an Egyptian art now inscribed in
Africa. Yet, as a non-African-American, I am forcibly alienated from the
presumed transparency of the dominant view of the past, and thereby
7
As has been discussed most frequently in recent years with respect to the cultural
goals of “biblical archaeology”: see, for example, Neil Asher Silberman, Digging for God
and Country: Exploration in the Holy Land, 1799–1917 (New York: Doubleday, 1982).
8
Gerald Early, “Understanding Afrocentrism,” Civilization 2:4 July–August 1995):
31–40.
9
Donna De Salvo, Past Imperfect: A Museum Looks at Itself (New York: The New Press
and The Parrish Art Museum, 1993).
10
Cited by Foucault, The Order of Things, xv, from Borges’ fabulous “Chinese ency-
clopedia” that classifies mammals into groups according to the following criteria: a)
belonging to the emperor, b) embalmed, c) tame. . . . . . . j) innumerable, k) drawn with
a very fine camel hair brush, l) et cetera, m) having just broken the water pitcher, n)
that from a long way off look like flies . . .—the apparent absurdity of which forces into
consciousness the possibility that our own system may be not only as impenetrable to
some other culture, but also apparently absurd!
from the past of my own cultural reading; and, as an archaeologist, I

am confronted with the problems inherent in the claim to a value-free,
scientific “archaeology” and its attendant typologies. In other words,
emphasis on the diversity of alternative constructions of history leads
inevitably to dis-ease and a sense of displacement.
Now, as a radical with a seditious mind, this may give me a certain
frisson, or indeed a great deal of pleasure. But I can no longer touch
“the past” through the glass of a museum exhibition. Or, I touch a
past that prohibits nostalgia.
This is perhaps most evident in Wilson’s site/institution-specific instal-
lations, whether for Boston, Seattle or the Maryland Historical Society.11
The powerful vision of nineteenth century wooden cigar-store Indians
turned, not as artifacts on display toward the viewer, but toward a wall
of photographs of actual, mid-twentieth century and contemporary
Native Americans from Maryland (fig. 4), speaks to essentialism and
stereotype in a voice not unlike that of Susan Hiller’s boxed A’shiwi.
In the same exhibition, the inclusion of nineteenth century steel slave
shackles in a case with ornate nineteenth century Baltimore repoussé
style silver tea service and wine goblets (fig. 5), or the masking of an
eighteenth century painting of a high-born white boy of about eight
years of age and his black servant to show only the servant’s head and
shoulders, makes it impossible to distance oneself through intellectu-
alization and thereby avoid a sense of complicity in what has been a
very selective writing, much less performance, of history.
The slave shackles further force the question, “What is Art?” upon
the archaeological artifact. Do the shackles become art because placed
in a case by a contemporary artist, like Duchamp’s bicycle wheel? Are
they art by virtue of certain formal properties, as a carved Northwest
coast fish paddle or a Guro mask is “promoted” from the category of
cultural artifact to the category of art by a Euro-American consumer?
Or because guilt pushes one to include the formerly excluded, regard-
less of inherent, or defined, properties attributed to a class called “art”?
That things have histories comes as a surprise to no one; indeed, recent
academic scholarship has been stressing the “biography” of things.12
11
See, for example, Lisa G. Corrin, ed. Mining the Museum: An Installation by Fred Wil-
son, ed. by Lisa G. Corrin (New York: The New Press; Baltimore: The Contemporary,
1994). (For which, my thanks to NB, Christmas 1995).
12
See the essays contained in The Social Life of Things: Commodities in Cultural Perspec-
tive, ed. by Arjun Appadurai (Cambridge: Cambridge University Press, 1986), and in
exhibit/inhibit 487
(Although I would prefer to retain the term history as a caution against

giving animate agency to the thing instead of the human makers/
users/movers of the thing.) But the question at what point and by
what measure the thing becomes art, not artifact, must, I would submit,
remain contingent upon the particular maneuver to which it is subject.
To that end, and because I do have a seditious mind, when Chair
of the Department of Fine Arts at Harvard, the two art installations
in the Chair’s office were a nineteenth century cast of a ninth century
b.c. Assyrian relief, reminder of my area of scholarship while diverted
to administration, and a four-tiered plexiglass shelf rescued from dis-
card, over which was placed an oversized wall label reading “. . . but
is it Art?” On the tiers, by dint of gift and scavenging, were set
objects ranging from a broken rear-view mirror, the cracks of which
could rival a Ming Chinese vase, to an Eskimo tourist doll, a highly
decorated paper box for henna (the application of which is itself an
art form in many cultures), and a 1950s “Painting by Numbers” of a
mallard on the wing. Many colleagues and students contributed to this
display; only one ever explicitly objected! But the display, like Wilson’s
shackles, was unsettling. And I have come to realize that what was
unsettling was precisely the instability of classification.
In the same way, that Fred Wilson chooses not only to collect but to
display Black Mammy cookie jars of the 1930s and 40s is also unsettling.
Does my own rather extensive collection of “Paintings by Numbers”
mock true Art? Or mock myself in my professional investment in Art?
Or does it simply serve as a healthy, if ironic, reminder that all defini-
tions are fluid, contingent, and demanding of contextualization? So with
Wilson’s Mammies. That he collects them as artifacts puts the cookie
jars in the same class as Andy Warhol’s bakelite: they are the products
of a sort of “historical archaeology,” in which the pieces have been
unearthed in a series of expeditions, although different from archaeology
in that the findspot is neither recorded nor important; what is important
is the “assemblage”—another archaeological term, signifying the massed
representatives of an artifact class. But that Wilson, unlike Warhol, elects
to display them as part of a museum installation says something more.
Through the formal variations of the “type,” one sees the artifact class
as a class; through the existence of the class, one sees a social category
particular, Igor Kopytoff, “The Cultural Biography of Things: Commoditization as

Process,” 64–91.
of a given time and place. To experience the full impact of the class,
it is conceivable that one also needs to know the collector/exhibitor,
for the class signifies differently, speaks to a different history, if collected
by an African-American artist, a southern middle-class housewife, or a
bemused Angelino with a fondness for the 1940s. Does the same class
of artifact, then, not have contingent histories, plural?
Here we are, back in Egypt. Or, on the Bowery, in the loft where
Wilson now resides, and where, it turns out, I myself lived during the
writing of my doctoral dissertation in the early 1970s. In my single
foray into artistic production, for a collective work called About 405 E.
13th Street, organized by Jean Dupuy sometime around 1972, I fixed
a potsherd from the site in Iran where I had been excavating to the
rough brick wall, and then set two labels—one below the sherd read-
ing: terracotta, Iran, second millennium B.C.; the other, floating free on the
wall itself, reading: terracotta, New York, second millennium A.D. Although I
am not certain I could have articulated it at the time, I was aestheti-
cally interested in the shared materiality of the two terracottas, the
brick and the sherd; but I was also concerned as a graduate student
about the spatio-temporal gap between the present and the past I was
endeavoring to reconstruct, and I wanted my intervention in the loft
space to force a confrontation with that divide.
I rather wish that Fred Wilson had been in the building for that show
(although the laws of physics are clear that two bodies cannot occupy
the same space at the same time). The very fact of an alternative space
exhibit called into question the more “legitimate” venues of the public
art institutions, museums and galleries, just as Wilson’s installations take
on the language and practice of museums and other cultural institutions.
But he now is far more subversive than I was then. Wilson’s “history” is
a history born of archaeology, clearly recognized as a discourse subject
to excavation and decipherment. The challenge he poses for governing
paradigms and their adherents is not an amateurish glimmer of early
post modern un-ease, but a direct, fully developed, late post modern
gauntlet flung: the familiar can be defamiliarized via context; the past
is a construct; archaeology is not free of ideology; value is contingent;
alternative histories can, and will, be written.
exhibit/inhibit 489
Figure 1. Fred Wilson, Claiming Egypt, 1993. Mixed Media, site-specific instal-
lation (at the Whitney Museum, N.Y. Courtesy Metro Pictures, New York).
Figure 2. Susan Hiller, A’shiwi/native, 1993. Mixed media. Courtesy Book

Works, London.
exhibit/inhibit 491
Figure 3. Fred Wilson, Installation at the Parrish Art Museum, 1992. Courtesy
The Parrish Art Museum, Southampton, NY. Photo: Bill Jacobson.
Figure 4. Fred Wilson, Mining the Museum, 1992. Installation at the Maryland
Historical Society, Baltimore. Courtesy Metro Pictures, New York.
Figure 5. Fred Wilson, Mining the Museum, 1992. Installation at the Maryland
Historical Society, Baltimore.
CHAPTER THIRTY-FIVE
CHANGE IN THE AMERICAN ART MUSEUM:

THE (AN) ART HISTORIAN’S VOICE
I was asked to participate in this challenging program largely as a result

of a conversation with John Walsh, director of the J. Paul Getty Museum
and president of the AAMD in 1989–90, in which I described some of
the recent directions of my own scholarship, as well as concerns that I
felt had contributed to my having been named to the Search Commit-
tee for a new director of the Harvard University Art Museums. The
question I was asked to address was: What does the current breed of
art historian want from the art museum, and how does that intersect
with some of the other issues before the museum today? In other words,
how do recent and continuing changes in the discipline of the history
of art translate into challenges for the museum in an age of ethnic,
cultural, and theoretical diversity?
I will not attempt to speak for the entire field of art historians, but
rather for myself—a scholar of the art of the ancient Near East, an
area perhaps less central to the discipline than some. Still, the ancient
Near East can claim not only the first “art collector” documented in
history (in the person of an Elamite ruler, who carried off to his capital
in Susa large numbers of Mesopotamian monuments, including the
famous Law Code of Hammurabi now in the Louvre); but also the
first “museum director” in history (a Babylonian king, who, with his
daughter as chief curator, purposively gathered and displayed works
of the ancients).
While I imagine that the targets of concern I flag here are shared
across many subfields in the history of art, some will be more of an
issue within certain fields than in others. It is certainly easier to isolate
* This article originally appeared as “Change in the American Art Museum: The
(an) art historian’s voice,” in Different Voices: A Social, Cultural and Historical Framework for
Change in the American Art Museum, M. Tucker, ed., New York: Association of American
Art Museums, 1992, pp. 30–57.
494 chapter thirty-five
areas of concern than it is to come up with a series of specific proposi-

tions for change in the art museum; but in most instances, responsive-
ness to those concerns will necessitate change—both in policy and in
practice.
I would begin by stressing that a number of the oppositional stereo-
types permeating art-related fields, such as “museum curators versus
academics” and “connoisseurs versus social historians,” are applied in
reductive, simplistic, and in the end counterproductive ways. One of
the more difficult public talks I have had to prepare in my professional
life was on “The Nimrud Ivories” at Harvard University in the spring
of 1983. I imagined Sidney Freedberg on one side of the room (the
right, of course) and T. J. Clark on the other, arch-paradigms of the
object-oriented connoisseur and the context-oriented social historian.
Given my perceived gap between the two, the problem was how to
hold both with my predominantly archaeological material. I pulled a
copy of Freedberg’s Italian painting catalogue, Circa 1600, and Clark’s
Courbet book, Image of the People, finding in fact that there was a con-
siderable degree of social and historical context in the former and an
equally considerable degree of careful visual analysis in the latter. It
was indeed secondary consumers in the field who had polarized the
two, and we were all the poorer therefor. Different legitimate scholarly
problems will call forth different approaches and degrees of focus on
the object and/or its context. I am persuaded that we do damage by
perpetuating false dichotomies and then assuming the alliance of the
museum with the “connoisseur,” to the exclusion of the “historian.”
And yet, to some degree there are fundamental differences between the
museum and the academy. More often than not, it is the history that
is left out of the art history in the museum exhibition space, with its
focus on the object. I will return to that as a major issue below.
This is clearly a period when the very premises upon which the great
art museums were built and continue to function are in question. I cite
only one manifestation of the dilemma: a recent issue of Art International
(vol. 10 [Spring 1990]), in which an article entitled “Art Is Everywhere
Except in Museums” is followed by one titled “Museums Have Never
Been More Central to Culture.” It is clearly a fruitful time to raise
issues of concern to the art historian and to work toward common
goals. In pursuit of my charge, then, to isolate areas of concern that
pose challenges to the art museum of the 1990s, I will focus on three
main issues: (1) acquisition policies; (2) exhibitions—choice and design;
and (3) relationship to targeted audiences.
change in the american art museum 495
At the same time, it is important to state at the outset that it is appar-

ent from recent installations and programs that a number of museums
have already begun to address these issues as well as changes in “the
canon,” in the field of art history, and in the world of art in general.
What follows is perforce not a comprehensive statement so much as a
highlighting of issues.
1. Acquisition Policies
In recent years, museum policy regarding acquisition has been the

subject of much discussion. Issues raised include the ethics of acquir-
ing objects whose trade across national frontiers is restricted, or objects
with no clear title, even when there are no constraints on trade in the
category of object per se, as they may be stolen property; the ethics of
using coercion, monetary or otherwise, in the gathering of ethnographic
art collections, particularly as functioning cultural artifacts have taken
on the status of “art”; and the ethics of underwriting the destruction
of historical sites in the pursuit of individual, exhibitable pieces.
I will restrict my comments to issues directly relevant to my own
scholarly work. For almost twenty years now, the archaeological com-
munity—individuals, as well as such professional organizations as the
Archaeological Institute of America and the Society of American
Archaeology—has been quite vocal on the problem of the illicit antiqui-
ties trade, but this has had little impact upon the major art museums.
For me, trained as an archaeologist as well as an art historian, there is
nothing sadder, more frustrating, more enraging, or more destructive
of the human history I am committed to studying than a bulldozed,
dynamited, or pot-hunted site covered with the detritus of illicit diggers’
work that represents the disturbed debris of human history; or a primary
cultural monument that has been desecrated by removal of integral
sculptural or fresco decoration to feed the appetites of museums and
private collectors.
The archaeologists’ rage has been too often dismissed as overzealous,
so let me emphasize that archaeology is a discipline that in the process
of its practice destroys an essential part of its own data base. All the
more essential, therefore, that recovery be systematic, well recorded,
and made public; for once an archaeological stratum has been dug
through, the particular constellation of evidence and context can never
be reconstituted. The self-serving operations of the illicit treasure hunter
destroy context: “unmarketable” objects are randomly discarded, often

destroyed in the process; assemblages such as tomb groupings, which
would permit chronological and cultural analysis, are dispersed. The
result is that the single “art” artifact is extracted and preserved at the
cost of the obliteration of its own history.
The institutional justifications for acquiring these objects are familiar
to all: how the museums and collectors of the past have been able to
preserve works that would otherwise have been destroyed; and how,
once the piece is extracted from its original context and place of
origin, there is nothing that can be done to salvage the site, so the
piece might as well be acquired by a museum, rather than buried in
a private collection.
Justifications such as these come very close to willful ignorance. Every
purchase fuels the continuing search for yet more marketable works,
which then contributes directly to the further disruption of sites and
the destruction of evidence in clandestine searches for the saleable
unitary piece. In short, these searches are destroying the very base from
which we can reconstruct any understanding of the original contexts
of the works that then turn up in collections, often leaving questions
of authenticity unanswerable.
While it is historically accurate that nineteenth-century and some
twentieth-century collectors and collections may have served to preserve
works, this is no longer relevant. Virtually all of the countries of origin
now have their own departments of antiquities and museums. More-
over, we are living in an international environment in which most host
countries and virtually all professional organizations have ratified the
UNESCO convention governing the protection of cultural properties.1
Thus, to purchase such works today means that a given museum and
its trustees and director are complicit in the illegal smuggling of such
properties out of their countries of origin and are either directly contra-
vening an international agreement by so doing, or using a not-so-subtle
dodge by moving a work through another country, often Switzerland,
to fudge its “immediate origin.” At the same time, the museum is
actively contributing to the destruction of human culture, for when
1
Drawn up in Paris, November 1970, as the “Convention on the Means of
Prohibiting and Preventing the Illicit Import, Export and Transfer of Ownership of
Cultural Property.”
the ancillary damage attendant upon illicit extraction and acquisition

is weighed against the piece that comes out, the balance is clear.
Indeed, the Joint Professional Policy on Museum Acquisitions Resolu-
tion of the American Association of Museums recognized the impor-
tance of establishing guidelines for the acquisition of cultural properties,
formally resolving that its member organizations cooperate in the pre-
vention of illicit traffic in such properties (see Museum News 52 [1973]:
48). Yet signatories to the document were essentially committee-chairs,
not museum trustees, directors, curators, or representatives of specific
institutions.
While neither I nor any other concerned scholar have any levers
with which to persuade a particular museum director of the virtues
of abiding by the UNESCO convention and/or other international
professional guidelines, I must say that the continued acquisition of
“hot” properties, be they African, South Asian, Mesoamerican, Greek,
or ancient Near Eastern, definitely has an effect upon the scholarly
expertise available to the collecting art museum; and this in turn has
an effect upon knowledge, both at the curatorial level and at the level
of public education.
One major art museum has recently mounted an exhibition of a
private collection of antiquities, and some of the best scholars in my
field, when invited to work on the show, decided not to do so. After
years of commitment to the history part of art history/archaeology and
the contextualization of works under scrutiny, it was deemed impossible
to devote one’s energies to the legitimation of works illegally torn from
their context—some of which had actually been physically hacked out
of architectural structures.
In my own case, I have turned down opportunities to publish interest-
ing objects from—even to talk to docents of—institutions notorious for
their active acquisition of antiquities, feeling that I would be lending
myself to acquisition policies that are tantamount to cultural and his-
torical destruction. I have discussed these issues with curators at these
museums, many of them otherwise good colleagues. In many cases,
they understand my position, but too often they feel forced into their
stances by premiums placed within their institutions on the glories of
the chase, and on success being defined as adding to the collections at
any cost, rather than as careful husbanding and scholarship.
I know of only one curator in a major art museum who, at consid-
erable personal cost, has been consistently in the forefront in raising
the issue of acquisition policies, who has consistently maintained an
exemplary distinction in his publications between museum objects of

excavated provenience and those of “attributed origin,” and who has
also pressed for “truth in labeling” when exhibited pieces come from
the art market and not from excavated contexts. I find particularly
offensive those curators and directors who pretend to more virtuous
policies than in fact they practice. These double standards become
quickly known within the scholarly community, as when several years
ago the director of a major urban art museum appeared on television
attesting to policies I knew from direct experience were commonly
disregarded in at least the two departments with which I was familiar.
His pious disavowal may have helped a public image (and even public
funding), but eventually the community of concerned scholars does
begin to close around an institution. In the case I cite here, when the
Archaeological Institute of America held its annual meetings in that
city, the museum was eliminated from the list of participating/hosting
institutions. Ironically, by then the museum in question had a new
director, one of the departments a new curator in charge, and more
rigorous acquisition policies had begun to be enforced; but it was still
felt that, in the absence of a public acquisition policy that met inter-
national guidelines, the gesture had to be made.
Some museums have been more active than others in collecting
antiquities, occasionally quite openly, other times hiding behind spurious
documentation that they and everyone else knows has been contrived
by interested parties involved in the sale. University museums have been
more willing to abide by the UNESCO agreement and refrain from
acquiring objects that cannot be demonstrated to have come out before
its implementation in the early 1970s. It has cynically been said that
in many cases these institutions have more to lose in punitive actions
from host countries issuing excavation permits, or that their acquisi-
tion budgets are minimal anyway. It has been my experience, however,
that in fact these institutions have come to better understand—through
faculty concerns and ongoing excavation projects—the damage done
in the name of the art market and the terrible cost of the destroyed
evidence that is the by-product of illicit extraction (see, for example,
Miller 1982, for a discussion of the consequences to Maya monuments
of Mesoamerica), particularly since, as university museums, their primary
commitment is often less to the object for its own sake than to it as a
trace, functional as well as aesthetic, of a reconstructable past.
Some public art museums and a number of university museums
therefore have laudable, published acquisitions policies that proscribe
the purchase of objects without provenience. The curators and/or

professors involved with acquisitions at these institutions have generally
adhered to the published institutional policies in making purchases.
Where I would target a problematic grey area, however, is in the realm
of gifts and bequests. In a hypothetical example, donor X is interested
in giving a major collection of South and Southeast Asian art to a
museum. Of the seventy objects in the collection, five or six are of
primary importance; thirty or so add significantly to the museum’s hold-
ings; the rest would contribute to the study collections. The museum
is therefore most desirous of acquiring the collection; however, of the
seventy works, some ten (or twenty, or half ) of the pieces were acquired
since 1970, in direct contravention of the UNESCO accord and the
policies of the countries of origin. What to do? Does the otherwise
scrupulous curator or director decide to finesse this one, in the face
of a greater good—the importance of these works to the collection?
Or does he/she risk losing the collection, by raising the issue with the
donor? If she/he were to say that the museum would be delighted to
accept sixty (or fifty, or thirty-five) of the pieces, but that its acquisi-
tion policies prohibit taking the others, then inquire whether the donor
would consider making that subgroup a gift elsewhere, what are the
chances that the whole group would be offered elsewhere? If the col-
lection is accepted and the questionable pieces made public, available
for repatriation or claim from the originating culture(s) (see Greenfield
1989; Murray n.d.), what sort of burden will this place on museum
resources—both staff and financial?
It seems essential that these issues be placed before you. I believe
there is a great deal a museum director can do: in setting institutional
policy and then assuring that those policies are in fact carried out, not
treated as public pieties to be ignored; in seeing the important costs
of certain kinds of collecting; and in establishing a tone in his or her
institution that will not reward success in the “hunt” over appropriate
display and study.
I understand that to ask art museums to restrict their acquisition of
at-risk cultural properties is asking a good deal, especially when the
unique piece comes along, or when there are holes in the collection,
the filling of which would serve ends perceived to be important. But,
given the destructive consequences of such collecting, I also see this
issue to be one of ethics, not just of divergent practices (see Messenger
1989). One constructive way in which some museums have been able
to continue to add to their collections is to turn the resources that had
been put to acquisitions from the market into support for excavations,
particularly in countries that still have a policy of dividing finds between
host and sponsoring institutions, so that legitimate, provenienced objects
can come to the museum for study and display. In addition, I believe
that there are things academic art historians and archaeologists can
do in concert with institutions whose policies follow guidelines con-
gruent with both ethical and didactic goals: for example, aiding in
the construction of bridges between institutions here and abroad that
would result in special exhibitions, long-term loans, and/or the legiti-
mate transfer of redundant objects that have been properly excavated
and studied and now pose enormous storage problems for their host
countries. But any such effort, on my part at least, would have to be
preceded by adherence to a serious public policy of nonacquisition of
cultural properties that have been illicitly excavated, stolen, or illegally
exported, in conformance with current international law. The potential
benefits are great, as witnessed by the recent successes of loan exhibi-
tions from several Italian museums presented at the Emory University
Art Museum, whose collection is predominantly ancient, but where
UNESCO guidelines have been followed to date. It is clear that without
such policies, the director would not have been able to negotiate these
exhibitions with the host institutions, to all of our loss.
2. Exhibitions—Choice and Design
The issue of the nature and quality of museum exhibitions is currently

under scrutiny from without and within the institutions (see Karp and
Lavine 1991). It must be divided into two subsets: what museums choose
to exhibit, in both permanent and temporary installations, and how
that material is shown. Turning first to the what, I would emphasize
the relevance of much recent discussion in art historical and critical/
historical literature on the founding of the great museums in the
nineteenth century and the establishment of a “canon” of Western
art. The introduction of non-Western works into this system has also
come under recent scrutiny, requiring the separation of “beauty” from
“social meaning,” rather than an integration of the two, as charted by
Sally Price (1989); then the channeling of works into what James Clif-
ford (1987) has called the “art-culture system,” through which objects,
decontextualized from the cultural life of their original making and
usage, have been promoted to the status of “art” and re-contextualized

in public collections.
While a consensual definition of art in the West has long eluded phi-
losophers, aestheticians, and historians, it has been possible to assume a
class of objects belonging to a category called “art,” to which has been
ascribed qualities of beauty and/or aesthetic interest that transcend
membership in any one culture. Many objects so considered come
from traditions that have no word or class for art per se. It is our own
post-Enlightenment/nineteenth-century tradition that has argued for
the possibility that a historical work can be affective, can be understood,
independent of its history.
Herein lies the crux of the issue. Along with an increasing number
of art historians and archaeologists (cf. discussion in Murray n.d.), I
question the “universalist” approach as itself a historical phenomenon
and inherently flawed; and I stand as a scholar on the necessity of the
contextual on the one hand and the acknowledgment of the nontrans-
parency of the lens of our own time and space on the other, if one is
to grapple with the material culture of another time and place. In this,
I attempt to bridge both sides of the modernist/postmodernist divide,
in that I argue for the possibility of retrieving native, categories and also
of mitigating at least the biases of cultural difference. If the museum
is to maintain the position that the work is accessible on formal and
universal grounds, transcending culture, context, and function in its
own time, then the museum itself becomes a monument to a particular
philosophical (and essentially historical) position, and my cry for context
will carry little weight. If, however, the museum claims as one of its
goals the understanding of another culture, time, or place, then the
nature of the contextual must enter into its discourses on acquisition
and exhibition policy.
This argument is best seen in recent historical perspective. While
World War II effectively brought the United States into a new phase
economically and politically, shaking loose many of the assumptions
and tenets of the nineteenth century, it was not until the challenges
to social values of the late 1960s that there was sufficient distance to
understand just how much the premises we had accepted as universal
and “natural” were in fact “cultural” (historical) constructs, and as such
were subject to question and to change. Many of the great museums of
this country and Europe were born in that nineteenth-century environ-
ment of universalist certainty. To date, they have perpetuated in their
massive, neoclassical architectural piles the immutability of the concepts

upon which they were founded. Later institutions of modernist persua-
sion, while introducing new styles to architecture and in artworks, still
adhered to fundamentally similar precepts of the purity and autonomy
of works abstracted from their cultural context.
Recent challenges, coming largely from studies of gender, race, and
class, have gone to the very underbelly of these values, undermining
their foundations through an insistence that the meanings of these
founding concepts are themselves historically grounded and, as such,
open to contextual analysis and redefinition. As feminist Joan Scott
(1988) and others have argued, meanings, values, and concepts are not
fixed in a culture’s lexicon, nor are they fixed cross-culturally; rather,
they are social variables, dynamic, always potentially in flux, subject to
change with changes in historical conditions.
Similar gauntlets have been thrown down by representatives of
various racial and ethnic groups. Mexican American artist Guillermo
Gómez-Peña (1988), in challenging what he calls the “dominant culture’s
view of itself,” speaks of a “new hybrid culture” in which “concepts like
‘high culture,’ ‘ethnic purity,’ ‘cultural identity,’ ‘beauty,’ and even ‘fine
arts’—concepts upon which many of our major art museums have been
founded and continue to exist—are absurdities and anachronisms.”
Although often articulated in arenas outside of the art museum, these
various critical challenges—whether the high/low distinction is or is
not inherent in all stratified societies; whether multiculturalism does or
does not permit scales to measure “quality,” and then, by whom?—have
important consequences for the museum: they challenge the authority,
indeed the very existence, of what until now has been thought to be the
center of dominant (high) culture in this country, and it is this center
that major art museums have up to now been addressing. Conservative
critiques of what museums tend to exhibit have called for an expansion
of the “canon” to include previously disenfranchised cultures and art-
ists. However, to some critics, like Russell Ferguson (1990), to call only
for the admission of a wider variety of voices in the canon tends to
leave many of the most fundamental assumptions unchallenged—rather
like finally letting Jews/blacks into previously all-WASP/white country
clubs, or women into men’s business clubs, but not questioning the
fundamental existence of such exclusive private organizations.
I would imagine that this is as frustrating for the museums as it is
for many scholars: just when it seems that the canon might be open-
ing up a little, to include works heretofore hors de discours, the whole
process of authorization for the canon is being called into question.
Now that one can at last get into the club, the club system itself is
being dismantled.
While often conflated, it is important to maintain a distinction between
a canon as a body of valued works and the criteria by which value is
ascribed. The issue of whether the canon itself must be expanded, or
whether it is culturally possible to dispense with the notion of a canon
altogether, is far from resolved. In any case, a more polyvalent, less
authoritarian system of judgment seems called for, one that does not
abandon criteria of quality, but equally does not scale those criteria
along a single dimension. We can no longer stop with questions like that
posed by Linda Nochlin ([1973] 1988): “Why have there been no great
women artists?” Rather, we are asked to question the very criteria by
which membership in the class of “great artists” is accorded, an issue
raised by Anne Wagner (1989) in a paper on the work of Lee Krasner,
whose work has so often been compared with that of her husband,
Jackson Pollock. As long as terms like energy, power, and vigor form the
criteria of value, they may perforce privilege the “male” in work and
thereby the male artist. Implicit in these questions is the consequence
that pluralist values will lead to a plurality of valuables; and without a
single standard of measure, the historical role of the museum as arbiter
of quality is seriously compromised.
At the same time as criteria of value in style are being called into
question, so also are the limits of acceptable subject matter (iconog-
raphy), as witnessed by the response to the recent Mapplethorpe and
Serrano exhibitions. Even at less sensationalized levels, major shifts in
the imagery of contemporary art in recent years have brought into
play issues of sexuality, gender, social practice, history, and memory
that have not been the material of predominantly abstract modernist
art exhibitions. Shown to date mainly in galleries and museums of
contemporary art, these new works, by their content, have mounted
the same attack on a targeted Eurocentric, sexist, and elitist canon that
has been articulated by many art historians (see Michael Kimmelman,
review of The Decade Show at the New Museum of Contemporary Art,
the Museum of Contemporary Hispanic Art, and the Studio Museum
in Harlem, in the New York Times, 27 May 1990). Eventually, the large
art museums will have to come to grips with this art as well—art that,
by its very existence, challenges many of the fundamental hierarchical
concepts according to which the museums themselves are structured.
Closely related to what is exhibited in the art museum is how works are
exhibited. Much has been learned from recent analyses of placement
and spatial allocation accorded exhibitions in public art museums that
demonstrate an inherent privileging of certain periods and cultures in

a given community’s heritage (see Duncan 1991). At another level, the
challenges arising from new perspectives on race and gender that have
led to questioning assumptions of the canon also have consequences
for basic concepts such as “the autonomy of the art product” that
has governed most museum display. With rare exceptions, such as an
attempt in the Landesmuseum, Hanover, in the 1920s to create period
environments for the viewing of Western Art (Bois 1989), the underlying
premise of most major art museums, as noted above, has been that works
of any time or place have qualities that can transcend their original
meaning and context and thus speak to a transcultural audience. This
led inevitably to what Clifford (1987) has called the “boutique style” of
display—what I had independently dubbed for some ancient Near East
galleries the “Henri Bendel’s Christmas window” display mode. In one
case, a half-dozen variants of admittedly quite beautiful Iranian grey-
ware pots in a single vitrine are provided with virtually no information,
just the visual rhythm afforded by the varying heights of rectangular
bases; one could easily imagine an affluent consumer at Bendel’s select-
ing one of this type for Aunt Susan and one of that type for Cousin
Wilfred. The rationale is that with minimal noise, the display lets the
objects speak for themselves; but this rationale is predicated upon the
premise that a work can speak across spatial and temporal boundaries
in anything other than an entirely self-reflexive tongue.
The field of art history has, I believe, preceded the art museums
in dealing with these issues of de- and recontextualization of the art-
work and the problematic of “context” in general. As a problem in
historiography, the challenge of Marxian theory, which insists upon the
embeddedness of the work in an originating sociohistorical context and
political ideology, has been absorbed into the discipline as a kind of
new historicism that maintains that one can recover historical mean-
ing and function, thereby reconstructing both “affect” and “effect.”
At the same time, a more sophisticated understanding of context (see
Bal and Bryson 1991) acknowledges that context is not “given,” but is
subject to the same interrogation as the work of art in examining the
social functions that give meaning to objects. Recent meetings of the
College Art Association have thus become increasingly contextual (and
interrogative) in panels and sessions, reflecting significant shifts in the
way art historians understand and interpret works.
Some art museums have likewise begun to address these shifts—with
innovative displays providing culturally meaningful groupings, extensive
gallery labeling, the inclusion of video and audio tapes to provide wider
information, and occasionally, living practitioners that afford the viewer
access to the making and experience of the work, not just the object.
These experiments have, however, been largely limited to temporary
exhibitions. The prevailing model remains one in which Western art is
seen essentially as a monolith, the viewer moving chronologically or geo-
graphically from gallery to gallery as if there were no contextual variants
from, say, the religious to the private domain; the “art” of non-Western
cultures is accorded the same purportedly neutral installation.
The result of such a mode of display is that the viewer is given no
information on original function or context; what is available to one
not informed enough to contextualize from prior information is, if not
total indifference, at best the projection of one’s own cultural and aes-
thetic values upon the works. As anthropologist Robert Garfias (1989)
has observed: “Cross-cultural outreach often stops with the discovery
of elements in other cultures that reflect our own.” In other words,
there is no bridge provided that will allow for understanding of another
system of values altogether. Thus, we see the ninth-century Chola
bronze figure of a dancing Siva in a museum gallery as a masterpiece
of metal-casting, as poetic movement, as form, but are given no help in
understanding that these are ritual, processional images that in worship
are clothed and cared for as manifestations of divinity.
This decontextualization in the art museum comes as no surprise. It
intersects directly with issues of acquisition, for the major collections
were built, after all, in that same nineteenth century whose colonial
arms reached across continents to appropriate and represent the
“other” on “our” terms. Clifford traces this objectification back even
further, to the seventeenth-century emergence of an ideal self as owner
of property and goods. The collection of art objects became part of
the “amassing of a domain of identity”—one initially tied up with
nationalist politics and subsequently extended out into the colonies.
The history of collecting has thus functioned within a ramified system
of symbols and values attached to “goods” (this may also be a useful
perspective on why “exchange value” has inevitably come to be rated
over “use value” as works of art themselves become commodities [see
Appadurai 1986]).
Further, the relationship of the Christian West to the colonized
others was definitely not a relativist position of equal-but-different value
systems. I wonder if this assumption of a moral hierarchy may have
governed in part the choice not to give value to the contexts of works,
but rather to those qualities that were judged to transcend their contexts
(primarily form, and sometimes content), doing so according to our own
classification system, which was purportedly rule-governed and rational,
neither idolatrous nor fetishistically fixated. It was precisely by ignoring
the context of non-Western works that the art museum and the West
in general could distance themselves from immoral, messy heathen
practices. Rescuing the object and thereby permitting its transcendent
aesthetic purity to shine through must have been comparable to the
virtue accorded the missionary effort of saving a heathen soul!
The decontextualization of the non-Western work, and even of the
historically distant Western work, has resulted in a particular sort of
interpretive paradox for the art museum: having first cut the object out
of its context, the illusion of adequate representation of the originating
culture is offered by making the object “stand for” the abstract whole. In
an example cited by Clifford, a Bambara mask becomes the metonym
for “Bambara culture” (1987, 220), just as for many, an archaic kouros
speaks for all of ancient Greece. This massively distorts the polyvalent
nature of culture itself, wherein no object is ever a perfect isomorph
for the entire system, but rather at most functions within a particular
set of customs and behaviors appropriate to a specific moment, or a
specific subsystem within the larger culture. It is thus the curator who
makes the interpretative decision regarding which part will speak for
the whole, often according to very distorted and self-reflexive notions
of what constitutes the “other.”
The abstraction of the work from its context has also led to significant
distortions in interpretation, effectively privileging those formal aspects
of the work that can be described in so-called objective terms and
undervaluing iconography, since deciphering meaning requires access
to the cultural code employed in making and perceiving. As a salient
example, one may cite the early research done in the West on collec-
tions of unprovenanced Chinese bronzes—particularly that by Max
Loehr at Harvard. Loehr’s primary claim was that changes in decora-
tive forms were tied to a chronological sequence and could therefore
provide a means of establishing relative dating for these works. He then
argued that these forms were independent of meaning and subject to
autonomous evolutionary development. The stylistic sequence Loehr
helped to determine has held up remarkably over the years, as new
archaeological evidence has been able to anchor objects in absolute,
not just relative, chronologies; however, the work of numerous scholars
more versed in Chinese text and language, such as K. C. Chang in
anthropology and Susan Cahill in art history, have shown that Loehr’s
view on meaning was quite misleading, that there is meaningful and
decipherable content in the decoration of Chinese bronzes: a meaning
carefully coded to clan identity and the shamanistic ritual practices of
pre-Buddhist China.
This historical perspective on interpretation in the field is relevant
for our purposes because most museum displays of these bronzes have
been frozen at the Max Loehr stage of understanding. For purposes of
illustration, I cite the gallery of Chinese bronzes in the Sackler Museum
at Harvard, my own institution (fig. 1). The pieces are arranged in
freestanding cases, organized by “type”—that is, by particular form
of vessel. Nowhere in the gallery is there a label indicating that such
vessels were made specifically for tomb or ritual use, or that a given
tomb was likely to contain not groupings by type, as exhibited, but a
cross section of a number of forms, each particular to some aspect of
the ritual process attendant upon, or provisioning, the deceased. Such
information has been available for decades from (legitimate) archaeo-
logical excavations. That more informative installations are certainly
possible is witnessed by a recent exhibition of the same material at the
University Museum of Archaeology and Anthropology of the Univer-
sity of Pennsylvania (fig. 2). One is forced to conclude either that it
is considered easier for the art museum to achieve what is thought to
be a visually satisfying installation by ignoring culturally meaningful
groupings, or that the museum is functioning several decades behind
scholarly interpretation and understanding.
To me, this avoidance of function and context severely limits the
ability of the art museum to adequately address the nature of the mate-
rial it chooses to exhibit. It need not be seen as an either/or situation.
Typology, chronological development, and abstract aesthetics surely
have their place in museum presentation. Nevertheless, with the various
studies in recent years that have brought collecting and display into view
as non-neutral, highly charged facets of Western identity formation,
it may be time to explore how the art museum can play with multiple
display modes (Levine 1991b) and engage context. I believe this is not
only possible, but highly desirable. It entails neither stepping over a
line into becoming an “ethnographic” museum, nor losing sight of the
fact that contextual information is often partial (witness the number of
archaeological objects in museum collections that originated in graves,
where preservation is best, thereby skewing information toward treat-
ment of the dead as opposed to the living), that context itself a product
of interpretation (Bal and Bryson 1991, 176), and that “in-context

approaches exert strong cognitive control over the objects” on display
(Kirschenblatt-Gimblett 1991, 24; and Grimes 1990).
In the controversy between the often articulated and rhetorically
polarized positions that art can be understood only in its original
context versus the argument that aesthetic qualities transcend their
original articulation and can reach the uninitiated viewer on purely
formal grounds, I sense the same seeds of false dichotomy as in the
opposition of connoisseurship to social history referred to above. I see
no reason why an art museum cannot choose to display a typical tomb
grouping of a range of Chinese bronzes, for example, or include the
raffia costume that was worn along with an African dance mask (as, in
fact, was done at the Center for African Art in New York in 1989). Or
to provide enough information through ancillary texts, photographs,
and interactive technology to evoke cultural function and meaning for
the viewer, without competing with the works on display. At the same
time as anthropology is addressing aesthetics and art anew, reaching to
art history for the tools of stylistic and iconographic analysis, art history
is approaching anthropology, sociology, and history in dealing with the
particulars of sociocultural function. As traditional domains separating
the disciplines are being bridged, the question that must be put to the
museum director and curator is how these intellectual directions can
be translated into museum display and the museum experience. These
issues ramify not merely in terms of when and why the art museum
got to be the way it is, but also in terms of how the museum reaches
its viewing audience.
3. Relationship to Targeted Audiences
The museum-going public has grown in the past thirty years, although
studies have shown that the social profile of attendants has not shifted
in proportion to the increase in numbers. According to Louis D. Levine
(n.d.), director of the New York State Museum, despite changes in the
educational programs offered by major art museums and the targeting
of wider audiences by particular installations and exhibitions, there
seems to be not only no statistically significant effect on the overall
profile of visitors but also very little evidence of a “return effect” on
the part of the specially targeted audience once the exhibition in ques-
tion comes down. Some museums have been more responsive than
others to a wider public; yet, despite the best of intentions, attempts

to increase the base of public membership and attendance have not
met with consistent success.
I have not engaged in a systematic study of this issue; however, I
permit myself a few observations, as I have come to believe that ques-
tions of targeted audience are directly related to the questions of display
and acquisition discussed above.
I referred earlier to Russell Ferguson’s publication for the New
Museum of Contemporary Art (1990), in which he writes of the “phan-
tom center” of dominant culture in this country. Ferguson cites the work
of Audre Lord, who defines the representative of this mythical norm
as “white, thin, male, young, heterosexual, Christian, and financially
secure.” Since the early 1960s, the mythical norm of the art museum-
goer has been (à la our then-ideal, Jackie Kennedy) white, thin, female,
college educated, and financially secure—part of the post-World War II
vision of affluent, well-educated women with professional husbands but
no career identities of their own, having the leisure to include Culture
with a capital C in their lives.
The capital C is important, as it implies a definition of culture as
comprising the most refined products of an artist/civilization. In the
more anthropological sense, culture with a small c stands rather for
the entire system of beliefs and practices out of which both the artist
and the work are produced.
Which definition of culture one opts for has immediate implica-
tions for the what and how of exhibitions in the art museum; it also
determines strategies and affects success in reaching audiences targeted
by the museum.
A number of recent authors, including Garfias, Gómez-Peña, and
Ferguson, have emphasized the demographic shifts, growth of cultural
diversity, and increased value placed on “cultural identity” that has
characterized the United States over the past twenty years. These
developments are important components in the challenges that have
arisen to dominant culture and the canon, discussed above. Garfias,
in particular, has noted that the “cultural literacy” movement is in
effect a backlash response to the changing demographic and cultural
patterns that constitute the country today. Referring to an article by
Joseph Berger in the New York Times of 12 April 1989 entitled “Ibn
Batuta and the Sitar Challenging Columbus and the Piano in Schools,”
Garfias argues for seeing cultural diversity as a strength of the nation,
rather than as one of its difficulties; and he worries that the cultural
and educational institutions of the country have been unable to keep

pace with these changes.
It is important to see this in the historical perspective of the gradual
breakdown of the myth of the melting pot, according to which the
diverse immigrant populations of the country would eventually social-
ize into a single, monolithic American culture; and the rise of a sort of
communalism in the late 1960s and early 1970s, in which subgroup
identity, values, and distinctive traditions were elevated and attached by
hyphens to American—an emphasis on the e pluribus of the U.S. motto,
rather than on the unum.
If this is, as I believe it to be, an ineluctable societal fact of the 1980s
and 1990s, then the issue for the art museum director is the degree to
which one’s institution should be responsive to it. There are two basic
attitudes embedded in this issue: the benevolent and the pragmatic,
both of which seem to point in the same direction. The benevolent
attitude is apparent in the wish of most museum directors to be seen
as progressive and (for all sorts of reasons) to increase the number of
people coming to their museums. The pragmatic attitude seems to be
a function of the decline in financial support among traditional, white
Anglo-Saxon donors in proportion to ever-increasing costs. Some, but
not all, of the shortfall has been taken up by (equally elite and private)
corporate sponsorship. If the art museum is to appeal to public, govern-
ment agencies for support, it must demonstrate an investment in public
programs and a wide base of public interest across varying socioethnic
groups. Indeed, such demonstrations have become more important
in recent years, as government support for the arts has declined and
competition between institutions for grants has become ever greater.
Realizing goals of expanding the museum-going public has posed
a challenge for museum administrators. Lip-service gestures to under-
represented groups have clearly not been enough, and well-intentioned
installations have sometimes resulted in controversy (see Levine 1991a;
Livingston and Beardsley 1991). Gómez-Peña (1988) is critical of
attempts to recognize heterogeneity by employing terms like primitive art,
minority art, ethnic arts, or Third World culture that are blatantly ethnocen-
tric, implying a vision of the world still based upon an Anglo-European
model. He is not surprised when exhibitions with such titles are more
successful with traditional audiences than with minority groups. If, as
Ferguson (1990) has argued, independent cultural traditions are an
essential factor in creating positive self-identity for its members, and if
public art institutions have set as goals the affirmation of varied ethnic
audiences, then the art museum must take those traditions and values
into account, not merely the material products thereof, if it is to reach
those audiences.
For the Latin community, for example, curators and directors need
to be informed about complex issues of nomenclature in relation to
identity (see, for example, Gómez-Peña on Latin-American versus Chi-
cano versus Hispanic, etc.). And, as Amalia Mesa-Bains emphasized in
her talk in this same program and Tomás Ybarra-Frausto has discussed
elsewhere (1991), they must be sensitive to the fact that Western distinc-
tions of “folk” and “fine” art are blurred in many indigenous traditions,
where visual intensity and aesthetic effect are not defined in formal,
intellectualized classification systems, but rather are rooted in cultural
experience, often centered around the church and the home altar. The
decontextualized art object may not, therefore, be an effective way to
reach the Latin community.
One socially responsible museum director of a major urban museum
in a city with a relatively large Mexican-American population spoke
with some urgency of his desire to reach that community through an
exhibition using pre-Columbian archaeological materials to illustrate the
high level of civilization attained by their originating tradition. While
his goal was laudable, the sad irony is that the collection in question is
one consisting almost entirely of recently acquired, illicitly extracted,
illegally exported—in short, plundered—artifacts. Now, the Mexican-
American community is not uninformed about the plundering of
antiquities from archaeological sites. Not only are such shows gener-
ally organized around chronology and formal matters, which may have
little intrinsic appeal to the community, they are also, when the objects
involved are the products of looting, witnesses to the desecration of the
very originating tradition the exhibition seeks to glorify. Another issue
of which art museum staff have often been woefully ignorant is the
importance of regional identities within areas we perceive as culturally
one: it is not at all clear that an exhibition of Aztec art would speak
equally to modern-day Mexicans from Durango or the Yucatan, for
example. Is it any surprise, then, that the exhibition is written off once
again as “for the Anglos, not for us”?
The issues are similar for any ethnically identified community. The
question to be asked is for whom has the program or exhibition been
set up? Murray (n.d.) cites recent legislation in at least one Australian
state asserting that the aborigines in effect “own” their heritage; while
Karp and Lavine discuss the recent debate in New Zealand—following
an exhibition of Maori art in the West—over how to allow the Maori

to define their own heritage (1991, 2–3). Steps are currently being
taken in this country to ensure that when the Museum of the Ameri-
can Indian turns its vast collection over to the Smithsonian and a new
house for a representative part of that collection is built on the mall
in Washington, Native Americans will participate at all levels of infra-
structure and consultation. Yet, when we opt to represent the material
culture of a group distant in time or space from our own, how does
one establish criteria of appropriateness? How does one respond to
Grimes’s challenge (1990) that any display of Native American sacred
objects is tantamount to a public statement of the white man’s successful
desacralizing of Native American tradition? Is it enough to have Maori
chant over sacred sculpture as it is installed in the Metropolitan Museum
of Art? Is it truly to consecrate the works, or has the ceremony been
organized to titillate the circus-hungry Western audience with a taste of
the exotic, the “authentic”? And, if the question is not as either/or as
the above implies, is it possible to strike a balance between the dignity
of the material and the originating culture on the one hand, and the
need for public appeal on the other?
The recontextualization of formerly sacred objects into institutions
of public display happens not only in the West, but also in many coun-
tries with highly developed cultural traditions (particularly those with a
colonial past, where the museums were often set up by the occupying
culture and exist today as much or more for the tourist trade as for
local audiences). Again we must ask: For whose benefit? For whose
viewing experience? I have seen pious Hindus in Madras, as well as in
California, avert their eyes from certain images in museums, because
even if deconsecrated, these were once sacred images embedded
in ritual practice, meant to be clothed and ornamented, not shown
unadorned in public.
For the African-American community, many museums have been
less than sensitive about terminology and appellation. Although most
institutions made the shift years ago from Negro art and population to
black and now put quotation marks around primitive if the term is used
historically to describe African art, there has been resistance to the black
community’s preference in some instances for the use of the broader
identifying term people of color. The question is whose terminology should
prevail if the institution is attempting to reach particular communities?
At the same time, words like outreach, while seeming generous and well
intentioned, are often perceived as offensive, implying one-way activity

and condescension.
There is also a string of seldom-raised questions that must be brought
out into the open, if underlying and often unconscious biases are to be
examined and, hopefully, exorcised. Is it easier to organize shows of indi-
vidual (Hispanic or black) artists than it is to organize group shows, and
does this not merely buy into the elevation of the single creative indi-
vidual by Western definition, as opposed to membership in group and
community? Are exhibitions of black artists always and only scheduled
in January (Martin Luther King Day), or during Black History Month?
Or put into summer slots, when large portions of the local elite white
population are away and the social season is slow? Are exhibitions timed
to coincide with the meeting of the Black Arts Alliance the product
of desirable group-identity maintenance, or are they actually the by-
product of a social reality in which room has not been made for black
artists in the (undesignated, of course) white, local Arts Alliance?
Finally, for my own beat, the student community, issues of the edu-
cational mission of the museum intersect directly with issues related
to other identifiable communities. In some respects, students represent
a distinct subset, because they are already motivated and partially
informed, and often themselves of the elite; nevertheless, student here can
stand for any individual who does not come already equipped to experi-
ence and/or understand the work on display and who wishes or needs
additional information. Many museums have confined their educational
mission to special programs (lectures, classes) outside of the gallery.
Some provide docents or taped guides for exhibitions, more often for
temporary than for permanent installations. Some have experimented
with ancillary materials, including more informative labels, photographs,
and videos, and “user-friendly interactive technology” (cf. discussion in
Museum & Arts Washington [May/June 1990]; and installations like the
Tibetan thanka project in the Asian Art Museum of San Francisco). But
still, the arguments rage over the degree to which the object should be
left to speak for itself and the degree to which the “contextualized”
exhibition is more appropriate in a museum of ethnography and/or
natural history than in an art museum.
What needs to be underscored is the degree to which uninformative
exhibitions buy into both the old myth that the objects speak for them-
selves and the old elitist conception of the normative museum-goer, one
who already has the tools of understanding. Even the headsets and the
highly informative catalogues now a major focus of many exhibitions

are differentially available to those with the means to rent or purchase.
Galleries relieved by reading rooms with catalogues in the midst of the
exhibition space seem to be highly successful, although I have seen no
effectiveness studies; but here again, they have been associated with
temporary, rather than permanent installations, often because it is only
in the temporary galleries that dedicated space can be designed into
the museum experience.
In the end, exhibitions like the Chinese bronzes at Harvard or the
Rockefeller collection of African art at the Metropolitan Museum
privilege viewers who share in the same understood classification and
value system as the installers, and who already have the background
necessary to understand the materials on display. The implications of
such exhibition policies are that students, as members of other (sub-)
cultures—often even descendants of the same cultures as the works on
view—are unable to derive much from the galleries without a guide.
The questions raised here relate not to which audience is targeted by
the art museum, but whether that audience has been successfully reached.
As a cultural historian, I would say that success is seriously impeded
by the limits inherent in eighteenth- and nineteenth-century holdovers
governing attitudes toward curiosities, the aesthetic, and the object; and
I would argue that the degree to which artworks and aesthetic issues
can be brought alive in an institutional setting has by no means been
fully explored. It would certainly seem that more consultation is needed
with targeted audiences than has been traditionally employed. At issue
is the social and educational responsibility of public institutions in the
face of cultural diversity, as opposed to the moral assumption of the
museum as repository both of “Art” and of cultural values.
Does this imply that there should be no art museums, that all works
should remain frozen in the amber of their original traditions? I can-
not argue thus, for museums do have a function in our society and, at
their best, can bring the viewing public into a particular relationship
with material products of human culture (see Alpers 1991); but I think
it important that art institutions neither proceed on the assumption
that they represent a universal, neutral, even benign stance, nor expect
unquestioning reverence toward their collections and presentations. As
we move into a new decade, directors of public art museums, large and
small, must know that they will be challenged. They will have to think
through where their own commitments lie; decide the degree to which
their institutions will be responsive to the community at large, as opposed
to a privileged elite. Into this must be factored a new dependence upon

public monies, and the degree to which the acceptance of those funds
commits the museum to different, less elitist policies from the past.
If it remains the museum’s goal to reach a broader community, then
it may be necessary to rethink not just acquisition and exhibition poli-
cies but also such fundamental issues as what the legitimate purposes
are of an art museum today; how the intellectual and the experiential
can be put together for different cultural subgroups; and what ethical
issues are raised along the way.
In this respect, I wonder whether the recent events surrounding the
Mapplethorpe and Serrano exhibitions have not had a positive side.
Most art institutions over the past three decades had been accustomed
to seeing themselves as representing the establishment, with challenges to
the museums coming from a radical fringe. In these two cases, however,
institutions have had to defend themselves from challenges coming from a
self-styled establishment—challenges from which the wealth and prestige
of their trustees and their long-term privileged position in society could
not protect them—and they have had to seek (and publicly acknowledge)
support from more liberal, activist constituencies. An impressive number
of institutions mounted admirable and effective campaigns to educate
the public on the issue of censorship and on the relationship between
de facto and de jure censorship in the withholding of funds based on
criteria of content. Galvanizing their constituents into action required
courage, integrity, and an articulated public policy, and I have profound
admiration for all that has been done on that front. Such an institutional
stance cannot be wholly convincing, however, if on the one hand the
art museum, when under fire, raises questions of ethics and morality
regarding freedom of expression and artistic integrity, and on the other
hand continues to direct a significant proportion of its resources and
programs at a narrow, privileged band of society and pursue policies
of acquisition that contravene international law and the integrity of
cultural properties, themselves issues of ethics and morality.
If this were not a time of reflection, the AAMD panel for which this
paper was initially presented would not have been organized. Under
consideration is the very nature of the art museum in American society.
I was asked to speak to changes in the field of art history and how
they reflect current views of and needs from the art museum. It has
become increasingly clear, however, that I cannot speak of changes in
the history of art without addressing changes in prevailing definitions
and perceptions of art, changes in our culture, changes in the world—all
of which have serious implications for changes in the art museum. The
consequences of maintaining old policies and old practices would be to
relegate art museums to the status of mausolea of the past—what Jack
Tchen in this program has referred to as “Anglo-American ancestral
shrines.” And yet, to move with the times would mean relinquishing,
or at least calling into question, many of the institutions’ founding
principles.
As I declared at the outset that I would speak only for myself and not
for all art historians, I have focused my discussion on issues of acquisi-
tion and exhibition policy that impact directly upon my own work. My
colleagues, many of whom have similarly distanced themselves from the
art museums, would address other issues or weigh those I have raised
differently. I have heard them speak, for example, of having been dis-
affected by the fetishizing of works of art as dollar-value commodities,
and of the reification of the concept of the object when for them the
object is part, but not the end of, their inquiry. I have heard concerns
regarding the privileging of certain media and of scale (bigger is better)
and of the lack of address given to whole subfields, such as architectural
history, in any systematic fashion.
To the extent that the phenomenon of the art museum retains its
value for our time and culture, it is essential to be cognizant of just
what it is the museum is saying, implicitly and explicitly, about time(s)
and culture(s). The challenges raised in recent years have significant
consequences for the way in which the art museum has traditionally
done its business. In response to these challenges, and as a result of
their own internal development, a number of institutions have under-
gone significant shifts in policy and practice; many more are projected.
Nevertheless, movement has been uneven and slow at best. Often it is
less the museum director or curatorial staff, than conservative boards
and trustees—many of whom are themselves collectors and represent
a narrow social band committed to particular privilege and world-
views—who have impeded change. Often, with the best of innovative
vision, change is retarded as a result of budgetary limitations. I am,
however, led to the conclusion that art museums—their boards, direc-
tors, and staffs—ignore the larger issues of change in our society and
the implications for their institutions at their peril.
Certainly, until I sense more consciousness of the terrible cost of
present acquisition policies and the limits of current exhibition poli-
cies, there is little of value for me in most major art museums but the
housekeeping services rendered to the collections—themselves of widely
varying utility in the pursuit of the history of art.
Acknowledgments
I am grateful to Norman Bryson, Robert C. Hunt, Louis D. Levine, Oscar White

Muscarella, Barbara and Michael Porter, Sally Price, Alan Shestack, Kapila Vatsyayan,
and Frederick A. Winter for pertinent criticism and observations on an earlier draft of
this paper, even—perhaps, especially—when I have not always followed their advice;
I have learned from them all.
Bibliography
Alpers, Svetlana. 1991. “The Museum as a Way of Seeing.” In Exhibiting Cultures: The
Poetics and Politics of Museum Display, edited by Ivan Karp and Steven D. Lavine,
25–32. Washington, D.C., and London: Smithsonian Institution Press.
Appadurai, Arjun (ed.). 1986. The Social Life of Things: Commodities in Cultural Perspective.
Bal, Mieke, and Norman Bryson. 1991. “Semiotics and Art History.” Art Bulletin 73(2):
174–208.
Bois, Yve-Alain. 1989. “Exposition. Esthètique de la distraction, espace de demonstra-
tion.” Cahiers du Musée National d’Art Moderne 29: 57–79.
Clifford, James. 1987. The Predicament of Culture: Twentieth-Century Ethnography, Literature,
and Art. Cambridge: Harvard University Press.
Duncan, Carol. 1991. “Art Museums and the Ritual of Citizenship.” In Exhibiting Cultures:
The Poetics and Politics of Museum Display, edited by Ivan Karp and Steven D. Lavine,
Ferguson, Russell et al. (eds.). 1990. Out There: Marginalization in Contemporary Cultures.
New York: The New Museum of Contemporary Art; Cambridge: MIT Press.
Garfias, Robert. 1989. “Cultural Diversity and the Arts in America.” Unpublished
paper given 6 September 1989.
Gómez-Peña, Guillermo. 1988. “Documented/Undocumented.” In Graywolf Annual
Five: Multicultural Literacy, edited by Rick Simonson and Scott Walker, 127–34. Saint
Paul, Minnesota: Graywolf Press.
Greenfield, Jeanette. 1989. The Return of Cultural Treasures. Cambridge: Cambridge Uni-
versity Press.
Grimes, Ron. 1990. “Breaking the Glass Barrier: The Power of Display.” Journal of
Ritual Studies 4/2: 239–62.
Karp, Ivan, and Steven D. Lavine (eds.). 1991. Exhibiting Cultures: The Poetics and Politics
of Museum Display. Washington, D.C., and London: Smithsonian Institution Press.
Kirschenblatt-Gimblett, Barbara. 1991. “Objects of Ethnography.” In Exhibiting Cultures:
The Poetics and Politics of Museum Display, edited by Ivan Karp and Steven D. Lavine,
Levine, Louis D. 1991a. “The Curator, the Museum and the Community: Getting
Beyond the Mapplethorpe Controversy.” The New York State Museum News 6(2) (Spring/
Summer): 3–6.
—— 1991b. “Museum Mission and Museum Exhibits—Getting it Right.” Forum 2
(May/June): 5–7.
—— n.d. Personal communication to Irene J. Winter.
Livingston, Jane, and John Beardsley. 1991. “The Poetics and Politics of Hispanic
Art: A New Perspective.” In Exhibiting Cultures: The Poetics and Politics of Museum
Display, edited by Ivan Karp and Steven D. Lavine, 104–20. Washington, D.C., and
London: Smithsonian Institution Press.
Messenger, Phyllis M. 1989. The Ethics of Collecting Cultural Properties: Whose Culture?
Whose Property? Albuquerque: University of New Mexico Press.
Meyer, Karl E. 1973. The Plundered Past. New York: Atheneum.
Miller, Arthur. 1982. “Archaeological Looting: A New Approach to the Problem.”
Expedition 24: 35–45.
Murray, Tim. n.d. “Communication and the Importance of Disciplinary Communities:
Who Owns the Past?” In Theoretical Archaeology: Who Sets the Agenda, edited by N. Yoffee
and A. Sherratt. Cambridge: Cambridge University Press. In press.
Nochlin, Linda. [1973] 1988. “Why Have There Been No Great Women Artists?” In
Women, Art and Power and Other Essays, 145–80. New York: Harper & Row.
Price, Sally. 1989. Primitive Art in Civilized Places. Chicago: University of Chicago Press.
See especially, “Objets d’Art and Ethnographic Artifacts” (pp. 82–99).
Scott, Joan Wallach. 1988. Gender and the Politics of History. New York: Columbia Uni-
versity Press.
Wagner, Anne M. 1989. “Lee Krasner as L. K.” Representations 25: 42–57.
Ybarra-Frausto, Tomás. 1991. “The Chicano Movement/The Movement of Chicano
Art.” In Exhibiting Cultures: The Poetics and Politics of Museum Display, edited by Ivan
Karp and Steven D. Lavine, 128–50. Washington, D.C., and London: Smithsonian
Institution Press.
Figure 1. Permanent installation, Chinese Bronzes Gallery, 1990 Sackler

Museum, Harvard University.
Figure 2. Temporary exhibition, Symbols of the Ancestors: The Power of Chinese Bronze
and Jade, November 1989–December 1991, University Museum, University
of Pennsylvania.
CHAPTER THIRTY-SIX
PACKAGING THE PAST: THE BENEFITS AND COSTS

OF ARCHAEOLOGICAL TOURISM
Archaeology, the systematic uncovering of the material remains of the

past, is a professional enterprise that generates products—the archaeo-
logical site and its artefacts—along with knowledge. To the academic
archaeologist, the joy of archaeology lies in the recovery of human his-
tory, which need not be linked to one’s own racial, ethnic, or historical
ancestors to be appreciated. The “systematic uncovery” emphasized
above is important, because the amount of human history revealed
to the well-trained team of archaeologists is significantly greater than
that romantic “trace” triumphantly unearthed by amateurs and/or
traffickers for personal gain.
Others at this symposium, in company with a number of current
scholars, have dealt with the complex issue of the various agendas
involved in exposing the past, and I shall not explore them here.1 In
many ways, the importance of archaeology for tourism actually fol-
lows logically upon such agendas. Traces of the past, once unearthed,
become objects of popular consumption—important to both host
and consuming cultures. Indeed, this has occasionally led to tension
between archaeologists’ desires to dig for purposes of “knowledge” and
nationalistic/tourist-industry desires to preserve the spectacular attrac-
tion—those sensational monumental structures that afford charismatic
“experience”—thereby halting investigation of lower (earlier) levels in
stratified sites and the information they contain.
In recent years, tourism has become a major world industry. Espe-
cially in developing countries, it often accounts for a large part of the
* This article originally appeared as “Packaging the Past: The Benefits and Costs
of Archaeological Tourism,” in Partnership in Archaeology: Perspectives of a Cross-Cultural
Dialogue. [14th Symposium of the Swiss Academy of Humanities and Social Sciences.]
Beat Sitter Liver and Christoph Uehlinger, eds., Fribourg: University Press, 1997,
pp. 127–145.
1
See Silberman 1982, McBryde 1985, Greenfield 1989, Messenger 1989, Winter,
F. A. 1991, Murray 1993.
522 chapter thirty-six
national income, and certainly a high percentage of foreign exchange.

The results of archaeological investigation have come to represent an
exploitable resource, consciously packaged, marketed, and advertised as
prime attractions.2 The field excavation uncovers architectural remains
that generally are left in situ, providing “monuments” to visit and experi-
ence. In addition, the excavation generates an artefactual assemblage,
the associated material culture, that is usually removed from the site,
with select examples presented to the public in national museums. With
the advent of the motor car and the rental agency for individuals, plus
organized transport facilities for groups, the tourist (as distinct from
the lone traveller) is now no longer limited to displays in the nation’s
capital institutions. Hordes of non-locals can be bused, ferried, even
helicoptered in to the most inaccessible sites. These hordes present an
interesting challenge to local resources and services, not least of which
is how to protect the sites themselves from degradation and spoliation.3
The intersection between archaeology and tourism, therefore, has
consequences that echo well beyond the initial scholarly enterprise of
the retrieval of the past.
In pursuing those consequence, I have titled my paper quite con-
sciously, inverting the usual order of costs and benefits. In most
economic studies, the costs come first and the benefits are realized
subsequent to payment. Here, I believe the pattern has been reversed.
2
See, for example, the linkage in Britain, discussed in Prentice 1993, and in Europe in
general, discussed in Ashworth and Larkham 1994. Indeed, in some countries, as in the
United Arab Emirates, antiquities are grouped with tourism in the same Ministry—cf.
Archaeology in the United Arab Emirates, Vol. 1 (1976/7), published by the Department of
Antiquity and Tourism, U.A.E., 1977; while the 1995 double-page advertisement for
the Turkish Tourism Office in the US focuses almost exclusively on archaeological
materials and sites of the past 10,000 years as inducements to visit Turkey (see, for
example, Civilization: The Magazine of the Library of Congress 2 ( July–August 1995, pp. 2–3).
India is noteworthy as having not only one of the earliest government departments of
antiquities (the Central Archaeological Survey, est. 1861), but also an “Ancient Monu-
ments and Archaeological Sites and Remains Act,” passed in 1959, which enabled the
Archaeological Survey to work with the Ministry of Tourism in establishing a planned
program of cultural tourism (see Vatsyayan 1972, p. 63).
3
Archaeological sites have always been subject to pillaging, purposeful desecration,
and well-meaning intrusive defacement. As charming as it may be to seek the name
of Lord Byron scrawled on the temple remains at Cape Sounion, Greece, the sheer
numbers modern tourism has introduced to monuments makes the problem a difference
in “kind”, not merely scale. As noted by John Romer in a BBC Television program
on Egypt’s Valley of the Kings (aired in the US on 9 May 1994), tourists “care,” but
even when they do not scratch their names on monuments, they bump and weaken;
and, in cases like the Palaeolithic Caves of Lascaux in southern France, even their
breath erodes (on which, see below).
packaging the past 523
The benefits of making archaeological monuments accessible to tourism

are many, and are relatively obvious: increased awareness on the part of
traditional cultures of their own past and heritage; increased awareness
on the part of consuming cultures of traditions other than or ancestral
to their own; increased economic benefit at the local level to monu-
ment-rich regions; and increased foreign exchange on a national scale,
as tourist dollars buy goods and services across a wide spectrum of the
national economy. In those terms, archaeology is indeed “important”
for tourism, as suggested in the guidelines for the present conference,
providing means of economic as well as cultural development.
Mindful of these benefits, host countries have begun to elicit com-
mitments from archaeological expeditions, through their respective
Departments of Antiquities, and indeed, committed archaeologists have
often instituted programs themselves, for both the training of indigenous
expertise and the conservation or restoration of monuments on desir-
able sites to complement the extraction of data.4
At the same time, there are serious costs to host countries in the
marketing and packaging of their national past(s).5 And if this confer-
ence is to lay the foundations for on-going discussion, for a vision of
maximizing benefits and minimizing costs, and for future policy, it
seems necessary to discuss worst-case as well as best-case scenarios.
Some of the apparent costs include:
1. financial benefits accruing only to urban (organizing), not local
interests, thereby increasing, not decreasing the unequal distribution
of wealth;
2. damage to the fabric of life in the traditional, host culture, as tour-
ism turns not only the site, but living practices and practitioners into
objects of voyeuristic consumption;
4
In a conversation with the then-Director of the Ankara Museum, Raçi Temizer,
in 1974 he spoke of the importance of the University of Pennsylvania expedition
to Gordion undertaking conservation work in the great Midas mound tomb, which
has since been completed; while in Mexico, archaeologists were required to devote
resources to the stabilization of the pyramid at Teotihuacan, in order that it could be
safely visited by tourists. Many expeditions are now conceived as “joint” projects with
scholars and institutions in the host countries—as for example the current excavation
at the site of Puerto-Escondido in Honduras, directed by Rosemary Joyce, University
of California at Berkeley, John Henderson of Cornell, and Rodolfo Pasto-Fasquelle,
an archaeologist who is now the Minister of Culture in Honduras.
5
See also the paper of Mario Navarrete Hernández in this volume.
3. an increased gap between local identity and the particular archaeo-

logical attraction if the site represents a dead civilization, such that,
without concomitant educational programs at the regional level,
both archaeological recovery and tourism can result in a lack of
identification with monuments, a sense of alienation, and sometimes
even the desecration of sites;
4. physical damage to sites and monuments not equipped to handle
large numbers of visitors—inadvertently because of traffic load,
and/or purposely by souvenir scavengers or the disrespectful;
5. an increased market demand for uncontrolled objects, as the focus on
monuments whets appetites for “authentic” souvenirs and archaeo-
logical materials become both subjects of tourist interest and objects
of consumption, thereby encouraging the spoliation of sites and the
illicit traffic in antiquities;6
6. substantial burdens placed on the carrying capacity of the local
environment, as a result of the addition of tourist facilities; and
7. jurisdictional struggles within countries between ministries with dif-
fering agendas, resulting in a tug of war in which the practice of
archaeology and the archaeological site become the focus of vying
ministers with competing and often incompatible interests.
To date, host countries have been largely willing to bear these costs,
since the benefits in touro-dollars, brought by visitors from wealthy,
consuming countries, is thought to offset the costs. Most of the con-
suming countries that generate the bulk of customers in foreign tour-
ism also package their own heritage, both for the internal and the
international tourist market.7 They, too, have experienced invasive
behavior as the down-side of tourism is being felt across the globe. In
developing countries, such costs echo across the entire social spectrum
of the host population.
In order to explore and develop solutions to the problems raised
above, one needs to take each so-called “cost” in turn, gather data for
specific case studies, analyse impact, and work toward devising strate-
gies for national development that take the nature and preservation
of the cultural heritage seriously. In addition to appropriate “policies”
6
In such cases, despite whatever antiquity laws may be in place in the host country,
what is actually effected is the depletion of the cultural patrimony rather than increased
respect and maintenance, along with the corruption of the local population (especially
when alienated, see no. 2, above) for immediate economic gain.
7
See Prentice 1993, and the publications of the European Forum of Heritage
Associations.
on paper, actual agencies for managing the cultural heritage, including

archaeological sites, are needed—agencies that are sufficiently funded
and empowered so as to be able to maintain those goals and strategies
developed through planning and also to review and implement new
goals and strategies as situations change.8
1. Unequal distribution of the economic benefits of tourism to archaeological sites.

This is a problem that is being faced with respect to all sorts of tour-
ism, not merely to archaeological sites. Wherever there are natural
or cultural phenomena in non-urban—often relatively undeveloped,
agricultural or rural—areas, it is urban (private or government) inter-
ests that come in to build hotels and provide tourist services.9 In many
cases, such services do provide employment for selected members of
the local population; however, all too often skilled labor is imported,
and/or local labor is exploited, with major profits accruing only to the
urban sphere. As a result, local social structures and labor patterns can
be disrupted rather than enhanced by new schedules.
The overall negative impact on rural communities has been exten-
sively demonstrated. As noted by Susan Guyette, who has worked
with the New Mexico Planning and Research Department studying
“Cultural Tourism,” i.e., the visiting of Native American communities
by tourists in the American Southwest, “in terms of gain, urban centers
secure the largest share of expenditures, while rural areas provide the
“attractions.”10 In this respect, cultural tourism to ethnic communities
with exotic lifestyles is comparable to, if potentially even more disrup-
tive than archaeological tourism, where the attraction is monuments
rather than people.11 In both domains, social and economic planning
is required that considers local impact, not just (urban) development:
i.e., plans for the engagement of local industry, as well as the revival or
development of new local industries, such as craft/souvenir production,
which will not just provide service positions (e.g., in hotels), but also
produce or revive local cultural traditions.
8
This point was also powerfully made in the paper of Safwan Tell in the confer-
ence that generated the present volume.
9
A rare case to the contrary was reported to me by Prof. Karen D. Vitelli, Indiana
University, in which archaeologists in Mesoamerica are working directly with the local
population to build and organize their own hotels and develop the site, after exacava-
tion, as a tourist attraction.
10
Guyette 1994, p. 1, and in press.
11
See Cornell 1989.
2. Disruption of living cultural practice at designated sites. If local use of “living”

sites is not to be seriously disrupted, non-invasive plans for selective
access must be developed. Here, a number of Native American com-
munities in the United States, Aboriginal communities in Australia and
New Zealand, and religious communities in India provide examples
of restricted access that helps to minimize the de-naturing of cultural
practice that results from turning both the site and participants into
objects of consumption.12 As Guyette noted, in tourism to Native
American Indian reservations, for example, planning should be guided
by three important factors: consideration of form, scale, and timing.13
If planning is not done well, and if the local population is not included
in the making of policy, tourists seeking the frisson of an “authentic”
experience can, in fact, be damaging to the fabric of life in the tra-
ditional, host culture—as when they crowd and alter the tone of the
performance of religious rituals on sites or in temples that are also
“archaeological monuments,” or even as the performances themselves
are diluted to accommodate the presence of outsiders.14
The one virtue of dealing with monuments associated with living
populations is that the indigenous groups themselves can organize to
protect both tradition and the cultural resources. In India, for example,
one of the explicit tasks of the Director General of Archaeology is to
balance current needs for worship against the designation of protected
status for a monument.15 In the US, recent legislation has sought to
provide Native American tribes with the ability to control not only
their own patrimony, but how their culture is presented to others.16
Negotiations have resulted in a sort of “Cultural Impact Statement,”
12
Archaeologists and anthropologists in Australia have led the way in discussing
these issues (see in particular, Langford 1983, Lewis and Rose 1985, Layton 1989, Ley
1991, Murray 1993). For other areas, see Vatsyayan 1972, Pommaret 1981, Cornell
1989, Guyette, in press.
13
Guyette 1994, p. 3.
14
See below, section 6, with respect to the environmental consequences of tourists’
desire for authentic natural experiences.
15
Vatsyayan 1972, p. 63. Some temples that have not been designated “protected
monuments,” and therefore do not fall under the auspices of the Archaeological Sur-
vey of India whose charter mandates public access, have taken steps to exclude all
non-worshipping visitors.
16
See the US National Historic Preservation Act (NHPA) of 1966, which addresses
historical preservation and asserts: “When an undertaking will affect Indian (Native
American) lands, the agency official (for the NHPA) shall invite the governing body of
the responsible tribe to be a consulting party and to concur in any agreement.” See
also, the Archaeological Resources Protection Act of 1979, which mandates “notifi-
cation of Indian tribes of possible harm to or destruction of sites having religious or
comparable to an “Environmental Impact Statement,” that must be filed

prior to tourism as well as to archaeological intervention.17 It should
be noted, however, that in many cases, governments are not willing
to place indigenous cultural needs above development potential, and
mechanisms for consultation are either non-existent or ignored.
3. Increased alienation of local population in the face of archaeological investigation

and subsequent tourism. If living populations can exercise some degree of
control over archaeological investigation and subsequent presentation
(as with Islamic period burials in Muslim countries or Native American
burials in the US), exposure of the sacred sites and practices of dead
civilizations is generally considered fair game. At the same time, where
the local population is no longer identified with the religion, ethnicity,
or culture of the (dead) archaeological tradition, both archaeological
recovery and tourism can actually increase the gap between local
identity and the particular archaeological/tourist phenomenon. While
this aspect has been little studied, possible results include a sense of
alienation from and/or a lack of identification with monuments. What
is more, such distancing makes it easier to desecrate sites for immediate
economic gain, as in the looting of antiquities.
One way to avoid this is the fundamentalist solution of simply not
uncovering, much less displaying, the not-currently-representative cul-
ture; however, this solution is increasingly impossible as more history
is known and access is demanded by a variety of interest groups. The
desecration of sites or artefacts by non-believers is nothing new: monu-
ments have been literally “defaced” since antiquity itself—as the face of
the Assyrian ruler, Assurbanipal, was carefully hacked out of his palace
reliefs at Nineveh, presumably by conquering Medes and Babylonians
in the 7th century b.c. Within the more recent time-frame of the 1960s,
I myself have experienced this distancing by guardians of archaeologi-
cal sites in the Middle East who felt no connection whatever to the
culture(s) exemplified by the pre-Islamic site or materials they were
guarding, and so were quite ready to sell objects from the site, permit
clandestine looting, or even use ancient monuments for target practice.
With recent educational programs about the importance of the
archaeological heritage and national campaigns about the importance
cultural importance” (cited and discussed in Treitler, Ruppert, and Stoffle 1994; and
Parker and King 1990).
17
See Winthrop 1994, p. 26.
of tourism, many countries have brought such damage to sites under

control—except, of course, where desecration is actively encouraged by
doctrine. In countries where subsistence is barely above, or oscillates
below, sustainable levels, economic pressures are such that it is virtually
impossible to control the looting of sites by locals, once a market value
for antiquities is established.18 It is essential, therefore particularly with
respect to “dead” civilizations, to institute educational programs at the
regional level to bridge the often-substantial historical, cultural, and
religious gaps between contemporary populations and archaeological
remains.
In addition, government support for local, rather than merely urban,
museums provides educational means for local inhabitants—children,
as well as adults—and contributes to a constructive view of the total-
ity of the regional past. This has been particularly effective in Turkey,
for example, where most major sites, like the ancient Hittite capital at
Boğazköy, have adjacent local museums with representative assemblages
of artefacts from the site and even comparative materials from other
sites; also in India, where the Archaeological Survey of India has estab-
lished “Site Museums” adjacent to monuments across the country.19
The combination of educational programs stressing the value of the
historical past and the construction of local museums, rather than
shifting all finds to the capital (in itself not much different from the
colonial creaming off of finds to European or American museums), can
significantly decrease both alienation and desecration, if addressed to
the local population and not merely to visiting tourists.
4. Damage to sites. If countries are in fact to preserve their archaeological

sites and monuments adequately, and ensure both their national/sym-
bolic and international/touristic value, then sufficient budgetary means
and planning must be allocated in order to provide for walk-ways,
18
See Brent 1994.
19
See Annual Report, Part II, of the Department of Culture, Ministry of Human
Resource Development, Government of India, 1993–94, p. 28. The Republic of
Ecuador has recently been seeking support for the construction of local museums in
proximity to archaeological sites, as well (source: visit of the wife of the President of
Ecuador to the Smithsonian Institution Museum of the American Indian, 1994—per-
sonal communication, E. Weatherford). The parallel plan to construct “model villages”
with traditional life-styles also in proximity to archaeological sites suggests, however,
that the goal of such construction is less for local benefit than to provide additional
attractions for tourists.
lighting, and carefully routed access that will not disturb unexcavated
areas, erode or destroy exposed phenomena. Cases in point are the
Lascaux caves of southern France, many of the tombs of Pharaonic
Egypt, or the Buddhist caves of China and India, where exposure to light
or human breath can cause irremediable damage to the color of wall
paintings, or introduce moisture, mold, and harmful micro-organisms,
thereby creating a destructive environment; but even large architectural
monuments are being seriously threatened. Recently, an International
Conference was convened in Athens that brought together government
officials and experts from 16 Mediterranean countries to discuss the
impact of tourism on sites like the Acropolis (Greece), the Pyramids of
Giza (Egypt), Carthage (Tunisia), and Ephesus (Turkey).20
In order to preserve the cultural heritage represented by such monu-
ments, host countries must either limit access altogether or develop
strategies for the nondestructive consumption of sites; otherwise, the
sites, like other “natural” resources, will themselves be consumed. In
short, archaeological sites are not a renewable resource!
5. Increased demand for illicit souvenirs, spoliation, and the depletion of the cultural
patrimony. As more tourists visit sites, and few are themselves aware of
the cost of even surface (or, in the case of underwater archaeology,
sub-surface) souvenir-hunting to knowledge, the very archaeological
remains they have come to visit are increasingly at risk. Indeed, I
would submit that the focus on monuments and the experience of the
past actually increases touristic appetites for the “authentic” souvenir.
At this point, moveable archaeological materials themselves become
subjects of touristic interest and objects of consumption. This, in turn,
leads to the spoliation of sites for objects, and thus, the depletion of cul-
tural patrimony on the one hand, and the corrupting complicity of the
local population for immediate economic gain (especially when already
alienated and insufficiently educated or rewarded by the tourist boom,
as above) on the other hand.
I cite but two examples from personal experience, the first in China,
the second in Turkey; but it should be noted that many other coun-
tries with an archaeological heritage could be substituted. Despite
the fact that antiquities laws in modern China look good on paper, a
20
The conference, organized by the Getty Conservation Institute and the J. Paul
Getty Museum, took place from 7–12 May 1995; the proceedings will be published.
colleague surprised and then photographed a tomb-robber crawling out

of a robber-hole in broad daylight when he happened to visit a site in
the company of members of the local department of antiquities in the
early 1980s. In 1988, when I visited Xian as part of a one-day tour,
merchants outside the tomb who openly displayed more recent artefacts,
like Imperial silks, would surreptitiously flash antiquities to tourists if no
official seemed to be about; and I understand that with privatization
in the past few years, the situation has become considerably worse.
In Turkey, a country with antiquities laws so stringent I had been
forewarned not even to bend over to tie a shoe lest one appear to be
picking up sherds, I was about to depart from a visit to an important
site in 1973 when the local watchman proceeded to offer me whole
pots from the site for sale. Overall, the individual and occasional sale
on site represents a relatively small percentage of antiquities being
plundered; far worse is spoliation on a large scale, where locals are
induced by urban dealers to systematically pillage entire sites as has
been the case recently in Mali.21 Nevertheless, without some parallel
planning, the visit of large numbers of tourists can actually lead to an
increase in the dispersal of the cultural patrimony and the destruction
of archaeological sites.
To meet the economic needs of local populations, and the ambi-
tions stimulated by the introduction of a new, affluent market (i.e.,
tourists), thoughtful alternatives are required, such as the develop-
ment of handicraft or replica industries to substitute for archaeological
souvenirs. Furthermore, to curb spoliation, educational programs are
needed, in concert with realistic and enforced internal policies related
to the export and illegal extraction/sale of antiquities. By education,
I mean the adequate preparation of tourists by prior information and
by tour guides of the damage done to sites by souvenir-hunting, as
well as, in host countries, programs whose aim is the development of
positive identification with the archaeological culture, geared to both
local and urban populations.
In the end, education is a crucial tool in the protection of archaeo-
logical materials. In the Sudan, for example, a National Trust was
established specifically to make the Sudanese public more aware of its
21
See Brent in press, pp. 3–4, 6. A similar situation prevailed in pre-1978 Iran;
and systematic looting of both sites and museums has been documented in Iraq in the
aftermath of the 1991 Gulf War.
national heritage.22 And in Israel, where half the population considers

itself amateur archaeologists23 and a prominent, charismatic general
permitted himself special access to archaeological trophies, a massive
educational campaign was mounted recently to stem the tide of trea-
sure-hunting. Large, attractive posters were distributed throughout the
country with the caption: “THE PAST IS LIKE A PUZZLE.” The text
went on to state: “All archaeological finds in our country are like pieces
in a puzzle. The more that we preserve them, the clearer the puzzle
will be.” The last line, again in bold, large letters, then read: “PRO-
TECT THE ANTIQUITIES OF THE COUNTRY!”24
To address issues of the illicit traffic feeding external consumption,
host countries are presently beginning to appeal to international bodies
for help in protecting their archaeological legacies. Prince Norodom
Sihanouk of Cambodia recently spoke at the ICOM Advisory Com-
mittee held at the UNESCO offices in Paris, describing the enormous
loss of cultural property in recent years, outlining government plans to
stop the pillaging of sites, and appealing for international cooperation.25
A number of countries, buoyed by the successful return of the Pathur
Nataraja bronze image of Shiva to India and the Kanakaria mosaics to
Cyprus, have sought legal redress through the exercise of laws of return
in the home courts of the consumers, but to date, such victories have
been rare.26 In addition, host countries are now in a position to demand
commitments from consuming countries by threatening economic
reprisals in reverse, whether by limiting desired resources or by limiting
access to markets. By invoking potential sanctions or embargoes, host
countries can go a long way toward insisting that consuming countries
ratify and bring into force the UNESCO Convention on the Means
of Prohibiting and Preventing the Illicit Export, Import and Transfer
22
See the paper of Salah Mohamed-Ahmed in this volume.
23
Wasserman 1994.
24
What has not been done, to my knowledge, is devising measuring standards to
assess the success of this or any similar campaign.
25
Reported in Aviso, the official Newsletter of the Association of Art Museums,
August 1994, “Committee Discusses Cultural Property,” p. 2.
26
On the Pathur Nataraja, see Greenfield 1989, and Weil 1989; on the Kanakaria
Mosaics, see Hofstadter 1992 and Karageorghis 1992. For a more recent case, see the
Wall Street Journal article of 21 June 1994, “Art World Shaken by Nations Seeking to
Reclaim Items,” about the return of 6th century b.c. Lydian antiquities to Turkey in
lieu of legal action by the Government of Turkey against the Metropolitan Museum
of Art in New York.
of Ownership of Cultural Property of 1970 and subsequent protective

policies, such as the UNIDROIT convention.27
Otherwise, consuming countries feel little pressure to take such
treaties seriously. The US Congress, for example, kept the issue of
the UNESCO Convention in debate for more than a dozen years,
ratifying it only in 1983. As a student in the early 1970s, I thought
this was reprehensible, and indeed it was; but I have since learned,
to my chagrin, that the US is the only major consuming country of
archaeological and cultural properties to have ratified the convention
at all (Canada and Australia have also signed, but are not themselves
major consumers); Great Britain, France, Germany, Japan, Switzerland
have as of this writing all refused to sign.28
In fact, with respect to consuming countries, the issue is not merely
bringing international protective conventions limiting the acquisition of
cultural properties into abstract force, but actually implementing such
nonbinding agreements by promulgating laws pertaining to imports,
plus the right of scrutiny and seizure at customs.29 Most effective to
date seem to be the specifically-negotiated bilateral agreements between
special-interest partners, such as the US and Mexico, where customs
now has the right of seizure if antiquities are discovered at points of
27
See UNESCO 1970, UNIDROIT papers 1991 and 1993—in draft and awaiting
ratification at the time of the writing of this article.
28
See Preserving Mankind’s Heritage: US Efforts to Prevent Illicit Trade in Cultural Property,
Washington D.C., United States Information Agency, 1991. As an aside, it should be
noted that most of the “hot” antiquities of the Middle East and Asia, at least, pass
through dealerships in Switzerland on their way to markets in Europe and America
(on the international trade in general, see Bator 1981, Coggins 1972, Meyer 1973,
Muscarella 1979, and various articles in ther International Journal of Cultural Property,
begun in 1992). One of the issues delegates were asked to consider in this conference
was: “What are the specific contributions Switzerland and Liechtenstein can offer in
the domain of the robbery of cultural assets and the trade in antiquities?” With the
greatest respect for the aims of the Swiss-Liechtenstein Foundation for Archaeological
Research Abroad and the Swiss Academy of Humanities and Social Sciences, I would
suggest that, as for all of us, there is work to be done right at home—in the Swiss
case, to continue to press for Swiss ratification of the UNESCO Convention as but a
partial measure, and to address in Switzerland the economic interests that perpetuate
this mega-industry in flagrant contravention of the preservation of cultural patrimonies
world-wide.
29
See discussions in Greenfield 1989 and Winter 1993; and the various draft
UNIDROIT conventions on the return of stolen and illegally exported cultural objects
being prepared by the International Institute for the Unification of Private Law, which,
when propounded, will complement (not supersede previous conventions—cf. H. Bur-
man, “Press Guidance for Queries on Treaty Negotiations for Return of Stolen Cultural
Objects,” US Department of State 29 June 1994) the 1970 UNESCO convention.
entry. The US and El Salvador have just signed a similar agreement

restricting the import of pre-Columbian materials, replacing the more
limited, “emergency restrictions” formerly in place.30 Honduras is pres-
ently urging a similar arrangement to protect their Mayan antiquities,
and has established a “Commission de Notables” to explore policy,31
while Peru has been successful in negotiating a five-year extension of
an emergency import restriction on cultural artefacts.32
In addition to these special, bilateral agreements, however, there is
a need for more comprehensive international treaties seeking to curb
the antiquities trade, along with further international legislative bodies
empowered to make binding decisions. UNIDROIT, an international
organization specializing in the unification of private (national) law,
has been working for several years to draft a convention that will bring
the market nations into a treaty regime with the principal “source”
countries.33 For, only if there is a body of international law that allows
for adjudication across national boundaries, can competing definitions
of “rights” and property be evaluated and host countries’ rights to
archaeological property be protected.34
Finally, to have all of these efforts taken seriously, host countries
must also make an effort to ensure that their own export policies are
enforced at departure points. Collectors and dealers who have no prob-
lems smuggling antiquities out of host countries will not be moved to
restrict their activities smuggling antiquities in to consuming countries!
Cooperation from scholars and scientists is essential in lobbying both
at home and abroad for enlightened policies of control. It is all too
easy for archaeologists to stand on so-called disinterested platforms and
preach the curbing of international appetites while dealers go merrily
about their business and host countries storm in impotent rage, or
while powerful interest groups within host countries openly contravene
national laws. Recently, several professional archaeological associa-
tions have propounded codes of ethics for archaeologists themselves,
with respect to indigenous populations and host countries, as well as
30
Cited in the “Field Notes” column of a recent issue of Archaeology 48 (1995), p. 26.
31
Personal communication from Rosemary Joyce, an archaeologist who works
primarily in Honduras, 25 July 1994.
32
See “US Extends Import Ban on Artifacts,” in Aviso, the Newsletter of the Ameri-
can Association of Museums, August 1994, p. 2.
33
See Prott 1989, Burman 1991 and 1994, Hoffman 1995.
34
See discussion in Winter 1993.
to artefacts.35 At the same time, realistic policies within host countries

can be encouraged by archaeologists who have access to policy-mak-
ing officials. These may even include the de-accessioning of redundant,
non-unique, and well-recorded objects to responsible public institutions
abroad to meet some of the international demand for historical materi-
als, thereby at least replacing unprovenienced, illicitly excavated, and
illegally exported objects by “known,” “controlled,” and “legitimately
exported” works. The development of contemporary “authentic” handi-
craft traditions, mentioned above, also enters into the picture here, as
legitimate souvenirs divert tourist interest from antiquities.
6. Environmental problems caused by tourism. Substantial burdens are placed

on the carrying capacity of the local environment as a result of the
addition of tourists and tourist facilities. To a non-specialist, the most
obvious problems seem to be the added demands for water and need
for waste disposal in relatively arid environments, additional need for
fossil fuel in colder climates, and food supply in relatively poor areas.
Some of these problems can be addressed by ecologically tailored,
conservation-conscious tourism that is mindful of limitations—the
“take nothing but your pictures, leave nothing but your footsteps”
sort practiced in some game preserves in Africa, where all supplies
are carried in and all waste carried out. However, such enterprises still
represent a relatively small proportion of tourism, and have been best
adapted to areas where the natural environment itself is the touristic
goal. And even in such cases, the touristic quest for the “authentic
natural experience,” like the quest for the “authentic souvenir” or the
“authentic ritual performance,” can have a profoundly negative impact
on a fragile ecosystem. One case in point is that of Nepal, where
organized trekkers seeking to experience the “authentic” campfire and
unwilling to sustain the smoke of more conservative dung-cake fuel
have brought about massive deforestation.
To the extent that most archaeological sites are located outside
of urban centers and in relatively under-developed areas, archaeo-
logical tourism often penetrates the natural or rural environment as
35
See the various publications of the Society of Professional Archaeologists (SOPA)
and issues of the Public Archaeology Review, of the Center for Archaeology in the Public
Interest, Indiana University. In particular, see Wylie and Levy 1994, for a bibliography
of “Archaeological Ethics in Print” through 1992. The issue of ethics in archaeology is
finally entering the archaeological curriculum at a number of universities as well—e.g.,
Boston University, Indiana University.
well; but since it is the ruin that is the primary focus of the journey,
penetration is often undertaken without much thought to the sur-
rounding area. Depending upon the location and terrain of the site,
tourists require varying support systems, from guest houses/lodging
to meals and recreation. In general, the average tourist of urban
origin requires supplies far beyond the needs of traditional popula-
tions, and leaves behind waste far in excess of absorption capacities.
The magnitude of these problems remains unstudied to date. In
order to truly assess the costs, systematic studies of the consequences of
touristic development in specific regions are needed; indeed, data on a
number of ecologically-diverse areas would be ideal. The east coast of
the Yucatan peninsula in Mexico, rich in archaeological sites, is one
area that would lend itself well to such a study, where the region around
the site of Cozumel, for example, has seen the construction of huge
numbers of tourist hotels in recent years. With such empirical studies,
the environmental impact of development resulting from archaeologi-
cal tourism could be assessed, so that informed policies that work to
lessen environmental degradation might be implemented. Indeed, such
studies are needed to address not just the environmental costs, but all
of the projected costs discussed above.
7. Internal struggles between competing interests for control over the fate of archaeo-
logical sites. Jurisdictional struggles within countries between ministries
with differing agendas—e.g., Economic Ministry, Ministries of Labor
and Tourism, of Agriculture, Development, Culture and/or Educa-
tion—can often result in a tug of war in which the archaeological site
or the archaeologically-rich region is subject to competing and incom-
patible interests, or becomes the locus of contested meaning.
Ayodhya in India, claimed by both Hindus and Moslems, is a case in
point of the latter. For the former, national policies of protection and/or
short-term investigation need to be established so that archaeological,
sites in the area of urban, agricultural or hydraulic development (as in
Turkey and Syria with the large dam projects, or in Sweden with urban
construction)36 can be saved, sounded, or at least recorded through
salvage work prior to disruption or destruction.
36
See, for example, the comments of Olaf Olsen in the publication of the European
Forum of Heritage Associations, Bolletino 3 of the XIII Congress U.I.S.P.P. (International
Union of Prehistoric and Protohistoric Sciences), held in Forli, Italy in September of
1996, Forli 1996, pp. 114–5.
An unfortunate counter case is that of Iran, where sites across the

entire plain of Khuzistan in the southeast have been wiped out by bull-
dozers to create gravity-driven irrigation channels. One archaeologist
who had surveyed the region in the late 1960s returned to Khuzistan
10 years later to find that virtually none of the sites he had mapped
remained. In fact, this is a double tragedy, because more modern and
in the end less costly technology in drip-irrigation systems were already
available, in which water loss through evaporation would actually have
been less. In such cases, it is clear that agro-business and short-term
development projects win out over claims for the importance of cul-
tural heritage.37
Recognition of the magnitude of these problems within central gov-
ernment is necessary, followed by planning with complex, if compet-
ing, issues in mind. Hopefully, more sophisticated investigations into
“sustainable development” will lead to the setting of more balanced
priorities and less-destructive policies. But equally important is a more
holistic notion of what constitutes the national/cultural “good.”
Such holistic constructs are difficult to achieve in the face of short-
term, usually economic, benefits. Thus, when Rajiv Gandhi convened
the second Asian Relations Conference in 1987, there were four discussion
groups running simultaneously: Economic, Political, Educational, and
Cultural. Yet it was clear, not so much from the point of view of the
conference organizers as from the invited participants, that the “Eco-
nomic” group was by far the weightiest, “Culture” was soft, while
“Politics” and “Education” were somewhere in-between. An alternative
to this ranked hierarchy is to see the systemic unity of the four areas,
each of which contributes in essential ways to the polity in question.
The marriage of “Archaeology and Tourism” actually takes place at
the intersection of all four of the above categories. What is needed is
recognition that culture and the cultural heritage (of which the sites and
artefacts produced through archaeological investigation are a part) is
not a luxury of the urban elite, but rather fundamental to national and
transnational identity as part of the historical record of our species. In
this way, archaeology takes on importance at home as well as abroad
for consumers. (And then, of course, the problems of archaeological
37
A similar situation is cited by Paul Sinclair in this volume concerning the damage
to cultural resources in East Africa from agricultural development.
tourism are compounded by the more recent inclusion of non-foreign,

urban nationals to the touring population!)
In conclusion, a number of areas can be identified where action is
important if we are to mitigate the costs of inscribing archaeological
monuments in the tourist circuit. Within the profession of archaeology,
the training of indigenous, not just foreign, expertise, plus strict adher-
ence to codes of ethics and active collaboration in planning; and in
host countries, the valuing of archaeology as a profession, plus internal
vigilance to bring about and maintain protective policies.38 In particu-
lar, long-range planning is required that takes into consideration the
powerful symbolic and image-making potential of archaeology, along
with its economic and educational potential. All of these domains—the
cultural, the economic and the educational—suggest that there is,
in fact, a powerful political dimension to archaeological remains as
representations of the past, since politics and elections are carried
by symbols, moved by promises of economic benefit, and informed
through education.39
But this is effective only if national identity does not implicitly contain
a hierarchy that places “culture”—as manifest in the archaeological
record—either in the domain only of the elite or at the bottom of a stack
of other values; and if modern national identity does not carry with it
a component of unequal value accorded to pre-modern history.
While archaeological monuments need to be understood as part of
a larger problem of the management of cultural and environmental
resources,40 it is equally important to acknowledge that “culture” is
not merely a resource to be exploited; rather, it constitutes the warp
and the weft, the very fabric of living. As policies initially developed
for the conservation of living traditions (e.g., the Navajo Nation in
the US and the Australian Aborigines) have shown, if a given cultural
38
Note that in at least one country in the Middle East, the highest achievement
in pre-university examinations automatically marks young people for professions in
engineering, law, and medicine. The study of history and archaeology is at the bottom
of a ranked list of other pre-professional programs, so we are doomed to the same sort
of ranked hierarchy of value as was manifest in the assumed hierarchy of the Asian
Relations Conference. What is needed is some of the best minds and managerial skills
to be professionally engaged in the cultural sphere.
39
Bourdieu 1980, Choay 1992.
40
Cleere 1984.
legacy is to survive, there is a need for protective policy at the highest

governmental levels.
For archaeology, this means both the protection of physical remains
and the protection of the precious knowledge of the human past pre-
served in sites. Archaeology is, after all, the only “social science” that
systematically consumes its own data base as it proceeds, since an
archaeological level, once dug through or otherwise obliterated, can
never be reconstituted for analysis. The knowledge contained in that
site, level, or locus is lost forever. For this reason, archaeologists cannot
distance themselves from tourism issues; they are engaged, and must
be! The initiation of collaborative ventures with national specialists
is crucial in an effort to balance the questions of patrimony and site
preservation with plans for site intervention and development; while
active participation in site preparation for tourism will go a long way
toward lessening damage.41
In the end, comparable to the discussions concerning sustainable
development, what is needed are policies committed to a sustainable cul-
tural heritage! And this means policies where the often-competing interests
of pure knowledge, symbolic/national identity, and economic gain can
be rationalized. This further implies enlightened, long-term economic
and social planning, educational programs, professional training and
management, along with an explicit acknowledgement of the intrinsic
value of the heritage of the past.42 In the absence of such planning and
programs, I fear that in the intersection of archaeology and tourism,
even with the best intentions, nations risk having the long-term costs
far outweigh the short-term benefits.
41
See also the papers of Fayza Haikal and Mario Navarrete Hernández in this vol-
ume. It should be noted that appropriate preparation of sites for large-scale visitation is
also important for the visitors. The collapse of a baulk killing a young Armenian visitor
at an American excavation site in Israel in the summer of 1994 is a case in point.
42
It makes sense to press for the establishment of national organizations for the
preservation of cultural heritage, both governmental and private, since lobbying groups
can have great power in the making and maintenance of both priorities and policy.
In the Sudan, noted above (note 22), as in a number of monument-rich countries,
national trusts have been established; in the Netherlands, the Cultural Heritage Policy
Directorate is located within the Ministry of Welfare, Health and Cultural Affairs (see
Boylan 1993, p. 67).
Acknowledgments
My thanks to Robert C. Hunt and Fred Winter for comments on various aspects of this
text; to Michel Brent, Clemency Coggins, Barbara Hoffman, Rosemary Joyce, Lyndell
Prott, Kapila Vatsyayan, and Karen D. Vitelli for information and bibliography to
which I would not otherwise have had access; and to Danton Char for hunting down
several invaluable references.
References
Archaeological Institute of America 1994, “Stolen Artifacts,” in: Newsletter 9 (3), 2.

Ashworth, G. J. and Larkhan, P. J. 1994, Building a New Heritage: Tourism, Culture and
Identity in the New Europe, London and New York: Routledge.
Bator, Paul M. 1981, The International Trade in Art, Chicago and London: University
of Chicago Press.
Bourdieu, Pierre 1980, “The Production of Belief: Contributions to an Economy of
Symbolic Goods,” in: Media, Culture and Society 2, pp. 261–93.
Boylan, Patrick J. 1993, Protection of Cultural Property in the Event of Armed Conflict (The
Hague Convention of 1954), UNESCO Doc. CLT-93/WS/12, Paris: UNESCO.
Brent, Michel 1994, “The Rape of Mali,” in: Archaeology 47, pp. 26–35.
——, in press, “A View Inside the Illicit Trade in African Antiquities,” paper given
at the Colloquium on Illicit Traffic, Gainesville, Florida, April 1993.
Burman, Harold S. 1991, “Issues Raised by the Draft UNIDROIT Convention,” 12
December 1991, Washington, D.C.: US Department of State, Office of the Legal
Advisor, Secretary of State.
—— 1994, “Press Guidance for Queries on Treaty Negotiations for Return of Stolen
Cultural Objects,” 29 June 1994. Washington, D.C.: US Department of State, Office
of the Legal Advisor, Secretary of State.
Choay, Françoise 1992, L’allégorie du patrimoine, Paris: Editions du Seuil.
Cleere, Henry (ed.) 1984, Approaches to the Archaeological Heritage: A Comparative Study of World
Cultural Resource Management Systems, Cambridge: Cambridge University Press.
Coggins, Clemency 1972, “Archaeology and the Art Market,” in: Science 175, pp. 263–6.
Cornell, Stephen E. 1989, Tourism and Economic Development: Considerations for Tribal Policy
and Planning, Cambridge, Mass.: Harvard University Project on American Indian
Economic Development, John F. Kennedy School of Government.
Greenfield, Jeanette 1989, The Return of Cultural Treasures, Cambridge: Cambridge
University Press.
Guyette, Susan 1994, “Strategic Planning for Culturally Appropriate Tourism Devel-
opment,” unpublished paper.
——, in press, Balancing Development for Cultural Survival: A Planning Guide for American
Indian and Rural Communities, Santa Fe: Clearlight Publishers.
Hoffman, Barbara 1995, “How UNIDROIT Protects Cultural Property,” in: New York
Law Journal, March 3, pp. 5–8.
Hofstadter, Dan 1992, “The Angel on Her Shoulder,” in: New Yorker, July 13 and
July 20, 1992.
Karageorghis, Vassos 1992, “The Trade in Cypriot Antiquities in London,” in: Inter-
national Journal of Cultural Property 1, pp. 331–2.
Langford, R. F. 1983, “Our Heritage—Your Playground,” in: Australian Archaeology
16, pp. 1–16.
Layton, Robert L. (ed.) 1989, “Introduction: Conflict in the Archaeology of Living
Traditions,” in: Conflict in the Archaeology of Living Traditions, London: Unwin Hyman,
pp. 1–21.
Lewis, D. and Rose, D. B. 1985, “Some Ethical Issues in Archaeology: A Methodology
for Consultation in Northern Australia,” in: Australian Aboriginal Studies 1, pp. 37–44.
Ley, John F. 1991, Australia’s Protection of Movable Cultural Heritage: Report on the Ministerial
Review of the Protection of Movable Cultural Heritage Act 1986 and Regulations, Canberra:
Australian Government Publication Service, for the Department of the Arts, Sport,
the Environment, Tourism and Territories.
Lipe, W. D. 1984, “Value and Meaning in Cultural Resources,” in: Cleere, H. (ed.)
1984, Approaches to the Archaeological Heritage, Cambridge: Cambridge University Press,
pp. 1–11.
McBryde, Isabel (ed.) 1985, Who Owns the Past? Papers from the Annual Symposium of the Aus-
tralian Academy of the Humanities, Oxford and New York: Oxford University Press.
Messenger, Phyllis M. 1989, The Ethics of Collecting Cultural Properties: Whose Culture? Whose
Property? Albuquerque: University of New Mexico Press.
Meyer, Karl E. 1973, The Plundered Past: The Story of the Illegal International Traffic in Works
of Art, New York: Athenaeum.
Murray, Tim 1993, “Communication and the Importance of Disciplinary Communities:
Who Owns the Past?” in: Yoffee, N. and Sherratt, A. (eds.) 1993, Archaeological Theory:
Who Sets the Agenda? Cambridge: Cambridge University Press, pp. 105–16.
Muscarella, Oscar White 1979, “Unexcavated Objects and Ancient Near Eastern Art,”
in: Levine, L. D. and Young, T. C., Jr. (eds.), Mountains and Lowlands, in: Bibliotheca
Mesopotamia XII, Malibu: Undena, pp. 153–207.
Parker, Patricia L. and King, Thomas F. 1990, Guidelines for Evaluating and Documenting
Traditional Cultural Properties, Washington, D.C.: National Park Service, Department
of the Interior.
Pommaret, Françoise 1981, “The Religious Culture of Bhutan and Tourism,” in:
Background Presentation Workshop on Development, Planning and Tourism in Bhutan, Aix-en-
Provence: CNRS, pp. 1–10.
Prentice, Richard C. 1993, Tourism and Heritage, London and New York: Routledge.
Prott, Lyndell V. 1989, “Problems of Private International Law for the Protection of
the Cultural Heritage,” in: Recueil des cours de l’Académie de droit international de La Haye
217, pp. 219–317.
Siehr, Kurt 1992, “The UNIDROIT Draft Convention on the International Protection
of Cultural Property,” in: International Journal of Cultural Property 1, pp. 321–32.
Silberman, Neil Asher 1982, Digging for God and Country: Exploration in the Holy Land,
1799 –1917, New York: Doubleday.
Treitler, Inga E., Ruppert, David, and Stoffle, Richard W. 1994, “The Changing
Culture of Resource Management,” in: Practicing Anthropology 16 (3), pp. 3–6.
UNESCO 1970, Convention on the Means of Prohibiting and Preventing the Illicit Import, Export
and Transfer of Ownership of Cultural Property, Paris: UNESCO.
—— 1983, Conventions and Recommendations of UNESCO Concerning the Protection of the
Cultural Heritage, Paris: UNESCO.
UNIDROIT 1991, “Committee of Governmental Experts on the International Protec-
tion of Cultural Property,” Study LXX, Document 23: Report on the First Session,
(Rome, 6–10 May 1991).
—— 1993, Draft UNIDROIT Convention on the International Return of Stolen or Illegally
Exported Cultural Objects, from the Fourth Session of the UNIDROIT Committee of
Governmental Experts on the International Protection of Cultural Property (Rome,
29 September–8 October 1993).
Vatsyayan, Kapila Malik 1972, “Some Aspects of Cultural Policies in India,” in:
UNESCO Series on Cultural Policy, Paris: UNESCO.
Wasserman, Stuart 1994, “Israel Visitors Can Dig Right In,” in: The Boston Globe, 20
February 1994, pp. 1 and 17.
Weil, Stephen 1989, “Who Owns the Nataraja?” in: Art News, May 1989, p. 188.
Winter, Frederick A. 1991, “Who Owns the Past? Ethical Dilemmas in Contemporary
Archaeology,” in: Explorers Journal 69, pp. 128–33.
Winter, Irene J. 1993, “Review of J. Greenfield, The Return of Cultural Treasures,”
in: Art Journal 52, pp. 103–7.
Winthrop, Robert 1994, “Conflicting Perceptions: Tribal and Regulatory Views of
Nature, Risk and Change,” in: Practicing Anthropology 16 (3), pp. 25–8.
Wylie, Alison and Levy, Janet E. 1994, “Archaeological Ethics in Print: An Annotated
Bibliography,” in: Public Archaeology Review 2, pp. 21–5.

Irene J. Winter - On Art in The Ancient Near East - Vol. 2 - From The Third Millennium BCE (Culture and History of The Ancient Near East) (2009)

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Irene J. Winter - On Art in The Ancient Near East - Vol. 2 - From The Third Millennium BCE (Culture and History of The Ancient Near East) (2009)

Uploaded by

Copyright:

Available Formats

On Art in the Ancient Near East

Culture and History of the

On the art in the ancient Near East / edited by Irene J. Winter.

Copyright 2010 by Koninklijke Brill NV, Leiden, The Netherlands.

All rights reserved. No part of this publication may be reproduced, translated,

Authorization to photocopy items for internal or personal use is granted by

printed in the netherlands

Introduction ................................................................................ vii

SCULPTURE AND THE EARLY STATE

Chapter Sixteen After the Battle is Over: The Stele of

EXPERIENCING ‘ART’ AND ARTIFACT

Chapter Twenty-Four Representing Abundance: A Visual

Chapter Twenty-Five Reading Ritual in the Archaeological

VIEWING (IN) THE PAST AND THE PRESENT

Chapter Thirty-Two The Eyes Have It: Votive Statuary,

The studies included in the present volume of collected essays range

role(s) played by belief systems as evident in ritual and religion no less

the prior oppositions are at least sometimes understood as the results

A number of cross-cutting threads related to work in the subfield

and observed empirically, is associated with meaning and argued to

Such arguments at least implicitly address the early modern::postmodern

of Chapter 25, for example, which yielded insight into Mesopotamian

AFTER THE BATTLE IS OVER: THE STELE OF THE VULTURES

To raise the question of the nature of narrative is to

Definitions of narrative, generally falling within the purview of liter-

all. For an investigation of visual representation, it seems important to

My aim is fourfold: to see whether this monument truly fits the

The anzu emblem appears on several “royal” monuments found at

triumphant over the vanquished enemy of Lagash and accompanied

god. It is important to note that no mention of the Zu-bird or of the

flounced garment probably made of sheepskin. The particular action

Only a small portion of the fourth register is preserved, the upper

question remains: why, after sacrifice and celebration, would Eannatum

first, as corresponding to the preamble of the text (cols. 1–3), which

divine acts and events of the mortal world.45 Groenewegen-Frankfort,

episode, and vertically, as a sequence of several events, makes the stele

To borrow Eco’s term, which Richard Brilliant has equally applied

The Historical Context

As Barbara Herrnstein-Smith has noted, “the fact that something is true

reconstruction of a history of the period and a profile of the workings

tract of fields and grazing lands along their twenty-eight-mile common

Although Lagash may have been dominant in the early stages of

It remains to be demonstrated how this historical background ani-

narrative is therefore selected as a mode of representation because it

Figure 1. Obverse, Stele of the Vultures of Eannatum, ensi of Lagash,

Figure 2. Reverse, Stele of the Vultures. Photo courtesy Musée du Louvre.

Figure 3. Reconstruction drawing, obverse, Stele of the Vultures

Figure 5. Silver vase of Enmetena of Lagash, c. 2420 b.c., Tello; Musée du

Figure 6. Votive plaque showing priest before the goddess Ninhursag,

Figure 7. Copper panel from temple of Ninhursag at al-’Ubaid, Early

Figure 8. Reconstruction drawing, reverse, Stele of the Vultures

Figure 9. Detail, reverse: Vultures and enemy heads, upper register.

Figure 10. Detail, reverse: Phalanx of soldiers, upper register.

Figure 11. Detail, reverse: Eannatum in chariot, soldiers behind, second

Figure 15. Votive plaque of Ur-nanshe, king of Lagash, c. 2500 b.c.

Figure 17. Boundary/land-grant stele, Early Dynastic I period, c. 3000 b.c.

EANNATUM AND THE “KING OF KIŠ”?

The Stele of Eannatum of Lagaš (the so-called ‘Stele of the Vultures’)

close examination of the stele, there are significant gaps—both in the

we see a hand holding a long spear-shaft, the point of which is directed

Let us now return to the so-called identification of the King of Kiš

Side 4 III 2´ k[a . . .]

for some imprecation regarding the consequences if the stele were to