Professional Documents
Culture Documents
(Culture and History of The Ancient Near East 62) Oscar White Muscarella - Archaeology, Artifacts and Antiquities of The Ancient Near East - Sites, Cultures, and Proveniences-BRILL (2013)
(Culture and History of The Ancient Near East 62) Oscar White Muscarella - Archaeology, Artifacts and Antiquities of The Ancient Near East - Sites, Cultures, and Proveniences-BRILL (2013)
Founding Editor
M.H.E. Weippert
Editor-in-Chief
Thomas Schneider
Editors
Eckart Frahm (Yale University)
W. Randall Garr (University of California, Santa Barbara)
B. Halpern (Pennsylvania State University)
Theo P.J. van den Hout (Oriental Institute)
Irene J. Winter (Harvard University)
VOLUME 62
By
Oscar White Muscarella
LEIDEN • BOSTON
2013
Library of Congress Cataloging-in-Publication Data
DS56.M848 2013
939.4–dc23
2013006440
This publication has been typeset in the multilingual “Brill” typeface. With over 5,100
characters covering Latin, IPA, Greek, and Cyrillic, this typeface is especially suitable for
use in the humanities. For more information, please see www.brill.com/brill-typeface.
ISSN 1566-2055
ISBN 978-90-04-23666-0 (hardback)
ISBN 978-90-04-23669-1 (e-book)
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.
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
PART ONE
SITES AND EXCAVATIONS
PART TWO
ARTIFACTS, CULTURES, FORGERIES, AND PROVENIENCE
1 The works I read included James Henry Breasted, A History of Egypt (2nd ed.; New York:
Charles Scribner, 1924), and W.M. Flinders Petrie, The Religion of Ancient Egypt (London:
A. Constable & Co., 1906), among others.
2 I attended the evening session because I had to earn my living in the daytime.
3 Grace eventually worked with Piet de Jong and became his assistant as well as a fellow
archaeological artist.
introduction 3
Agora because of unbearable class snobbery and went east. Given the dif-
ferences in these two encounters, it is no wonder that I chose to remain in
the world of ancient Near Eastern archaeology instead of classical.
My first engagement with Iranian archaeology came in 1960, the year I
finished my coursework in classical archaeology with Young and was hired
to teach history at CCNY, beginning in the fall. Solely because of my Gordion
background (I had taken only one course in ancient Near Eastern archaeol-
ogy at Penn), Robert H. Dyson Jr. invited me to come to Hasanlu, in north-
western Iran, for which deed I am indebted to him. Thereafter I excavated at
Hasanlu and eight other Iranian and Anatolian sites, and through this work
I became an ancient Near Eastern archaeologist. My doctoral dissertation
(finalized in 1965 and published in 1967) was of course on Phrygian fibulae,
my first artifactual love. In 1964 I left CCNY for a position in the Department
of Ancient Near Eastern Art at the Metropolitan Museum of Art (MMA),
which I left upon my retirement in 2009.
In my early years at the MMA I continued to excavate and to publish
articles derived from those activities. But after a few years of working con-
tinuously with purchased antiquities in the museum, I began to realize
that something was utterly wrong regarding the nature of how scholars
in the discipline, both at the museum and in academia, published antiq-
uities. The very first time I empirically learned about plundering and its
attendant destruction of ancient sites was in the early 1970s, when I heard
lectures by the Harvard Mesoamerican scholar Clemency Coggins and the
Iranian archaeologist Ezat Negahban on this major but for me hitherto fully
neglected issue. I first then noticed that not only my museum colleagues but
many others failed to make the obvious and necessary distinction between
an excavated artifact and a purchased antiquity acquired from collectors,
antiquity dealers, or the auction houses ubiquitous in Europe, the United
States, and in modern Near Eastern countries. I also learned that this activ-
ity was in fact a worldwide cultural problem. My mind stumbled but I soon
came into contact with a few other similarly minded colleagues, mainly
James Wiseman of Boston University and Ross Holloway of Brown Univer-
sity. In 1973 we formed a new archaeological group, the Association of Field
Archaeologists, and soon thereafter we began to publish The Journal of Field
Archaeology (mashallah, still in existence) through Boston University. The
battle had begun, and it has not ceased. My first article on plunder and
the role of museum trustees, officers, and curators as well as private collec-
tors in the destruction of this planet’s history appeared in the JFA (1/2, 1974:
221–222), and the publications have continued nonstop; my most recent
appeared in 2012.
4 introduction
Following upon this discovery, I also came to realize that a very large
number of antiquities in museums, collections, and manifold publications
worldwide were in fact modern forgeries, commonly baptized and accepted
as ancient artifacts by (too) many scholars and introduced into their discus-
sions, reviews, and analyses of ancient Near Eastern cultures. Alas, I did not
learn of this profound flaw in the archaeological discipline at large (i.e., not
only in ancient Near Eastern studies) from my educational and excavation
background, or from my MMA work. I never heard the word “forgery” men-
tioned in graduate school or in the field, or by colleagues. What I learned of
these issues derived from my normal research and my hands-on archaeolog-
ical background. I did discover that in the late nineteenth and early decades
of the twentieth century a few scholars had addressed these concerns, but
then the subject ceased to be discussed until recent times.
The volume presented here follows the evolution of my scholarly work
and interests and is divided into several categories, which, as noted above, I
see as interrelated fields and not isolated subsets. The first part deals primar-
ily with excavations and associated artifacts, issues in ancient geography and
the identification of ancient sites in northwest Iran, my research involving
the culture and chronology of the Phrygian capital at Gordion in Anatolia,
and the chronology and Iranian cultural relations of a site in the Emirate
of Sharjah. Part two is wide-ranging, as its title makes evident. Represented
are my interests in Aegean and ancient Near Eastern cultural and political
interconnections, the role of fibulae in revealing cultural and chronological
matters, and the gender-determined usage of parasols and their recogni-
tion in excavated contexts. There follow articles specifically concerned with
what I call the “Plunder Culture” and the forgery of both objects and their
alleged proveniences. In all cases, bibliographies will, inshallah, lead to fur-
ther investigation by scholars and (to me more important) by students.
PART ONE
Iran
chapter one
On August 25, 1936, Sir Aurel Stein completed a six-day excavation on the
mound called Dinkha Tepe, situated in the Ushnu valley in northwestern
Iran, and moved his camp about three miles to the east across the Gadar
River to a location near the modern village of Cheshmé Göl. There he
examined “a curious succession of conical mounds stretching in a straight
line at short intervals, known as Seh Gird.” Stein first thought that the
mounds might be “a series of burial tumuli.” After examining them and
finding them to be composed of a hard gravel, the same type of soil found in
the adjacent area, he concluded that they were “natural.”1 The mounds were,
therefore, not excavated, and Stein moved his camp to another area.2
In the summer of 1966, the writer and Robert H. Dyson, Jr., accompanied
by several members of the Hasanlu Project excavating at Dinkha Tepe,
visited the mounds and concluded that they were in fact tumuli and not
natural formations. At that time nine tumuli were counted; in 1968 a total of
eleven were recorded.3
The tumuli lie about five kilometers east of Dinkha Tepe and may be seen
with the naked eye from that mound. They were built about one kilometer
below and west of the foothills that form the eastern boundary of the valley.
The most important site recognized in the immediate area is a recently
discovered Urartian city located at a place in the same foothills now called
* Excerpted from Metropolitan Museum of Art Journal. Copyright © 1969 by The Metro-
“a fat elderly American” (i.e. someone who spoke English) who came to the area with an
Indian and his wife about 30 years ago after excavating at Dinkha. He also claimed that the
Indian found two vessels in one of the tumuli. It would seem that these vessels must actually
have come from one of the tepes sounded in the area by Stein (e.g. Stein, Old Routes, p. 377).
Tumuli G and H show unmistakable signs of having been excavated to a limited extent, and
they are probably the ones tested by Stein. However, two or three other tumuli also show
signs of excavation, infra.
3 Tumulus K, the eleventh recorded in 1968, was recognized as a tumulus by Christopher
Qalatgah, just slightly to the northeast (Figure 1).4 That site is known today
in the area for a pair of magnificent springs that gush from the rocks and
irrigate the fields below. All the tumuli are clearly visible from Qalatgah and
also from the modern Nagadeh-Ushnu road, which passes just below at the
base of the hills. The village of Cheshmé Göl lies about half a kilometer to
the northwest of the tumuli.
In 1968 the Hasanlu Project, under the joint sponsorship of the Uni-
versity Museum of the University of Pennsylvania and The Metropolitan
4 Qalatgah means “place of the fortress.” The site was first visited by the author, Agha
Z. Rahmatian, and two members of the staff, Christopher and Carol Hamlin, on August 31,
1968. We were guided by a local landlord who promised to show us the place where an
inscribed stele was allegedly found in 1967. A major stretch of Urartian-type walls, Toprak-
kale-type pottery (highly polished red ware), and Iron III Iranian sherds were discovered. A
second trip a week later led to the discovery of an Urartian inscription on a building stone
and an Urartian stone stampcylinder seal. Collectively, the evidence suggests that Qalatgah
is an Urartian site.
the tumuli at sé girdan: a preliminary report 11
5 For a report on the first campaign, see R.H. Dyson, jr., “Dinkha Tepe,” in “Survey of
Excavations in Iran During 1965–1966,” Iran 5 (1967) pp. 136ff.; Oscar White Muscarella,
“Excavations at Dinkha Tepe, 1966,” The Metropolitan Museum of Art Bulletin 25 (1966) pp. 16ff.
The 1968 campaign was codirected by the author and Professor Robert H. Dyson, Jr. The
author directed the work at Sé Girdan. He was ably assisted by Mr. Christopher L. Hamlin,
who planned the site and contoured Tumuli I, II, III, and E; Mrs. Carol Hamlin, who excavated
and planned Tumulus II; Miss Elizabeth Stone, who contoured Tumulus III; and Mr. Arthur
Smith, who was the photographer. Agha Z. Rahmatian was our most cooperative inspector,
who helped us in many ways. Agha Nozar Supheri was, as always, our right hand in all matters.
Our foreman, Beshir, deserves much credit and thanks for his intelligence, honesty, and help,
both to his Kurdish- and Turkish-speaking workers and to the excavation staff; through his
endeavors we were able to accomplish much work in little time.
12 chapter one
Tumulus I
This is the largest tumulus at Sé Girdan (Figures 1, 2, 5): its height is 8.25
meters; its diameter, limited by an irrigation ditch and the pea field, mea-
sures about 60–65 meters. Figures 6, 7, 11 show the present border of the
tumulus and the pea field. The tumulus was divided into quadrants using
compass directions as dividing lines. The northwest quadrant was further
subdivided into two sections, and the southern section was excavated. While
excavation progressed in the main cut, four narrow test trenches were exca-
vated around the tumulus (Figure 5).
The upper part of the tumulus fill was composed of a mixture of gravel
and earth, characterized by masses of small pebbles (1 in Figure 7). Below
this level the fill was composed of hard and firmly packed clay in combina-
tion with small pebbles, but not in the same quantity as in the upper level
(2 in Figure 7). After we reached this level, which continued until we sus-
pended digging, work proceeded very slowly.6 A color change was noticed
in the clay at a certain point, and there was also a lens of clay, but otherwise
nothing distinctive about the density and composition of the clay could be
detected (2A, 2B, and 2C in Figure 7). The color change may represent only
different sources for the clay.
After some days of digging, vertical cleavages were noticed in the north-
south scarp; the faces of some of these cleavages had a southeast, others
a northeast, orientation penetrating into the undug scarp. At first it was
assumed that these cleavages represented the drying and cracking of the
clay in the sun, but soon other cleavages were noticed in the surface of
the clay in the main cut. Work ceased for several days while these cleav-
age lines were cleaned and recorded. It was soon evident in the main cut
that a series of roughly concentric circular units were present (Figures 7,
8, 9, 10), and moreover, the cleavages joined neatly with some of the ver-
tical ones in the north-south scarp. Other cleavage lines were noticed in
addition to the circular ones: longitudinal lines that divided the circular
units into partitioned sections. Some of these partition-cleavage lines also
joined neatly with some in the north-south scarp, thus explaining the dif-
ferent angles observed there for the vertical cleavages. Examination of the
cleavages showed that there was a grain pattern on all the faces cleaned,
indicating that wood fences, stockade-like, had at one time served to con-
tain the clay and left their impressions upon it.
6 Each of the four strong pickmen found it necessary to strike the clay four to six times
before a section could be removed. The clay seems to have been packed in while still wet.
the tumuli at sé girdan: a preliminary report 17
Tumulus II
This is the smallest of the tumuli excavated in 1968, being about 47 cm.
in height and about 14 meters in preserved diameter (Figures 3, 4). The
mound was divided into quadrants, and the southwest one chosen for initial
excavation. The surface of the tumulus showed remains of recent hearths,
but no other features that might suggest disturbance were noticed.
The fill was very shallow and was composed of gravel and gray-brown
earth. At a depth of from 10cm, to 30cm., at different parts of the quadrant,
portions of a circular pile of rubble stones 10cm. to 30 cm. in diameter were
uncovered. In two areas there were gaps in the rock pile, and below the
larger gap we encountered the top of a well-built stone wall and a section
of another; this turned out to be the tomb. At this stage the rock pile was
completely excavated (Figure 12). It consists of a round mound of stones
several courses thick at the center and diminishing to one or two courses
20 chapter one
at the edges. The center of the rock pile was not under the center of the
tumulus but was actually some meters to the west. The tomb was placed at
the center of the rock pile.
Gaps in the rock pile noticed early in the excavation made it evident
that the tomb had been plundered: one gap in the southwest area seems
to indicate an abortive attempt; the large gap in the area directly over the
tomb represents a successful one. The stones scattered in this area, and
extending beyond the rock pile, at a higher level, represent debris from the
robbers’ trench, which was apparently dug from the east. This trench was
subsequently refilled with the same gravel and earth that covers the rest
of the tumulus. There was no evidence of the robbers’ trench in the north-
south balk despite the fact that there is a break in the line of stones. This
could indicate that the robbing occurred soon after the completion of the
tumulus and that the refill had consolidated with the undisturbed fill in
the course of time. Nothing on the surface of the tumulus gave any hint
that plundering had occurred, which further suggests that the robbery took
place in antiquity. Several of the other tumuli in the necropolis (I, III, F, G,
and H) have noticeable hollow depressions that indicate relatively recent
attempts at excavation. The depressions in two of these, F and G, represent
the soundings of Stein, according to some local inhabitants.
In the excavation of the tomb, the rubble stones of the overlying rock
pile were removed, and the south and east tomb walls were freed from
earth and rubble. The tomb (Figures 13–16) is built of neatly cut large flat
stones, set into a thick mud mortar layer. The same mud used in the mortar
was also applied as a plaster to the outside walls of the tomb. The walls
form a rectangular structure one stone thick and about 1.5 meters by 3.1
meters. Around the top (except on the robbed south side) the flat stones
are bordered by additional large flat stones, which are themselves bordered
by rubble stones. On the three sides not disturbed by the robbers, the depth
of the tomb is preserved fully to a height of 1.2 meters. There is no evidence
to suggest of what material the roof was constructed.7
Below the walls to a depth of about one meter the tomb was filled with
rubble stones mixed with gravel and earth. Either these stones were thrown
in by the robbers or they fell in from the disturbed pile above. Under this
rubble fill, a layer of well-packed pebbles and some fist-sized stones were
encountered, apparently representing a deliberate packing. Among the
7 A wood roof could have been employed over the stone tomb, cf. M. Gimbutas, Bronze
Age Cultures in Central and Eastern Europe (The Hague, 1965) p. 284.
22 chapter one
Figure 16. Sections of the tomb, north and west walls, Tumulus II.
the tumuli at sé girdan: a preliminary report 25
pebbles were two thin lenses of fine gray ash whose origin remains a mystery.
Under the packing was the tomb floor proper, constructed of large flat stones
of the same type used in the walls. The center of the tomb floor contained
no slabs but only a smooth and hard gravel surface spotted with red stains,
samples of which were collected. A test trench dug 25 cm. into this surface
demonstrated that it was virgin soil. It may well be that the pebble packing
rather than the partially slabbed floor served as the surface on which the
dead person was placed, and that the slabs were missing in the original tomb
construction.
Attempts were made to find out whether a tomb pit had been dug, but no
evidence of one was found. Sections were cut at both the north and south
ends of the tomb area, and no pit lines could be recognized. The present
surface of the valley is 2 meters above the stone floor of the tomb, and it
would indeed be possible to assume that a pit had been dug for the tomb.
Perhaps, since the gravel used as fill was of very much the same consistency
as the neighboring virgin soil, the outlines of the pit have been obscured.
The contents of the tomb consisted of a few small bone fragments in very
poor condition, found in the northwest corner at a depth of 40 cm. from the
top of the wall, and a small, nondescript disc-shaped shell bead. A small,
coarse sherd was found below the rock pile overlying the northeast corner
of the tomb, but it unfortunately yields no information.
Three sherds, each incised with part of a triangle, were found in three
areas of the tumulus fill, in each case just beyond the circular rock pile. We
will return to a discussion of these sherds shortly.
It may be seen from the plans and photographs that Tumulus II has not
been completely excavated. It would therefore be premature to arrive at a
negative conclusion concerning the presence of a circular stone revetment,
as was excavated in both Tumulus I and Tumulus III. The tomb area and
the overlying circular rock pile have been cleared, but not the outer areas of
the tumulus. Perhaps during a future season conditions in the pea field will
allow a test trench to be dug in a search for a revetment.
Tumulus III
southeast section was decided upon for initial excavation. The fill through-
out was solid clay mixed with a few pebbles. Aside from the upper area,
which had been softened by plowing, this clay was quite hard and compactly
laid down. No horizontal lines were visible in the balk, nor were there any
cleavages such as those recognized in Tumulus I (Figure 18).
Approximately 50cm. below the surface around the outer perimeter of
the tumulus, a section of a sloping ring of rubble stones one or two courses
thick and of varying sizes was uncovered (Figures 18, 19, 20). The ring is about
5.75 to 6 meters in width (measured horizontally) and seems to have served
as a revetment. Although many stones are missing as a result of plowing,
there is no evidence of plundering. A test trench was cut into the north slope
of the tumulus (Figure 17) and a sloping stone surface thereby uncovered,
indicating that the revetment encircled the tumulus in the same manner as
recorded in Tumulus I.
Some 2.5 meters below the center of the tumulus a small section of a
rubble rock pile was encountered. The outer border of the pile was one stone
thick, and the pile increased in depth toward its center, thus forming a low
mound; the outer edge, which was curved, indicated that the pile was round
in form (Figures 19, 21). The rock pile rested on a well-made floor of clay,
smooth and hard, only part of which could be cleared. Since it was evident
that the center of the rock pile was located in the southwest quadrant, a
trench extension was made in that direction; another extension was made
to the north to expose more of the rock pile and to allow extra room for
excavating the tomb.
A hollow dome in the clay over a depression in the stones 1.1 meters deep
indicated clearly that the rock pile had collapsed into the tomb chamber.
When this area was cleared, the top of a grave pit completely filled with
fallen rubble stones was revealed. Powdery remains and small fragments
of wood, which apparently belonged to the now decayed roof, were also
recovered mixed with the rubble. No rubble stones found in the area could
have served as a roof for the grave pit.
The grave pit was a neatly excavated rectangular area 2 meters wide, 3.5
meters long, and 1.2 meters deep. The walls sloped outward slightly and were
coated with a thin mud plaster. The level gravel floor of the pit was carefully
covered with a layer of small pebbles to a depth of 7–8 cm. (Figure 22).
Excavation into the gravel floor showed it to be virgin soil.
On the pebble surface was found the badly crushed skeleton of an adult
male (Figures 23, 24). It lay on its side with the head facing southeast and
the legs drawn up in a contracted position. The left arm was awkwardly
positioned under the right arm, and the back was twisted. The original
28 chapter one
Figure 22. South and west sections of the tomb pit, Tumulus III.
Figure 24. Skeleton on the floor of the tomb pit, Tumulus III.
position of the skeleton may have been distorted by the rubble collapse, but
the placement of the arms seems surely to have been original. All bones of
the skeleton were completely covered with a deep red color, specimens of
which have been taken for analysis.
The following objects were found on the pebble floor beside the skeleton
and clustered in an area to the west and northwest of the head (Figures 23,
24):
1. A perfectly preserved whetstone-like object terminating in a feline’s
head. The stone is finely grained and very smooth so that if it is a
whetstone it does not appear to have been used; perhaps it served
as a baton or scepter. Length 37cm., diameter 3.2 cm. (Figures 25, 26,
27).
2. A very fragmentary silver drinking vessel, the metal of which was in
excellent condition when found (Figure 28 for a reconstructed draw-
ing; Figure 23, 2 in plan).
32 chapter one
Figure 29. Bronze celt blade, Tumulus III. Iran Bastan Museum, Tehran.
34 chapter one
Summary
Tumuli II and III share certain features: each of their tombs was situated
away from the center of the overlying tumulus; each tomb was built into a
pit (not absolutely certain for II); each tomb had a pebble floor; and each
tomb was covered by a mound of rubble stones. Until a trench is cut into
the outer area of Tumulus II, we are not in a position to conclude that it
had a stone revetment like Tumulus III, but this is very probable. The major
difference between the two tumuli, aside from size, is that II contained a
well-built stone tomb whereas III contained only a simple pit as the grave
chamber. Whether this fact may be interpreted as reflecting a difference in
wealth between those who were buried in the tomb or a chronological gap
between the erection of the tumuli cannot yet be established.9
A feature Tumulus I shares with Tumulus III is the stone revetment.
With respect to the unique technique used in the construction of Tumulus I
(wood fencing), one may presume that the relatively small size of Tumuli II
and III precluded such an elaborate system.
8 Such markers were found within some Phrygian tumuli: G. and A. Koerte, Gordion
Ergebnisse der Ausgrabung im Jahre 1900 ( Jahrbuch des deutschen archäologischen Institut,
Ergänzungsheft V; Berlin, 1904) p. 39; R.S. Young, “Gordion 1956: Preliminary Report,” Ameri-
can Journal of Archaeology 61 (1957) p. 325 (Tumulus P); idem, “The Gordion Campaign of 1959:
Preliminary Report,” AJA 64 (1960) p. 228 (Tumulus W).
9 The low height of Tumulus II might have resulted to some extent from the plundering
activity, but there is no conclusive evidence that it was ever as high as Tumulus III.
the tumuli at sé girdan: a preliminary report 35
10 A fragment from the Stele of the Vultures of Eannatum, now in the Louvre, has a scene
that could represent the erection of a tumulus over a mass burial. Two men carry earth up a
ladder or slope in order to cover a group of dead men, the defeated enemy. The fact that they
are climbing seems to preclude the suggestion that we are witnessing a regular inhumation
burial. However, there is at present no known tumulus burial from the Mesopotamian area of
this time, or indeed later. For a drawing of the fragment and a suggestion that the scene does
represent the erection of a tumulus, see G. Perrot and C. Chipiez, A History of Art in Chaldaea
and Assyria (New York, 1884) pp. 177 ff., fig. 93.
11 Ankara: M. Schede in Archäologischer Anzeiger 1930, col. 480; T. Özgüç: and M. Akok,
“Die Ausgrabungen an zwei Tumuli auf dem Mausoleumshügel bei Ankara,” Belleten 11 (1947)
p. 59, where the excavators’ statement that the tomb of Tumulus 1 was directly under the
“Gipfel” seems contradicted by fig. 5; on p. 69 they state that the tomb of Tumulus 2 was
“unter der Mitte des Hügels …” Gordion: Koerte, Gordion, pp. 99, 105, 139ff. (Tumuli IV, II, V);
the Koertes specifically state, p. 40, that the tomb was under the center of Tumulus III, which
appears to be an exception at Gordion; R.S. Young in various preliminary reports: AJA 61 (1957)
p. 325 (Tumulus P); AJA 64 (1960) p. 228 (Tumulus W); AJA 62 (1958) p. 147 (Tumulus MM);
Bulletin of the University Museum 16 (1951) p. 11, pl. V (Tumulus G); Archaeology 3 (1950) p. 200,
fig. 7 (Tumulus B); AJA 70 (1966) pp. 267 ff. (Tumuli X and Y). Kerkenes Dagh: E. Schmidt, “Test
Excavations in the City of Kerkenes Dagh,” American Journal of Semitic Languages 45 (1929)
pp. 250 ff.
12 H.H. von der Osten, Explorations in Central Anatolia Season of 1926, Oriental Institute
Publication, V (Chicago, 1929) pp. 48 ff., fig. 78; M. Schede in AA 1930, cols. 479ff., fig. 23;
M. Forrer in Mitteilungen der Deutschen Orient Gesellschaft 65 (1927) abb. 20 with caption.
36 chapter one
tombs are usually constructed of wood, but this may only reflect ecological
differences between Anatolia and western Iran.
Several Lydian tumuli excavated in the region of Sardis also have the
tomb placed off-center, a feature, incidentally, that one may interpret as an
example of Lydia’s cultural dependence on Phrygia.13 In some tumuli there is
also a pile of rubble stones placed over the tomb. While there is no evidence
of a stone rubble revetment in the tumuli at Sardis, the elaborate stone wall
found in the large tumulus called Karniyarik Tepe may actually be related in
some manner to the type known at Sé Girdan.
Still another area where there are tombs placed off-center is on the
island of Cyprus at the necropolis at Salamis, recently excavated by V. Kara-
georghis.14 One of the tumuli, called Tomb 77, is a fourth-century bc ceno-
taph. The pyre, with its contents and covering rock pile, was excavated intact
only because it was missed by grave robbers who had tunneled straight for
the center of the tumulus and thereby missed their goal. In addition to these
parallels with Sé Girdan, the tomb off-center and the overlying rock pile,
there is another feature of some importance: thin stone rubble walls were
found that radiated out from the center of the tumulus, dividing the area
into sections in order to facilitate the orderly dumping in of the earth fill.
This employment of stone partitions was also recorded in Tumulus 3 of the
same necropolis.
Some European tumuli also present interesting parallels to those at Sé
Girdan. Tumuli of the second millennium bc excavated in the western
Ukraine, the Baltic area, and central Europe often have a tomb pit that is
covered with a pile of rubble stones. In addition, some have a simple stone
ring encircling the tumulus at the base.15 These stone rings do not seem to
have functioned as a revetment, as we suggested stone rings did at Sé Girdan,
13 Perrot and Chipiez, History of Art in Phrygia, Lydia, Caria, and Lycia (New York, 1892)
p. 262, fig. 159, p. 270; H.C. Butler, Sardis, I (Leyden, 1922) p. 10; G.M.A. Hanfmann, “The Fifth
Campaign at Sardis (1962),” Bulletin of the American Schools of Oriental Research 170 (1963)
p. 56, fig. 39; idem, “The Ninth Campaign at Sardis (1966),” BASOR 186 (1967) p. 39, fig. 25,
pp. 42 ff.; on p. 47 there is mention of a tumulus called Cambaz Tepe that has a tomb placed
under the center. For a brief discussion of Phrygian influence on Lydia see Oscar White
Muscarella, Phrygian Fibulae from Gordion (London, 1967) p. 44, notes 30, 31.
14 V. Karagcorghis, “Chronique des Fouilles en Chypre,” Bulletin de Correspondence Hel-
lénique 90 (1966) p. 377 (Tomb 77); BCH 89 (1965) p. 283 (Tomb 3, seventh century bc).
15 Gimbutas, Bronze Age Cultures, pp. 285, 308, 319ff., 420, 460, figs. 190, 212, 219, 273, 301.
Parenthetically I would mention in this context the “tumulus” covering the House of Tiles
at Lerna. A circle of stones surrounded thc earth mound, and a layer of stones covered its
surface. No burial was found, but the construction is certainly similar to that employed in
grave tumuli: Hesperia 25 (1956) p. 150, fig. 3, pp. 165 ff., fig. 5.
the tumuli at sé girdan: a preliminary report 37
but the same general idea—an encircling of the tumulus with stones—
seems to be in evidence. And it is this feature that particularly relates the
Sé Girdan tumuli to those known in Europe. An important difference may
be seen in the fact that it was normal for a European tumulus to have a tomb
built directly under its center.
One tumulus known to me from England that has a feature recognized
at Sé Girdan is a long barrow at Skendleby in Lincolnshire. Its excavation
produced evidence that upright wood posts and fences were employed in
its construction. The system of partitioning, or dividing areas into units, was
also recorded.16 This technique is the same as that employed in Tumulus I,
and, as already discussed, in tumuli from Ankara and Cyprus.
Tumuli in the Altai area do not present direct parallels with those at
Sé Girdan in that many are actually rock cairns rather than earth tumuli.
The rock pile covering the tomb pit was itself the tumulus and was usually
centered over the tomb below.17
An outer ring of stones surrounding a centrally positioned tomb is fairly
common in the Caucasus region.18 These stones, however, form a simple ring
and not a revetment, and as such may be more closely related to European
tumuli. I am told that some of the tumuli ruthlessly plundered in the Ardebil
area in recent years have an outer ring of stones like those in the Caucasus.
The objects recovered from the rubble-filled tomb at Sé Girdan do not
yield as much information about culture and chronology as one would wish;
especially lacking is pottery. Nevertheless, certain general comments may
be set down. The silver drinking vessel, although badly crushed, can be
partially reconstructed on paper (Figure 28) and has several parallels at
Ziwiye belonging to the Iron III period, there not earlier than the late eighth
century bc and continuing through the late seventh century bc19 The knife
and celt, however, are not distinctive enough to allow them to be placed
chronologically with any certainty.
16 C.W. Phillips, “The Excavation of the Giant’s Hills Long Barrow, Skendleby, Lincoln-
shire,” Archaeologia 85 (1936) pp. 60 ff., fig. 7 on p. 61, and pl. XXI, fig. 2.
17 M.R. Griaznov and E.A. Golomshtok, “The Pazirik Burial at Altai,” AJA 37 (1933) pp. 32ff.,
fig. 1; K. Jettmar, Art of the Steppes (New York, 1964) pp. 120ff., fig. 105.
18 J. de Morgan, Mission Scientifique en Perse (Paris, 1896) p. 43, fig. 45; H. de Morgan,
“Recherches au Talyche Persan,” Délégation en Perse, Mémoires 8 (Paris, 1905) pp. 256ff.,
figs. 339–342, p. 260, fig. 346, p. 262, figs. 347, 348, etc. For similar types of tombs in Italy see
A. Minto, Marsiliana D’Albagna (Florence, 1921) pp. 22 ff., 30ff., fig. 2, pl. VI.
19 T. Cuyler Young, Jr., “A Comparative Ceramic Chronology for Western Iran, 1500–500bc,”
Iran 3 (1965) P. 60, fig. 4, no. 10. There are examples from Ziwiye still unpublished.
38 chapter one
The most important object from the tomb, and the only one completely
preserved, is the feline-headed whetstone-scepter. The head of the feline is
stylized and simple in execution, and seems to be pre-Achaemenid in style.
Whetstones, often with detachable metal animal heads, are known from the
Achaemenian period and earlier.20 A few whetstones of a similar but smaller
type are reported from the Minusinsk area in Russia. These have animal
heads and were made in one piece. Tallgren states that they are difficult to
date.21 I would tentatively suggest a pre-Achaemenid date for the whetstone-
scepter found in Tumulus III.
The few sherds with incised triangles found in the fill of Tumulus II
(Figure 30) may fit within the Iron III period, perhaps late eighth to early
sixth century bc Painted and incised triangles were common motifs in that
period (hence Iron III’s original field name “Triangle Ware Period”) at such
sites as Hasanlu, Susa, Ziwiye, and Zendan, and at sites in the Caucasus.22
Recently there have appeared on the antiquities market vessels with incised
triangular decoration allegedly coming from northwest Iran (Figure 31);
these also seem to belong to the Iron III period.23
Sherds and other material found in the fill of a tumulus do not date its
construction except in the form of an ante quem non date,24 that is to say, the
objects may be interpreted as either contemporary with the erection of the
tumulus (a workman scattered a pot he accidentally broke) or earlier than
20 7000 Ans d’ Art en Iran (Paris, 1961) no. 689; R. Ghirshman, The Arts of Ancient Iran
(New York, 1964) p. 67, figs. 84–36; A. Godard, Bronzes du Luristan (Paris, 1931) pls. XI, XII, an
example from Susa and others from Luristan. In this context compare a “door socket” made
of a finely grained stone with a stylized ram’s head at one end, found at Hasanlu, dated to the
ninth century, R.H. Dyson, “Treasures From Hasanlu …,” Illustrated London News, Sept. 30,
1961, p. 536, fig. 12. The scepters discussed by D. Berciu, “A Zoomorphic ‘Sceptre’ Discovered
in the People’s Republic of Bulgaria …,” Dacia 6 (1962) pp. 397ff., may be related to but are not
of the same type as the one from Sé Girdan. For this last reference I wish to thank Professor
T. Sulimirski.
21 A.M. Tallgren, “Some North Eurasian Sculptures,” Eurasia Septentrionalis Antiqua 12
XXXIII, GS. 3; Young, “Comparative Ceramic Chronology,” pp. 68ff., p. 56, fig. 2; R.H. Dyson,
Jr., “Problems of Protohistoric Iran as Seen from Hasanlu,” Journal of Near Eastern Studies 24
(1965) pp. 204 ff., figs. 7, 9, 10, 11, 13; idem, “Iran, 1956,” Bull. Univ. Mus. 21 (1957) p. 36, fig. 27;
W. Kleiss and R.M. Boehmer, “Die Ausgrabungen aufdem Zendan-i-Suleiman,” AA 1965, cols.
763 ff., figs. 74–76; J. de Morgan, Mission Scientifique au Caucase (Paris, 1889) p. 148, fig. 155,
p. 151, fig. 162, p. 155, fig. 170. (I am not unaware of the fact that the sherds from Tumulus II
could be second millennium in date.)
23 See also Trésors de l’ Ancien Iran, Musée Rath (Geneva, Ig66) no. 672, fig. 64. Many others
In July of 1970 the Hasanlu Project, under the joint sponsorship of The
Metropolitan Museum of Art and the University of Pennsylvania, began its
second campaign at Sé Girdan, situated in the Ushnu valley in northwestern
Iran.1 It will be recalled that Sé Girdan is a cemetery consisting of eleven
tumuli of various sizes near the modern village of Cheshmé Göl and below
the recently discovered Urartian site of Qalatgah.2 The campaign of 1968 had
been basically a survey resulting in the partial clearing of the largest tumulus
there, designated I, the excavation of a small plundered tumulus, II, and the
excavation of an intact tumulus, III.
The aim of the second campaign was to excavate three additional tumuli
and to complete the excavation of Tumulus I with the view to learning some-
thing about the culture and chronology of the people buried in the cemetery,
since information of this sort had not been firmly established in the first
season. Our season did not begin until two weeks later than planned, and
we were therefore not able to complete the clearing of Tumulus I. We were,
however, able to excavate three of the other tumuli in the area, called IV, V,
and VI.3 On the plan published in Sé Girdan I, fig. 2, these tumuli are labeled
E, F, and G.
* Excerpted from Metropolitan Museum of Art Journal. Copyright © 1971 by The Metro-
Museum Journal 2 (1969) pp. 5 ff. (hereafter Sé Girdan I). Credits for the drawings in this first
report are as follows: figs. 12, 15, 16 are by Carol Hamlin; figs. 6, 7, 10, 17, 18, 19, 22, 23 are by the
author; the tracings from field notes and subsequent inking are by Maude de Schauensee. In
the present report initials are included with the drawings; Marie Miller did the inking of the
drawings. I wish to thank all for their help and cooperation in these undertakings.
2 Sé Girdan I, p. 5, note 4; a discussion of Qalatgah by the author will appear in a future
issue of Expedition. For a discussion of other Urartian sites in Iranian Azerbaijan see W. Kleiss,
“Bericht über zwei Erkundungsfahrten in Nordwest Iran,” Archaeologische Mitteilungen aus
Iran 2 (1969) pp. 20 ff.; and W. Kleiss, “Urartäische Plätze in Iranisch-Azerbaijan,” Istanbuler
Mitteilungen 18 (1968) pp. 1 ff.
3 The season began on July 17 and ended on August 27. The staff consisted of the author
as director and Michael Nimtz (University of Pennsylvania), Karen Rubinson (Columbia Uni-
versity), and Betty Schlossman (Briarcliff College) as archaeologists; Robert Lewis surveyed
and oriented the tumuli. Agha Nozar Sepheri was the able assistant to the director, and Agha
44 chapter two
Tumulus I
Ardeshir Ferzanegan was the representative of the Iranian Archaeological Service. My aim in
both Sé Girdan I and the present report is to publish the archaeological results as quickly as
possible. I therefore do not claim to have exhausted all the evidence available for purposes
of seeking comparisons and relationships.
4 Unfortunately, in Sé Girdan I, p. 9, fig. 5, the lower horizontal course was inadvertently
not recorded correctly in test trench 3. It is interesting to note that when we redug the test
trenches filled in by us in 1968 we found the gravel and earth to be hard and without any
indication that they had been dug two years previously.
the tumuli at sé girdan: second report 45
a horizontal course of stones was found; this course did not appear in
the three other test trenches dug. In 1970 we checked this information by
redigging test trenches I and 4: no horizontal course was found. Why this
course only occurs in the long trench and in one test trench is not known.
Perhaps certain parts of the ground needed leveling.
Tumulus II
An attempt was made to dig a trench outward from the area cleared in 1968
in order to confirm that there was a stone revetment (Sé Girdan I, p. 16).
Unfortunately, the landlord refused us permission although we promised to
refill the trench. There can be little doubt, however, that this tumulus had a
revetment since all the other tumuli excavated had such a feature.
Tumulus IV
1970 were given a true north orientation).5 Excavation began in the upper
part of the southwest quadrant within an arc forming a wedge-shaped area
extending 11.15 meters from the top of the mound.
Beginning about 1.50 meters west of the center of the mound, and just
below the surface, a closely packed mass of small stones—10 to 20 cm. in
diameter—was uncovered (Figure 4). The mass was approximately 3 by 6
meters in area and did not form any recognizable plan; it extended partly
into the northwest quadrant and was one to two layers thick. A coarse,
carinated bowl, dark gray in color, and showing evidence of burning, was
found nestled within the stones on the southeastern edge; a stone was found
inside the bowl (Figure 27). I will return to this bowl later.
5 This change was inadvertent. I assumed mistakenly that our surveyor in 1970 would
use magnetic north as orientation. When I discovered the change, it was too late to make a
correction.
48 chapter two
Figure 4. Plan of stones below the surface at the top of Tumulus IV.
meters, then turning westward until it disappeared into the undug balk
(the west balk of the north-south trench). The tunnel was more than one
meter in height, although we could not measure it exactly because its course
was directly over a hollow cavity and it was not considered safe to work
there.
The cavity began at a depth of about 5.50 meters, just below the tunnel.
When the loose fill at the bottom of the cavity was cleared, the upper part of
the tomb was exposed; the tomb chamber itself was completely filled with
earth. The cavity extended over the whole area of the tomb (Figure 8), and at
the southwest corner it became a tunnel that continued southwestward and
upward from the tomb, beginning at a place where the latter was damaged,
several stones having been torn away (Figures 9, 10). This part of the tunnel-
cavity could safely be explored only superficially, but loose slabs of stone
were seen there, slabs that certainly had been torn away from the walls of
the tomb chamber.
The relationship of the cavity over the tomb and the vertical shaft and
horizontal tunnel seems quite clear. Whoever dug the shaft, i.e., the tomb
robbers, began it in an attempt to reach the tomb. (Because the shaft is so
narrow in section, it may be assumed that we cut into it near its perimeter
rather than at its widest part.) At a depth of 3.50 meters they decided to dig
a tunnel, first going south, then west, and finally north, moving downward
until they reached the tomb at its southwest corner. It seems plausible
to assume that the robbers had a general idea where to find the tomb
but were uncertain about its exact position. The clay over the tomb had
been removed laboriously through the tunnel and up the shaft, work that
must have been slow and hard, and we may assume that there were many
helpers. The firmness of the clay kept the cavity and part of the tunnel
intact, but the roof of the tunnel where it left the shaft had collapsed; the
vertical shaft also filled up with earth and stones in the course of time—
in fact, may have been deliberately filled in so as to cover any traces of the
robbery.
The tomb lies in the western part of the tumulus, mostly in the southwest
quadrant, partly in the northwest quadrant (using either true or magnetic
north as orientation). It was placed so that, but for part of the short east
wall, it was away from the center of the tumulus (infra) (Figure 10). The
cavity created by the robbers’ digging activities extended over the entire
tomb and cleared part of the upper surface of all the walls. It took us
several days to clean out the earth fill in the tomb, mainly because, as we
approached the bottom, we encountered thick, wet mud, the water table
being at hand.
52 chapter two
Figure 9. The western and part of the northern and southern walls of the
tomb, Tumulus IV. Note the robbers’ tunnel and entry at the left.
the tumuli at sé girdan: second report 53
Figure 10. Plan of the tomb and outer stone ring, Tumulus IV.
6 Although not too clear from the drawing in Sé Girdan I, fig. 16, the corners of the walls
Figure 12. Section of the east wall of the tomb, Tumulus IV.
the tumuli at sé girdan: second report 55
In the southwest corner of the tomb five courses of stone were missing for
about 1 meter to the east; in the western wall two courses were missing for
about 40cm. to the north (Figures 9, 10). This destruction occurred when the
robbers entered the tomb from their tunnel, where, as already stated, some
of the slabs were to be seen.
Although the tomb had been plundered, the robbers left some objects
because of either haste or carelessness. No skeleton was found, but we did
recover a few fragments of human bone, all showing definite red coloration;
some of the pebbles from the tomb also had this red color. A small fragment
of a smoky-clear obsidian blade was found in the fill above the tomb, and
one of the stone pebbles removed from the mud was a red chert core
from which blades had been chipped. Presumably this was not part of
the tomb equipment but just another rubble stone. If we are correct in
this observation, the stone must have come from a local field and suggests
that a neolithic or earlier settlement existed in the area. A few coarse,
nondiagnostic sherds and a few scraps of nondescript bronze were also
found in the tomb fill.
More important objects were also recovered from the tomb fill. At the
western part of the southern half of the chamber and close to the floor,
within the mud, were found 565 gold beads of varying types and 38 stone
beads.
The beads are all quite small, as may be seen in Figure 13. There were 431
flat, round beads (labeled 3), 3.5mm. in diameter and .5 mm. in height; 87
round beads with double carination (5), 5mm. in diameter and 1.5 mm. in
height; 40 hollow, spherical beads (6), 7mm. in diameter, with walls .5 mm.
thick; 4 very thin, flat, round beads (8), 5mm. in diameter and 5 mm. thick; 2
round, lentoid-shaped beads (4), 2.25mm. in diameter and .5 mm. thick; and
1 round, narrow-walled bead with a relatively large hole (not numbered),
4.5mm. in diameter and 3mm. thick.
Those of stone included 31 round carnelian beads with a slight double
carination (10), similar to but slightly larger than some of the gold examples,
6mm. in diameter and 2mm. in height; and 7 solid, round carnelian beads
(9), 7.5mm. in diameter and 4mm. in height. In addition there was one
simple flat bead apparently made of tortoise shell (11), 5 mm. in diameter
and 2.5mm. in height.
In the eastern part of the southern half of the chamber we found one flat
bronze adze and three bronze axe heads, all of the same type, but each made
in a separate mold (Figure 14). The bronzes were in excellent condition,
albeit they were found in the mud. The edges of all the blades were quite
sharp, and it would therefore seem that they belong to the original tomb
56 chapter two
Figure 13. Gold and stone beads from the tomb, Tumulus IV.
contents and were not the tools used by the robbers to dig into the tumulus.
Those tools were not left behind as they were needed to dig into the other
tumuli in the area!
The adze is 13.8cm. long and 3mm. thick; it flares out slightly from a width
of 3cm. at the base to 4.35cm. at the outer edge.
Each axe has a shaft hole close to the back part of the weapon, a single
oblique point forming the rear, and an outward-flaring blade. The three axes
have slightly different measurements: (12) length 14.3 cm., width 4.6 cm.;
(13) length 14.5cm., width 4.3cm.; (14) length 13.9 cm., width 4.5 cm.
The north-south trench was excavated for a length of 10 meters, measur-
ing south from the tomb edge, down to the level of the top of the tomb. At
a distance of 4.5 meters south of the inner edge of the tomb we cleared an
irregular section of rubble stones, three stones and 90 cm. wide, extending
east-west across the trench and disappearing into both undug balks (Fig-
ures 8, 10); 1.75 meters north of these stones was a single stone sticking out
of the west balk.
There can be little doubt that the section of stones represents part of a
ring that encircles the tomb, rather than the remains of a central rubble
the tumuli at sé girdan: second report 57
Figure 14. Bronze axes and adze from the tomb, Tumulus IV.
58 chapter two
7 It has of course occurred to me that the revetments at Sé Girdan may actually have been
originally exposed and not covered with earth as they now are: that is, they are covered now
by earth from the upper part of the tumulus. However, the upper borders of the revetments
are never uniform, and there is no regularity in the manner in which they are laid down: gaps
and depressions, and shifts in levels, occur on all tumuli, as may be seen by looking at the
sections. One might conclude that early stone robbing would account for these irregularities.
I prefer to leave the matter open but suggest that the revetments were meant to be covered,
as concluded in the text.
the tumuli at sé girdan: second report 59
Tumulus V
This tumulus lies about 100 meters to the northwest of Tumulus IV, in a
row with Tumuli I, II, and IV (Sé Girdan I, fig. 2, F on plan). Its present
height is about 5 meters, its diameter 48 to 50 meters. At present it is
assymetrical in shape with a deep pit at the top (Figure 15). There were
no clear indications as to the location of the precise high point of the
tumulus, so we arbitrarily chose the center of the pit as our center point
(infra); we assumed that whoever dug the pit picked the highest point as
the center. After the usual division of the tumulus into quadrants we chose
the upper part of the southwest quadrant for excavation, using true north as
orientation.
60 chapter two
The fill consisted of gravel and clay; about 20 cm. below the surface we
began to encounter scattered stones. They covered a good part of the south-
ern area of the excavation but presented no pattern. Stones continued to be
found throughout the fill (in the southern area). The northern part of the
excavation, on the other hand, consisted of hard clay. After a time it became
clear that the softer gravelly clay mixed with the stones represented a dis-
turbed area, and we could see the faint outlines of an irregular pit (Figures 16,
17); the pit penetrated to a point just above the tomb subsequently discov-
ered.
In the northern part of the excavated area, the part consisting of hard clay,
five distinct cleavages were recognized in the section, and we were able to
isolate portions of them on the horizontal surface (Figure 18); other cracks in
the section may be cleavage lines or cracks from the sun, but we could not
tell. The five cleavage lines mentioned here are distinct and unmistakable
the tumuli at sé girdan: second report 61
and, as with Tumuli I and IV, reflect the use of portable partitions.8 No
cleavages were recognized in the north-south section.
At a depth of about 4 meters large rubble stones mixed helter-skelter with
flat stones appeared in the west, north, and south areas of the excavation.
Unfortunately these stones turned out to be the remains of three sides of the
tomb (we did not excavate the fourth side) and the disturbed rubble overlay
(Figure 19).
The tomb was apparently rectangular in shape, about 2.25 meters in
width, and formed from a pit dug into the earth. It was oriented roughly
east-west, with the southern wall entering the B–C section, the northern
wall entering the A–B section. The sides were the earth walls of the pit
itself, but because of the havoc we could not tell if they had been plastered
or smoothed. The upper edges of the pit apparently had been lined with
irregular flat stones or slabs. We were able to surmise this information
because some slabs were found on the edge of the pit and also because of the
analogy with Tumulus VI (infra). Within the tomb pit some more slabs, also
irregular in shape, were found (Figure 19), but we are not able to conclude
8 The distances between cleavages were 17, 30, 12, and 15cm.
the tumuli at sé girdan: second report 63
whether they represent a floor that was torn up or fallen slabs that originally
lined the edge of the tomb (cf. Tumulus VI). Otherwise, no floor could be
recognized; the deeper we excavated, the muddier the earth turned.
Soft grayish white ashy deposits were found mixed with the stones and
perhaps are the remains of a wood or reed roof, but we cannot be certain.
64 chapter two
The tomb had been ruthlessly torn apart by the robbers, making it impos-
sible to draw a plan. Originally a rubble-stone overlay covered the tomb, but
since this had been torn away, we found the rubble stones jumbled together
with the flat stones. Within the tomb were found fragments of bone scat-
tered about and part of the skull of a young adult male (Figure 19); a long
bone was found on top of some stones outside of the tomb at the north-
west edge. The only other objects recovered were a small gold bead, flat
and like one of the four gold beads found in Tumulus IV (Figure 13, no. 8); a
small, carinated, black and white stone bead, 1.2 cm. in diameter and 7 mm.
in height; and small, nondiagnostic fragments of bronze. All were found in
the disturbed fill around the tomb. Some pottery sherds were also found in
the tumulus fill. They are red-buff wares and generally non diagnostic (Fig-
ure 29) except for one sherd that was once part of a carinated bowl similar
to the one found in the stones on top of Tumulus IV (Figure 28).
Whether or not the tomb lies away from the center of the tumulus cannot
be stated because of the disturbance caused by the large pit. Since the high
point of the tumulus is now missing and the adjacent areas corrupted, we
have no objective guide. I will return to this matter shortly.
Test trenches were dug in the north and west quadrants, and a long trench
was dug from the main cut (Figures 15, 17, 20). These trenches revealed the
expected revetment of small stones that encircled the lower slopes of the
tumulus.
At the upper border of the revetment stones revealed in the test trench
in the western quadrant (Figure 15, X on plan) and just below the surface,
we found a redbuff-colored jar with an everted lip and a raised ridge at the
shoulder (Figure 31). Within the jar, which was in fragments, were found
badly crushed human bones, apparently those of an infant. The jar was
placed at the edge of the stones just touching them, implying perhaps that
those who deposited the burial knew about the revetment. Yet we cannot
rule out the possibility that the deposition at this particular place was acci-
dental. The vessel could be called an Iron II or III vessel, but I am reluctant
to make a more definite decision on the basis of a coarse, undecorated jar.
No other burials (except for an Islamic burial in Tumulus VI) were found
within the fill of the tumuli at Sé Girdan, but since we have not cleared
away all the upper fill of the tumuli, we are not in a position to make defini-
tive statements on the matter. In any event, even if the jar was buried at
the time of the erection of the tumulus, we do not know if the burial was a
significant event or simply an instance of someone taking advanage of the
tumulus as a convenience. I can see no reason to bring in a discussion of
sacrifice.
the tumuli at sé girdan: second report 65
Sir Aurel Stein mentioned that in his excavations of the tumuli in 1936
“… shafts were sunk on the top of a couple of these mounds …”9 Stein did
not mention which tumuli he tested with shafts nor how deep his shafts
penetrated. We therefore do not know if the pit recognized in Tumulus V
is Stein’s work, although this is quite possible. For what seems fairly certain
to me is that the pit does not represent the work of those who plundered
the tomb: the pit does not penetrate as far as the tomb (Figures 16, 17). In
fact, it seems very probable that the robbing and destruction of the tomb
occurred before the erection of the tumulus. Hard clay exists directly over the
tomb, and the cleavages, surely representing a technique of construction,
were in situ in the fill over part of the destroyed area, the area not disturbed
by the later pit. And directly below the hard clay and the cleavages was
found the destroyed tomb. The only conclusion possible, it seems to me, is
that the tomb had been plundered and destroyed after the interment and
deposition of the grave goods, and that the mourners of the dead person
decided to erect the tumulus nevertheless. Perhaps we may assume that
the tomb was robbed as a result of an enemy or bandit raid. Following this
act of desecration the survivors decided not to dishonor the dead man by
leaving him unburied, and erected the tumulus; why they did not arrange
his scattered bones eludes us.10
An interesting parallel (archaeological, not historical) for the erection of
a tumulus after its tomb was plundered apparently exists in Tuekta, about
120km. west of Pazyryk, in the Altai region of eastern Russia. S.I. Rudenko
excavated two stone kurgans, dated to the late sixth century bc, neither
of which exhibited any signs of disturbance. Yet when the tombs were
reached and cleared, it became obvious that they had been robbed. The
conclusion seems to be that they were plundered before the tumulus was
erected.11
pp. 93 ff., pls. XIII, XIV (I wish to thank Prof. Ann Farkas for helping me with translation); K.
jettmar, Art of the Steppes (New York, 1964) pp. 12 ff., figs. 104, 106.
the tumuli at sé girdan: second report 67
Tumulus VI
This tumulus is the last one in the row of seven counting southeast to
northwest (Sé Girdan I, fig. 2, G on plan). It is a relatively small and low
mound, with a preserved height of 2.5 meters and a diameter of about 30
to 38 meters (Figure 21). Like the other tumuli it also is surrounded by
cultivated fields.
In the top part of Tumulus G there was a large depression. Although
the depression could have resulted from Stein’s work, I was certain it was
evidence of plundering. Therefore I wished to excavate Tumulus H, less than
190 meters to the northeast of Tumulus IV. But because there was confusion
on the part of the local authority about whether or not H was part of the Sé
Girdan cemetery, I reluctantly had to excavate G.
The tumulus was divided into quadrants, and we excavated most of the
southwest quadrant and parts of others while clearing the tomb. It was
68 chapter two
not possible to be sure about the location of the original high point of the
tumulus because of the disturbed nature of the area. We arbitrarily chose
the center of the depression as the center point of our quadrants.
Surface features, aside from the depression, consisted of many stones 30
to 50cm. in diameter lying around the lower edges of the tumulus. They
appeared to have been loosened from the revetment stones encircling the
tumulus. The upper part of the rubble revetment was exposed for the whole
length of the southwest quadrant, and the complete length of the revetment
was exposed in a narrow test trench in the northeast quadrant (Figures 21–
23). The stones are of mixed sizes, 5 to 20cm. in diameter, laid down in two
or three courses. However, in the western half of the southwest quadrant the
upper part of the revetment consisted of one or two courses of rather large
slabs, similar in type to those lining the upper edge of the tomb.
The tomb is a pit cut into the earth and is an irregular oval in plan. It
is oriented northwest by southeast with an interior measurement of 4.20
by 2.25/50 meters. It had (as surmised) been plundered in the past, and a
section at its northwest end was destroyed. Because of the plundering and
the tumuli at sé girdan: second report 69
partly of flat slabs but mostly of large and small stones, that formed a
rough circle around the tomb. No evidence in any of the sections exposed
suggested that the rubble stones ever extended over the tomb. Needless to
say, we do not know the extent of the area cleared by the robbers, and the
sections yielded no evidence in this matter, but it is doubtful to assume that
they cleared away the stones neatly and uniformly down to the level of the
tomb on all its sides. Therefore, it would seem that the rubble mass was
laid down around the tomb and never functioned as an overlay, otherwise
so common at Sé Girdan, as seen in Tumuli I, II, III, and V. In this respect,
Tumulus VI reminds us of the fact that the tomb within Tumulus IV (another
plundered and disturbed tomb) also seems not to have had a rubble overlay.
The upper part of the tumulus fill, judging from the section in the north-
south trench (Figure 26), consisted of light-colored clay and gravel. Below
this was a layer of compact gray clay with pebbles that partly overlay the
rubble stones, and next was a layer of compact light gray clay that was
72 chapter two
packed firmly against the outer border of the rubble mass. Below this were
still another layer of clay, tan in color, and then dark and moist earth that
must be virgin soil. We observed no cleavages in the sections or in the
surfaces excavated.
A burial of an adult male was found about 38 cm. below the surface in a
section that partly overlapped the west wall of the tomb. The skeleton was
lying on its side and faced southwest; there were no objects with the burial,
but two stones had been placed about 20 cm. above the head. This burial
dates from Islamic times and has nothing to do with the tumulus and its
construction.
An assumption has been made both in Sé Girdan I and in this report that the
present high point of the tumulus is its “center” as understood by the ancient
architects. Most archaeologists who have excavated tumuli and discussed
the tomb position seem to have taken this viewpoint without specifically
defining their terminology. Yet it is important to realize that we do not know
what shape a particular tumulus had in its original state, given more than
two millennia of wind and rain erosion, not to mention human activities.
Nor do we know if the tumulus was originally built so that the shape was
uniform in its dimensions, i.e., whether it had uniform contour lines on all
sides, or whether this effect was not required. And we do not really know if
the highest point of the tumulus was understood to function as the “center,”
and that this point was kept in mind after a tomb was built. Another item to
be remembered is that the original edge of the tumulus will always be buried
under the present level of the surrounding fields.12 And it seems probable
to assume that this burying did not occur uniformly on all sides, so that the
plan of the tumulus will have been altered. In other words, the original shape
and geographical center of the tumulus may actually elude us.
The excavator of the tumulus at Takht-i-Suleiman in Iran stated that its
original center, and high point, had moved about 3 meters to the north-
northwest and did not correspond to the present high point (“Spitze des
Hügels”).13 I am hesitant to either reject or accept this conclusion because
to my mind it is apparently possible from reading the published section to
conclude that in fact the original and present high point are the same. This
would mean that the stone pile and wooden marker excavated there, not
under the present high point, were meant only to be a guide for the builders
up to a certain stage of the construction and were not meant to mark the
final high point of the completed tumulus.14 However, this conclusion is not
based on direct observation of the excavated section. In any event, no tomb
was found either under the stone marker or under the present high point,
thereby establishing that, whatever unit is used as a modern guide, the tomb
was in fact placed off-center.
Reexamining the evidence of the early excavations at Gordion, we find
that the Koertes used such terms as “Gipfel,” “Mitte,” “Mittelpunkt,” and “Zen-
trum” when discussing the geography of the tumulus.15 I assume that “Gipfel”
must signify the present high point. How they arrive at the term “Mit-
telpunkt” is not discussed, but they do state that the tombs within Tumuli II,
III, and IV were not under the “Mittelpunkt” but under the “Gipfel,” that is,
under the high point. At the same time, the grave of Tumulus I was almost
exactly under the “Mittelpunkt,” and that of Tumulus V was three meters
from the “Mittelpunkt”; the “Gipfel” is not mentioned.16 However, the later
1965, cols. 788 ff., especially 789. Unfortunately I did not become aware of this important
article until Sé Girdan I was published. W. Kleiss sent me an offprint of his and R.M. Boehmer’s
contribution (but not Wiegartz’s) on the excavations at Takht-i-Suleiman, “Die Grabungen
auf dem Zendan-i-Suleiman,” Archäologischer Anzeiger 1965, pp. 759ff., and I assumed it was
the only report: see Sé Girdan I, p. 4, note 22.
14 Wiegartz, “Die Ausgrabung,” cols. 795 f., Abb. 79, where the fill described as “Kies-
Bänder” would be the final course of earth laid down over the regular bands of “Lehm,” “Kies,”
and “Brauner Löss.” The “Humus” would be accumulated fill resulting from erosion and would
not be part of the original tumulus. Note that my conclusion would better fit the suggestion
that the stone circle surrounding the tumulus was originally exposed; see Wiegartz, “Die
Ausgrabung,” col. 792.
15 G. and A. Koerte, Gordion Ergebnisse der Ausgrabung im Jahre 1900 ( Jahrbuch des Deut-
schen Archäologischen Institut, Ergänzungsheft V; Berlin, 1904) pp. 38f., 99, 104f., 129f., 139f.
16 Koerte, Gordion, pp. 129, 139. My comments in Sé Girdan I, p. 22, note 11, about the tomb
of Tumulus III being under the center should be corrected to say under the “Gipfel.” The
74 chapter two
confusion is of course the reason for this present discussion on terminology. Note also that
both the Koertes and R.S. Young found wood masts or markers over some of the tombs at
Gordion: the Koertes found them under the “Gipfel,” the latter over the tomb. Did they mark
off the high point and the tomb, or only the tomb? Following Young, they presumably marked
off not the peak but the tomb. See p. 22 and note 8 of Sé Girdan I.
17 The references are in Sé Girdan I, p. 22, note 11. In the same footnote appear two different
words used by T. Özguç and M. Akok for describing the position of tombs within Tumuli I
and 2 at Ankara: under the “Gipfel” for I, under the “Mitte” for 2. In 1969 a Phrygian tumulus
was excavated near Ankara by archaeologists from the Middle East Technical University.
The tomb seems to have been placed off-center, away from the present high point; it is still
unpublished. Note that the Koertes, Gordion, p. 129, refer to the tomb of Tumulus I being in
the southwestern quadrant of the tumulus fill, using the “Kuppe des Hügels” as the center.
18 Sé Girdan I, pp. 22 f., notes 11, 13.
19 Sé Girdan I, p. 23, note 14; now also see V. Karageorghis, Excavations in the Necropolis of
Salamis (Nicosia, 1967) pp. 25 ff., 121 f.; V.`Karageorghis, Salamis: Recent Discoveries in Cyprus
(New York, 1969) pp. 71 f., 151 ff.
the tumuli at sé girdan: second report 75
The three tumuli excavated this season share general features with the two
excavated in 1968: Tumulus IV and probably also Tumulus I (unexcavated)
contained tombs built off-center (we can say nothing definite about the
tombs in the disturbed Tumuli V and VI); all the tombs were built into
pits cut into the earth; all the tumuli have encircling stone revetments;
Tumulus V had a rubble-stone overlay covering the tomb; Tumuli IV and
V contained cleavage lines; and the stone tomb in Tumulus IV is of the same
type as that in Tumulus II.
Within this area of agreement, differences do occur, demonstrating that
variety did exist and that no rigid system of tomb architecture obtained.
Tumuli IV and V contained pit tombs, the top borders of which were lined
with slabs of stone. The plain pit tomb excavated in 1968 in Tumulus III did
not have a stone-lined border. The tomb ofTumulus IV apparently did not
have a rubble-stone overlay, but rather it had a feature unique in the Sé
Girdan series, namely, a narrow rubble wall that must have encircled the
tomb. And Tumulus VI apparently also did not have a true rubble overlay
covering the tomb but a variety in the form of a packing laid down around
the tomb. Finally, Tumulus VI is also unique at Sé Girdan for its roughly
ovalshaped tomb pit; the tomb plans of all the other tumuli are rectangular.
Some general comments about tumuli as well as foreign parallels for the
Sé Girdan tombs and tumuli have been presented in the first report; a few
comments will be added here, although I make no claim that all sources
have been covered.
20 I originally thought that by using the upper border of the stone revetment as a circle in
each tumulus, I could arrive at a true center point. I subsequently realized that this method
would not work as there was no regularity in the position of these stones around the tumulus,
and therefore I could not get a true circle; see, for example, Sé Girdan I, p. 9, fig. 5. For a brief
discussion of the possible relationship of Lydian and Phrygian tumuli (and Cypriote tumuli
also), see my article “Phrygian or Lydian?” Journal of Near Eastern Studies 30 (1971) p. 63.
76 chapter two
Within Iran itself one must refer to the two tumuli at Takht-i-Suleiman,
one of which, Tepe Majid, has been partly excavated (supra). This tumulus is
larger than any at Sé Girdan. Aside from the conical rubble pile and wooden
marker mentioned previously, a circle of stones, 1.50 meters wide, extended
around the base of the tumulus. According to the excavator this circle was
originally exposed. This is a feature shared with some tumuli in Europe and
the Caucasus.21 The technique of tumulus construction was not the same as
that recognized at Sé Girdan: at the latter site there were no central rubble
piles with wooden masts and no outer circle of stones, and the earth was not
laid down in the uniform manner observed at Tepe Majid. The significance
of this will have to await the excavation of the tomb that no doubt lies within
the tumulus. It has already been mentioned that no tomb was found at the
center of the tumulus.
W. Kleiss recently published a plan of a tumulus from the Ardebil region
west of the Caspian Sea.22 The tomb was constructed of stone and built into
the center of the tumulus; it was oval in plan, reminding us of the plan of the
tomb in Tumulus VI. No rubble-stone overlay covered the tomb, but there
was a stone circle around the perimeter of the tumulus.
Another tumulus in Iran on which Kleiss reported lies at the foot of
the Iranian-Urartian site of Bustam, 35km. north of Khoy; it is still to be
excavated and we have no data on it.23 One wonders if there can be any
significance in the fact that the cemetery at Sé Girdan also lies close to an
Urartian site, Qalatgah.
At Boǧazköy in Anatolia a tomb was excavated in 1958 that may have been
originally placed under a tumulus, although this is not certain because of
disturbances in the area.24 The tomb is brought into discussion here because,
21 References in Sé Girdan I, pp. 22 f., notes 15, 18. N.G.L. Hammond, “Tumulus-burial in
Albania, the Grave Circles of Mycenae, and the Indo-Europeans,” The Annual of the British
School at Athens 62 (1967) pp. 77 ff., has written about Albanian and Greek Bronze Age tumuli.
He discusses the House of Tiles at Lerna (see Sé Girdan I, p. 23, note 15) and grave circle B
at Mycenae as tumuli with encircling stones. See also N. Yalouris, “A Mycenaean Tumulus
at Samikos,” Deltion 20 (1965) pp. 6 ff. (Greek with French resumé), a tumulus with a stone
circle; M. Ervin, “News Letter from Greece,” American Journal of Archaeology 74 (1970) p. 264,
for Bronze Age tumuli (Early Helladic date?) at Marathon.
22 Kleiss, “Bericht über zwei Erkundungsfahrten,” p. 19, fig. 16; Sé Girdan I, p. 24.
23 Kleiss, “Urartäische Plätze,” p. 23. The conical mound at Tusikarn on the road from
Kangavar to Jowkar looks to me as though it may be a tumulus, but it has not yet been
excavated. There is a puzzling reference to a tumulus burial in Persia where Clearchus
perished: Plutarch Artaxerxes 18. 5. Does Plutarch record an actual tumulus burial?
24 Peter Neve, “Die Grabungen im Wohnviertel J-K/20,” Mitteilungen der Deutschen Orient-
The finds from the three tumuli excavated at Sé Girdan in 1970 were few, as
we have seen. Fortunately, several of the sherds found in the fill of Tumuli V
and VI furnish us with some information about chronology. Several of the
25 E. Heinrich, “Die Vorbereitung: Bericht über … die im Euphrattal bei Aleppo begonne-
205ff. I should also like to call attention to the fact that the tumuli excavated on Bahrein
Island have their tombs placed under the center, and they also have a surrounding stone
wall: E. Mackay et al., Bahrein and Hemamieh (London, 1929) pp. 3ff., pl. iv; G. Bibby, Looking
for Dilmun (New York, 1969) pp. 59 ff., pl. ii.
27 B.A. Kuftin, Archaeological Excavations in Trialeti, I (Tiflis, 1941) (in Russian) pp. 101ff.,
figs. 107, 108; Claude Schaeffer, Stratigraphie comparée et chronologie de l’Asie occidentale
(London, 1948) p. 506, fig. 41; O.M. Japaridzi, Archaeological Excavations in Trialeti (Tiflis,
1969) p. 76, fig. 47; l owe this last reference to Ann Farkas.
78 chapter two
sherds come from deep bowls with plain incurving or outcurving sides
(Figures 29, 30), and one sherd from Tumulus V (Figure 28) has an incurved
rim and concave sides, representing a shallow bowl. Parallels for the vessels
represented by the sherds occur in levels of the Iron III period at several sites
in Iran, viz., Baba Jan, Godin, Hasanlu, Nush-i-jan, Zendan, and Ziwiye.28 On
the assumption that the sherds in the tumulus fill represent either earlier
or contemporary material that was inadvertently dumped as fill, we have
a terminus post quem date of Iron III for the tumuli. The vessel used as a
container for the child’s bones found in Tumulus V (Figure 31) seems to fit
into an Iron II or III background, but I am reluctant to state this in absolute
terms.29
28 Bowls: C. Goff Meade, “Luristan in the First Half of the First Millennium bc,” Iran
6 (1968) p. 122, fig. 10, nos. 1, 12; T. Cuyler Young, Jr., “A Comparative Ceramic Chronol-
ogy for Western Iran, 1500–500 bc,” Iran 3 (1965) p. 56, fig. 2, no. 6, p. 54, fig. 1, nos. 1, 2, 4,
p. 58, fig. 3, nos. 1, 6, 11, 17; R.M. Boehmer, “Forschungen am Zendan-i-Suleiman in Persisch-
Azerbeidschan 1958–1964,” Archäologischer Anzeiger 1967, p. 580, fig. 8; D. Stronach, “Excava-
tions at Tepe Nush-i-Jan, 1967” Iran 7 (1969) p. 17, fig. 6, nos. 1–3. Bowls with incurved rim and
concave sides: Goff Meade, “Luristan,” p. 122, fig. 10, no. 3; T. Cuyler Young, Jr., “A Comparative
Ceramic Chronology,” p. 54, fig. 1, no. 6, p. 56, fig. 2, no. 10, p. 60, fig. 4, nos. 7, 12; T. Cuyler Young,
Jr., Excavations at Godin Tepe: First Progress Report (Toronto, 1969) p. 119, fig. 43, nos. 4, 5, 10,
p. 123, fig. 44, nos. 6, 7, 9, 11, 14, 15, 17; Kleiss and Boehmer, “Die Grabungen,” pp. 759f., fig. 72,
nos. 4, 5, 6; see also some close parallels in R. Ghirshman, Village Perse-Achéménide, Mémoires
de la Mission Archéologique en Iran 36 (Paris, 1954) pl. xxxvii, nos. G.S. 1219d, G.S. 1224 from
level 2; there are also good examples from Agrab Tepe and Pasargadae, not yet published.
29 See, for example, T. Cuyler Young, Jr., “A Comparative Ceramic Chronology,” p. 56, fig. 2,
The nearly complete bowl found at the top of Tumulus VI (Figure 27)
within the stone debris close to the robbers’ shaft is indeed a good Iron III
vessel similar to the sherd from Tumulus V mentioned above (Figure 28).30
Its presence near the robbers’ shaft and its broken state surely indicate that
it is associated in some manner with the robbers’ activity. But did it come
from the tomb itself as booty, then to be dropped and abandoned? Or was it
the personal bowl of one of the robbers, brought with him to hold his yogurt?
We do not know, of course; but at least we have a terminus ante quem date
for the tumulus, also Iron III or earlier. In this respect we have been able to
reinforce the suggested dating for the Sé Girdan cemetery proposed in the
first report (Sé Girdan I, p. 24).
The axes (Figure 14) present a more difficult problem in terms of chronol-
ogy and foreign parallels because I cannot find any other axes of exactly the
same shape with the single sloping rear point. Axes with flaring blades and
multiple rear points are quite common in the Near East from very early times
continuing into the first millennium bc.31 At present it seems to me that it
would be correct to date the blades tentatively to the seventh or sixth cen-
tury bc on the basis of the archaeological interpretation reached for the date
of the tumuli.
The gold beads from Tumulus I (Figure 13) are very similar to those found
in the tomb of Tumulus III in 1968. However, they are not characteristic of
any one particular period, and we cannot discuss them chronologically.
No evidence exists that would allow us to decide which tumuli are earlier
and which later. Looking at the plan in Sé Girdan I, p. 6, fig. 2, we see that
seven tumuli exist in a row placed roughly east-west. They are all spread out
one from the other except for Tumulus II, which seems to have been tucked
in between Tumulus I and IV, implying perhaps that it was built after those
two were in existence. But we do not know which tumulus in the row was
built first.
Tumuli H, I, J, and K exist outside of the row and are spread out in no
apparent order. What their chronological relationship is to the others is of
course not known, and guessing will not help. We must, therefore, conclude
that the cemetery at Sé Girdan is an Iron III creation, perhaps seventh or
sixth century bc, and not make any finer distinctions.
chapter three
In 1968 and 1970 six of eleven surveyed tumuli were excavated by the author
at a site called Sé Girdan, a few miles northeast of Dinkha Tepe and about
15 miles southwest of Hasanlu and Lake Urmia in northwestern Iran (Fig. 1);
the reports of the excavation were published in 1969 and 1971.1 The six tumuli
(labeled I to VI) were laid out in a roughly straight row, NW–SE. Four of
their tombs had been plundered at some time in the past. All the tombs
were neatly cut pits; all were rectangular, except VI, which was an irregular
oval; II and IV had stone lined walls; III was a plain pit, while V and VI had
only the upper border of their pits lined with courses of flat stones; II and III
(IV?) had pebble floors; all the tombs but IV had a stone rubble overlay; and
all tumuli were encircled at their perimeters by a rubble stone revetment.
The tomb of Tumulus I (the largest in height, 8.25 m, Fig. 2) was not found,
and only Tumulus III and its tomb was recovered intact (Fig. 3). Relatively
few artifacts were recovered. From the plundered Tumulus IV (Fig. 4) were
retrieved three socketed axes and one adze (Fig. 5), and 565 small gold and
38 flat and rounded carnelian beads. The intact Tumulus III contained a
feline-headed stone scepter or whetstone, a fragmented silver vessel, an
adze, a knife, many stone and paste beads, and two fragmented silver rods.
A few nondescript sherds and vessels were recovered from the surface and
fill of Tumulus II, IV, V and VI; also, from the fill of Tumulus I, a circular lid
with a solid knob made of a reddish fabric with inclusions (Fig. 6). On the
basis of these artifacts, for which I could find few if any comparanda, and
depending primarily on tumuli comparisons in Anatolia, and elsewhere, I
dated the tumuli to the Iron III period of Iranian archaeology, the 7th–6th
centuriesbc.2
* This article originally appeared as “The Chronology and Culture of Sé Girdan: Phase III,”
figs. 13, 14; for the proposed chronology see Muscarella 1971, 22–24; and Muscarella 1971, 23–28.
For some unexplainable reason I neglected to mention the lid in Fig. 6 in these publications.
It may supply an ante quem non date for the tumulus.
84 chapter three
3 Deshayes 1973, 176–178. Deshayes generously shared his article with me in advance of
its publication.
4 Muscarella 1973, 178–180.
88 chapter three
were 3rd millennium parallels more available than my proposed later dated
examples, and which I continued to accept. I recognized no observable
relationships between the Caucasian axes,5 the adzes, or with the (plain!)
scepters he brought forth as comparanda. Nor did the nature of the tumuli
constructions he presented reveal relevant comparative information. They
did share the off-center from the tumulus peak tomb placement, but the
rare rubble walled tombs were not the same as the neat, well-cut slabs at
Sé Girdan; and rubble revetments surrounding each tumulus—a distinctive
feature of the Sé Girdan tumuli (not mentioned by Deshayes; see below)—
were absent; further, ochre continued in use over the millennia.
After re-reading the original reports and the 1973 articles I believe that
given what was presented as evidence by Deshayes for an earlier chronology,
my response was not archaeologically unreasoned. Perhaps in hindsight
more attention should have been paid to the absence of iron in the tombs
and relevance of the presence of ochre (for copper, see below). But at that
time, of course, Scythians and Cimmerians were on my mind, and I did not
have the evidence now available—significant data unknown to Deshayes
and me.
In May 2002, three decades after the last discussion, Elena Izbitser, a Rus-
sian scholar at the Metropolitan Museum of Art, brought to my attention
an article by B.A. Trifonov and a book by A.D. Rezepkin, whose works, pub-
lished in the same year (2000), challenged my late dating and argued that
the Sé Girdan tumuli were constructed in the Early Bronze period.6 Trifonov
called attention to the uniqueness of tumuli in Iran and the lack of com-
parative material available at the time of excavation, both, he recognized,
causing difficulties for recognition of a correct chronology. But now, how-
ever, new information demands that the site’s chronology be reevaluated
and addressed anew. Both authors, Trifonov in more detail and analyses,
demonstrate that within the Maikop cultural period in the northern Cau-
casus tumuli at several sites share a crucial feature with those at Sé Girdan:
the outer encircling stone revetment (Fig. 7). As for the tombs, they share
5 He cited examples from his book (Deshayes 1960), one of which, no. 2103, was not exca-
vated and has little to do with the Sé Girdan examples: compare the axe parallels suggested
by Deshayes with those mentioned below and in note 8, which are indeed acceptable paral-
lels. Note that the closest parallel I could then find was an axe from Nimrud, Muscarella 1971,
25, footnote 31; Muscarella 1988, 81.
6 Trifonov 2000 (in Russian—translated for me by Dr. Izbitser); Rezepkin 2000. Trifonov,
246, footnote 1, cites the scholar G.N. Kurochkin and an earlier (1989) article by Rezepkin
where this early chronology for Sé Girdan had previously been expressed.
the chronology and culture of sé girdan: phase iii 89
a simple pit or stone lined walls, pebble floors, the tomb covered by rubble
stones, also body positions, and the presence of ochre on the bones.7
Further, and fundamental to the issues of chronology and culture under
review, both authors presented examples of socketed axes from several
Maikop period sites, Lechinkai and Pyatigorsk, that are obvious and rele-
vant parallels to the Sé Girdan examples, a bent butt and a blade with a
curved base (Fig. 8).8 Among these parallels is a hoard of socketed axes and
7 Trifonov 2000, 244, 255–256, fig. 1 for a neat juxtaposition of the tumulus construction
of Tumulus III at Sé Girdan and one from the north Caucasus (here Fig. 7); Rezepkin 2000,
20–21, 54, Taf. 10,11, 20, 26, 27, 32, 38, 49, 51, 63, 64, 74, 75, 76; Munchaev 1994, 184, fig. 45, 1.
For similar tumulus and comparable tomb constructions see Batchaev 1984, 116, 118, 120—
fig.; Chechenov 1984, 185, 195, 207—fig. Trifonov 2000, 256—exaggerated the extent of the
agreements of the body and head orientation between the two regions under review. Note
that neither author was aware of the 1973 Deshayes and Muscarella articles.
8 Trifonov 2000, 257: compare his fig. 3 (here Fig. 8), nos. 1–3, axes from Sé Girdan with
those from the northern Caucasus, nos. 6, 8 (his adze comparison is not so close); no. 7 is from
Tianeti in the northern part of the southern Caucasus (see Kushnareva 1997, fig. 18, no. 23).
See also Rezepkin 2000, 6, 20–21, Abb. 2, nos. 64, 65; Chernÿkh 1992, fig. 24: 24, 25. Rezepkin
2000, 20: cites the article by Batchaev and Korenevskii 1980, 79–83 on the Lechinkai site that
shows the same relevant Maikop period axes noted by Trifonov, including an example from
Pyatigorsk in the Northern Caucasus, fig. 3: 3 (neither these authors, nor Chernÿkh, refer to
90 chapter three
adzes from a site near Erevan (called Prierivanian hoard in Russian), Arme-
nia, south of the homeland and half the distance between Sé Girdan and
the northern Caucasus.9 Trifinov also cited the silver vessel from Tumulus III
as fitting into a Maikop background (also A. Sagona, personal communica-
tion).10
In my 1971 publication I assumed that the Sé Girdan axes and adzes
were bronze, but in the 1980s an axe and adze from the site in the Metro-
politan Museum were tested and disclosed to be 98.6 and 99.3 % copper
the Sé Girdan or Deshayes publications). Trifonov correctly notes that the Sé Girdan adzes
are not identical to the Caucasian examples.
9 Trifinov 2000, 257; see also his satellite photo, p. 245, and Munchaev 1994, 171, fig. 3 for a
map of sites. The Erevan axes, not the adzes, closely parallel those from Sé Girdan. Chernÿkh
1992, 64–65, fig. 21, had previously discussed these axes but he could find no parallels and
tentatively attributed them to the later Kura—Araxes, Transcaucasian period.
10 Trifinov 2000, 258; see also Munchaev 1994, 201, fig. 51.
the chronology and culture of sé girdan: phase iii 91
11 Muscarella 1988, 81, nos. 143, 144, footnotes 1, 2; arsenic was not mentioned. The two
objects came to the Museum as part of the excavation division with Iran.
12 Trifonov uses the term bronze to describe not only these artifacts but also those from
the culture in general; Chernÿkh also does the same throughout his book. I suggest this is
incorrect and misleading—the metal is copper or, arsenical copper, not arsenical bronze.
Lyonnet 2000, 202, correctly uses “cuivre arsenic.” See Chernÿkh 1992, 74 on arsenic as a
component of Maikop period copper artifacts.
13 See Muscarella 1969, figs. 13–16; Muscarella 1971, figs. 5, 8–12, 24, 25.
14 Munchaev 1994, 192, fig. 47.
15 Burney and Lang 1971, 77 mention without references a pig/horse headed whetstone
from a tumulus near Lenkoran. Trifinov 2000, 258–259 notes that there are no Maikop
examples, and he does not present relevant parallels. The animal headed “Tierkopfzepter”
shown in Rezepkin 2000, 14, Abb. 5 are not relevant.
16 Sotheby Parke Bernet Inc. New York, May 20, 1982, no. 29: the Sé Girdan example is
cited as a parallel. Indeed, it could have derived from one of the six unexcavated tumuli at Sé
Girdan from recent, post 1970 plunder.
92 chapter three
And there was no pottery recovered in the single intact tomb at Sé Gir-
dan (Tumulus III), and none recovered (abandoned) in the plundered ones,
which is certainly strange, an anomaly. Whether these differences collec-
tively reflect a cultural or, rather a chronological divergence between the
two distant regions warrant future research and discussion.
Tumuli are rare in Iran but two further examples are archaeologically
recorded in northwestern Iran. Although not fully excavated, they neverthe-
less preserve important evidence that allows them to be cited, if tentatively,
as culturally and chronologically relevant to our tumuli. Their existence
not too distant from each other or from Sé Girdan, suggests that the lat-
ter tumuli were not an isolated occurrence, although the latter alone may
be characterized as a planned necropolis. At Zendan-i Suleiman, north of
Takab, southeast of Sé Girdan, a partially excavated tumulus (the tomb was
never discovered) has an encircling revetment wall and the tomb is surely
off-center.17 The other tumulus, only surveyed, is 27 km west of Ardebil, west
of the Caspian, to the northeast of both Sé Girdan and Zendan; it too has
an outer revetment wall.18 This noteworthy feature characterizes the three
tumuli sites recorded to date in northwestern Iran.
This phase III of the investigation thus allows us to identify the Sé Gir-
dan (and perhaps the other two mentioned) tumuli as components of the
Early Bronze/Maikop cultural period. Their historical reality is all the more
complex and noteworthy when situated in the local historical background.
To date no Maikop period sites have been identified in northwestern Iran,
but in fact Maikop period material even in its homeland is archaeologically
known primarily from burials. However, a very large number of excavated
and surveyed sites, about 35, were identified by an Italian team in the plain
and foothills on the western shore of Lake Urmia as having Early Bronze
Age Transcaucasian (ETC) wares of its various sequential phases. The exca-
vated sites include Geoy Tepe (Period K), Haftavan, and Tappeh Giljar.19 On
the relevant ETC sites they discovered: numbers 2, 3, 7, 12, 16, 19, 20–23, 33, 34, 36, 42, 51, 55, 57,
60, 65,72–75, 79–81, 87, 88, 97, 98, all placed on the map opposite on 179 (no. 16, Tappeh Giljar,
sometimes written Gizler, is a major long lasting ETC settlement). For other discussions of
ETC sites in NW Iran see Burney and Lang 1971, 45, 53, 59–61; Dyson 1973, 697–699; Sagona
1984, 60–65, 241–242, 246–247; Kushnareva 1997, 43 ff., especially 65, 69, 70 (but missing here
the chronology and culture of sé girdan: phase iii 93
the eastern side of the lake is the excavated site Yanik Tepe, and about 9
unpublished sites surveyed by W. Kleiss that yielded ETC wares.20 Some 45
or more ETC sites have also been identified to the south of Lake Urmia,
including Hasanlu Period VII, which has a couple of “classic” ETC sherds,
and also in an earlier context, pottery of the earliest ETC, phase, and Godin-
Tepe.21 They occur east of the Mahidasht, close to Malyer; the range for Iran
is ca. 3000–2200 bc.22 In general, then, they may all disagree more or less
with the incipient date, but not the terminal one.
New chronologies for the Maikop period from C 14 data suggest a. range
from ca. 3700/3500 to 3000/2900bc, a long time. Trifonov dates Sé Girdan to
the second half of the 4th millennium bc, ca. 3500–3200bc, i.e. presumably
within the culture’s middle to late phases (and earlier than the initial ETC
dates suggested by Sagona and Kushnareva). And he suggests that rather
than being contemporary to the local ETC sites of Urmia, the tumuli are
more likely to have been constructed before their arrival.23 If his chronology
is essentially correct, and noting some of the incipient ETC chronologies
suggested, the Sé Girdan tumuli are consequently artifacts of an intrusive
cultural occupation in northwestern Iran at a time earlier than the ETC
incursion there; they represent its precursor. An alternative interpretation
might propose that the late Maikop and ETC cultures overlapped in part in
northwestern Iran, that both cultures co-existed side by side for some time.
This view surfaces because the proposed chronologies suggest that the ETC
are the important Pecorella, Salvini et al. and Sagona publications, any mention of Godin
Tepe; and map on p. 46 is useless on the NW Iranian ETC sites—none are cited; the problem
exists also with Lyonnet 2000, fig. 1); Meyer 2001 also omits Pecorella, Salvini et al. 1984; on
314 he lists an incomplete 43 sites on the west and northern areas—up to the Araxes.
20 Meyer 2001, 314, n. 50; see also Edwards 1986, 68, map, fig. 3, and App. A. 75. Both
publications omit the important evidence of Pecorella, Salvini et al. 1984, and the latter omits
references to several surveys to the south listed above.
21 Dyson 1973, 699; Sagona 1984, 122, 242–243; Meyer 2001, 313, n. 43; more information
International sur l’ Archéologie du Proche-Orient Ancien, Paris, April 15–19, 2002 (given
to me by E. Izbitser), he gives the full range of Maikop as ca. 3700/3500–3200/2900 bc; in
the same internet site S. Korenevskii gives ca. 3950/3600–2900 bc, both based on corrected
C 14 analyses; Chernÿkh 1992, 78–79, dates the Maikop kurgans to the 2nd half of the IVth
millennium bc, continuing into the IIIrd.
94 chapter three
24 R.M. Munchaev in his book (Munchaev 1975, 394) refers to his fig. 81, nos. 16, 17as axes
from Georgia. These are surely related to those from Sé Girdan but no information regarding
their proveniences or acquisitions are given: no. 16 seems closer (its long butt) to the Erevan
hoard examples than to those from Sé Girdan; see footnote 8 above.
25 I want to thank Karen Rubinson, Elena Izbitser, Alex Bauer, and Tony Sagona for biblio-
graphical information, and the latter two especially for valuable discussions on chronological
and cultural issues and implications.
the chronology and culture of sé girdan: phase iii 95
Bibliography
Precise Dating of the Tumuli at Sé Girdan. In: The Scholar’s Destiny, to the Cente-
nary of Boris A. Latynin (St. Petersburg), 244–264 (in Russian—translated for me
by Elena Izbitser).
Wiegartz, W. 1965: Die Ausgrabung am Tumulus Tepe Majid. Archaeologischer An-
zeiger 4, 788–799.
Young, T.C. 1966: Survey in Western Iran, 1961. Journal of Near Eastern Studies 25,
228–239.
———. 1975: Kangavar Valley Survey. Iran 13, 191–193.
chapter four
Modern archaeological interest in Urartu and its culture has several phases.
The first, which lasted until around 1945, started in the 1870’s, when, after
several objects reported to have come from Urartu appeared on the mar-
ket, the British Museum began to dig at Toprakkale in modern Turkey.
After a short time, D. Raynolds, an American missionary, and E. Clayton,
a British vice-consul, resumed work at Toprakkale for the Museum. Mean-
while, objects were continuously being purchased by various people from
local inhabitants, and these, together with the few objects and architectural
elements being excavated, formed a corpus of Urartian art. When in 1898
C.C. Lehmann-Haupt and V. Belck began what may be considered the first
attempt at scientific excavation of Toprakkale—the others could only be
called treasure hunts—they had a good idea what they expected. In 1911–
1912, two Russian scholars, M.J. Marr and I.A. Orbelli, did some more digging
at Toprakkale, and at nearby Van, but interest soon petered out. Serious
excavation did not begin again until 1937–1938, when an American expe-
dition led by Kirsopp Lake undertook a campaign at both Toprakkale and
Van.
Russian scholars became interested in that part of ancient Urartu pres-
ently situated within the Soviet Union’s borders, after 1930. At that time
survey teams began to seek out and record Urartian sites and inscriptions.
In 1939, B.B. Piotrovskii began one of the most significant excavations ever
undertaken in Urartian archaeology: the site of Karmir Blur (Red Hill), near
Erivan in Soviet Armenia. Work there continues to the present and one may
argue that the results achieved by Piotrovskii have played a large role in
reviving interest in Urartu.
We see, then, that British, American, German, and Russian archaeologists
were initially involved in the discovery of Urartu’s past.
The period after the Second World War begins the second, and more
intense, phase of Urartian scholarship. This phase is marked by very active
the Ushnu valley of northwest Iran. For several days after the beginning
of excavations, workmen pointed to a cluster of trees to the northeast, on
the slopes of a mountain that formed the eastern boundary of the Ushnu
valley. There, they said, was a very important place, an ancient city, and that
the area was called Qalatgah, or place of the fortress. On July 23 one of the
local landlords also talked of the site, and he added that there was a road
there and that in 1967 local peasants had found a large stone, broken into
three pieces, all of which had writing engraved. His description of “stick
like” writing suggested that he had seen cuneiform. Yes, he had himself
seen the stones and he was able to tell us that one or two pieces had been
taken by a local antiquities dealer (who never paid the promised money),
who in tum sold the stones to a dealer from Tehran; the third piece was
taken to the authorities in Ushnu, the valley’s chief town. He knew the
qalatgah: an urartian site in northwestern iran 101
exact place where the stones were found and would be glad to take us there.
At last on July 30, I first visited Qalatgah, along with Agha Z. Rahmatian, the
representative of the Archaeological Service of Iran, and Carol and Christo-
pher Hamlin, at that time graduate students at the University of Pennsylva-
nia.
Qalatgah is a large and steep site consisting of several high spots. A few
hundred meters up from the modern road, two magnificent springs gush
forth from a vertical rock outcrop situated behind a cluster of willow trees.
The remains of one or more buildings are still visible on the surface to the
north of the springs but we do not know their date. Many thousands of
stones litter the site over a wide area, and in at least two places on the slopes
to the northeast of the springs are stretches of fortification walls formed of
large, well-cut boulders, running east-west. (These walls were the “road” of
the landlord.) Further up the steep slope holding the walls, we found a level
area at the top and evidence of more walls just protruding from the surface.
From this highpoint we could see the whole Ushnu valley and a great part
of the Solduz valley to the east, as far as the modern town (and ancient Iron
Age tepe) of Nagadeh.
102 chapter four
On the eastern slope of the site, not visible from the Ushnu side, there
is a large open rectangular rock chamber, apparently natural in origin but
showing signs of having been worked by man; the open area is now used as
a shelter by shepherds. In his recent article “Urartäische Plätze in Iranisch-
Azerbaidjan” Wolfram Kleiss reports that at the Urartian site of Kale Waziri,
at the northwest corner of Lake Rezaiyeh, there is a “Felshöle,” an open rock
chamber, that was natural in origin but showed signs of human working.
Surely these two chambers had similar functions, which may be known to
us after Qalatgah’s excavation and more research.
Of the many sherds we picked up over an extensive area of Qalatgah,
all appear to be Iron III types (post-9th century bc), painted wares with
plain or hatched triangles decorating the inner rim of cups and bowls and
monochrome plain buff and red-burnished wares, well known from nearby
Hasanlu in Period III. Hollow-based bowls, simple, outward curving plain
bowls, carinated rim bowls, and cups, are among the shapes represented by
qalatgah: an urartian site in northwestern iran 103
Painted triangle ware sherds from (a) Qalatgah, (b) Ziwiye, (c) Hasanlu.
the Qalatgah sherds. Similar inner-rim painted triangle decoration has been
found at such sites as Hasanlu and at Ziwiye farther to the southeast, and in
Urartu at Altintepe, in the final upper level, and in the Van area at Van itself
and at nearby Tilke Tepe. The red-burnished wares may also be related to
similar types of ware found in several Urartian sites. Moreover, in addition
to the painted-triangle ware and red-burnished ware sherds, at home in
northwest Iran and Urartu, we also found some very fine, highly polished
redware sherds known to archaeologists as “Toprakkale Ware,” so-named
after the site in Urartu where they were first found, but since recognized at
other Urartian sites, and also recently found by C. Burney at Haftavan Tepe,
near Shapur in Iran.
A second trip on August 9th to the site with the whole staff of the Hasanlu
Project proved to be more rewarding. Not only did we find more diagnostic
sherds of the types just discussed, but I found a white stone stamp-cylinder
seal of characteristic Urartian type and motif. More important than this was
the find made by Christopher Hamlin: an Urartian inscription on a broken
104 chapter four
stone block. The stone was found, together with other plain stones, built
into a modern dam holding back a pool of water, fed by the two springs of
Qalatgah, and used for local irrigation.
The seal is a concave cylinder, free of decoration, with a stylized horned
animal running to the right carved at the base; a loop for suspension is at
the top. Seals of this type have been called stamp-cylinders by R.D. Barnett,
and they have been excavated at Karmir Blur and Toprakkale in Urartu, and
at Igydr and Kelankran in the southern Caucasus, made there, no doubt,
under Urartian influence. I believe the example from Qalatgah is the first
from northwest Iran.
Although they have been handicapped both by the weathering of the
stone and by its fragmentary state, Christopher Hamlin and Maurits van
Loon have been studying the inscription. Van Loon has sent me a prelim-
inary translation which I am quoting here; he will shortly publish a more
extensive commentary and discussion in the Journal of Near Eastern Stud-
ies:
1) When, relying upon the god Haldi and upon the god T[eisheba?, Ish-
puini,]
2) son of Sarduri, king of Urartu [and Meinua.]
3) son of Ishpuini, of the country Sapaya […]
4) the king? they too [k? …] both the god Hal[di and …]
5) for the god Haldi they change[d to hi]s? city Uishe? of the country […]
6) to the trees of … they carried …
The text, therefore, had been set up for public viewing by Ishpuini and
his son Menua, sometime during the co-regency of these kings, that is
sometime around 810 to 805bc. It should also be recalled here that during
the co-regency of Ishpuini and Menua they set up an inscription written in
both Assyrian and Urartian within the Kel-i Shin pass. That inscription is
therefore from the same period as this Qalatgah inscription found in 1968
and published above.
In the 1969 issue of the Archaologische Mitteilungen aus Iran, J. Friedrich
published an incomplete stone inscription found in the Ushnu valley in
1967. Friedrich says the fragment, now in the Tehran Museum, is one of
two stones recovered and that the other is in the hands of an antiquities
dealer. I suggest that there can be little doubt that this fragment is one of the
inscribed stones reported to me as being found at Qalatgah in 1967, the find-
spot of which we were seeking when we began exploring Qalatgah. Whether
Friedrich, or his source, or the Cheshmé Göl landlord who mentioned three
stones, is correct, is not presently known. Friedrich’s inscription was written
106 chapter four
for Menua, son of Ishpuini, and records the erection of a special building as
well as a city, not named in the fragment. Therefore, this inscription was set
up by Menua after the termination of the co-regency period with his father
Ishpuini and is thus to be dated between about 805 and 786 bc, contempo-
rary with a well-known inscription set up by Menua at Tash Tepe some fifty
miles to the east of Qalatgah. The Tash Tepe inscription mentions the found-
ing of a city in the area (not yet discovered), and is farther southeast than
any other inscription of Urartian origin presently known.
The Qalatgah inscriptions—the one found by the Hasanlu Project team,
and the one published by Friedrich—document the historical fact that
by the late ninth century bc and the earliest years of the eighth century
an Urartian city existed in the Ushnu valley, close to the Kel-i Shin pass.
The Tash Tepe inscription further establishes that Menua’s penetration into
northwest Iran extended east across the Solduz valley and the southern
shore of Lake Rezaiyeh. In short, this inscription, taken together with Kleiss’
recent discoveries, enables us to state that Iranian territory west and south of
the lake was under Urartian control at the beginning of the eighth century bc
The pottery found at Qalatgah cannot yet give us exact information about
the duration of the site because there is still much work to be done in sorting
out the sequences of Iron III pottery both in northwest Iran and in Urartu.
Kutlu Emre, a Turkish archaeologist, has recently noted that monochrome
burnished wares existed throughout the history of Urartian cities, that is
in the eighth and seventh centuries bc. The painted-triangle ware sherds
from Qalatgah are the same as those found at Hasanlu IIIB and at Ziwiye.
Also, they appear to be very close to, or the same as, the painted pottery
from Altintepe in western Urartu, there found only in the upper level, not
in the lower where only monochrome pottery was in use. The lower level
is dated by the excavator T. Özgüç to the eighth century bc (the founder of
the city may have been Arghisti II, about 714–685 bc), the upper level to the
second half of the seventh century bc, a date consistent with the terminal
dates suggested for Hasanlu IIIB and Ziwiye (new C-14 adjusted dates for
the terminal date of Hasanlu IIIB average to 679bc with a half-life of 5730
years). Painted-triangle ware is also reported at Van and nearby Tilke Tepe,
as already stated, but the exact relationship of this ware to the monochrome
ware found at these sites is not clear to me, as it is stated by some archae-
ologists (including the excavator K. Lake) that the painted ware was in use
earlier, during the eighth century, than the monochrome ware, dated there
to the seventh century. If true, the situation would be exactly opposite the
pottery sequence reported at Altintepe. Moreover, at Toprakkale, which is
close to Van and Tilke Tepe, monochrome but not painted pottery is found.
qalatgah: an urartian site in northwestern iran 107
Agrab Tepe is one of three mounds lying close to the modern village of
Dalma in northwestern Iran (Figure 1), southwest of Hasanlu in the Solduz
Valley near the low ridge that separates the Hasanlu plain from the mod-
ern town of Nagadeh. Of the three mounds, one called Dalma Tepe was
excavated in 1961 and yielded Neolithic remains.1 The third mound is still
unexcavated.
The distance between Hasanlu and Agrab is approximately two miles,
about a forty-five-minute walk from one site to the other; the sites are also
visible to each other, but the low ridge to the south cuts off the view of the
Nagadeh plain from Agrab. A modern road that connects Nagadeh to points
north passes by Agrab about a half-mile to its west; it cannot be established if
this road follows an ancient track. Before excavation the mound was about
52 meters in diameter and about 6 meters in height. A spring that caused
the surrounding land to be swampy and impassible in 1964 exists about 100
meters to the southwest. The site was built directly over a rock outcrop, the
only one visible in the area (Figure 2).
In 1964 the Hasanlu Project, a joint project of the University Museum
of the University of Pennsylvania and the Metropolitan Museum of Art,
decided to excavate a second mound at Dalma as part of its ongoing plan to
collect archaeological and cultural data of the Solduz Valley, south of Lake
Urmia (Rezaiyeh). Work at Hasanlu had to be suspended while we worked
at Agrab, and we were able to devote a total time of three full weeks to the
completion of our task. The mound had no local name. Since it was thought
* Excerpted from Metropolitan Museum of Art Journal. Copyright © 1973 by The Metro-
years …,” ILN (Nov. 3, 1962) pp. 707 ff.; R.H. Dyson, Jr. “Excavating the Mannaean Citadel of
Hasanlu …,” ILN (Sept. 30, 1961) pp. 534 f. References to Agrab Tepe: Dyson 1965, p. 212f., pl.
xliii, fig. 12; R.H. Dyson, Jr., “Hasanlu and the Solduz Valley,” Archeologia Viva 1 (1968) p. 85;
Young 1967, p. 30; W. Kleiss, “Zur Ausbreitung Urartus nach Osten,” Istanbul. Mitt. 19–20 (1967–
1970) pp. 127, 129; also his “Bericht über Zwei Erkundungsfahrten in Nordwest-Iran,” AMI 2
(1969) p. 26.
110 chapter five
Figure 1. View north from south ridge, past Dalma village. At center,
Agrab Tepe. To the right, in the plain, Dalma Tepe. The village
of Shetanabad, just west of Hasanlu, is at the upper right.
Figure 2. Agrab Tepe from the south, after completion of the excavations.
that the use of the name “Dalma No. 2” would cause confusion with respect
to the Neolithic mound, it was decided to refer to the site as Agrab Tepe.2
Agrab Tepe consists of a single large building built over a rock outcrop
(Figures 3, 4). The building is a fortified structure with thick outer defensive
2 Scores of scorpions were killed, hence the name “Scorpion Mound.” The staff consisted
of T. Cuyler Young, Jr., and the author as co-directors, Louis D. Levine and Ted Rathbun as
archaeologists, and Ed Keall, architect. The director of the Hasanlu Project, whose advice in
the production of this report I here acknowledge with thanks, was R.H. Dyson, Jr. I also take
pleasure in expressing my thanks to T. Cuyler Young, Jr., and Louis D. Levine for commenting
critically on many items discussed in this report. For a good map of northwestern Iran, Kleiss,
“Bericht tiber Zwei Erkundungsfahrten,” fig. 1.
excavations at agrab tepe, iran 111
walls whose interior surface also functioned as walls of the rooms. The
defensive wall consists of a foundation of large, roughly cut stone blocks
with small stones used for chinking (Figures 5, 6, 9), and with a brick super-
structure. Eight buttresses or piers and a massive tower-entrance flanked
by two piers project from the walls. In plan the structure is irregular, look-
ing like a flattened oval at one end, with no obvious compass orienta-
tion. It measures about 31 meters east-west and about 28 meters north-
south.
The walls are 1.90 to 2.00 meters thick, except for the two units east of
the entry tower; the first unit varies from about 1.80 to 2.00 meters, the
second from about 1.50 to 1.65 meters. The eight piers are irregularly spaced,
varying from 4.5 to 5.7 meters distance from each other. Their width also
varies from 3.00 to 3.70 meters; and their projection from the wall varies
from 1.10 to 1.30 meters. If the structure were not so well made one could
assume some haste in the building activities reflected by these irregular
measurements.
112
chapter five
No obvious means of entry into the shaft exists, and presumably a ladder
was employed. Nor is there any evidence available to suggest how one
entered the tower from the plain below. Presumably a door existed in the
now destroyed brick superstructure, and one has to assume that a ramp, or
ladder, probably portable, allowed access from the plain below. No other
primary function for the tower other than that of an entrance unit, or
gateway, comes readily to mind.
Directly in the center of the defensive wall facing the tower, the north
wall of the shaft, was a doorway 1.25 meters wide, with a large stone used as
a sill (Figures 5, 6). The door was eventually filled in with stones and a new
sill was built at what seemed to be level 2 (counting from the bottom up)
of room A2/B6. Whether this door served only as a passage to and from the
tower, or served also in some manner as an entry to the shaft, is not clear.
In any event, there is no other indication of a passage from the tower to the
main structure.
The structure as preserved consists of thirteen rooms or areas, most of the
walls of which are constructed of brick, sometimes set on a single course, 10
to 20cm. in height, on stone foundation, other times set right on the floor
surface. In rebuilding, sometimes a stone foundation layer was placed over
an earlier wall stub, other times the new wall was placed directly on the stub.
Brick size is uniform throughout the structure, 10/12× 49/50× 49/50 em. The
walls were coated with a thick layer of mud plaster.
A rectangular room, A2/B6, 5.30 × 3.60 meters, led to most of the other
rooms and to the stairway area to the west. To the right is a small room,
B10, 3.30 ×3.00 meters. In Period 2 it had a drain constructed of a small
sunken pithos surrounded by flat stones; this was connected to a draining
system originating in B4 (Figure 7). During Period 2 the doorway of B10 was
narrowed by the addition of two stubs of brick.
Abutting B10 is a small triangular room, B8, 3.10× 2.75 meters. In Period
1 this room had a door in its northeast corner (not shown on the plan)
that entered into the area called B5. During the rebuilding of Period 2 this
doorway was blocked. B8 was either abandoned in Period 2 or was entered
from above.
West of room A2/B6 is a rubble-surfaced area, A3, enclosing a rectangular
brick pier that must be interpreted as a support for a stairway. A3 flanks the
stairway support on three sides and continues up to the fortification wall
at the west; the area south of the stairway support was unsurfaced and had
a round terracotta hearth in Period 1. Except for those rooms entered from
above by a ladder, one presumably had to use this stairway to communicate
between stories, and perhaps to reach the battlements. The stairway was
116 chapter five
Figure 7. Draining system in B10. The wall has been cut away.
entered directly from A2/B6, which in turn served as a passage to the other
rooms. One is here reminded of the entrance system used at Hasanlu IV,
where stairways were placed in a room to one side of the anteroom.
Directly to the north of room A2/B6 is room B3, 4.20 × 3.60 meters. It is
connected by doors to rooms D1 and D2. A curious stone-edged semicircular
stepdown, about .25 deep and 1.25 meters in diameter, exists in the room’s
southeast corner, taking up about half the room’s space (Figure 8). Nothing
was found to suggest what its function might have been.
Room D1 is triangular in shape and abuts onto the defensive wall (Fig-
ures 8, 9); its two walls are each about 3.50 meters in length. This area no
doubt functioned as a kitchen and storage room, since many animal bones
and occupational trash were found in the fill and on the floor; also, five pithoi
excavations at agrab tepe, iran 117
were excavated in situ resting on the floor (Figures 3, 8, 9), four of these
against the defensive wall, and sherds of others were found in the fill.
Room D2, 2.50× 1.85 meters, also abutted onto the defensive wall (Fig-
ures 10, 11). It apparently functioned as a storeroom, for a large pile of stones,
presumably slingstones, was found resting in the northwest area against the
defensive wall. Remains of a late wall, apparently Period 2 or 3, ran north-
south; the slingstones were found partly under it, resting on the primary
floor. The south wall of D2, seen on the plan as a double wall, actually con-
sists of several wall periods.
Room C1, 4 ×8.5 meters (Figures 10, 11), abutted against the defensive wall.
In Period 1 a long northsouth wall (no door is visible in the low wall stub,
but it could have eroded away) divided the area from D2 and continued
118 chapter five
further south. In Period 2 a wall running along the line of the earlier wall, but
slightly displaced to the east, was built; only a stump, which projects into C1,
now remains (the stone foundation is visible in Figures 10, 11); presumably
it continued to the defensive wall. In the fill and on the floor of C1 were
found charred grain, stone pounders, several smashed pithoi, and sherds,
suggesting a work and storage room.
The rooms to the south of C1 are the most unusual of the structure. A large
rectangular room, 6.00× 9.00 meters, was subdivided into four rectangular
rooms of unequal size by a cross-wall of brick resting on a stone foundation
(Figures 3, 4, 12). The cross-wall was built against a well-made outer wall,
70cm. thick, constructed of small stones, which in turn was built against the
brick walls of the neighboring rooms, and also against the defensive wall.
excavations at agrab tepe, iran 119
This stone wall was constructed as one unit before the subdivision of the
area. In addition, the floors of the whole area were paved with flat stones
one layer thick. A layer of earth about 15 to 20cm. thick separates the paved
floor from the foundations of the cross-wall (Figure 12); it is therefore clear
that the paved room with its stone-lined walls existed as one large area
for a time before it was subdivided. The floors of the four rooms now all
slope slightly toward the center of the area. It is not clear if this represents a
sagging (presuming the flooring was not laid directly on the bedrock) or an
original plan. A stone-lined and capped draining system, 3.30 meters long,
25cm. high, and 15cm. wide, was constructed through both the southwest
corner of the stone wall in B4 and the neighboring brick wall. It emptied
into the drain in the northwest corner of B10 (Figure 7); no drains existed
120 chapter five
in the other stone-paved rooms. Presumably the drain was built before the
subdivision occurred.
The particular time in the history of the building when the area was
given a stone lining, was paved, and was subsequently subdivided seems
fairly clear. The stone lining wall was built against the brick wall of Period 2,
which itself rests on the stub of the earlier Period 1 wall, and which blocked
the doorway entering into B8. Therefore, the lining and paving belong to
the second construction period and the subdivision to a later phase of this
period. No doorways exist to connect the four rooms to each other or to the
neighboring rooms, nor is there any indication that the abutting brick walls
once had doors. Thus here, as with rooms C1 and B8, entry was doubtless
from above. This feature at Agrab plus the stairway makes it certain that
another story existed over the level preserved to us.
excavations at agrab tepe, iran 121
The nature of this particular area within the Agrab structure remains
a mystery. Surely the elaborate walling, paving, partitioning, and draining
system reflect a function not shared by the other rooms. Was it originally
built as a bathing area? Could it have been built as a rodent-proof, moisture-
free storage room? Pithoi fragments were found in Room 1, and animal bones
were found in the fill of areas 1, 2, and 4, but these could have fallen from a
higher story. The problem remains unresolved.
Several test trenches were dug outside the defensive wall, in the north,
south, east, and west. Nothing but bedrock was encountered in three of
these trenches, but in the east trench a small enclosed space was excavated.
It was formed by two short brick walls projecting from the front of the two
piers, and creating an open doorway 1.75 meters wide. The walls are two
122 chapter five
bricks wide, or from 1.00 to 1.10 meters thick. The sill was also of brick and
exited to the bare rock below. There was no visible means of communication
between this space and the main unit, nor was there any visible means of
closing the door in the space. Three arrowheads, a bone spatula, a bead,
and a grinding ball were found here in the fill. Perhaps the unit served as
a temporary postern-gate area.
The building at Agrab Tepe was destroyed at least twice by fire and rebuilt
using the same basic plan, thus creating three periods. During most of
the course of excavation this fact was not recognized for several reasons.
The rebuilt walls were placed directly over the earlier walls (Figures 8–10).
The heat generated by the fires was quite intense and vitrified the plaster,
which made it difficult to clear wall faces. This situation prevented us from
seeing wall stratification and offsets of later walls over earlier ones, which
occurred in a few cases. Moreover, in only a room or two were there any
floors preserved from Period 2, as these floors were not hard, nor were
they regular (Figure 13). Thus we assumed that we were digging a site with
one occupational level. When we were able both to read the sections and
examine the walls closely, we concluded that Agrab had several levels. By
this time the digging was basically finished and it became difficult in some
excavations at agrab tepe, iran 123
cases to divide the pottery into the three levels. However, Period 3 was close
to the surface and had no recognizable floors or remains as such, only sherds
considered to be surface finds. An Islamic level, much denuded, had been
cut into it and destroyed it. Also, many of the finds were in the fill over the
primary floor level and presumably came from Period 1. Yet, one cannot be
sure in all cases. In room D1 several pithoi were found on the primary floor,
supplying us with important information.
The lack of good, firm floors in Period 2, plus the rebuilding on original
plans, suggests that only a short amount of time elapsed after the first
destruction before rebuilding, and that it too may have been rapidly rebuilt
and only briefly inhabited in Period 3.
Objects Excavated
Pottery
Bowls with rolled rims: Figure 14, 1 and 2, plus more sherds; coarse ware,
buff, lightly burnished. Related bowls come from Hasanlu IIIB and A (Young
1965, fig. 6, 5); Bastam (Kroll 1970, fig. 6); Godin II (Young 1969, fig. 44, 13,
15); Baba Jan A (Goff Meade 1968, fig. 10, 4); see also Zendan (Boehmer 1961,
pl. 60,5; pl. 55, 14).
Carinated bowls with square rims: Figure 14, 3, two examples; coarse
ware, buff, unburnished. Related bowls are from Hasanlu IIIA;3 Godin II
(Young 1969, fig. 43, 14); Pasargadae, unpublished; Geoy Tepe A (Burton-
Brown 1951, fig. 36, 357).
Pots with oblique shoulder spouts: Figure 14, 4, with handle, 5; three
more examples, one with handle; coarse ware, buff, unburnished. Related
pots are from Hasanlu IIIB (Young 1965, fig. 2, 5); Geoy Tepe A (Burton-
Brown 1951, fig. 35, 126; fig. 40, 1644; fig. 41, 113); Godin II (Young 1969, fig. 42,
17); Norşuntepe (Hauptmann 1970, fig. 16, 12; fig. 17, 3; fig. 18, 1, earliest Iron
Age level); cf. Altintepe, later level (Emre 1969, pl. v, 1); Achaemenid Village
(Ghirshman 1954, pl. XXIX, G.S. 1206b, G.S. 959).
Pots with one handle: Figure 14, 6 and 7; coarse ware, buff, unburnished.
Similar vessels come from Hasanlu IIIB; Geoy Tepe A (Burton-Brown 1951,
fig. 35, 106).
3 When Hasanlu IlIB and A appear without a reference it means that I found the vessel
Lip spout: Figure 14, 8, one example; buff, smoothed surface, medium grit
interior. Related spouts are from Hasanlu IIIB and A (Dyson 1965, pp. 205, 212,
note 36); Geoy Tepe A (Burton-Brown 1951, pl. 39, 219); Zendan (Boehmer
1961, pl. 45, a1; pl. 51, 21); Ziwiye, unpublished (Young 1965, fig. 10 chart);
Achaemenid Village I (Ghirshman 1954, pl. XXIX, G.S. 2242); Luristan (Goff
Meade 1968, p. 123, fig. 11, 14, Iron III; Vanden Berghe 1967, pl. 59, no. 2);
Masjid-i-Suleiman (Ghirshman 1970, p. 184, pl. IVb).
Plain bowls with in curving sides, pinched rims: Figure 14, 9; one buff,
red-slipped, burnished with fine paste; two others, without hollow base,
buff, smoothed surface. Related bowl shapes are known from Hasanlu IIIB
and A (Young 1965, fig. 1, 1); Ziwiye (ibid., fig. 3, 1); Bastam (Kroll 1970, fig. 1, 1;
pl. 2, 1, 2); Godin II (Young 1969, fig. 43, 2); Achaemenid Village II (Ghrishman
1954, pl. XXXVII, G.S. 1219f); Norşuntepe (Hauptmann 1970, fig. 23, 1, middle
Iron Age); Armavir Blue (Barnett 1963, fig. 19, bottom). Hollow-based bowls,
not to be confused with omphalos bowls, occur at Agrab in the earliest
period and in the fill. They also occur at Hasanlu IIIB (Young 1965, fig. 1,
2); Sé Girdan (Muscarella 1971a, fig. 29); Qalatgah surface (Muscarella 1971b,
p. 46); Godin II (Young 1969, pl. 44, 6, 7; also earlier in Godin III, fig. 32, 7, 8);
Altintepe, earlier level (Emre 1969, p. 295, fig. 12); cf. also Karageorghis 1962,
p. 114, and pls, 144, 145, 148, 149, 156, 166, 171, dated to the early sixth century
bc. The hollow base seems to be a variant of the omphalos, which occurs in
Iron III also. Bowls from Igdyr (Barnett 1963, fig. 15) may be hollow based,
rather than omphaloi.
Bowls with wide flaring collar and pinched rim: Figure 14, 11; one
example is coarse, buff, unburnished, another is buff, with a smoothed
surface, a third is burnished buff orange ware, with a fine paste. Similar
bowls occur at Hasanlu IIIA; Ziwiye (Young 1965, fig. 3, 11; cf. 6, 9, 12, with
omphalos); Zendan II (Boehmer 1961, pl. 50, 7, 8, with omphalos; Boehmer
1967, pp. 577, 580, fig. 8, A–C; also fig. 8 for Ziwiye and for 7th-century Nimrud
examples); Yanik Tepe (Burney 1962, pl. XLV, 36, but painted); Qalatgah
surface (Muscarella 1971b, p. 47); Altintepe, later level (Emre 1969, p. 299,
pl. v, 1); Norşuntepe (Hauptmann 1970, fig. 21, 4, middle Iron Age); Persepolis
(Schmidt 1957, pl. 72, 1); Pasargadae, unpublished; Gordion, Persian level
(R. Young 1962, pl. 41, fig. 1a, b); Samos, wood (Kopcke 1967, p. 119, fig. 6, 7th–
6th century bc.)
surfaces: coarse ware buff; burnished buff; coarse buff, slightly burnished;
buff, smoothed surface; two (like Figure 14, 12 and Figure 15, 1) are red-slipped
buff; one (like Figure 15, 2) is gray burnished. Similar bowls are found at
Hasanlu III B and A (Young 1965, fig. 1, 2, 4; fig. 2, 6; cf. also Hasanlu IV, fig. 6,
2, 4); Bastam (Kroll 1970, fig. 1, 2, 8); Ziwiye (Young 1965, fig. 3, 14); Zendan I, II
(Boehmer 1961, pl. 56, 23–25); Godin II (Young 1969, fig. 44, 17, 18); Haftavan
(Burney 1970, p. 170, 6); Baba Jan I (Goff 1970, fig. 8, 7); Armavir Blur (Barnett
1963, fig. 19, top); Patnos (Ankara Museum); Kef Kalesi (Öǧün 1967, fig. 16);
Van (von der Osten 1952, pl. VIII, 3, 4).
Bowls with everted rims: Figure 15, 3; red-slipped buff ware. Similar bowls
occur in Hasanlu IIIA; Ziwiye, unpublished; Zendan (Boehmer 1961, pl. 50,
9); Godin II (Young 1969, fig. 43, 3); Yanik Tepe (Burney, 1962, pl. XLV, 33,
painted); see also Igdyr (Barnett 1963, fig. 16, bottom) red polished; and Van
(von der Osten 1952, p. 325, pl. VI, I).
Bowls with inward-curving rolled rims: Figure 15, 4, coarse, buff ware;
5, red-slipped, burnished; 6, coarse, buff; 7, red-slipped, burnished; others,
like 4 and 5, red-slipped, burnished. Such bowls are known from Hasanlu
IIIB and A (Dyson 1965, p. 204, fig. 13; Young 1965, fig. 1, 3, 5; common in IIIB);
Ziwiye (ibid., fig. 3, 19); Bastam (Kroll 1970, fig. 1, 10; fig. 3, 1, 7; fig. 4, 3); Baba
Jan I (Goff Meade 1968, fig. 8, 9); Norşuntepe (Hauptmann 1970, fig. 23, 3, 4);
Altintepe, earlier period (Emre 1969, fig. 8, 10; pl. II, 1, 2); Van (von der Osten
1952, pl. VIII, 1).
Bowls with rolled and grooved rims: Figure 15, 8, burnished gray ware,
fine paste; 9, buff, smoothed surface. Similar shapes occur at Hasanlu IIIA;
Ziwiye (Young 1965, fig. 3, 4, 5); Godin II (Young 1969, fig. 43, 11).
Carinated bowls with grooved collars: Figure 15, 10, buff, smoothed
surface; 11, burnished gray ware. Similar forms occur at Hasanlu IIIB; Geoy
Tepe A (Burton-Brown 1951, fig. 36, 643); Godin II (Young 1969, pl. 43, 12);
cf. Norşuntepe (Hauptmann 1970, fig. 18, 4; fig. 21, 3); Achaemenid Village
(Ghirshman 1954, pl. XXXVII, G.S. 1219g).
Carinated bowls with slightly rolled rims: Figure 15, 12 (64-4), 13 (64-3)
from fill over A3 primary floor; buff, smoothed surface, with slip on interior
and exterior; 14, with square hollow base, from primary floor of B6, red-
slipped, burnished. Similar bowls are found at Hasanlu IIIB, also, in one case
at least, with a squared hollow base; Ziwiye (Young 1965, fig. 3, 3); Zendan
128 chapter five
(Boehmer 1961, pl. 56, 16, 17); Bastam (Kroll 1970, fig. 4, 2); Godin II (Young
1969, fig. 43, 4).
Trefoil pitchers: Figure 15, 15; buff, burnished, medium grit. Similar pitch-
ers, with or without shoulder grooves, occur at Hasanlu IIIB and A (Dyson
1965, fig. 13, lower right; Young 1965, fig. 2, 7); Bastam (Kroll 1970, fig. 1, 7; fig. 2,
6); Baba Jan I (Goff 1970, fig. 8, 1); Achaemenid Village II (Ghirshman 1954,
pl. XXXVIII, G.S. 1221d); Altintepe (Emre 1969, fig. 17, 18); Karmir Blur (Piotro-
vskii 1959, fig. 51; Piotrovskii 1969, pl. 49, 50); Igdyr (Barnett 1963, fig. 20, 2);
also at other Urartian sites; Norsuntepe (Hauptmann 1970, fig. 22, 10, middle
Iron Age).
Jar with rolled tab handles: Figure 15, 16; buff ware (Dyson 1965, 213,
fig. 13, listed in the IIIA section). Similar handles on pots occur at Hasanlu IV
(Young 1965, fig. 2, 3); Geoy Tepe A (Burton-Brown 1951, fig. 36, 102; fig. 37,
120); see also Tresors de l’Ancien Iran (Geneva, 1966) fig. 64, catalogue no. 672.
One-handled pitchers: Figure 16, 1 (64–24), 2 (64–23), both from the fill
over the floor of B10, and of Period 2; coarse ware, buff; 2 has a smoothed
broken rim suggesting it was used after breaking. Similar pitchers occur at
Hasanlu IIIB, Ziwiye, Giyan I, Sialk B (Dyson 1965, fig. 7); the neck of 1 is
missing but the general shape looks like pitchers from Bastam (Kroll 1970,
p. 73, fig. 1, 7), Altintepe (Emre 1969, fig. 17), Karmir Blur (Piotrovskii 1959,
fig. 51), and at other Urartian sites; see also Norşuntepe (Hauptmann 1970,
fig. 23, 7). At the Urartian sites this shape usually has a fine red polished
surface.
Jars with two handles: Figure 16, 3; buff, burnished, fine paste. Compara-
ble jars may be seen at Godin (Young 1969, fig. 42, 1); Nush-i-Jan (Stronach
1969, fig. 6, 9; fig. 7, 2). None of these is an exact parallel.
Small pots with plain or slightly rolled rim: Figure 16, 5 (64–35),
from primary floor of D1, and belonging to Period 1; buff, burnished, slightly
130 chapter five
hollow base; 6 (64–16), fill over primary floor of B6, also Period 1; buff,
medium paste. Similar small pots occur at Bastam (Kroll 1970, fig. 5, 5); Geoy
Tepe A (Burton-Brown 1951, fig. 38, 20); Zendan (Boehmer 1961, pl. 50, 4);
Ziwiye (Young 1965, fig. 3, 7); War Kabud (Vanden Berghe 1968, fig. 29, 3, 5).
Tripod jar with hole at shoulder: Figure 16, 7 (64–26), fill over floor
of B10, and belonging to Period 2; buff, smoothed surface. The neck has no
opening. Tripod vessels occur in Iran in the Bronze and Iron Ages: L. Vanden
Berghe, Archéologie de l’Iran ancien (Leiden, 1959) pls. 111b, c, 115b, c, 145e,
146c, 165b, 167 (middle), 173a, b.
Large pot with rolled rim: Figure 18 (64-8), ht. 33.5 cm.; from primary
floor of A2, and belonging to Period 1; buff, burnished, medium grit paste.
Parallels in shape occur at Haftavan (Burney 1970, fig. 8, 3, red burnished);
Altintepe (Emre 1969, fig. 3); Igdyr (Barnett 1963, fig. 21, 4, 6,7).
Tray: one fragment of coarse ware, buff. Trays occur at Hasanlu III
(Young 1965, 75, fig. 12, apparently IIIA); Ziwiye (ibid.); Bastam (Kroll 1970,
excavations at agrab tepe, iran 131
fig. 6, 5); Zendan (Boehmer 1961, pl. 60, 5, 16, 17, 18); Sé Girdan (Muscarella
1971a, fig. 30, upper left, and second from bottom); Qalatgah surface, unpub-
lished. In the fill of B7 were found two base sherds of undetermined shape
each of which has holes made before firing; they are probably not trays. A
tray from Qalatgah, referred to above, has a hole in its center; and a vessel
from Baba Jan (Goff Meade 1968, fig. 10, 26) has a hole in its center.
High-necked jars: Figure 16, 8–12, sherds of about five others; all are buff,
smoothed surfaces, except 10 which is buff, burnished. Similar jars occur at
Hasanlu IIIB and A; Ziwiye (Young 1965, fig. 3, 10); Bastam (Kroll 1970, fig. 3, 4,
5); Zendan (Boehmer 1961, fig. 30, 2–4, pl. 54, 6); Godin II (Young 1969, fig. 42,
3, 4, 6, 9, 14, 16); Pasargadae, unpublished; Altintepe (Emre 1969, fig. 2, 3, 5,
6).
Bowls or basins without handles: Figure 19, 1, coarse ware, buff; 2, buff,
lightly burnished; fragments of two others that are coarse buff, and one that
is buff, smoothed surface. Similar vessels occur at Hasanlu IIIB and A.
132 chapter five
Small pithoi, storage jars: Figure 19, 3–10, and fragments of others; most
are coarse, buff, and unburnished; 7 is red-slipped, 8 and 9 are buff with
smoothed surfaces, and they have grooved rims. Parallels are found at Ha-
sanlu IIIB; Ziwiye (Young 1965, fig. 4, 3); Geoy Tepe A (Burton-Brown 1951,
fig. 40, 1648); Zendan (Boehmer 1961, fig. 31, 2; pl. 54, 7); Godin II (Young 1969,
fig. 43, 1, 6, 15, 19); Bastam (Kroll 1970, fig. 5, 1, 4). Grooved rims on storage
vessels occur at these sites also.
Tab handle: Figure 19, 11; from fill in B1, belonging to Period 2; dark brown,
smoothed surface, medium grit paste; only one example found. Parallels
for these handles occur at Hasanlu IV (Young 1965, fig. 6); Hasanlu IIIB,
Ziwiye (ibid., fig. 3, 17; fig. 4, 9; fig. 10); Zendan I (Boehmer 1961, pl. 57, 8–11;
Boehmer 1965, fig. 75b, 74a); Godin II (Levine 1970, p. 43, drawing); cf. Geoy
Tepe A (Burton-Brown 1951, fig. 35, 284) and Nush-i-Jan (Stronach 1969, fig. 7,
2).
Horizontal handle: Buff, coarse ware, from fill in B7; only one example
found. This type of handle occurs at Hasanlu IIIA, Ziwiye, Khorvin, Giyan I
(Dyson 1965, 206, fig. 7); Godin II (Young 1965, fig. 34, 16); Baba Jan II, III
(Goff Meade 1968, fig. 10, 12–14, 18; Goff 1970, fig. 7, 4–6; fig. 8, 11–12, Period 1);
Nush-i-Jan (Stronach 1969, fig. 6, 3–6).
Pithoi: Many pithoi fragments were found in the fill and on the primary
floor of D1. Figure 20, 1a, b, Figure 21 (64-1), from high in the fill of A2—
and apparently either Period 2 or 3—a fragment of a brown buff, coarse
pithos, apparently handmade; stamped into the clay are two round sealings,
each a skidding horned creature with erect tail; features cannot be made
out but both creatures seem to be of the same species. To their right is a
stamped curved-sided square with a round depression in the center; below
the impressions are hand-impressed inverted V marks. I cannot find exact
parallels for these sealings, but one should compare sealings from Urartu
(van Loon 1966, p. 156, F11; p. 159, E15, EI6; p. 161, G2; Barnett 1959, fig. 6, 15;
see also Mallowan 1966, p. 198f., fig. 134, 7, 7th century bc). Three practically
complete buff pithoi with scraped surfaces were found on the primary
floor of D1: Figure 20, 2–4, Period 1; 2 and 3 have raised triangles on the
shoulder, 4 has a rope or corded design, Parallels for the triangle decoration
occur at Karmir Blur (Piotrovskii 1969, pl. 61, sunken triangles); Kayalidere
(Burney 1966, fig. 15, sunken triangles); Patnos (Ankara Museum); Kef Kalesi
(Biliç, Öǧun 1965, pl. VIII, sunken triangles, corded also). Vessels with corded
decoration on the shoulders also occur at Urartian sites, viz., Kef Kalesi, op.
134 chapter five
cit.; Barnett 1959, pl. IV. Other pithoi have coarse, buff, unburnished surfaces:
Figure 20, 6, 8; or slightly burnished buff surfaces, 7; or plain buff, smoothed
surfaces, Figure 22, 1, 3, 4; or a red-slipped buff surface, 2 (two of these
were found). The flat-ledged type, Figure 22, 2, has parallels at Hasanlu
IIIB and A; Kayalidire (Burney 1966, fig. 15, fig. 16); the other pithoi have
parallels at Hasanlu IIIB and at Urartian sites, viz., Altintepe (Emre 1969,
pl. VI).
excavations at agrab tepe, iran 135
Figure 21. Pithos sherd with seal impressions, AG 64-1. Ht. 12.5cm.
Pithos: Figure 23, 2 (64–42), Figure 24 (Muscarella 1971b, fig. on p. 44; Dyson
1965, fig. 13, lower left in IIIA section); found on top of and apparently within
a pith os on the floor of D1—the southwest pithos shown on the plan;
deep red, well burnished, red-slipped; inside plain and uneven; triangles
and bands on surface scraped and lighter than rest of vessel; traces of
white paint on the bands and triangles; ht. 56 cm., rim diam. 54 cm., base
diam. 19cm., carination diam. 47, 5cm. A very similar, but larger vessel
was published from Patnos (van Loon 1966, fig. 3), where the triangles
are also filled with white paint; others, unpublished, are also from Patnos.
See also a similar vessel and decoration, but cruder, from Armavir Blur
(Piotrovskii 1969, pl. 69); compare for general shape a vessel from Karmir
Blur (Piotrovskii 1969, pl. 55).
Pot stand: Figure 23, 1 (64–27); from the fill over the floor of B5 and belong-
ing to Period 2; burnished red-slipped ware. Rolled upper and lower rims;
four oval cutouts around the middle.
Askos: Figure 20, 5 (64–34), Figure 26; from the fill over the floor of C1; red-
slipped buff ware. The vessel is egg-shaped with upright spout and handle;
handle is grooved with a clay rivet at the base, and two rivet-like impressions
over this. Askoi occur at Karmir Blur (van Loon 1966, fig. 11; Piotrovskii
1969, pl. 58, but painted). See Vanden Berghe 1968, p. 117, fig. 144, for an
example from Luristan; examples in Copenhagen and the Louvre are said
to have come from Luristan (Contenau 1935, pl. XVIII, top; M.L. Buhl, Acta
Archaeologica 21 [1950] p. 197f., fig. 46, 47). Other askoi are reported from
Patnos (Mellink 1965, p. 142). More examples are known further west: Hama
(Riis 1948, fig. 84); Cyprus (Karageorghis 1969, fig. 31, 7th century sc). Related
shapes occur at Persepolis (Schmidt 1957, pl. 71, 9, 72, 13); Achaemenid
Village (Ghirshman 1954, pl. xxxv, G.S. 1270; pl. XXXVIII, G.S. 1176). False
terracotta rivets occur at Hasanlu IIIB and A (Young 1965, fig. 2, 1–3); Geoy
Tepe A (Burton-Brown 1951, fig. 37, 121); Yanik Tepe (Burney 1962, pl. XLV,
fig. 30); Baba Jan I, II (Goff Meade 1968, fig. 6, 19); Zendan (Boehmer 1965,
fig. 75, c); Achaemenid Village (Ghirshman 1954, pl. XXXIX, G.S. 1249d). For
a later shape related to the askos, see Schmidt 1957, pl. 72, 13.
Nipple-base vessel: Figure 23, 3, Figure 25 (64–38); also 64–43; both exam-
ples from the fill of D1, burnished red-slipped buff ware. I can find no pub-
lished examples but am informed that similar vessels have been found at
Cavustepe.
excavations at agrab tepe, iran 137
Stem and base of a goblet: From fill of D1; red-slipped burnished; light
brown interior; fine clay core; clearly the finest red burnished vessel at
Agrab. The goblet is one of the most characteristic shapes in the Urartian
repertory. They are found at Bastam (Kroll 1970, p. 73 for bibliography, fig. 1,
4); Haftavan (apud Kroll 1970, p. 73); Kayalidere (Burney 1966, pl. XV, b);
Altintepe, earlier level (Emre 1969, fig. 19); Karmir Blur (Piotrovskii 1959,
fig. 50); Toprakkale (C. Lehmann-Haupt, Armenien Einst und Jetzt [Berlin,
1931] p. 567).
From the foregoing we see that the preponderance of the pottery at Agrab
Tepe consists of buff wares, with coarse, burnished, and smoothed surface
types in approximately equal proportions; buff unburnished wares seem to
follow next in quantity. Medium and fine paste interiors occur mixed among
these groups with no correlation to surface features other than that coarse
wares do not have fine paste. In lesser quantity, but still considerable and
very noticeable, are the red-slipped wares, usually slightly or well burnished.
Aside from the unique vessels—the askos, nipple vessels, the pot stand,
and the triangle-decorated pithoi—a few pithoi and small bowls are of this
fabric. Rare, but in evidence, are a few burnished gray bowls.
140 chapter five
The buff wares fired from yellowish through pinkish to orange. Often, gold
flakes (mica?) are visible on the surface. Some of the red-slipped bowls also
have these gold flakes on the surface, suggesting they were made locally,
from the same clay source as the buff wares.
This ceramic collection makes it quite clear that Agrab Tepe was an
Iron III site as defined by Dyson and Young, belonging to the “late buff ware
horizon.”4 Many of the sites of this period, as we have seen in referring to
pottery parallels above, have not only the characteristic buff wares, but also
the red-slipped wares; and also a small quantity of burnished gray wares viz.,
Ziwiye, Achaemenid Village, Giyan I (Young 1965, pp. 59, 66, 68), Baba Jan
(Goff Meade 1968, p. 116), Bastam (Kroll 1970, p. 70), Godin II (rare, personal
communication with T. Cuyler Young, Jr., who also informs me that there
are also a few red-slipped wares at Godin II). This configuration of pottery
is characteristic for Iron III sites and need not be elaborated on here. Future
research will have to define the cultural relationship and significance of the
occurrence of red-slipped wares at Iron III sites and at most Urartian sites.
Noticeably lacking at Agrab Tepe are the fine wares recorded at Hasan-
lu IIIB, Ziwiye, and Qalatgah (Young 1965, pp. 55, 59 ff.; Muscarella 1971b,
p. 46f.), Yanik Tepe (Burney 1962, pl. XLV) and Pasargadae, unpublished;
also the incised wares found at Ziwiye and Zendan (Boehmer 1965, fig. 75;
Boehmer 1967, fig. 9). Painted pottery is lacking at Bastam, Godin II (except
for three sherds, personal communication from T. Cuyler Young, Jr.), Geoy
Tepe A, and Zendan. What, if any at all, are the chronological implications
of this lack of local painted pottery at Agrab Tepe cannot yet be established.
Actually it may have no chronological significance; rather, it could simply
mean that it was a luxury product, and not needed at Agrab (but what of its
lack at other Iron III sites?).
Metal
Bronze helmet earflap: Figure 27, 1 (64-9), Figure 28; fill over floor of B8.
The border has two grooves and an exterior flat ledge with holes. Traces of
thread found in situ on obverse, connecting holes; traces of leather found
on both sides, on the obverse under the thread. Earflaps of the very same
shape have been found at Hasanlu IV, but with decoration (unpublished).
T.A. Madhloom, in The Chronology of Assyrian Art (London, 1970) p. 38, says
separately made earflaps began in the 7th century, but this is contradicted
by the Hasanlu evidence.
4 Young 1965, pp. 53 ff, 72 ff.; Dyson 1965, pp. 203 ff. In addition see the pottery analysis of
Bronze trilobate arrow: Figure 27, 2 (64-6); upper fill of A2; remnants of
wood in the shaft. The distribution of these arrows is quite extensive in the
Near East and in Europe, see T. Sulimirski, “Scythian Antiquities in Western
Asia,” Artibus Asiae 17 (1954) p. 295f.; R.V. Nicholls, “Old Smyrna: The Iron Age
Fortifications …,” BSA 53–54 (1958–1959) p. 12; P.R.S. Moorey, Catalogue of
the Persian Bronzes in the Ashmolean Museum (Oxford, 1971) p. 87; Boehmer
1965, p. 773f., note 98 (n.b., Boehmer says they are found in the Phrygian level
at Gordion but surely this is an error: they are found only in post-Phrygian
levels). There is as yet no certain evidence that these arrows predate the 7th
century in Iran (Dyson 1965, p. 207).
Bronze arrow with two wings: Figure 27, 3 (64–57). Similar arrows, with-
out the side hole, occur at Hasanlu IV; Achaemenid Village II (Ghirshman
1954, pl. XLIV, G.S. 2104); Karmir Blur (Piotrovskii 1959, fig. 81, right; fig. 84,
left); Gordion (R. Young 1953, p. 164f., 166, fig. 10, 6th century bc); and Smyrna
(Nicholls, “Old Smyrna,” 130f., pl. 6d, right, about 600bc).
excavations at agrab tepe, iran 143
Bronze arrow: Figure 27, 4 (64–40); fill in D1; leaf shaped, flat on one side,
with median strip on the other; solid ridged tang; found bent from use.
Iron arrows: Figure 27, 8, 9, 11, 12, 13 (64–10), from the fill of B8, and probably
from Period 2 (a total of seventeen arrows were found here); 10 (64–37), from
the fill of D1; 5, 6, 7 (64–51), East Trench; a) flat blade, solid tang, 5–7; b)
blade oval in section, solid tang, 8–10; c) blade oval in section, with collar
and solid tang, 11; d) blade oval in section, tang hollow but with iron core,
12, 13: traces of reed binding remain here. A total of twenty-one iron arrows
were found, many very corroded. Similar arrows occur at Hasanlu IV; Bastan
(Kleiss 1970, p. 54 f., pl. 34, 1, iron and bronze); Ziwiye, unpublished; Sialk B
(Ghirshman 1939, pls. LXXI, S892e; LXXV, 5923c, d; LXXVII, 5973a); Haftavan
(Burney 1972, pl. VIII, b); Kayalidere (Burney 1966, fig. 21, 8, 9; pl. XIII); Karmir
Blur (Barnett 1952, fig. 13); Altintepe (Özgüç 1966, pl. XXXIV, 1–6); Toprakkale
(Barnett 1954, fig. 15); Igdyr (Barnett 1963, fig. 37, 4–7, 11); Nimrud (Mallowan
1966, 11, fig. 332, a-c).
Iron lance head: Figure 27, 14 (64–31); from fill in C1; very corroded; traces
of wood in shaft.
Iron shafthole ax: Figure 27, 15 (64–21); fill over primary floor of B7; very
corroded; traces of wood in shaft.
Iron tool, ferrule?, plowshare?: Figure 27, 21 (64–39); from fill in D1;
hollow but for inner 9cm. of tip. A similar, but smaller, object was found in
Hasanlu IV; cf. also J. Deshayes, Les Outils de Bronze … II (Paris, 1960) pl. XVI,
16, 1235; Achaemenid Village (Ghirshman 1954, pl. XLIV, G.S. 2109).
Bronze wood clamps (three): Figure 27, 16 (64–14a, b, c); fill over primary
floor of B3; pieces of wood were found adhering to the inside of one example.
Bronze boss (two): Figure 27, 17 (64–20); fill over floor of B10. Cf. Zendan
(Boehmer 1965, p. 773f., fig. 77a); Sialk B (Ghirshman 1939, pl. LVI, S819);
Karrnir Blur (Piotrovskii 1969, fig. 79).
Bronze stud: Figure 27, 18 (64–41); fill in D1; ends bent out intentionally.
Iron knife blade: Figure 27, 19 (64–30); fill in C1; cracked; flat in section at
rounded end; tapers at blade end; a rivet pierces the metal at the grip end.
144 chapter five
Figure 29. Stone, bone, and terracotta objects from Agrab Tepe.
Bronze hook (?): Figure 27, 20 (64–48); from East Trench fill; thick, twisted
circular shaft, divided and flattened at both ends.
Bronze bracelet fragment: Figure 27, 22 (64–47); from fill in East Trench;
animal or snake head at ends.
Bronze bracelet fragment: Figure 27, 23 (64–32); fill in C1; may have
snake-head ends.
Bronze bracelet: Figure 27, 24 (64–15); from primary floor of B3; probably
a child’s bracelet; plain ends slightly overlap.
Bronze bracelet fragment: Figure 27, 25 (64-2); upper fill of A2; the ends
are flattened with square corners and hollow depression; arc decorated with
two rows of shallow lines.
Two terracotta beads: Figure 29, 4 (64–28); fill of C1; (64–18), fill in B8;
buff, perforated for stringing. A third bead found in fill over primary floor of
B7.
Stone quern: Figure 29, 6 (64–44); fill in D1; smooth on flat surfaces; break
may be unintentional.
Stone object: Figure 29, 7 (64–46); unstratified; gray stone, rough on upper
surface, polished on lower; a projection below is broken. A pivot stone?
Whetstone: Figure 29, 8 (64–13); fill over B area; very fine grained dark
brown stone; hole at one end.
It is obvious that the elaborate building at Agrab was built to serve as a for-
tified structure. The massive walls, tower-gate, and piers preclude another
interpretation. And the hoard of slingstones, the helmet earflap, and the
arrows and iron lance, reinforce this opinion; the pithoi would have served
as storage vessels for the presumed garrison’s food.
On the other hand, one might conclude that a fort at Agrab makes little
sense. It was built not on a high place on one of the ridges, but in an
exposed position on the plain, in which position it could not have served as a
watchtower. Moreover, the building is relatively small and presumably could
not have contained many soldiers comfortably, even with an upper story.
Agrab is also within sight of Hasanlu and may have had some relationship
with that city; why then should a fort have been built so close?
At Bastam, Kleiss excavated an isolated building situated about 700 me-
ters from the citadel. He suggested that this building, or castle, might have
been built by an enemy force besieging Bastam.5 Such a conclusion cannot
be presently proven, nor can we make a similar interpretation with respect
6 van Loon 1966, p. 38, states that Urartians built fortified outposts to guard water sup-
plies.
7 Kleiss 1970, pp. 40 ff. Note that Altintepe in Urartu was built on a low hill in the plain:
Özguç 1966, p. 37 f.
8 Kleiss 1970, figs. 34, 36; Kleiss, “Bericht uber Zwei Erkundungsfahrten,” figs. 18, 18a; van
What of the Medes and Manneans? Here, too, insufficient evidence pre-
vents a conclusive answer. Given the geographical problems, and recogniz-
ing that different peoples and armies must have traversed the region in the
seventh century, any suggestion becomes a mere guess. I therefore see no
alternative to leaving the question of the ethnic identity of the builders and
occupiers of Agrab open for future research. At the same time one is tempted
to suggest that the occupiers of the site could have been either an Urartian
garrison, using local help to build the fort (but who planned it?), or a local
garrison of Manneans. But, to repeat, we do not know.
Whoever the people were who lived or worked at Agrab Tepe, they used
the same basic types of pottery for their kitchen needs as that used by
the inhabitants of contemporary cities and towns in western Iran. They
also used a pithos type common to Hasanlu IIIA and to several Urartian
cities. Moreover, some of the inhabitants owned a few exotic pottery vessels,
apparently, all imported from Urartu. Therefore we may presume that they
were in familiar contact both with Iranian and Urartian cities.
The inhabitants at Agrab Tepe stored their grain in large and small pithoi,
at least one of which was impressed with seals, and they ground their grain
on the premises. Aside from one possible agricultural tool, there are no other
indications from the material remains to suggest that they were farmers.
However, we do not know what was not preserved for us to find. Their water
was obtained from the neighboring spring, and, as stated, they were within
walking and viewing distance of a fortified city, Hasanlu. They used bows
and arrows, slings, and lances as weapons, and they had body armor.
No luxury items aside from a few terracotta beads, and a few small (for
females?) bronze bracelets were recovered. Some kind of wood furniture or
apparatus was used, of which only the bronze clamps now remain.
We turn now to a discussion. of the chronology of Agrab Tepe within
the Iron III period. To begin with, it must be pointed out that specific
dates for the beginning and end of most sites of this period have yet to
be firmly established. Speaking generally for northwestern Iran, Iron III
begins sometime after the destruction of Hasanlu IV in the ninth century
bc, presumably after a hiatus of still undetermined length. But the complex
and still unresolved chronological difficulties surrounding the beginning,
flourishing, and end of the Sialk B culture to the south play a crucial role
in any discussion of the end of Iron II and the beginning of Iron III, and not
only for central Iran, but also for the north.
A brief discussion dealing with the opinions of various scholars con-
cerned with Sialk: Ghirshman and Porada see Sialk B as an Iron II site
both in culture and date, terminating about 800 bc, about the same time
148 chapter five
as Hasanlu IV.9 Dyson and Young accept the possibility that Sialk began in
the late ninth century, contemporary with the last phase of Hasanlu IV, but
see the culture continuing to exist until about 700bc (Young), or to about
650bc (Dyson).10 Goff Meade seems to agree with this, preferring Young’s
final dating to that of Dyson.11 She and Young also still use the term Iron II
to define the flourishing of Sialk, Goff Meade calling the eighth century “late
Iron II,” which suggests that Iron II continued to exist at Sialk for at least a
century later than in the north.12 Boehmer attempts to divide Sialk B into
an earlier and a later period (B1 and B2), the former beginning in the late
ninth century, the latter beginning about 770–760bc, after the destruction
of Hasanlu IV, and ending about 690– 680bc.13 The question to be answered,
considering these various opinions, is: do we consider Sialk B to be strictly an
Iron II culture, contemporary with but divergent from Hasanlu IV (Porada,
Ghirshman), or initially an “Iron II” culture that began in the ninth century
and continued (uninterrupted) into the eighth century (or slightly later)
(Dyson, Young, Goff Meade, Boehmer), developing into what archaeologists
call Iron III culture, and perhaps even having been the precursor of that
culture? How one interperts the nature of the anomalous Sialk B material
(only cemetery remains, let it be noted), and also perhaps the early phase
of Zendan, will determine whether one sees Sialk as Iron II, late Iron II,
or Iron II/III. Surely a chronological and cultural distinction for the terms
Iron II and III may have to be established.
Whatever the outcome of discussions concerning the culture and chro-
nology of Sialk B,14 Agrab Tepe remains an Iron III site, and to that site we
now return.
9 R. Ghirshman, The Arts of Ancient Iran (New York, 1964) p. 279; Ghirshman 1939, p. 95;
some of the Sialk B material (but which?) must be considered as Iron III, op, cit., pp. 27f.; and
Goff Meade’s late Iron II overlaps with Young’s Iron III.
13 R.M. Boehmer, “Zur Datierung der Nekropole B von Sialk,” AA (1965) pp. 802ff. Note
that some of the pottery in his earlier-period tombs occurs also in his later tombs. To add
to the confusion about the dates of Sialk B, note that of the five ceramic parallels Young
finds between Hasanlu IV and Sialk B (Young 1965, pp. 76f.), only one, the gray ware spouted
pitcher, is to my mind a strong parallel; and of the nineteen nonceramic parallels he cites
between the two sites, at least fifteen are in Boehmer’s B2 late tombs (Young 1965, p. 76,
note 28)!
14 My present opinion is that Sialk B existed into the 8th century, but I have no strong
opinion as to whether or not it can be stated that Sialk existed in the 7th century, nor if it
began to exist in the 9th century. But note that if, in fact, the designation Iron III is to be used
excavations at agrab tepe, iran 149
At Agrab Tepe five C14 samples were tested for dating.15 The results given
here use a half-life of 5730 years and should be corrected by a MASCA cor-
rection factor of +50 years (as of 1972). From the floor of C1, P-895, char-
coal: 79±556bc (845 bc); from Period 1 of area D, P-980, charcoal: 667 ± 58bc
(717 bc): ave. 781bc ±57. From Period 2 fill of area D, P-979, charcoal: 581 ±
53 bc (631bc). From high in the fill of Area A, P-893: 408 ± 48 bc (458 bc).
From high in the fill of Area A, P-894, charcoal: 710 ± 57bc (760 bc); and a
sample, probably from Period 2 of Area B, P-879, a burned beam: 513 ± 56bc
(563bc): ave. 597bc±55. We thus have an outer range of dates for Period 1 to
be 838–724bc; for Period 2, 652–542 bc; and possibly for Period 3 (assuming
P-893 to be from this time), 506–410 bc P-894 is assumed to be an aberrant.
Given the chronological range of over three hundred years it seems that we
should accept these carbon dates as a guideline rather than as data pointing
to specific historical dates.
Pottery comparisons with other sites in Iran and Urartu allow us to make
better judgments about chronology. We have seen that there are ceramic
parallels between Agrab and most of the known Iron III sites, In terms of
quantity, which of course could be accidental, the strongest ties are with
Hasanlu IIIB and A, Bastam, Ziwiye, Zendan (I and II), and Godin II; other
ties, less strong, are with Haftavan, Baba Jan (I and II), the Achaemenid
Village (I and II), and Geoy Tepe A. A brief summary of the chronology of
these sites is in order here; it will be seen that most were seventh century bc
sites that ceased to exist about 600bc
Very little of the Hasanlu IIIB and A material is available for study, and
basically the two periods remain unpublished.16 What may be said at present
only or mainly as a cultural term—to define the appearance of painted wares and oxidized
buff wares—and not simply as a chronological term signifying a period following upon the
destruction of Hasanlu IV, then Sialk B could be called Iron III from its inception. This would
obtain even if we accept that part of Sialk was contemporary to Hasanlu IV. An important
task for archaeologists is to learn if the painted wares of Sialk influenced the “triangle ware”
pottery of the north, and when this influence occurred (see Goff Meade 1968, p. 125). This
problem is the more important because similar painted wares occur in Urartu, viz., Van (von
der Osten 1952), Altintepe (Emre 1969, fig. 21, 22, pl. IV, v), Karmir Blur (rare; Piotrovskii 1969,
pl. 58), and in Anatolia from the 8th century on at many sites, viz., T. Özgüç” Kültepe and its
Vicinity in the Iron Age (Ankara, 1971), figs. 8 ff., pIs. XIV, XXI, 2; T. and N. Özgüç, Ausgrabungen
in Karahüyük (Ankara, 1949) pl. XXXI, 3, XXXII, 3; G.E.S. Durbin, “Iron Age Pottery from the
Province of Tokat and Sivas,” Anat. Stud. 21 (1971), pp. 104 f., fig. 3, 1–5, 13; cf. also pottery from
Gordion and Bogazkoy.
15 R. Stuckenrath, W.R. Coe, E. Ralph, “University of Pennsylvania Radiocarbon Dates IX,”
Radiocarbon 8 (1966) pp. 348 f. Recent information (1972) suggests that we may have to revise
past C14 dates upward again.
16 For general remarks and bibliography on Hasanlu III, Dyson 1965, Young 1965, pp. 53f.,
57ff., 72 ff.
150 chapter five
about the two levels is this: the beginning date of IIIB is still not known;
it could be from about 750 to 700, or perhaps even later, as it is clear that
IIIB followed upon a squatters’ settlement over Period IV (called IVA by
Dyson) of unknown duration. The destruction of IIIB (on parts of both the
western and eastern areas of the mound ash and charcoal layers document
a fire) occurred sometime in the seventh century; this is not in doubt. On
the western part of the mound IIIB walls were partly reused in Period IIIA,
and some new walls were built over the earlier ones; on the eastern side of
the mound there is an ash and trash layer 50 cm. wide over the IIIB level.
IIIA’s beginning, and more important, its termination date are still major
problems. It is possible, to my mind, that IIIA ceased to exist (abandoned?)
close to 600 bc; more excavation and analysis of pottery, however, may
make it necessary to extend this date well into the sixth century, beyond
585bc.17
The end of the settlement at Bastam has been dated by Kleiss and Kroll
to the late seventh century or early sixth century bc on the basis of Urar-
tian pottery comparisons.18 This date seems acceptable on the evidence pre-
sented (and neatly ties in Bastam’s destruction with that of Agrab’s [Period 1,
at least], especially since both sites depend a great deal on Urartian remains
for their chronology).
As Dyson has stated, any discussion of Ziwiye must distinguish the
archaeological site itself from the so-called Ziwiye treasure.19 He has pro-
posed a dating of about 750 for the beginning of the site and a terminal date
of about 600 for its abandonment.20 Young and Boehmer generally agree
with this range of dates.21 There seems little doubt but that the final period
at Ziwiye occurred either in the seventh century, probably toward the end
of that century, or possibly early in the sixth century.
17 Dyson 1965, pp. 211 f., and Young 1965, pp. 81 f., have IIIA continue into the Achaemenid
period. Kroll 1970, p. 76, note 105, suggests that Hasanlu IIIA ended ca. 600 bc on the basis of
the triangle-pithoi found at Hasanlu and Urartian sites. If C14 dates are to be pushed back in
time, this situation would support an earlier date for the termination of Hasanlu IIIA than
suggested by Young and Dyson. For a hiatus between Hasanlu IV and III, Young 1965, pp. 57ff.,
80.
18 Kleiss 1970, p. 57 f., accepting the possibility for an 8th-century beginning date; Kroll
1970, p. 76.
19 Dyson 1965, p. 206.
20 Dyson 1965 and R.H. Dyson, Jr., “Archaeological Scrap Glimpses of History at Ziwiye,”
p. 206.
excavations at agrab tepe, iran 151
Zendan had two settlements. The beginning of the earlier one, period 1, is
dated by Boehmer close to 800bc, on the basis of parallels with Hasanlu IV;
he dates the end of the second settlement, II, to the late seventh century.22
Young dates the beginning of the earlier period later than Boehmer, pre-
ferring a date between 750 and 650 bc, but he also believes that the site
continued to the end of the seventh century (for the second period), being
contemporary with the end of Ziwiye. Thus he agrees with Boehmer that the
late seventh century was the final date at Zendan; Dyson has also supported
this dating.23 The strong parallels between Zendan II and Ziwiye pointed out
by Young make it clear that about 600 bc is the probable date for the end of
period II.
The excavator of Godin Tepe, T. Cuyler Young, Jr., has cautiously given a
wide range of almost 200 years, 750 to 550 bc, as the time within which the
columned hall and fortress were built.24 It would seem from the pottery evi-
dence, however, that the end of Godin II could be placed in the late seventh
or early sixth centuries, given the parallels with Ziwiye and Zendan II (not
to mention Agrab, to avoid a circular argument). Nor do I think it can be
demonstrated on the evidence available that Godin II was built much before
the beginning of the seventh century bc.
Geoy Tepe A is a mixed complex, and it cannot help in dating any Iron III
site; rather, it must be dated by comparisons with other sites. Nor can the
limited remains from the upper levels at Haftavan at present be of help to
us in chronology.25 It would seem that the levels could be dated only from
evidence available at other sites.
Goff Meade has compared the ceramics from Baba Jan to those found
at Pasargadae, Godin II, and Nush-i-Jan.26 She suggests that Baba Jan II is
eighth century, not earlier, and that Baba Jan I is probably sixth century bc,
because of parallels with Pasargadae (unpublished). The Agrab parallels
with Baba Jan seem to be with both periods, but aside from a general Iron III
relationship, we get little specific chronological help from this site.
22 Boehmer 1961, p. 82; Boehmer 1965, pp. 736, 740, 763f.; Boehmer 1967, pp. 576, 579.
23 Young 1965, p. 82; Young 1967, p. 271; Young 1969, p. 50; Dyson 1965, pp. 201f., 211, agrees
that the early period is Iron II.
24 Young 1969, pp. 31 f.; see also Goff 1970, pp. 155.
25 Burney 1970, p. 182, suggests a late 8th- to early 7th-century date; Burney 1972, p. 142,
suggests that Haftavan was destroyed by Sargon II in 714bc, but presents no objective evi-
dence.
26 Goff 1970, p. 155; Goff Meade 1968 pp. 121 f.; actually, little of the pottery has been
published.
152 chapter five
The pottery parallels between Agrab and the Achaemenid Village are
basically in the levels I and II. These are dated by Ghirshman to the seventh-
sixth and sixth-fifth centuries bc, and he has been supported in general by
other scholars.27 None of the shape-parallels from the Achaemenid Village,
except the lip spout of level 1, are crucial enough to basically affect the
chronological relationship to Agrab Tepe.
Two other Iranian Iron III sites with a few parallels to Agrab Tepe are
Nush-i-Jan and Yanik Tepe. Both have been dated by their excavators to a
time in the seventh century bc.28
Turning now to the Urartian sites outside of Iran that have ceramic
parallels with Agrab, we find that the strongest ties, not necessarily with
respect to quantity, but to a very characteristic shape, are with Karmir Blur,
Altintepe, Norşuntepe, Patnos, Çavuştepe, and Igdyr; lesser ties are with
Toprakkale, Kayalidere, and Kef Kalesi. Some of these Urartian sites are not
yet completely published so that we are not always able to discuss Urartian
pottery types, nor to discuss their chronology with certainty.
Although there is continued discussion concerning the precise time
when Karmir Blur was destroyed, it seems that the event must have occurred
close to 600bc, apparently before 585.29 The finds from the final, second,
period at Altintepe also seem to date to this time; and it appears that the
material from the end of the first period may also be dated within the sev-
enth century, although the site may have been built in the eighth century.30
Kayalidere has not been more closely dated than to the eighth-seventh
century. But if we can use the pithoi decorated with triangles as a guide, it
could be that the destruction here occurred around 600bc, close to that of
Karmir Blur’s destruction.31
At Patnos inscriptions of several Urartian kings have been found; they
date from the late ninth through the middle eighth century bc. Later archae-
ological material is in evidence, however, and it seems clear that the site
existed through the seventh century, possibly even a little later.32
27 Ghirshman 1954, p. 20; Dyson 1965, pp. 205, 211; Young 1965, pp. 79, 82; Goff Meade 1968,
p. 125.
28 Stronach 1969, p. 16, Burney 1962, pp. 147 ff.
29 Muscarella 1965, p. 237, notes 34–36; Piotrovskii 1969, pp. 198f.
30 Özgüç: 1966, pp. 38, 46; Emre 1969, p. 291 f.
31 Burney 1966, pp. 55 ff., 79. Burney links the destruction to the Kimmerians, who first
appear in Urartu in the last years of the 8th century. Note that triangle-pithoi occur at the
termination of Hasanlu IIIA and Agrab I, events no doubt close in time but not necessarily
simultaneous; see my note 17.
32 K. Balkan, “Ein Urartäische Tempel auf Anzavurtepe bei Patnos …,” Anadolu 5 (1960) pp.
middle period; it is a type that cannot be earlier than the late 8th century and continued to
be used for centuries.
154 chapter five
37 R. M, Boehmer, “Volkstum and Städte der Mannäer,” Deutsch. Arch. Inst. Abt, Baghdad
3 (1964) pp. 11 ff.; also his “Zur Lage von Parsua in 9. Jahrhundert vor Christus,” BJV 5 (1965)
pp. 187 ff.; Young 1967, p. 14 f.; Kleiss, “Zur Ausbreitung Urartus,” pp. 130ff.; Louis D. Levine,
“Contributions to the Historical Geography of the Zagros in the Neo-Assyrian Period,” Ph.D.
dissertation, University of Pennsylvania, 1969. One also wonders if the Assyrians might not
have called the area around the south and west shores of Lake Urmia “Urartu” because of
their control over that area, Muscarella 1971b, p. 49.
38 Dyson 1965, pp. 208 ff.; Young 1967, pp. 12 ff.; Stronach 1969, pp. 5ff.; van Loon 1966, p. 23;
from east of Lake Urmia; Kleiss, “Zur Ausbreitung Urartus,” pp. 124f., 127ff.
excavations at agrab tepe, iran 155
only two historical events, themselves related, that can be brought forth
for consideration. One is the exodus of the Scythians from Iran, the other
is the northward expansion of the Medes, through north Iran, Urartu, and
eventually west to Anatolia. Again, this suggestion deserves further study,
but no other large-scale action occurred in Urartu and Iran at this particular
period.
The second hypothesis, also to my mind viable in that it does not abuse
the limited evidence, is that the sites in question were destroyed or aban-
doned over a slightly longer period of time than suggested by the first
hypothesis. Thus, one could assume that some of the sites could have ceased
to exist about 600–585 bc, while others could have ended about 550, say at
the time of the Achaemenid revolt against the Medes. One would then have
a time differential of twenty-five to forty years between the end of one par-
ticular site and another. In this context it must be understood that we do
not yet have an idea of what early sixth-century bc and early Achaemenid
pottery looked like, and it is quite possible that there was no major ceramic
change between about 600–585 and about 550bc. The fact that pottery
analysis at our present state of knowledge might not allow us to detect a
chronological difference between pots used at different neighboring sites
over a period of a few decades is the crucial factor here. This hypothesis,
incidentally, might also cover the problem of the difference between the
destructions at Agrab 1 and 2, and Hasanlu IIIB and A, but it would be pre-
mature to push this idea now.
In any event, these are nothing more than working hypotheses, to be
challenged or supported as more ceramic, archaeological, and historical
information comes forth. The end of the seventh and the first half of the
sixth century bc in Iran and Urartu was a time of chaos, destruction, and
abandonment for its people, and it is a time of chaos for modern historians.
Bibliography
Barnett 1952
R.D. Barnett, “Russian Excavations in Armenia,” Iraq 14 (1952) pp. 132–147.
Barnett 1954
“Excavations of the British Museum at Toprakkale, Addenda,” Iraq 16 (1954) pp. 3–22.
Barnett 1959
“Further Excavations in Armenia (1949–1953),” Iraq 21 (1959) pp. 1–19.
Barnett 1963
“The Urartian Cemetery at Igdyr,” Anatolian Studies XIII (1963) pp. 153–198.
excavations at agrab tepe, iran 157
Goff 1970
C. Goff, “Excavations at Baba Jan, 1968,” Iran 8 (1970) pp. 141–156.
Goff Meade 1968
C. Goff Meade, “Luristan in the First Half of the First Millennium B.C.,” Iran 6 (1968)
pp. 105–134.
Hauptmann 1970
H. Hauptmann, “Norşuntepe …,” Istanbul. Mitt. 19–20 (1969–1970) pp. 23–78.
Karageorghis 1962
V. Karageorghis, Excavations in the Necropolis of Salamis 1 (Nicosia, 1962).
Karageorghis 1969
Salamis (New York, 1969).
Kleiss 1970
W. Kleiss, “Ausgrabungen in der Urartäische Festung Bastam,” AMI 3 (1970) pp. 7–65.
Köpcke 1967
G. Köpcke, “Neue Holzfunde von Samos,” Ath. Mitt. 82 (1967) pp. 100–148.
Kroll 1970
S. Kroll, “Die Keramik aus der Ausgrabung Bastam 1969,” AMI3 (1970) pp. 67–92.
Levine 1970
L.D. Levine, “Of Medes and Media,” Rotunda 3 (1970) pp. 36–44.
Mallowan 1966
M. Mallowan, Nimrud and its Remains I, II (New York, 1966).
Mellink 1965
M. Mellink, “Archaeology in Asia Minor,” AJA 69 (1965) pp. 133–149.
Mellink 1966
“Archaeology in Asia Minor,” AJA 70 (1966) pp. 139–159.
Muscarella 1965
O. White Muscarella, “A Fibula from Hasanlu,” AJA 69 (1965) pp. 233–240.
Muscarella 1971a
“The Tumuli at Sé Girdan: Second Report,” MMA Jour. 4 (1971) pp. 5–28.
Muscarella 1971b
“Qalatgah: An Urartian Site in Northwestern Iran,” Expedition 13, 3–4 (1971) pp. 44–
49.
Öǧün 1967
B. Öǧün, “Die Ausgrabungen von Kef Kalesi …,” AA (1967) pp. 481–503.
Özgüç 1966
T. Özgüç, Aliintepe I (Ankara, 1966).
Piotrovskii 1959
B.B. Piotrovskii, Vanskoe Urartu (Moscow, 1959).
excavations at agrab tepe, iran 159
Piotrovskii 1969
The Ancient Civilization of Urartu (New York, 1969).
Porada 1965
E. Porada, The Art of Ancient Iran (New York, 1965).
Riis 1948
P.J. Riis, Hama, Les Cimetières à Crémation (Copenhagen, 1948).
Schmidt 1957
E. Schmidt, Persepolis II (Chicago, 1957).
Stronach 1969
D. Stronach, “Excavations at Tepe Nush-i-Jan, 1967,” Iran 7 (1969) pp. 1–19.
Vanden Berghe 1967
L. Vanden Berghe, “La Nécropole de War Kabud,” Archeologia 18 (1967) pp. 49–67.
Vanden Berghe 1968
Het Archeologisch onderzoek naar de Bronscultur van Lurisian …, (Brussels, 1968).
van Loon 1966
M. van Loon, Urartian Art (Istanbul, 1966).
von der Osten 1952
H.H. von der Osten, “Die Urartaische Topferei aus Van …,” Orientalia 21 (1952)
pp. 307–328.
R. Young 1953
R.S. Young, “Making History at Gordion,” Archaeology 6, 3 (1953) pp. 159–166.
R. Young 1962
“The 1961 Campaign at Gordian,” AJA 66 (1962) pp. 153–168.
Young 1965
T. Cuyler Young, Jr., “A Comparative Ceramic Chronology for Western Iran, 1500–
500BC,” Iran 3 (1965) pp. 53–85.
Young 1967
“The Iranian Migration into the Zagros,” Iran 5 (1967) pp. 11–34.
Young 1969
Excavations at Godin Tepe (Toronto, 1969).
chapter six
* Excerpted from Metropolitan Museum of Art Journal. Copyright © 1974 by The Metro-
consisted of the writer and Robert H. Dyson, Jr., as Co-Directors, Christopher Hamlin, Carol
Hamlin, Matthew Stolper, Elizabeth Stone, William Sumner, and Harvey Weiss as site super-
visors, and Marie Sherman Parsons as Registrar. Most of the drawings were made by Mary
Voigt and Maude de Schauensee (1966). John Alden and Elizabeth Hopkins inked the draw-
ings; their expenses were paid for by a generous grant from the Schimmel Foundation. I wish
to thank all the individuals mentioned as well as the Schimmel Foundation for their coopera-
tion in the production of this report. I also wish to thank Robert H. Dyson, Jr., Louis D. Levine,
and T. Cuyler Young, Jr., for discussions and opinions exchanged over the years about Iron
Age problems, and for reading this report in manuscript. Of course, I alone assume respon-
sibility for the format and the conclusions expressed, and for not always following their
advice.
162 chapter six
This paper first reports on the Dinkha III cemetery, its burials and their
contents, and its relations with contemporary sites. Following this is a report
and discussion on the Dinkha II architecture and burials. No attempt is
made here to write a history of the Iron Age or a definitive summary of
that period. Not enough information is available at present and several
good summaries already exist (Dyson 1964a, pp. 34–40; 1965, pp. 195–213;
1968a, pp. 29–32; Young 1965, pp. 55–59, 62–68, 70–83; 1967, pp. 22–29; Bur-
ney, Lang 1972, pp. 113–126). Rather, the emphasis here is on Dinkha Tepe
itself.
The mound was first divided into large grid-squares one hundred meters
to a side, and these were then subdivided when necessary into ten-meter
excavation squares. In addition to these squares, shorter test trenches and
wells were opened at various parts of the mound (Figure I; Stein 1940, p. 369,
fig. 23 for a contour plan).
the iron age at dinkha tepe, iran 163
One hundred and five burials were excavated on the mound. Thirty-three
of these are of the Dinkha III or Iron Age I period (Muscarella 1968, p. 189,
incorrectly listed twenty-six; see Table I). The majority of the burials were
excavated in the four northern excavation squares, the main cemetery area
of the mound, but some were found in the south and west. Whether these
latter burials were originally thinly scattered away from the main center at
the north, or whether they were part of a regular cemetery area encircling
the mound is not known, as extensive excavations were not conducted
in these areas. If there was a settlement on the mound that belonged to
Dinka III it could have been in the center and eastern sections, but no
architectural remains attributable to this period were recognized.
The dead were buried in individual graves with no markers; the brick
tombs generally opened to the east. Men, women,2 and children were buried
in the same area and apparently given the same burial rites. All the burials
of Periods III and II were placed within pits, which were then refilled; in a
few cases we were able to recognize the pit lines (Muscarella 1968, p. 190,
fig. 7).
Dinkha III burials were recognized primarily by the associated grave
goods, artifacts quite familiar to us from the Hasanlu excavations. In general,
the burials were stratigraphically lower in the fill than the later Dinkha II
burials; in some cases they were in the same stratum or were only slightly
lower than the later burials. Some Dinkha III burials were recognized as
being lower in the fill than others of the same period and these might be
early—although the possibility exists that some pits were dug deeper than
others (but compare TT VII, below). In a few cases the pottery types of these
deep burials seem to support a conclusion for a suggested earlier deposition
(see below).
Twenty-three of the burials were simple inhumations while ten were
associated with built brick tombs. Of the latter, four consisted of a horizontal
row of mud bricks to one side of which was placed the body. Three tombs
(B9a, β22, B9b, β11, B10b, β13) consisted of a horizontal row of mud bricks
with a projection or arm at each end, forming a three-sided tomb that
2 No professional physical anthropologist examined the bones when they were exca-
vated, and therefore it is not certain that the sexing was always accurate. The bones are
currently being studied by Ted A. Rathbun of the University of South Carolina. For contem-
porary skeletal material see Rathbun’s A Study of the Physical Characteristics of the Ancient
Inhabitants of Hasanlu, Field Research Projects (Coconut Grove, Miami, Florida, 1972).
164 chapter six
Figure 2. Burials 11, 12, 16, 18, 20, and 21 are of Period III.
enclosed the body and goods; one of these tombs (B9a, β22) had a mudbrick
floor. One tomb had an offset at each corner of the arms (Figure 2:21, partly
excavated; and Muscarella 1968, fig. 2), a feature common in the next period;
two tombs were disturbed. The main horizontal wall had two to four courses
and the arms two to three courses, the latter lower than the former. The top
course of the horizontal wall overlapped the grave area, often dug deeper
than the lower level of bricks, and in a few cases collapsed onto the body.
The bodies were oriented N-S or E-W, the former in the majority, and
although heads faced all points of the compass, those facing E predomi-
nated. The body was placed on the back or side; legs were flexed, with three
the iron age at dinkha tepe, iran 165
exceptions; arms were usually flexed before the face, chest, or pelvis, or
placed at the sides. Eight skeletons had one arm flexed across the body while
the other was bent back tightly, touching its own shoulder. No relationship
with regard to age or sex seems to exist in orientation or body position (for
details, Table I).
The characteristic ceramic objects of Dinkha III are the bridgeless
spouted pouring vessel, never with a handle; the pedestal-base goblet with
a vertical loop handle; and a flaring-sided bowl, either with a raised crescent
on the interior surface—colloquially called “worm” bowls—sometimes
with holes for suspension, or the same type bowl but without the crescent.
These types of vessels are classic diagnostic objects from the Iron I period.
Seventeen of the thirty-three burials did not contain a spouted vessel, but
nine of these had either the worm bowl or goblet; the eight others, contain-
ing only one or two vessels, were low enough in the fill to allow for a Period III
designation. Not a single burial contained all three of the diagnostic vessels
together (compare below, Geoy Tepe and Hajji Firuz).
Other Dinkha III shapes include deep carinated bowls, carinated jars with
relatively large mouths, and basket-handled teapots. These shapes continue
into Period II and by themselves are not easily distinguished into Iron I or
II.
Eighty-one vessels were recorded from the Dinkha III period, eighty from
the burials, one from the fill (Muscarella 1968, p. 193, fig. 17, left). Of these,
fifty-six were gray, twenty-three buff,3 and one was painted: thus the per-
centage of gray to buff is 71 percent to 29 percent. Whether the surfaces
were intentionally fired to these colors by controlling the oxygen within the
kilns, or whether the colors resulted fortuitously from firing to firing, or even
from uneven control in a given firing, is not clear. However, the fact that in
3 Buff is a term used for the oxidizing firing that produced non-gray (reduced) surfaces.
The colors of the buff pottery at Dinkha range from buff to light orange, orange, reddish-
orange, and red. The problem is not significant if one realizes that the Dinkha kilns produced
both reduced grays and oxidized buff colors. In the text I use the word buff in a general sense,
for the non-gray pottery, and in those particular instances where no specific color other
than “buff” was registered. Surfaces are categorized macroscopically as matt (A): no luster;
smoothed (B): a slight luster, with some stroke marks visible; burnished (C): stroke marks
quite visible and a definite luster. These divisions grade into one another. Interior paste is
categorized macroscopically as I: smallgrit inclusions of sand size; II: grit inclusions smaller
than sand to no grit inclusions visible; and III: coarse, with grit inclusions larger than sand
and visible. For convenience I use the abbreviated forms, e.g., IA, IIB, when describing a vessel
in the text. Next to each field number referred to in the text is a letter that gives the present
location of the object: M: Metropolitan Museum of Art; P: University Museum, Philadelphia;
T: Teheran Museum; D: discarded in the field.
166 chapter six
Period II buff pottery predominates might suggest that the coloring was con-
trolled (Young 1965, p. 55).
Within the gray pottery repertory, burnished surfaces outnumbered
smoothed surfaces more than two to one; only one had a matt surface.
Concerning the buff pottery, of which orange predominated, twelve were
smoothed, two burnished, eight matt, and one was red-slipped. About a half-
dozen vessels, gray and buff, had traces of mica flakes on the surface. Most
of the vessels were made of a paste that had no visible inclusions; about a
half-dozen had medium-sized grit, and only one (a worm bowl) was made
of coarse ware. Thus, the vessels may be categorized as neither of fine nor
of coarse ware, but rather what has been called common ware (Young 1965,
p. 55). Note that gray and red-slipped wares, and vessels with mica flakes,
occur in small amounts in the preceding Dinkha IV period.
The number of vessels placed within a burial varied from one to four,
and there seems to be no connection between the number of vessels, or,
indeed of burial goods in general, to inhumation or brick tomb, or to age
and sex (Table I). Some of the vessels, including all types, had obviously been
damaged in antiquity. But this fact did not deter their inclusion in a burial
and suggests that vessels placed in a burial were the same ones normally
used in the contemporary households.
Four burials contained weapons; thirty contained jewelry, worn by men,
women, and children, indicating that the dead were adorned as well as
clothed. In no burial of the Dinkha III period was an iron object found.4
An exception could be B10b, β11, discussed below, and which I consider to
belong to the Dinkha II period. Only one burial contained gold, B9a, β26,
and only one burial (B9a, β23) a cylinder seal (Table I).
Food remains in the form of sheep/goat bones were found in only three
burials, but it is quite possible that boneless meat and even liquids, all now
disappeared, were placed in some burials (see B9a, β15).
As stated above, it was possible to recognize that a few burials were
deposited earlier in time than others. To these examples we now turn. In
TT VII, Grid L, a square 2.5 × 1.5m, two period III burials were discovered, and
by a stroke of luck one had been deposited directly over the other (Figure 3).
β1, the later, found in stratum 2, was an inhumation of a young adult female
placed in an extended position on the R side, oriented N-S, head to N. The
4 When a metal object is mentioned in the Dinkha III section it is bronze (not analyzed),
except for the gold earrings; and when a spouted vessel or goblet is mentioned, it is a
bridgeless spout and a pedestal-base goblet. In the Dinkha II section a spouted vessel always
means a bridged one.
the iron age at dinkha tepe, iran 167
Figure 3.
168 chapter six
body had a pin, plain loop rings, bracelets, and a torque,5 (all corroded, so
discarded,) a spouted vessel (933P, like 234 in Figure 16), a bowl with two
holes (882D), and a carinated jar (938T), all gray IIC ware.
In stratum 6, but apparently cut from 4, was found β2 (Figures 3 and 4).
This was a young adult male in an extended position with the legs slightly
flexed, E-W, head to E, placed in a brick tomb (whether the tomb had arms or
not we do not know). The skeleton had a plain penannular bracelet on the
R wrist (316P), and a tanged dagger, with wood remains on the tang and a
wood peg still in the tang hole (1000P), placed behind the head. In the same
position was a red-orange IIB spouted vessel with a missing tip (237P), and
a tall gray IIB goblet (229P). No other goblet found at Dinkha has the same
shape, with straight walls, nor does any other spouted vessel have the same
5 By torque I mean a penannular necklace, at Dinkha made from one piece of metal,
and not necessarily twisted. In two burials, both of Period III, B9a, β22, B9b, β16, originally
penannular necklaces had their ends tied together, I have not considered them as true
torques.
the iron age at dinkha tepe, iran 169
body structure, with relatively tall and straight inner vertical section of the
spout—except a vessel from B9a, β24 (Figure 6, 936) also early. The vessel
also lacks a “beard” below the spout, and has a narrow mouth. Doubtless this
is one of the earliest Iron Age burials at Dinkha.
In B9a a number of period III—but no II—burials were excavated in
strata 5a and 6; six burials were also found below these in stratum 7, cut
into the Bronze Age deposit. These latter could be early period III burials.
One of these burials stands out from the rest because it contained the only
painted vessel from the Iron Age at Dinkha.
B9a, β25 contained the inhumation of a mature adult male flexed and
placed on his back, E-W, head to E. He wore a toggle pin with a finely
decorated top (Figure 3, 473P), and a necklace of paste beads (1006T); a gray
IIC goblet (696T) and a polychrome jar (420T) were the other grave goods.
The latter has a cream surface overpainted on the upper body with reddish
brown hatched triangles outlined with dark brown lines (Figure 5).
170 chapter six
6 Tests on the bead were made by J.H. Frantz and Suzanne Heim in the Research Labora-
tory of the Metropolitan Museum of Art. The tests show that the bead is not glass, leaving the
following possibilities: a glazed soft-stone, faience, or glazed earthenware. Only the surface
and the inlay appear to be vitreous. Compare A. von Saldern, “Other Mesopotamian Glass
Vessels (1500–600bc),” in Glass and Glassmaking in Ancient Mesopotamia, ed. A. Leo Oppen-
heim (New York, 1970), p. 217.
the iron age at dinkha tepe, iran 171
holes (599P); broken plain rings found by the teeth (D); two gold earrings
consisting of a cluster of hollow balls with a loop: one was found by the
left leg, the other under the skull; associated with the earrings are gold
loops (629T); 76 flattened carnelian beads, 21 similar-shaped copper ones,
plus 150 round paste beads; and one calcite disc (622T), at the back of the
neck. Vessels included a bronze omphalos (542T) by the chin; a broken gray
burnished bowl (881D); a broken, buff, smoothed basket-handled teapot
with mica flecks (792T), and a broken, gray-brown burnished spouted vessel
(922D), same type as 234 in Figure 16, by the feet.
β27: Male, mature adult, inhumation, flexed on back, N-S, head S. Right
arm bent back to touch its own shoulder (Figure 8). Furniture (Figure 7):
plain bracelet with overlapping tapered ends (605D) on R wrist; a stone
button with drilled designs (616P) by L foot; assorted beads by throat: 833a,
coarse faience; b, fine faience (glass?); c, paste; d, e, f, carnelian; g, a lotus-bud
shape, fine faience; h, j, glass (P); also, a socketed spear on L leg so that the
shaft must have passed over the body (1045T; compare Dinkha II burials B9a,
βg, and B10a, β12, Figures 24, 36). At the feet, a dark gray burnished spouted
vessel (921D, same as 234 in Figure 16), and a gray IIC worm bowl with two
holes (889P).
We now proceed to some of the other Dinkha III burials; these do not
allow themselves to be distinguished as early or late on the basis of stratig-
raphy or artifact comparisons. Space limitations forbid publication of all the
burials, but no important features of the period will be omitted (see Table I).
The burials are presented according to their grid positions:
b9a, β15: Female (?), mature adult, flexed, on back, N-S, head N, in hor-
izontal brick tomb (Figure 9). Furniture: one round bracelet with overlap-
ping ends on R arm (307P), two on L (308T, 354D); two plain rings with
overlapping ends, one on R hand (595T), one on L (601P); two pins with
simple knobbed heads at throat (Figure 52, 400P), one was found stick-
ing up in the fill; another pin with one knob by L arm and another by R
(Figure 52, 385T); a needle over the chest (460P); paste and copper beads
(391T); and five bronze buttons found on the skull probably from a cap or
diadem (617P; B8e, β8, a Period III tomb of a mature adult, also had five
bronze buttons on the head. The buttons as shown in the photograph may
be in their original position; there is no comment in the field notes to the
contrary). At the feet, a highly burnished gray spouted vessel with ridges
uniformly arranged around the body (334M; Muscarella 1968, p. 193, fig. 17,
top), a gray IIB bowl with two holes (358T), and a gray IIB jar with two ridges
at mid-body (404T) sealed with a stone; this vessel probably held some
liquid.
172 chapter six
Figure 6.
the iron age at dinkha tepe, iran 173
Figure 7.
174 chapter six
b9a, β17: Mature adult, inhumation, flexed on L side, N-S, head to N; R arm
missing, L bent back onto its own shoulder (Figure 10). Furniture (Figure 11):
a flattened bracelet with overlapping ends on L (310T) and R (309P) wrists;
an anklet with overlapping ends on each foot (539T, 540P); a plain toggle pin
at L shoulder (326D), fragments of another in the fill; a ring of twisted wire
with overlapping ends on L hand with cloth impression (466T); a needle in
the fill, with top bent back to form the hole (325D); a plain torque on the
neck (1038D). Touching the forehead was a gray IIB spouted vessel (337T);
by the feet, a broken gray IIC bowl with two holes (893P), and a gray IIC
carinated jar (903D).
b9a, β19: Child, inhumation, flexed on R side, N-S, head S; L arm flexed
across body, R bent back onto its own shoulder. Furniture (Figure 12): a
plain, not quite round bracelet with overlapping tapered ends (541T), on R
wrist; a plain ring, also with overlapping tapered ends (462D), on R hand.
A buff IB tripod bowl, feet of which were broken (982D), resting on a gray
burnished jar (952D), at the forehead; at the feet, a buff smoothed carinated
bowl (871D) and a gray burnished goblet (717T).
176 chapter six
Figure 11.
b9b, β11: Child, flexed on L side, N-S, head N; in brick tomb with projecting
arms (Figure 2; Levine 1971, p. 40, top); the fill in the grave was packed in
very hard. Furniture (Figure 13): a plain flat band bracelet with overlapping
ends (369P); a plain round bracelet with overlapping tapered ends on R
wrist (319D), two on L (320P; like Figure 7, 532); two plain loop earrings (?)
(351D); a bronze spiral object (bead? pendant?) at the neck (618P); two plain
flattened anklets with overlapping ends on L foot (311P), two on R (312D); a
shell bead necklace (299P); and a plain torque (538T). Placed at the feet: a
gray IIB spouted vessel decorated with ridges around the upper body (84T),
an orange IIB carinated jar (83T), and an orange-red IIB bowl (85P).
b9b, β12: Child, inhumation, flexed on the back, N-S, head S (Muscarella
1968, p. 192, fig. 15). Furniture (Figure 13): a plain bracelet with overlapping
ends (350D) on R wrist, the sole jewelry. Covering the head was a gray IIB
tripod worm bowl with mica flecks, and one hole (88M; Muscarella 1968,
p. 193, fig. 17, right), in which were three astragals; by the feet was a gray IIB
goblet, missing the handle (87P), and an orange IIB carinated jar (86T).
b9b, β16: Young adult, inhumation, flexed on R side, NE-SW, head S (Fig-
ure 2). Furniture (Figure 14): a plain round penannular bracelet on R wrist
(300D); a bronze necklace with its hooked ends linked together (therefore
not a true torque), with cloth impressions, at the neck (1037T); a necklace of
paste disc beads (301P); a broken bone awl in the fill (242T); and a tanged
dagger in the fill (241P). At the head was a gray IIB spouted vessel with
a hatched design on the base, exhibiting mica flecks (269T), and a gray
smoothed jar, also exhibiting mica flecks (949D); at the feet was a gray IIC
tripod bowl with two holes (359T).
178 chapter six
Figure 12.
the iron age at dinkha tepe, iran 179
Figure 13.
180 chapter six
Figure 14.
Figure 16.
b10b, β10: Female, mature adult, flexed tightly on L side, N-S, head N; in
horizontal brick tomb (Figure 15; Muscarella 1968, p. 189, fig. 2). Furniture
(Figure 16): a plain pin at L and R shoulder (138P, 137T); a pin, square in
section, with the top twisted into a loop, at R shoulder (200P); a needle
at the chest (198T); plain loop penannular earrings (148T); a flattened ring
with tapering, overlapping ends on R hand (199P); fifty small round paste
and bronze beads at the neck (896T). Clustered at the feet: a gray IIB–C
spouted vessel (234P), a gray burnished carinated jar (939P), and a broken
red-slipped worm bowl with two holes (357P); animal bones were found in
the bowl.
182 chapter six
Figure 17.
b8e, β7: Both the brick tomb and skeleton were disturbed. Furniture
(Figure 17): an orange matt basket-handled teapot with a broken spout
(937T), and a buff IIC worm bowl with two handles and two holes (891D).
Only these two vessels were found.
Not enough material from Hasanlu V has yet been published to permit a
comprehensive comparison of the material from both sites. As more Has-
anlu material becomes available we will no doubt recognize more parallels
and connections than are given here.
The sites are about fifteen miles apart, separated by ridges, but with no
impediments to travelers from one site to the other. That travelers, mer-
chants, and perhaps potters and other craftsmen did indeed travel freely
and often back and forth is documented by the obvious strong ties between
the sites, evidenced by the material culture that was basically the same in
many cases, and very close in others (Muscarella 1968, pp. 189, 194). And not
only does this closeness obtain in the Iron I period, but, as will be seen, it
the iron age at dinkha tepe, iran 183
vertically pierced handles, jars with one handle, and cups like those found
in Hasanlu V (Dyson 1965, fig. 13; Young 1965, pp. 72–73, fig. 11), do not occur
in the Dinkha graves.
jewelry: The types of pins, bracelets, anklets, etc., from Hasanlu have yet
to be published. We can state, however, that torques were not found in the
Hasanlu V graves and that a few toggle pins were found (Dyson 1968a, p. 23).
Two Dinkha graves contained bronze buttons or studs that belonged
originally to a headband, diadem, or cap (Figure 9). At Hasanlu a Period V
burial (VIF, β8) contained a plain bronze band, curved to fit the head and
pierced at both ends, presumably for attachment to another, perishable,
material. Headbands were also reported from Period IV graves at Hasanlu.
The gold earrings from B9a, β26 (Figure 7) are similar to gold earrings
found in Hasanlu IV’s Burnt Building II, attached to an ivory statuette frag-
ment (Muscarella 1966, pp. 134–135, fig. 36). This earring has the cluster of
hollow gold balls, but placed under a buttonlike form attached to a twisted
gold wire. Another gold earring, consisting of hollow carinated balls in a
pyramid cluster, and attached to a loop, was found at Hasanlu in 1947 (Rad,
Hakemi 1950, fig. 90b). This earring is more elaborate than, but related in
form to, the Dinkha earrings.
weapons: Four of the Dinkha III burials contained weapons, a spear and
four tanged daggers in all. In the same Period V burial at Hasanlu that
contained the painted jar, Stein (1940, p. 402, pl. xxvi, 2) found a bronze
spear; and Dyson (1964a, pp. 34–35, fig. 2:1) published a bronze dagger with
a lappet-flanged hilt (57–129) that came from a Period V burial.
As stated above, iron was not found in any Period III burial. At Hasanlu
only one iron ring was found in a Period V context (Dyson 1964a, p. 39; 1965,
p. 196; 1967, p. 2957).
Two Dinkha III burials each had among the grave goods three astra-
gals placed in a bowl. Were they from meat, or were they game pieces?
There is certain evidence in Period II at Dinkha that astragal game pieces
were placed in tombs (see below), but it seems to me that in these cases
the astragals were probably simply the remains of meat placed as food in
bowls.
of these sites, the late neolithic or chalcolithic mound of Dalma Tepe, just
south of Hasanlu, had a number of Iron I burials deposited in its upper
level. One grave, Operation IV, β2, contained a gray pedestal-base goblet
(MMA 62.173.9; Young 1962, pp. 707–708, fig. 8), but little more can be said at
present than that an Iron I extramural cemetery existed here and that there
may have been a settlement somewhere in the vicinity.
In 1968 at Hajji Firuz, a neolithic mound southeast of Hasanlu, an Iron I
inhumation of an adult was found (unpublished). The arms and legs were
flexed, and the body was oriented NNW-SSE, head NNW. The grave con-
tained all the classic diagnostic pottery of the period: a bridgeless spouted
vessel, a worm bowl, and a goblet (information from Mary M. Voigt). It
will be remembered that none of the Iron I burials at Dinkha (or at Has-
anlu) contained all three diagnostic vessels together (but see Geoy Tepe,
below).
geoy tepe: The B period, in particular the contents of a single tomb,
Tomb K, represents the sole published evidence for an Iron I occupation
here (Burton-Brown 1951, pp. 141ff., figs. 28, 29, 32, 34; Dyson 1965, p. 196,
fig. 2; Young 1965, pp. 70–72, fig. 11, p. 78). Recent tests made at the site
indicate that Iron I trash deposits occur; thus evidence of occupation from
that period exists although never extensively excavated (personal commu-
nication from Robert H. Dyson, Jr.). Geoy Tepe B shares with Dinkha III
the bridgeless spouted vessel, the goblet, and the worm bowl, all found
together; in addition, there are toggle pins very close to those at Dinkha
(Figure 11, 326), and tomb architecture, albeit not bricks but stone (compare
Dinkha II).
haftavan: Here settlement on the mound seems to be indicated along
with an extramural cemetery just below the citadel, as at Hasanlu. In the
settlement area were found bridgeless spouted vessels and worm bowls; and
a spouted vessel and a cup were found in an inhumation burial (Burney 1970,
p. 170, figs. 8:1, 7, pl. iii; 1973, pp. 155, 162–164; Burney-Lang 1972, fig. 40). No
more data are presently available.
yanik tepe: No settlement occupation was found, but a cemetery at the
foot of the west side of the mound was located. Here eight Iron I burials were
excavated, of which only one has been published, A6 (Burney 1962, pp. 136,
146–147, pl. xliv, figs. 24–29). The flexed bodies have no particular orienta-
tion and were placed on the left or right side; the graves were “sometimes
lined with mud brick,” a feature at present recognized in the Iron Age to my
knowledge only at Dinkha. No spouted vessel occurs in A6, but a vessel with
a vertical loop handle, similar to the Iron I goblets, was a clue to the grave’s
approximate date; this burial may belong to a late stage of Iron I.
186 chapter six
A burial from Trench P, in which were found toggle pins and a painted
jar, along with two hand-made vessels, was dated to the Iron I period, about
1000bc (Burney 1964, p. 60, pl. xv, 14–19); this date is not certain, but if
correct, the grave is then surely of the Iron II period. Toggle pins, for example,
occur in the Bronze Age and throughout the Iron Age.
tashtepe: Dyson (1965, p. 196) referred to Iron I pottery from Tashtepe
based on Ghirshman’s claim (1954, pp. 61–62) that gray wares similar to those
from Giyan were found there on survey. These sherds remain unpublished
and therefore prevent independent acceptance of Ghirshman’s statement
(compare Young 1967, p. 22, note 70).
khurvin: This site was plundered by local inhabitants and only a few
graves were excavated by vanden Berghe (1964, pp. 6 ff.). The graves are not of
Iron I date (Dyson 1965, pp. 196, 206) although a particular type of bridgeless
spouted vessel of late type is common; Goff Meade (1968, p. 125, note 50)7
dated the burials to Iron II based on analogies with Sialk B. True Iron I
vessels said to be from Khurvin, but without archaeological contexts, exist
in private collections (vanden Berghe 1959, pp. 123–124, pl. 153, and p. 124,
pl. 158 for Chandar; 1964, passim; Ghirshman 1964, figs. 15, 16); others are
from controlled field surveys (Young 1965, fig. 9). The bridgeless spouted
vessel of Iron I type, goblets, bowls on tripods—similar to worm bowls,
but without the worm—are part of the repertory. Also reported are familiar
bronze torques, tanged swords, needles, toggle, incised, and plain pins, and
pins with curled tops (vanden Berghe 1964, pls. iv, xiv, xv, xxiii, xxvi,
xxxiv, xxxix, xli, xliii). While we cannot control the information enough
to actually know if this metal material is Iron I rather than later, given the
material itself and the pottery configuration, it is quite possible that part of
it, at least, is early (Moorey 1971, p. 25; for pins and needles, pp. 172–215).
Vanden Berghe also relates (1964, p. 3) that the burials were all inhuma-
tions without any particular orientation; presumably he is talking about the
burials he excavated, but he implies that he is also discussing those burials
dug by the local inhabitants.
marlik: Without doubt some of the Marlik tombs belong to the second
millennium bc while others must be later (Muscarella 1972, pp. 42–43).
Bridgeless spouted vessels in metal and pottery, as well as toggle pins and
tanged swords, are attested there (Negahban 1964, figs. 25, 29, 41, 108, 121,
7 Claire Goff (Meade) considers the unbridged spout with curled ornament to be late
Iron II, eighth century bc (1968, pp. 115, note 17, p. 121), while Dyson considers it to be Iron III,
eighth century bc (1965, p. 206, fig. 11), and Young (1965, p. 73, fig. 11) lists it as Iron I. The
evidence from Sialk B suggests that it was in use in the eighth century.
the iron age at dinkha tepe, iran 187
135). Two published gold earrings are not dissimilar to the clustered hollow
balls on examples from Hasanlu and Dinkha.
kizilvank: Bichrome vessels closely paralleling the Hasanlu V and
Dinkha III painted vessels discussed above were excavated here (Schaeffer
1948, p. 500, fig. 270; Muscarella 1968, p. 194). Moreover, aside from the spe-
cific paint parallels, two of the bridgeless spouted vessels have a short spout,
one has a rather narrow mouth, and both have a straight interior wall on
the vertical spout, all features in evidence on the vessel from B9a, β24 (Fig-
ure 6). A goblet from the site is similar to those of Iron I type, but it has a
flat base. Schaeffer (1948, p. 500) dated these vessels on typological grounds
to between 1350 and 1200bc. Monochrome red and gray pottery of the Early
Iron Age, bronze daggers similar to those from Dinkha III, and a flanged dag-
ger of Iron I type are reported from the site (Burney-Lang 1972, p. 169, fig. 43a,
b; compare Dyson 1964a, figs. 1:5, 2:1, and p. 34).
sialk a: Moving to the south, to central western Iran, we see that the Iron
Age culture extended as far southeast as Sialk and as far south as northern
Luristan (Goff Meade 1968, pp. 127–132; compare Dyson 1968a, p. 25, for a
similar situation existing in the Late Bronze Age).
The necropolis of Sialk VI, Necropole A, provides the relevant informa-
tion (Young 1965, pp. 61–62, 73, fig. 11). Here only an extramural cemetery
is available for study. Some Iron I vessels, however, do come from limited
excavation on the South Hill, where we are also told related architecture was
cleared (Ghirshman 1939, p. 11).
The skeletons were flexed in single burials, with no particular orientation,
except that most of the heads pointed north, as at Dinkha. The pottery
covers the range of familiar Iron I wares and shapes (Young 1965, pp. 61–62;
Dyson 1965, p. 195). The clothed dead wore bracelets, pins, rings, and at least
one needle was found. One tomb contained gold; another, probably late (see
also Moorey 1971, p. 316), an iron tanged dagger and an iron point, along with
bronze weapons (Ghirshman 1939, pl. xxxix). Note that Young (1965, p. 62)
suggests that Necropole A lasted a long time.
giyan: Young (1965, pp. 62ff.) has reorganized the subdivision for Giyan I,
a system accepted by Dyson (1965, p. 195, note 5). At Giyan we have basically
a cemetery with no definite related settlement—except it is possible that
Construction A may be contemporary to some of the burials, but this is by
no means certain (Young 1965, p. 66). Graves of Giyan I4–I2 are the ones of
concern to us.
These graves are simple flexed inhumations with no apparent orienta-
tion. Except for one bridgeless spouted vessel of a type also found at Sialk
B and at Khurvin, and which may be later than Iron I, the shape is not
188 chapter six
represented at Giyan (see note 7). The pedestal-base goblet is fairly com-
mon, however, especially in I4 and I3. The dead were buried with bronze pins,
needles, bracelets and anklets, and occasionally with a tanged dagger, in one
case (late?) iron. One skeleton wore a headband of bronze loops, and a sin-
gle cylinder seal was found (Contenau, Ghirshman 1935, pp. 23, 26, pls. 14,
18). Another seal, of Mitannian type, was found low in Construction A and
could have come from a tomb.
godin: Three isolated burials containing Iron I material, but with no rela-
tionship to any settlement on the adjacent Godin mound, were discovered
in a Bronze Age cemetery (Young 1969, p. 19, figs. 24, 25). They are all simple
flexed inhumations, oriented E-W, on their R or L sides, facing N or S. Each
grave contained a typical Iron I goblet. One grave contained a ring and two
pins; another a bracelet and a bronze cup; the third an arrow and a sword
with an open crescent handle. Two skeletons held vessels in their hands
(compare p. 48 above).
Interestingly, each of the goblets is slightly different in base type and
outline, which does not necessarily signify that they were deposited over
a long period of time. It should also be noted that several toggle pins with
decorated tops from a Godin III, Bronze Age, burial (Young 1969, fig. 30) are
quite similar to an early example from Dinkha (Figure 3, 473).
tepe guran: In an occupational context of Level VII, the latest settlement
at Guran, a bridgeless spouted vessel with a handle (unlike Dinkha) was
excavated (Thrane 1964, pp. 122, 123–124, figs. 23, 24; 1965, pp. 158–159, note 6).
Cut into this level, and therefore later, was grave 4, which contained a bronze
spouted vessel of a type similar to those from Hasanlu IV and Sialk B (Thrane
1964, p. 129, figs. 30, 31; 1965, pp. 158–159, note 6; Moorey 1971, pp. 276–280).
Thrane, nevertheless, dates Level VII to the Sialk B period, that is, to the early
first millennium bc (also Thrane 1970, p. 31, 850–750bc; Moorey 1971, p. 21).8
It would seem that the stratigraphically later grave 4 is Iron II in date, and
that Level VII may belong to the Iron I period.
A word should be said about the ram’s head bead from B9a, β24 (Figure 6,
note 5; Muscarella 1968, p. 194, fig. 19). Similar “frit” and “glass” beads were
found at Nuzi (Starr 1939, pl. 120), Alalakh (Woolley 1955, pl. lxviii), and
al-Rimah (Carter 1965, p. 51), all approximately mid-second millennium
8 Moorey inadvertently placed the bridgeless spouted vessel from Thrane 1964, fig. 24, in
Grave 4 (Moorey 1971, p. 21); in fact, it was found in situ in Level VII (Thrane 1964, pp. 122–131,
figs. 23, 25; 1965, p. 159, note 6; 1970, p. 31, fig. at top). As stated in the text, Grave 4 is related
to Sialk B material and is later than Level VII, making the latter possibly pre-ninth–eighth
century bc, probably Iron I as argued here.
the iron age at dinkha tepe, iran 189
9 The only anomaly at Geoy Tepe is the use of a multiple burial, whether or not we
Hasanlu IV, the torque was used in Luristan (Godard 1931, pl. xxvi, 78, 80),
at Sialk B, and in the Caucasus and Talish regions (Schaeffer 1948, figs. 254,
298, 301; Hȧnčar 1934, p. 97; Godard 1931, p. 64, fig. 34; Barnett 1967, pp. 177,
174, fig. 27:3; Herzfeld 1941, p. 146, pl. xxx, says some were found at Giyan).
We do not really know if the torque was first used in the Talish area, or in the
Urmia basin area—whence it could then have moved north and south—or
whether the Khurvin examples are contemporary to those at Dinkha, having
been a basic element in the Iron I culture from the first years of settlement
in Iran.
The painted pottery from Kizilvank is difficult to evaluate. One asks:
does this site represent the first stage of the new Iron Age in its incipient
phase, thus affording us a clue about the area of origin (see Burney, Lang
1972, p. 116), or was it a backwater, being later than, or even contemporary
with, the Iron I culture to the south? It is preferable to leave the questions
unanswered at this stage.
The people in the central plateau, at Sialk, Giyan, and Godin, had similar
burial practices and included artifacts in their graves similar to those found
further north. Tombs at Sialk and Giyan contained tanged daggers of the
same type found at Dinkha. But they also occur later at Sialk B (Ghirshman
1939, pls. l, lvii, lxviii; Moorey 1971, pp. 66–68 for a late dating for some
examples). In the north this tanged dagger is clearly earlier than examples
with cast hilts.
Finally, we have the three graves from Godin. One wonders if they are
in fact isolated and were deposited by a people on the move (as Young
1969, p. 19), or whether there are other burials at Godin still unexcavated
that might perhaps indicate a nearby settlement, or more intense use of the
area. Without any more information at hand the graves offer us merely a
tantalizing glimpse, rather than a substantial view, of the Iron I period at
Godin.
1949, p. 109, refer to heavy bronze torques from Geoy Tepe and Iranian Azerbaijan, based
on a report from C.C. Lehmann-Haupt. These objects cannot be the same objects we call
torques that were found at Dinkha and Hasanlu. Ghirshman 1964, p. 113, fig. 148, following
Godard, refers to a gold fragment allegedly from Ziwiye as a torque, but this is not certain. For
Achaemenian torques see J. de Morgan, “Découverte d’une Sépulture Achéménide à Suse,”
MDP VIII (Paris, 1905) pp. 43–44, pl. iv; E.L.B. Terrace, “Sumptuary Arts of Ancient Persia,”
Boston Museum of Fine Arts Bulletin 13 (1965) p. 27, with references; see also Schmidt 1970,
pp. 111–116, and my comments in a review of Schmidt in AJA 75 (1971) p. 444. Note that a torque
with twisted ends, similar to Figure 32, B9a, β14, 1040, seems to be worn by a youth on a relief
from Marash: E. Akurgal, The Art of Greece (New York, 1968) pl. 29.
the iron age at dinkha tepe, iran 191
It has been stated many times that the Iron I culture represents a new
phenomenon in western Iran, a major break with the past and a new age.
The information available from the Dinkha excavations reinforces this con-
clusion both from stratigraphical and cultural evidence. There is a definite
break, a hiatus, after the termination of the last Bronze Age settlement. A
build-up of debris and erosion material covered this destroyed settlement,
creating a hard-packed, easily distinguishable stratum. Ash layers, debris,
and erosion material, containing Iron I sherds, coming from the southeast,
and thinning at the north, were laid down. It was into these layers that the
first Iron Age graves were deposited (Figures 18, 19, 20). Perhaps these Iron
Age layers came from the earliest Iron Age occupancy of the mound, from
a time before the primary use of the area as a cemetery (it will be recalled
that one of the earliest burials came from a test trench in grid L to the west,
TT VII, β2). In any event, trash and ashes continued to be deposited during
the Iron I and II periods.
Culturally the break is equally clear and dramatic, notably in the pottery
and in the burial customs, where single inhumations in an extramural ceme-
tery replace intramural multiple burials. That this new culture represents a
“shifting of population,” to quote Dyson, and that it represents at the same
time a “cultural uniformity,” pointing to a “common origin for the Iron I cul-
tures,” to quote Young, is beyond dispute. Indeed, all the bricks are not yet
available for archaeologists to build a fine structure of full understanding
about the nature of the historical events leading to the change. But further
excavations, conducted scientifically, will continue to supply the necessary
information and slowly put into focus the picture we all seek.12
An item of some importance is the fact that at Dinkha we have been able
to isolate a few burials and their contents that are of the early Iron I period.
At Hasanlu, aside from the polychrome vessel excavated by Stein, we have
no recognizable early material. Thus, although it would be rash at present to
conclude that the Iron I period began earlier at Dinkha (we still do not know
what is in the unexcavated ground at Hasanlu and at other unexcavated
local mounds) we can at least illustrate the earliest excavated Iron I material
there.
yielded a cemetery of Iron Age date: K. Fard, “Fouilles dans les Tombes ancien de Gheytareh,”
Bastan Chenassi va Honar-e Iran 2 (1969) pp. 26–30. A pedestal-base goblet of Iron I type,
said to come from Kalé Dasht, near Saveh, southwest of Teheran, is in the Teheran Museum,
no. 872.
192 chapter six
Figure 20.
194 chapter six
As sharp as the break was with the past, it seems almost certain that the
Iron I people had some knowledge of the earlier cultures, probably from
scattered pockets of survivors in the penetrated areas. The use of multiple
burials at Geoy Tepe, and the use of toggle pins and painted wares (perhaps
also of gray and red-slipped wares?), reflects a continuity with the past
within Iran, even if not of major proportions.
chronology: A C14 sample from the terminal Bronze Age deposit gave a
reading of 1435±52 bc (P-1231, half-life of 5730 years; Dyson 1968a, p. 22). This
gives us a rough terminus for the end of this settlement and an ante quem
non date for the following Iron Age.
Three C14 charcoal samples exist for Dinkha III; each came from separate
pits overlying the Bronze Age deposits, from the Iron Age fill. One gives
a reading of 1146±37bc (P-1475); the second, stratigraphically earlier than
the first, gives a reading of 1302±57bc (P-1474); and the third from a pit
resting directly on the Bronze Age level, a reading of 1243 ± 37bc (P-1449).
These readings suggest a general date for the end of the Bronze Age in the
fifteenth century bc, and a range of about the late fourteenth to the late
twelfth century for at least part of the overlying Iron Age I burials. They also
suggest that the gap between the end of the Bronze Age and the beginning
of the Iron Age may have been about one hundred years. To be sure, the
pit samples do not necessarily date the earliest nor the latest Iron I burials.
(Note that if the recently published MASCA correction dates for C14 readings
prove to be stable, it will be necessary to push all the dates further back in
time [Ralph et al. 1973, p. 11 and passim]. Thus, the end of the Bronze Age
will have occurred about 1600 bc, and the beginning of the Iron Age about
1500bc. And this correction factor would then change all the dates presented
here by 100 or more years.)13
Dyson (1968a, p. 31) suggested “a working date of 1350 ± 50 bc for the
beginning of the [Iron I] period,” close to the date of 1300/1250 of Young
(1965, p. 83; 1967, p. 12). This date, about 1350 bc, is also proposed by Burney
(Burney-Lang 1971, pp. 106, 113, 115–117; also Muscarella 1968, p. 196). Thus
the tombs that I suggest are the earliest at Dinkha, B9a, β25 and VII, β2,
would presumably have been deposited in the late fourteenth century bc,
13 I use the standard C14 dates in the present report and will continue to use them
until more information and discussion on the MASCA corrections are available. If these
corrections are eventually proven correct, the dating of the many Iron II objects from the
destruction level at Hasanlu IV—the ivories, bronzes, the gold and silver bowls, not to
mention the pottery and architecture—shifts dramatically from a late ninth- to a late tenth-
or early ninth-century date, with important implications also for the dating of much material
not from good archaeological contexts.
the iron age at dinkha tepe, iran 195
at least close to 1300bc. Later than these would be burials B9a, β23, and
24, followed by B9a, β26, and 27, and perhaps we could accept a general
thirteenth-century date for these in the order given. The other burials do not
allow themselves to be defined more precisely and presumably span some
centuries, if we can accept the fact that a certain conservatism obtained in
the middle and later stages of the Iron I period.
The terminal date for the Iron I period can be determined at present only
by reference to the large amount of data from Hasanlu. The evidence there
suggests that in the eleventh century bc (or earlier, given the MASCA correc-
tions) major developments occurred on the Hasanlu mound: the building
of fortification walls and large structures, and an expansion of new pottery
forms (Dyson 1965, pp. 197–199, 211; Dyson 1968a, pp. 31–32; Young 1965, p. 82;
1967, p. 24).
Evidence for Iron II architecture was found in several areas of the mound: in
the main cemetery area and in squares G9a–c.
The architecture in the main cemetery area consisted of three kilns and
fragments of walls and rooms. The walls were much destroyed by stone
gathering and burial activities and therefore no complete structure was
preserved.
In square B9b one section of a wall was found under β17 (Dinkha II period,
Figure 2), a fact that established the existence and abandonment of some
structure here before the burial was deposited.
In square B9a in stratum 2, near the surface, a kiln was excavated (see
below). It had evidently been cut into a stratum in which there were at
least two structures; one, at the northeastern corner of the trench, enclosed
by Walls A and B, the other at the northwestern part called Area 1. In the
southwest area of the trench, and a level or two below these structures,
was a single wall with a threshold preserved, called Wall C. It had been
cut into by β4, which in section was seen to be partly under some stones
from Wall B. Urn β5 was under the room area formed by Walls A and B,
and urn β6 was under the kiln. The sequence here would appear to be:
Wall C, followed by β4, and 5; then Walls A and B, and Area 1, and finally the
kiln. Burials were found in all the strata of B9a, but it is not clear just what
the relationship of the structures was to the use of the area as a cemetery,
chronologically and culturally. It may be that structures not considered
convenient for inclusion in a settlement area were built in the cemetery
196 chapter six
area. Fred Matson has suggested that the kilns were built here because they
would have represented a fire hazard if they had been near a residential
section (compare Stein 1940, p. 394, for a kiln near the Hasanlu cemetery;
compare also the biblical Potter’s Field, and the Athenian Kerameikos).
The kiln of stratum 2 was roughly oval in plan with a hard earth floor, over
which was an ash layer, and a wall of vertical bricks set on edge (Figure 21);
its entrance faced south. Within the chamber, originally domed, was a firing
unit touching the east wall: a N-S wall of bricks (about .40 × .35 × .12 m.),
made up of two rows of four bricks each laid flat and with an upright at each
end, that abutted a small chamber of two upright bricks supporting a brick
and a half, which forms the roof. Nothing was found inside the kiln.
the iron age at dinkha tepe, iran 197
Figure 22.
Square B10a had two architectural features, a kiln and the remains of an
interesting building called Structure A (Figure 22). The axis of Structure A
was roughly N-S. Its S wall, A, extant length 5.5m., consisted of large outer
stones filled with smaller ones; the eastern most preserved stone was a
threshold. To the south of Wall A was a pebble pavement, partly preserved,
which was perhaps an outside area. On this pavement were found two buff
IA pots (on the plan 16: 109D, 18a: 104D). Abutting Wall A at the north was
a rubble packing in an L-shape that was obviously a stairway support. One
meter further to the north of the stairway was a series of stones set on edge
bordering a stone paved area, .35 to .50m. in width, that in turn bordered
another, wider, paved area of larger stones 1.30m. in width. Another wall, B,
bordered both the paved unit and the stairway at the east and joined Wall A;
thus it separated the stairway-paved area from a room to the east that made
use of the threshold of Wall A. The paved area seems to be a unit consisting
of a jube (a water channel) set next to a narrow pavement, placed within
a room—or court—that also contained a stairway to a second story. The
juxtaposition of jubes and pavements exists at Hasanlu IV in the area just to
the west of the fortification walls, in the northwest quadrant of the citadel
area.
198 chapter six
This structure was built and abandoned before the kiln was constructed.
It seems also that some urn burials were deposited in the area after the
abandonment. Stone tomb β15 is partly under the stone pavement at the
south of the structure, but we cannot be certain that the tomb was earlier: it
may have been later and the burial pit could have undercut the remains of
the pavement.
The kiln was dug into the fill of level 2, just below the top soil (Figure 23).
In plan it was a rough oval built of clay and apparently originally domed.
Its entrance on one of the long sides faced southwest. Within the chamber
were two units, a lower chamber for firing, and an upper one for the pots,
both now collapsed. In the center of the lower chamber was a pillar of three
bricks with a single brick on end touching them; this helped to support the
upper chamber. The floor of this upper chamber consisted of large bricks or
slabs, one of which was found on edge, having slipped. Holes in the floor of
the upper chamber were made to carry the heat to the pots.
Within the chambers was found a broken buff IA jar with three nipples
on each side (107D), and inside the jar was a fragment of a plain bronze ring.
On top of the collapse was a broken, buff spouted vessel (13T).
the iron age at dinkha tepe, iran 199
In square B7f, a small L-shaped trench, a third kiln was partly excavated.
It too was very close to the surface and was exactly like the kiln in B9a, with
a brick wall and similar firing chamber; it opened to the north, unlike the
other two kilns, which opened to the south.
In squares G9a, b, and c, we excavated the poorly preserved remains of a
large building; recent stone gathering and burial activity had badly denuded
this part of the mound. The building as preserved consists of two rectangular
rooms bordered at either end by smaller squarish rooms (Figure 24). The
walls, 1.15m. wide, are made of large stones on the outside faces with smaller
stones used as filler, similar to the construction of Structure A in B10a. The
brick superstructure was no longer extant but was made of sun-dried clay
bricks, to judge by the wash adjoining the walls. The two excavated squarish
200 chapter six
rooms are the northern limits of the building and they are of uneven size;
the easternmost one is about 3.60 × 3.75m. (N-S × E-W); the western is about
3.60×3.20m. There is definite evidence for the existence of a third room to
the west, but very little of it has been excavated.
To the south a rectangular room about 7.40 × 3.70 m. was cleared; this is
the eastern limit of the building. To its south is a partially cleared area that
probably represents a squarish room, balancing the one to the north. The
total excavated length of the eastern facade is about eighteen m. To the
west of the rectangular room is a partially cleared room that could be either
another rectangular room, approximate in size to its neighbor, or a larger
central room or hall. In the latter case we would expect a balancing rectan-
gular room to the west, in the former case we would expect another rectan-
gular room further west. Thus one could conceive a plan that included three
rectangular rooms, bordered at north and south by smaller rooms, or a cen-
tral hall bordered east and west by rectangular rooms, all bordered by side
rooms. Unfortunately, too little was excavated to carry speculation further.
Exterior and interior doors are no longer extant but surely they must have
existed. Floors were hard-packed earth, and no artifacts other than Iron II
sherds were recovered; there was no evidence of burning.
Of special interest are the two flat stones, about 50 × 30 cm., preserved in
situ set into the floor of the easternmost large room. The northern stone is
about 75cm. from the northern wall, and about 13 cm. from the eastern wall;
the southern one is about 1.05m. from the southern wall and about 13 cm.
from the eastern wall. These stones clearly appear to be bases for now lost
wooden posts. How many other bases originally existed in between the two
extant ones is not certain, but there could not have been more than four
or five stones as a total number. Was the whole room filled with “columns”
at one time? It would seem from the narrow width that the answer is no,
although this idea cannot be categorically ruled out. However, it seems
easier to visualize a room with a series of posts set around the perimeter,
posts that may have held a balcony. In this respect one may make a formal
comparison to the posts in the Burnt Buildings at Hasanlu, there set flush
against the walls (Young 1966, figs. 1, 2), but nevertheless probably serving
the same function. Perhaps we may call the Dinkha building a manor, in
the same sense that Claire Goff called the building excavated at Baba Jan
in Luristan a manor. This building, slightly later in date than ours, had a
columned rectangular room about twice the width as the room at Dinkha
(Goff Meade 1968, pp. 112–115, figs. 4, 5; 1969, pp. 117–122, figs. 2–4). The Baba
Jan manor’s rectangular rooms were also flanked by smaller side rooms, and
in plan is not altogether dissimilar to the Dinkha manor.
the iron age at dinkha tepe, iran 201
We may conclude that the Dinkha manor was more than eighteen meters
north to south, and more than twelve meters east to west; that it contained
at least nine rooms, that the walls were thick enough to hold a second story,
and that one of the long rooms had posts or columns. In short, there is
evidence for a major building in the Iron II period at Dinkha.
Burials
multiple burials. The two other burials were in stone tombs and contained
adults (B8a, β1, B8e, β5I).
As in the earlier period, men, women, and children were buried in the
same cemetery area, with no apparent difference in funeral rites or treat-
ment of the corpse recognized with regard to age and sex—except that
infants were sometimes buried in urns.
In two burials of old adults arthritic lipping of the vertebrae was noted
(B10b, β7, β8), and in one burial, that of a child, a partially healed hole in the
skull was detected (B10b, β3). One burial consisted of disarrayed bones and
seems to represent a secondary burial (B10a, β13).
Burials were recognized as Dinkha II–Iron II in date sometimes by depth,
more often by the nature of the contents. This often consisted of a bridged
spouted vessel (both with and without handles, and with a “beard” project-
ing from the base of the spout), or a hydria (a medium-sized storage or water
vessel with three handles). In addition to these classic shapes, the various
jars, cups, carinated bowls, deep bowls with animal-head handles, and many
metal objects, jewelry and weapons, many made of iron, and all well known
to us from Hasanlu IV, made attribution fairly easy.
Thirteen of the burials did not have a spouted vessel, but in about eight
of these attribution to Dinkha II could be made on the basis of other shapes.
As was the case with Dinkha III burials, both complete and damaged pottery
were considered as possessions adequate for the dead.
Gray and buff pottery continued to be used side by side. In this period,
however, buff pottery predominated. The total number of vessels from
Period II was two hundred and fifty-two: two hundred and twenty-nine from
the burials, nineteen urns, two from the kiln, and two from Structure A. Of
these, sixty-seven are gray and one hundred and eighty-five are buff (nine-
teen of these are the urns): the percentage of gray to buff is therefore about
27 to 73.
Among the gray pottery, burnished and smoothed surfaces are even,
twenty-eight recorded for each, four are matt, and six were not recorded by
surface treatment. Among the buff pottery fifty-five are smoothed, twenty-
four burnished, sixty-nine (counting the urns) are matt, and ten are red-
slipped; the rest were not recorded by surface treatment. (Thus, as in
Period II, gray vessels were more likely to have been burnished than buff
vessels.) In color, sixty-seven vessels are orange, six are red-orange, eleven
are red; the rest were simply listed as buff.
In both the gray and buff pottery, common-ware paste, with few or no
inclusions visible, predominated about two to one over medium-sized grit.
Only two vessels were recorded as having mica flecks (Mica flecks exist not
the iron age at dinkha tepe, iran 203
only in Dinkha IV and III, but also on the Iron III pottery from nearby Agrab
Tepe, Muscarella 1973, p. 65).
The number of vessels associated with a burial varied from none (usually
incompletely excavated burials) to twenty-six, the majority having four or
five (Table II), and there appeared to be no special relationship between
type of tomb, and age and sex, to numbers of vessels or grave goods. The
only notable exception was that most of the stone tombs, but not all, were
among the richest of the burials.
Thirty-seven burials contained some form of jewelry; torques were found
in four burials. Seven burials contained weapons and only one burial con-
tained horse bits. Jewelry and weapons were made from both iron and
bronze, but the latter clearly predominates. A count14 of the available inven-
tory yields the fact that there are about one hundred and seventy-two bronze
pieces of jewelry and eighty-one of iron, and among the weapons there are
sixteen made of iron and three of bronze. Sheep/goat bones were commonly
found in the burials, and it is possible that liquids were placed in some of the
closed vessels.
We have seen that it was possible to isolate a few Dinkha III burials as hav-
ing been deposited at an earlier stage than other burials of the same period.
In a few cases this differentiation was also noted among the later Dinkha II
burials. However, in these examples the distinction was suspected primar-
ily on the basis of relative depth and does not seem to have independent
support on typological grounds. (Unidentified so far from both Hasanlu and
Dinkha is a “transition” grave from Iron I to II.) These possible early graves
include B9a, β9; B9b, β19; B10a, β16; and B10b, β11:
b9a, β9: Male, mature adult, flexed on back, E-W, head W; arms at sides
touching pelvis; in brick tomb (Figure 25). Furniture (Figure 26): a plain
round iron penannular bracelet, broken (415P), and a bronze corrugated
band type with overlapping ends (820T) on R wrist; an iron and a bronze
penannular ring, both plain (413P), and two plain iron ones, broken (419P),
all on R hand; a plain corroded bronze pin (383T); also a necklace of car-
nelian, paste, and Egyptian blue beads at the neck (389P). A bronze spear
with short, ovate blade was placed point up along the left side of the head so
that the shaft crossed over the body (221P); and an antler ax with remains of
the wood shaft along with a bronze and iron stud in situ (1042T) was placed
14 It was not possible to give the absolute number of metal objects; some had disintegrated
and were not given catalog numbers, and in some cases “rings” in the inventory were given
one number, while I counted them as two objects.
204 chapter six
Figure 26.
Figure 27.
206 chapter six
Figure 28.
next to the spear. A broken dark gray IIC spouted vessel, with a horned ani-
mal in relief on both sides (335T), resting on a stone, and a broken carinated
orange IIA jar (904D), were found on the R side; a sharply carinated orange
burnished bowl (870T) was at the head, and two jars one, buff IA (173D), the
other orange IIB (252P), one with a sherd over its mouth, were placed at the
R shoulder.
b9b, β19: Adult, flexed on L side, N-S, head N; in brick tomb. Furniture
(Figure 27): two bronze pins with decorated grooved tops at the shoulder
area (375P, 382P) a third on the chest (374T); two bronze earrings consisting
of a large plain loop with connected hooked ends attached to a smaller loop
(1009M, 1010T), a plain bronze ring with overlapping ends (352D), in the fill; a
group of beads (394T), and a plain bronze torque with bent ends, at the neck
(1039P). At the feet was an orange IIB spouted vessel with an animal-head
handle (238M, Muscarella 1968, p. 190, fig. 9), and at the back was a red-
orange IIB bowl with flaring sides and two holes set within grooves (236T),
and an orange IIB jar (422T).
b10a, β16: Infant, flexed on L side, most of the bones missing; N-S, head N;
in brick tomb partly destroyed by B10a, β6 (see below). Furniture (Figure 28):
two round, coiled, and twisted bronze bracelets, very corroded (181T, 203T);
a simple bronze torque on the neck (187T), along with some beads (439P).
At the feet were a gray IIB basket-handled teapot with mica flecks, in fact,
probably a milk bottle (224M, Muscarella 1968, p. 190, fig. 8), a gray IIB tripod
bowl, with one foot missing (230T), and a small gray IIB jar, also with mica
flecks (255D). (Note, in Muscarella 1968, p. 191, fig. 12, the caption should read
that the vessels came from B10b, β16, not B10a.)
the iron age at dinkha tepe, iran 207
b10b, β11: Female, mature adult, extended on the back with the arms
flexed across the chest; N-S, head N; in brick tomb (Figure 29). Furniture:
a plain, blunttopped, bronze pin at L shoulder (142P) and R (155T) (compare
Figure 7, B9a, β26, 607, Dinkha III); a bronze pin with a hooked end at the R
(154P) and a plain bronze pin (619T: like Figure 16, B10b, β10, 137, Dinkha III,
and Figure 45, B8a, β1, 709, Dinkha II) over R shoulder; and similar pins in
iron, one at R shoulder (444T) and one at L (445P), for a total of six pins. A
flat iron ring with three grooves, on L hand (147P), an iron archer’s ring on
the R (427P; compare Figure 35, B10a, β6, 195); a necklace of stone (carnelian,
jasper) round beads, others round “frit,” paste, “glass,” and copper (827T); one
is a lentoid antimony bead (437P). In the northeast corner, an orange IIB jar
(257P), by the hip a gray IIB one (179D).
At first I was inclined to place this burial in the Dinkha III period, mainly
because there are no other extended burials in the II period, and because
the pottery was not distinctive. But Stein (1940, p. 400) excavated at Hasanlu
an extended burial of Period IV date. Moreover, iron does not exist in any
bona fide Dinkha III burial, and archer’s rings (iron) occur only in bona
fide Dinkha II burials, and in Hasanlu IV. I therefore believe that this burial
belongs to Period II.
208 chapter six
Finally, there were two burials that were found high in the fill and may
be considered to be later than most of the others (B10a, β1, B10b, β2). One is
published herewith:
b10b, β1: Young adult, inhumation, flexed on R side; E-W, head E; R arm
extended, L arm touches knees (Figure 30). Furniture: a gray burnished
spouted vessel by chest (845T); another spouted vessel, gray IIB, with the
spout broken, and with three nipples in an inverse triangle on each side
and two vertically placed nipples on the back (191D), at the head; and a
gray IIB bottlelike jar (25T) between the other two vessels. There was no
jewelry.
the iron age at dinkha tepe, iran 209
We now return to the other burials of this period, i.e., those not distin-
guished by stratigraphy as either early or late in deposition. Since this period
is relatively well known from the Hasanlu excavations, only some of the buri-
als need to be presented here. They are representative of the whole corpus:
b9a, β14: Mature adult, flexed on back, N-S, head S; in damaged brick tomb
(Figure 31). Furniture (Figure 32): three plain penannular bronze armlets
on R arm (306T, 365T, 372P); four bronze penannular rings by hands (463T,
594P); a bronze torque with curled ends at neck (1040T); a stone button
(591T); a necklace of plain round paste beads (327T). Beneath the skull, sixty-
six astragals and two bronze buttons (1005T). At the feet, a gray bowl (D)
placed under an orange IIB spouted vessel (336T); a few feet from the face,
an orange IIB vessel with two animal-head lugs (401T). Sheep/goat bones at
the knees.
b9b, β13: A stone tomb (Figures 33, 34, top): only a few bones extant;
apparently N-S. Furniture: outside the entrance slab, an orange IIB jar (89T).
Inside, two plain bronze bracelets with overlapping ends (370T), a red-
slipped spouted vessel (850P), an orange IA jar (I70D), a gray matt carinated
bowl (874D), and a coarse, disintegrated vessel. This was the poorest of the
stone tombs.
b10a, β6: A stone tomb with hard-packed earth floor (Figure 35; Mus-
carella 1968, p. 191, figs. 13, 14). The few bones suggest a N-S orientation of an
adult. Furniture (Figures 36, 37): fifty-three objects; this was one of the rich-
est burials in the cemetery. Objects probably belonged to a male warrior. Fif-
teen vessels (not sixteen as Muscarella 1968, p. 189) were outside the closing
slab: buff: 81D, a fragment of a wide-mouthed pot with an oblique spout and
no handle; also a disintegrated vessel; buff, IIB: 29D, a broken carinated jar;
332D, a jar; orange IIB: 31P, a carinated jar; 32P, a sharply carinated jar with
incisions on the upper body; 33D, a brokenjar; 34T, a spouted vessel; 35T,
a sharply carinated deep bowl with a handle, now missing; 48T, a spouted
vessel; red-orange IIC: 40T, a gadroonedjar; red-orange IC: 423P, a spout-
ed vessel; red-slipped: 30T, an asymmetrical jar; gray IC: 46T, a vertically-
bridged spouted vessel, with broken spout; gray IIB: 28P, a gadrooned jar.
Placed among these vessels was the dismembered, incomplete skeleton of
a horse: skull, mandible, humerus, a pair of radii, two cannon bones, two
femera, one tibia, and a third cannon bone. Just outside the tomb, by the
northwest corner, was an iron socketed spear (118P, visible in Muscarella
1968, p. 121, fig. I4). Just inside the entrance of the tomb were five vessels: an
orange matt hydria (248D), two buff carinated bowls (858P, 910D), a small
buff IA jar (169D), and a gray IIB spouted vessel (239T). Two bronze penan-
nular anklets (761P, 762T) at the southern part of the tomb give the feet
210 chapter six
Figure 32.
the iron age at dinkha tepe, iran 211
Figure 34.
the iron age at dinkha tepe, iran 213
Figure 36.
the iron age at dinkha tepe, iran 215
Figure 37.
216 chapter six
Figure 39.
the iron age at dinkha tepe, iran 217
Figure 43.
the iron age at dinkha tepe, iran 219
Figure 44.
220 chapter six
Figure 45.
the iron age at dinkha tepe, iran 221
position. With the pottery at the entrance was a corroded iron point (D),
a corroded iron object with a rounded head and spike (182D), apparently
a mace head with a solid head; also an iron blade (151P), an iron pin with
traces of eight layers of cloth (706P), a thick-knobbed iron pin, also with
traces of cloth (1031T), a plain round penannular bronze bracelet (129T),
and remains of a bronze and iron chain (1054D, compare Figure 44, B8a, β1).
Along the western wall, N-S, were a bronze spiked or star mace head (119P), a
plain bronze bowl (114P), an elaborately decorated flat-band bronze bracelet
(112T), and a plain concave-sided band bronze bracelet with overlapping
ends (113T), along with two plain round iron bracelets with overlapping ends
(417P), plus a plain broken round iron (134T) and a plain penannular iron
bracelet (120T), for a total of six. There were also clusters of plain iron (426T)
and clusters of penannular bronze rings (593D), a bronze, two-piece, jointed,
horse bit with a solid ring (1026P), and an iron fragment of another horse bit
(69T), an iron shaft-hole ax fragment (1033D), a broken iron archer’s ring
(195T), a bronze needle (470D), a bronze boss (150P), a bronze stud (1007T),
a limestone disc (64P), and a pin consisting of an iron hooked-top set into
a bone button and attached to a reed, with traces of thread (755T, compare
Figure 47, 756, 757), two bone awls (222T, 223P).
There were also many beads (994T): carnelian—a, n-r; paste—st; “chalky
material”—s; amber—b; “glass”—e-g (the latter blue and yellow)—u;
cowrie shell—h; Egyptian blue—i–l; cast antimony—c; bronze—m.
b10a, β12: Male, flexed on R side, N-S, head S; in brick tomb (Figure 38).
Furniture (Figure 39): a plain round bronze and a plain round iron bracelet,
both with tapered overlapping ends (123P, 124T), by wrists; carnelian, “frit,”
and paste beads (117D); a beaded cast bronze torque with hooked ends, at
neck (115T); a bronze socketed spear resting along L side of head, point up
(125T); shaft would have rested along side of the body. By the knees, two
gray burnished spouted vessels (835T, 846P); by thighs, an orange smoothed
carinated bowl (873D); by feet, an orange burnished carinated jar (965D).
b10a, β13: Female, adult, inhumation. Bones were found disarticulated,
probably representing a secondary burial (Figure 40). Furniture: a bronze
stud (925P) was found inside the skull cavity; bronze hemispherical beads
with a loop, corroded together in sets of three (923P), and carnelian and frit
plain beads, all in the fill (436T); a simple iron ring was under the skull (130T).
Also under the skull was an orange IIA jar (98P); other vessels included a gray
IIC spouted vessel (261T), an orange IIA, and two buff IIA jars (192D, 193D,
94P). Sheep/goat bones.
b10a, β15: Stone tomb; part of the floor covered with stone slabs (Fig-
ures 41, 42). Bones disintegrated; N-S orientation, head N. Furniture: three
the iron age at dinkha tepe, iran 223
splayed pommel and a straight grip, and with wood fragments of sheathing
and hilt insets evident (1046T); a bronze chain (1034P) next to a large iron
object (found exploded) that might have been a staffor baton—it seems
to be too big for a pin, which it resembles (1032D); a corroded iron point,
possibly a large pin, with a bronze chain (1035T); a bone cosmetic container
open at both ends and decorated with incised circles with a dot, empty
(1047P); and under bowl 872 a handful of beads (1049T): a, b—amber; c-
f—colored glass (yellow, brown, white, and black); g—glazed material; also
beads of Egyptian blue and carnelian. In the fill of the chamber were found
two bronze needles (1017P); four plain bronze penannular rings (1018T), and
a fragment of a bronze coil (1002P). A total often vessels were placed in the
tomb, one of which had disintegrated: a gray burnished tripod bowl (894P),
a gray burnished hydria (917T), a gray matt jar (824D); a burnished orange
carinated bowl (872P), and buff matt vessels: 821P, 822D, 825T, 895P; 841P is
burnished.
b8e, β5: A stone tomb with stone floor, containing the scanty remains of
two individuals, male and female (I). Body positions, not clear, seem to have
been N-S. Outside the tomb’s western entrance or closing slab was a large
pile of pottery partly covered by a broken pithos (Figure 46). Here were
twenty-three vessels and under them were two skeletons, a female young
adult and a child (II), separated from the stone tomb by a mud brick wall
of one course. Wall ran N-S, paralleling the stone tomb, interrupted where
it touched the western wall; its total excavated length (measuring the area
occupied by the stone tomb section) 2.40 m. Both ends continue into the
unexcavated balks, so we do not know the total length. The female and child
were placed head to feet in a line, separated by about 45 cm.; they were flexed
on their R sides, N-S, heads S, facing the wall, E. Contents of stone tomb (I)
(Figure 47): there were five vessels: two bowls, one gray matt (863P), one
gray burnished (864T); one gray IIB spouted vessel (278M); one large buff IIB
jar with three nipples in triangular form on the sides (1055D); and one red-
slipped broken jar with narrow neck and two handles (806D). Body furniture
(Figure 48): about twenty plain bronze and iron rings (like 489D, 218D); a
spiral bronze ring (490D), and a flat-band iron ring attached to a round one
(188P); four figure-eight hairrings (219P, 488D, 1025D), a bronze coil (769T);
four bronze penannular bracelets (432P, 455P, 456P); two iron penannular
bracelets or anklets (220P, 710T); one iron needle (1021D). Also, three iron
pins with looped heads (486T, 714T, 715P); an iron pin with a knob head
(416P), and two iron pins with ribbed heads (699D, 700T); three plain iron
pins with blunt heads (407T, 409P, 410P); also two iron-reed pin-hooks, with
a bone collar (756T, 757P, of the same type as Figure 36, 755). There were also
the iron age at dinkha tepe, iran 225
Figure 47.
226 chapter six
Figure 48.
the iron age at dinkha tepe, iran 227
two iron archer’s rings (483T, 485P); two socketed iron spear heads (650T,
702D broken); and an iron knife with a curved tip (704P). Each of the two
outside skeletons had associated grave goods. The northernmost one, the
child, had a pottery jar that disintegrated, an iron bracelet with overlapping
ends (412T), two iron bracelets of the same type found together (625P), and
a plain bronze bracelet (318T). The female had two buff IA bowls (865P,
875D) containing sheep/goat bones; also two bronze penannular bracelets
on one arm (367T), and one on the other arm (368P); also a dark-stone,
pear-shaped mace head (1019P). Vessels found outside the stone tomb over
the skeletons of burial II (Figure 47): gray IIC: spouted vessels (403P, 848P,
845T, 857D, 849P, 333T), bottle (790P), cup (805P); gray weathered: hydria
(919D); gray IE: carinated jar (963D); orange IIC: spouted vessel (840P); buff:
hydria (912D); buff IB: hydria (911D); buff IIA: jar (964D); buff IA: jars (812T,
1058D), cups (800D, 807D), carinated bowl (876P); red-slipped: jar (809D).
A few objects were inadvertently not recorded as specifically coming from
the burials in I or II and are listed here together: an orange jar (262D), a buff
carinated bowl (D), and a disintegrated vessel. Also, a bronze tack or stud
(707T), a stone ax or pestle (1056D), a bronze coil (769T), an obsidian blade
(705T), and many beads: 397P—paste; 713P—bronze; 708T—glass; 997aP—
shell; c, j—paste; g, a spacer bead—paste; h—stone; i—bone or shell; 998—
paste.
The problem of the relationship between the stone tomb (I) and the
burials outside (II) remains to be discussed. What is clear is that the twenty-
three vessels were placed partly over both the west wall of the tomb and the
brick wall, and the skeletons of burial II. Therefore, both burials I and II were
in place and were exposed at the time when the pottery was deposited as a
final act. Yet, what is unclear is whether or not one of the burials was already
in existence before the second was deposited, i.e., whether the diggers of
the second burial pit inadvertently disturbed the earlier burial, or, whether
both burials were deposited simultaneously, with different treatment given
the respective bodies. If we prefer the first suggestion we can assume that
the burials in II existed first, and that it was accidentally encountered by the
stone tomb builders, who, upon completion of their funeral tasks, piled the
many vessels over both burials as a pious gesture. If we prefer the second
suggestion we must assume a unique occurrence at Dinkha: the fact that
at one time four people were buried, two in a closed stone tomb, and two
outside. It should be noted that the brick wall was only one course high and
its length very long, features not encountered with typical brick tombs at
Dinkha. Moreover, the first suggestion implies that when the earlier burial
was encountered, instead of recovering it and going elsewhere, the stone
228 chapter six
Urn Burials
Nineteen urn burials were excavated at Dinkha Tepe. None of these could be
attributed to the Dinkha III period either by low position or by grave goods,
but it is not impossible that a very few might have belonged to that time.
Most, if not necessarily all, were obviously laid down in the Iron II period
as they were usually found high in the Dinkha II fill; in a few cases they had
characteristic pottery associated with the urn.
The burial urns were either buff matt storage vessels or large cooking pots,
with both wide and narrow mouths (Figure 49: 108D, 111D, 284D). One urn
was blackened on the outside and inside, no doubt from use; often the urns
were broken or incomplete. Two urns were buff hydriai, and in another case
the top of a large pithos was used as an urn; often the mouths of the urns
were covered with large sherds. Urns were usually placed on their sides, but
a few were found upright, or upside down. Those on their side were oriented
N-S or E-W, following the same pattern practiced in the burials. In most
instances few or no bones were recovered from the earth fill inside the urns.
When the bones were recognizable they were usually those of infants, but
in one case (B9b, β2), an adult tooth was found in an urn (compare Stein
1940, p. 374; see also pp. 397, 400). Seven urns had pottery placed outside,
and four, two of them with gifts outside the urn also, had pottery and jewelry
inside.
A sampling of the urn burials:
b9a, β3: The buff urn (D), tilted up, was covered with a broken buff matt
bowl (909D) (Figure 50). Close to the mouth of the urn were a red-orange
IIC spouted vessel decorated with a crescent and two nipples on both sides
(259T), two small gray IIB jars (79T, 80T), and a large buff jar (102D). Inside
the urn were a plain bronze band penannular bracelet (Figure 52, 620P),
two plain bronze (464T, 467P), and one iron ring (425D), thirty-nine bronze
(386T), stone, paste, and shell beads (376P), and a clay button (D). This was
the “richest” urn burial excavated.
the iron age at dinkha tepe, iran 229
b10a, β2: (Figure 49): The buff IA urn (108D) was lying on its side with the
mouth to SE; the mouth was broken away. Outside was an orange IIB two-
handled jar or flask that had a short upright spout (10T).
b10a, β3: (Figure 49): The buff IA urn (284D) was placed with the mouth
up and sealed by sherds. An infant’s tooth was found inside. Near the urn
was a red-slipped carinated jar (905D).
b10a, β5: (Figures 49, 51): The buff urn (D) was lying on its side, roughly
E-W, mouth E; a large sherd sealed the mouth. To one side were an orange
IIB spouted vessel (38T), an orange IIB carinated jar (36T), and an orange-
brown IB miniature asymmetrical jar (15T).
230 chapter six
test trench 111, β1: (Figure 49): the buff-yellow IB (281D) urn was on its
side, facing NW; it contained infant’s bones and a bronze ring (D); outside
was a gray jar (D) and an orange IIB footed bowl with a hole below the rim
(231P).
Dinkha III–II
Much has been written about the Iron II period in western Iran, so it would
serve no useful function to repeat that information here except in those
instances where it relates to Dinkha Tepe. That there was cultural continuity
in the Iron Age, that Iron II followed Iron I peacefully and without any
observable interruption from outside forces, is well supported by the Dinkha
excavation. The most obvious evidence is that the Dinkha II burials were
deposited in the same cemetery area as the earlier ones, that both simple
232 chapter six
inhumation and the use of brick tombs continued, that the same body
positions and orientations continued, and that earlier customs, such as the
extended burial and the placement of an arm bent back, touching its own
shoulder, were not forgotten, although rarely practiced. In addition, the
same types of grave goods, pottery, jewelry, weapons, and food continued
to be placed with the dead. Not least in significance is the fact that the
same buff and gray pottery were produced in both periods. But it is of some
interest to note that whereas in Dinkha III gray vessels were more than twice
as common as the buff wares, in Dinkha II the ratio is strikingly reversed and
buff vessels were nearly three times as common as gray ones (see pp. 38, 59).
Obviously, as is to be expected with a dynamic culture that existed for such
a long time, and as is the case at other Iron Age sites, the pottery shapes
changed—the worm bowl and pedestal-base goblet disappear—and the
variety of shapes increased. Yet even within the changing pottery repertory
we are able to observe a continuity between Dinkha III and II: the ubiquitous
spouted vessels, the basket-handled teapot, the carinated bowls, and many
of the jar types.
Only one vessel from the corpus of Dinkha Iron II pottery might be singled
out as a possible import: the gourd-shaped red-slipped vessel from B10a,
β15. Three other vessels of this type, two red and one gray, were found in
Hasanlu IV (Rad, Hakemi 1950, pp. 59–60). Another, exactly the same in all
details, and also red, was seen in the Rezaiyeh market by Kleiss, who related
it, incorrectly, I think, to seventh century Urartian ceramics (Kleiss 1971, p. 71,
fig. 22, pl. 10:3, left). If these vessels were not locally made at Hasanlu or
Dinkha, we do not yet know their source.
Among the similar kinds of jewelry placed in the graves of both periods,
bracelets, anklets, rings, necklaces, pins, we might single out torques for
special mention. Of thirty-four Period III burials, seven contained a torque
(see note 5); of fifty Period II burials (not counting urns), four contained
torques, a drop from about one-fifth to about one-twelfth. Thus, although
still used, fewer people wore them—at least to their graves. It seems that
men, women, and children wore torques, although the evidence for this is
clearer in the earlier burials.
There is one example of a disarticulated burial (B10a, β13) in Period II,
none in the earlier period, but whether this is culturally significant or merely
an occurrence reflecting a local situation is not known. Also, as mentioned
above, in Period II four burials contained two individuals each, a feature not
encountered in the earlier period.
And one burial in Period II (B10a, β6) was associated with a fragmen-
tary skeleton of a horse placed outside the tomb. Within the tomb, it will
the iron age at dinkha tepe, iran 233
be recalled, was found a complete bronze horse bit and an iron fragment of
another. No other burial at Dinkha yielded either a horse bit or a horse skele-
ton. It is therefore not rash to conclude that the horse probably belonged
to the occupant of the tomb and was dispatched as part of the burial cere-
mony. But where were the other bones of the skeleton? Does the fragmen-
tary nature of the skeleton suggest that the horse was in fact an ordinary
animal merely meant for the funeral feast, and that the bones by the tomb
represent the dead man’s share? It seems to me that the juxtaposition of the
horse bit within the tomb and the horse skeleton outside is not fortuitous,
and that the horse did have some special relationship to the tomb’s occu-
pant. At the same time it would appear that the horse was eaten and that
some joints were kept for the mourners. I suggest, therefore, that the horse
belonged to the occupant of the tomb and also that the survivors ate it: both
ideas need not be mutually exclusive. Keeping this in mind, I believe that we
may correctly refer to the existence of a horse burial at Dinkha, as opposed
to the idea that the bones represent simply a food deposit. What remains
puzzling is the uniqueness of the occurrence of horse bones at Dinkha, even
if one disagrees with the conclusion presented here and believes the bones
are food.15
Horse burials associated with a human burial occur at Hasanlu in a
unique grave excavated in 1947; to date no other example has been found
even though the cemetery area has been extensively excavated (Ghirshman
1964, pp. 24–27, 99, fig. 131; Dyson 1965, pp. 208–212). Unfortunately, the
contents of this grave have never been identified and published and the date
is not known, which makes it impossible to bring it into a discussion of Iron
Age horse burials (Muscarella 1968, p. 192).16
15 S. Piggott, “Heads and Hoofs,” Antiquity 3 (1962) pp. 110–118, summarizes information
concerned with the burial of a horse’s head and feet, presumably along with the hide. The
Dinkha horse burial is of a different type. The description of Scythian horse sacrifices given
by Herodotus, IV: 62, 72, does not reflect light on the customs at Dinkha.
16 When discussing the horse burial at Hasanlu, Dyson (1965, p. 211, and also Young 1967,
p. 33) stated that a Scythian-like cheek piece found at Hasanlu came from Period IV (see
also Dyson 1964c, p. 372, fig. 3). This is an error; the piece came from a Period III context.
The error was repeated by M. van Loon in JNES 29 (1970), p. 69, and by P.R.S. Moorey 1971,
p. 109, and in Iran IX (1971) p. 121. Horse burials occur in the Hittite period, K. Bittel, Die
Hethitischen Grabfunde von Osmankayesi (Berlin, 1958) (only skulls and leg bones pp. 16, 24,
63, 65, 72, 73), and in the Mycenaean period, E. Vermeule, Greece in the Bronze Age (Chicago,
1965) pp. 298–299, P. Kabbadias, Proistoriki Archaiologia (Athens, 1909) p. 290; A.J.B. Wace,
“Chamber Tombs at Mycenae,” Archaeologia 87 (1932) p. 14; AA 1930 p. 170 for a buried
terracotta horse; see also the Iliad XXIII, line 170 et seq. Horse burials also occur in Gaza, in the
second quarter of the second millennium bc, F. Petrie, Ancient Gaza I (London, 1931) pp. 4–5,
pl. lvii: Hyksos? Horse burials occur in Phrygia at Gordion and Ankara, see R.S. Young “The
234 chapter six
Nomadic Impact: Gordion,” Dark Ages and Nomads (Istanbul, 1964) pp. 55–56, and T. Özgüç,
Belleten XI (1947) p. 80, all from post-destruction tombs. Of apparently contemporary date
are three horses buried in a stone-lined tomb at Norşuntepe, H. Hauptmann, “Norşun-tepe,
1970,” Anatolian Studies XXI (1971) p. 20. For horse burials in the Caucasus see F. Hančar, Das
Pferd in prähistorischer und frühen historischen Zeit (Munich, 1955) pp. 180ff. Horse burials
also occur in Cyprus in the eighth and seventh centuries bc, V. Karageorghis, Excavations in
the Necropolis of Salamis (Cyprus, 1967) Tombs 2, 3, 47. A. Hakemi, “Kalaruz,” Archaeologia
Viva 1 (1968) p. 65, mentions horse burials but gives no details. That such burials continued
in later times in Iran is attested by horse and human bones found together in a first century bc
tomb at Shahr-i Qumis excavated by John Hansman and David Stronach in 1967 (Journal of the
Royal Asiatic Society [1970] I, pp. 41–48). An account of horse burials in the Altai region is given
in S.I. Rudenko, Frozen Tombs from Siberia (University of California, 1970). The widespread
occurrence of horse burials over a wide chronological and geographical range indicates that
no one ethnic group had a monopoly on the practice.
the iron age at dinkha tepe, iran 235
of posts in the Dinkha II building, and the juxtaposition of jube and pave-
ment in Structure A, features shared by both cities, practically every pot-
tery shape used by the people at Dinkha II was used by the people of
Hasanlu IV: spouted vessels, plain, or decorated with crescents, nipples, or
animals in relief, or decorated with flutings, ridges, “crow’s feet,” or with an
animal head at the handle, and sometimes with a vertically bridged spout;
hydriae, which seem to occur only at Hasanlu and Dinkha; basket-handled
teapots; carinated bowls; deep bowls with animal-head protome handles;
the (imported?) pear-shaped gourd; the many jar types; knobbed vertical
loop handles; asymmetrical handmade bowls; gadrooned jars (Dyson 1964a,
figs. 118–121; 1965, fig. 13; Young 1965, figs. 6, 7; vanden Berghe 1959, figs. 144–
146; Stein 1940, fig. 109, pls. xxiv, xxx, xxxi; Rad, Hakemi 1950, pp. 59–60,
Burney, Lang 1972, p. 125).
Equally shared are the many metal and miscellaneous objects, such as
jewelry and weapons: pear-shaped stone mace heads and metal-spiked or
star maces are very common at Hasanlu (Rad, Hakemi 1950, fig. 78b; Dyson
1960, fig. on p. 128); so were iron knives with curved tips, and iron socketed
spears, found there by the hundreds (compare Moorey 1971, pp. 88–90;
compare no. 87, to my Figure 26, B9a, β9, 221). Bronze and iron archer’s
rings were excavated in many Hasanlu burials (Stein 1940, pl. xxv, 2; seven
iron examples were found at Dinkha) and two bone-antler axes (exactly
the same type as in Figure 26, B9a, β9, 1042) were found at Hasanlu, one
in a burial. To my knowledge only one dagger—actually the hilt alone is
preserved—exactly paralleling the sole example from Dinkha (Figure 45,
B8a, β1, 1046) comes from Hasanlu; but another similar example was also
found there (Dyson 1964a, p. 41, fig. 2:2, pl. ix, 2; see also Moorey 1971, pp. 70–
71).
Plain, jointed horse bits of the same type as Figure 36, B10a, β6, 1026, as
well as twisted and elaborate examples occur at Hasanlu in bronze and iron
(Ghirshman 1939, pl. cc: 17; 1964, fig. 338, left; compare also fig. 338, right).17
The iron and bronze chains from B10a, β6 and B8a, β1 (Figure 45, 1034,
1035, 1041) may have been originally attached to pins as was the case with the
many lion pins from Hasanlu (Dyson 1964c, p. 374, figs. 9, 12) and at Haftavan
(Burney 1970, fig. 7, middle). Chains occur at Hasanlu not only in connection
17 The number of horse bits found at Hasanlu gives evidence for the use of cavalry and
perhaps chariotry there, a fact corroborated by the scenes represented on ivories found at
the site, Muscarella 1966, figs. 11, 12 (but no bits are depicted on these horses); see also fig. 10.
It is not possible to make any comments about the extent of cavalry and chariotry at Dinkha
from the two examples—one a fragment—found in one tomb.
236 chapter six
with lion pins, but individually (although they too may have been connected
to other pins), and in a Period III context, attached to a fibula.18
It was mentioned before that torques were known at Dinkha III and II,
and that they occur in small quantities at Hasanlu IV. Bronze and iron plain
round rings, single and doubled, and flat-band rings are common at Hasanlu.
Several plain band bracelets, some with concave sides (Figure 36, B10a, β6,
113), and several examples of the elaborately incised band types (Figure 36,
B10a, β6, 112), were at home at Hasanlu. The dead at Hasanlu were also
dressed, and wore anklets, plain loop and figure-eight hairrings, and pins,
of exactly the same types as those from Dinkha; they were also furnished
with needles (Stein 1940, p. 401).19
Literally scores of thousands of beads of all typical materials, including
antimony, amber,20 and Egyptian blue, were found in the graves and on
the citadel at Hasanlu. Astragals, polished from use, were found at both
sites; Hasanlu produced some that were pierced. We may assume that the
same games were played at both cities, which is not surprising inasmuch
as knucklebone games have a long history in the ancient Near East, and
in modern history as well.21 Another type of bone object from Dinkha, the
18 Compare also M. Mallowan, Nimrud and Its Remains II (New York, 1966) p. 114, fig. 58.
19 Many of the figures on the Hasanlu ivories wear bracelets, and so do those represented
on the gold bowl; the nude female figure there also wears anklets.
20 It does not yet seem possible to be certain about specific proveniences of ancient
amber: Curt W. Beck, “Analysis and Proveniences of Minoan and Mycenaean Amber,” Greek
Roman and Byzantine Studies 7, 3 (1966) pp. 191–211.
21 Our workers always asked for discarded astragals from the ancient burials to give to
their children; see also C.L. Woolley in LAAA 26 (1930) p. 20, note 1, where it is reported
that astragals were placed in modern children’s graves. Their occurrence in ancient times
is widespread, as the following incomplete listing makes clear: in Iran they are reported,
besides those from Hasanlu and Dinkha, from Geoy Tepe, Burton-Brown 1951, p. 175, note 15,
pl. xxii, A Period; from Sialk B, Ghirshman 1939, p. 245, pl. lxxviii; from Ghalekuti, N. Egami,
et al., Dailaman I (Tokyo, 1965) pl. xlviii, no. 28; from early Susa, J. de Morgan, MMA en Iran
XXIX (Paris, 1943) pp. 46 ff. In Anatolia they occur early at both Hacilar and Çatal Hüyük,
J. Mellaart, “Anatolia Before 4000 BC,” CAH Fascicle 20 (1964) pp. 10, 14; at Troy, H. Schliemann
Ilios (New York, 1881) pp. 263, 426; at Alishar, E.F. Schmidt and H.H. von der Osten, in OIC XIX
(Chicago, 1932), p. 274, fig. 374; OIP XX (Chicago, 1933) pp. 82–83, fig. 129; OIP XXIX (Chicago,
1937) p. 433, fig. 488; OIP XXX (Chicago, 1937) pp. 105, 174–175, figs. 101, 196 (late); at Boǧazköy,
MDOG 72 (1933) p. 77, fig. 12; R.M. Boehmer, Die Kleinfunde von Bogazköy (Berlin, 1972) pp. 35,
181, 203. They were very common in Phrygian Gordion: a large vase filled with astragals was
found in one of the burnt buildings, and many were found on the floor of another, R.S. Young
in AJA 61 (1957) pp. 321, 327; 446 astragals were found in Tumulus P, a child’s tomb, ibid.,
p. 327. In North Syria they occur at Zincirli, F. von Luschan and W. Andrae, Die Kleinfunde von
Sendschirli V (Berlin, 1943) pp. 122–124, fig. 173, pl. 59: p, q; at Hama, P.J. Riis, Les Cimetières
à Crémation (Copenhagen, 1948) pp. 30, 35, fig. 22, p. 176; at Carchemish, C.L. Woolley, LAAA
26 (1939) pp. 20–21, 23 f.; note also the relief there with children playing an astragal game,
the iron age at dinkha tepe, iran 237
incised cosmetic container from B8a, β1 (Figure 45, 1047), has many relations
at Hasanlu (Stein 1940, pl. xxv, 6; Ghirshman 1939, pl. c, 24; see also Dyson
1964c, figs. 14–17).
In short, the two sites shared a common culture. This conclusion is not
contradicted by the fact that there were some differences between the
sites, some traits that were not shared in common. For example, burials at
Hasanlu continued to be simple inhumations, while at Dinkha, alongside
inhumation, the earlier use of brick tombs continued, augmented by the
innovation of stone chamber tombs; and at Hasanlu only a few urn burials of
undetermined date have been found (compare Stein 1940, pp. 397, 400, date
not clear). Certain pottery types, very much in evidence at Hasanlu, do not
occur at Dinkha: tripod stands for supporting spouted vessels, which were
found in many Hasanlu burials, and spouted vessels with animals sculpted
on the spout, or vessels with an animal at the handle (Dyson 1968b, figs. 118,
121; vanden Berghe 1959, pl. 145, c-e); bowls with tab handles, solid and
looped, sometimes with animal-head protomes on the body (Dyson 1964c,
fig. 13; Young 1965, figs. 6:3; 7:3; Boehmer 1967, p. 580, fig. 7); and vessels on tall
hollow stands (Dyson 1964a, figs. 4:7, 9, 10, 11). Nor do we have any evidence
at Dinkha for the fine wares with polished gray surfaces, and for glazed wares
(Young 1965, p. 55).
Lion pins, metal bells, belts, animal figurines, armor, metal and pottery
rhyta, not to mention ivories and vessels made of precious metals, were
not found at Dinkha. But it must be stressed that many of the Hasanlu
objects mentioned come from the destroyed citadel, whereas at Dinkha we
are dealing with a cemetery alone and have only the evidence from material
placed in burials. However, from the sophisticated and massive architecture
preserved at Hasanlu, and from the vast quantity of material remains, both
of local and of foreign manufacture (Muscarella 1971a, pp. 263–265), there
can be no doubt that Hasanlu was culturally and economically the richer
E. Akurgal, The Art of the Hittites (New York, 1962) fig. 122. For Assyria see A. Haller, Die Gräber
und Grüfte von Assur (Berlin, 1954) pp. 18, 21–22, 103. For Palestine, see for Lachish, O. Tufnell,
Lachisch II (London, 1940) p. 194; for Ugarit, C.F.A. Schaeffer, Ugaritica IV (Paris, 1962) pp. 80–
82, 103–105, figs. 64, 65. In Mesopotamia we find them at Tepe Gawra, E. Speiser, Excavations
at Tepe Gawra (Philadelphia, 1935) p. 33; Nuzi, Starr 1939, pp. 378–379, 414, 450, and Vol. II,
pl. 117, n; at Kish, in the Field Museum of Natural History in Chicago. They also were used
in Egypt, H. Schäfer, Aegyptische Kunst (Berlin, 1913) fig. 122, and Nora Scott, BMMA, Spring
1973, fig. 39. In the West they were common from Bronze Age to Roman times at too many
sites to mention here. See, for example, L. Deubner, “Zum Astragalspiel,” AA 1929, pp. 272–282;
Pauly-Wisowa Supplement, IV (1924) “Astragalomanteia,” pp. 51–55; R. Hampe, “Die Stele aus
Pharsalos in Louvre,” Winckelmannsprogram der Arch. Gesellschaft zu Berlin (1951); G. Bass,
Cape Gelidonya: A Bronze Age Shipwreck (Philadelphia, 1967) p. 133.
238 chapter six
site, perhaps even the main seat of government and trade in the area.
Dinkha, on the other hand, while obviously not poor, was quite clearly a less
important site, perhaps because it was closer to the western border.
The strong cultural connections demonstrated to exist between Hasanlu
and Dinkha over such a long period of time suggest an hypothesis: that both
the Solduz and Ushnu valleys were part of the same ancient state, of which
Hasanlu may have been the major city, with Dinkha one of several provincial
towns (there are still several unexcavated large mounds in Solduz) governed
by a prince or governor. It is also possible that the same language was spoken
at both sites. To be sure, we know nothing about ancient place names or
languages in the area and therefore can go no further than hypothesizing.22
But with respect to the material evidence of the two valleys in the Iron Age,
they must be treated as one cultural region.
Visible from Dinkha Tepe to the east is the still unexcavated Urartian
site of Qalatgah (Figure 19; Muscarella 1971b, pp. 44–49). During survey
work conducted by the Hasanlu Project an Urartian inscription, written for
Ishpuini and his son Menua sometime about 810–805 bc, was found. This
important inscription dates the entry of the Urartians from the north into
the southern Urmia basin, specifically, into the Gadar and Ushnu valleys.
And it is at this very time,23 as established by independent archaeological
research, that Hasanlu IV was violently destroyed and Dinkha II was ter-
minated, probably by abandonment—for it is quite clear that the chronol-
ogy of Dinkha II depends completely on that of Hasanlu. Surely, these two
events, the end of the Iron II culture and the entry of the Urartians into
the area are related: the Urartian invasion of the west and south of the
lake is the historical event that aborted the flourishing Iron II culture. A
few years later King Menua alone set up a stele at Tashtepe, about fifty
miles to the east of Dinkha, demonstrating the southeastern limits of the
invasion.
22 As stated in the text, there are unexcavated mounds in the area of Solduz, and future
work might alter the suggestions made here. At present the Solduz valley is inhabited largely
by Turkish-speaking Shia Moslems, the Ushnu valley by Kurdish-speaking Sunni Moslems.
Future archaeologists might not be able to surmise from the remains of their material
culture as represented by house plans, burial customs, and household goods, that they were
two different cultural groups with different languages and histories, and sometimes mutual
hostility.
23 If the Iron II period ended sometime before 800bc, according to possible interpreta-
tions of the MASCA correction dates, then the building of Qalatgah had no direct connection
with the end of Hasanlu and Dinkha, which would presumably have been in ruins.
the iron age at dinkha tepe, iran 239
We need only present here a few brief comments about Dinkha’s material
relationship to contemporary sites, since much has already been written
about this period. The ties between Hasanlu IV and Sialk B, Geoy Tepe
A (in part, for Iron III remains exist there also: Muscarella 1973, p. 72),
Khurvin, Giyan I1 (part), and the Zendan I (part) are well known and have
been discussed often (Young 1965, pp. 61–68, 70–72; 1967, pp. 24–27; Dyson
1965, pp. 197–203; Boehmer 1967, pp. 576–585; Burney, Lang 1972, pp. 122–
126). And because of Dinkha’s close relationship to Hasanlu, the same ele-
ments in the discussion obtain for Dinkha. Although pottery has been the
main element referred to in discussing relationships, we might expand
this by including other objects. Thus, at Sialk B several multiple burials
existed, and chains, plain jointed horse bits, flat-band rings, decorated band
bracelets, and torques were placed in burials there (Ghirshman 1939, pls. l,
lvi, lix, lxviii, lxxxv, lxxvii, lxxviii, lv, etc.; see also Young 1967, pp. 76–77,
note 28).
At Khurvin, in addition to the typical Iron II vessels, several metal objects
are of interest to us: torques (see above), tweezers, decorated band bracelets,
and plain bracelets with tapered ends (vanden Berghe 1964, pls. iv, v, xi, xii,
xvii, xxii, xxxix, xli, pp. 29–30, pl. xlii).
Grave 4 from Tepe Guran should be mentioned again in this context for
it contained bronze vessels of a type found at Sialk B and similar to some at
Hasanlu IV. The sword also found in the tomb indicates, perhaps, a tenth–
ninth century dating for the grave rather than ninth–eighth (Thrane 1964,
pp. 158–160, note 6; compare Moorey 1971, p. 21).
A few more Iron II sites may be added to the growing list of Iron II sites
in western Iran. Yanik Tepe is said to have yielded gray wares of Hasanlu IV
type, but no details are yet available (Burney 1964, p. 60). On the western side
of the lake at Haftavan Tepe, we are informed that an Iron II settlement was
partly uncovered. Moreover, part of an extramural cemetery was excavated
and Iron II burials were uncovered. In one was found a red bridged spouted
vessel, but in other burials dating is not so clear-cut (Burney 1970, pp. 165–
168, figs. 7, 8:2). Some of the burials had chains attached to pins, and figure-
eight hairrings (earrings?), anklets, bracelets, rings, and beads; there was
also one Mitannian-type seal (Burney 1970, pp. 165–168, fig. 7; 1972, pp. 134 ff.,
figs. 8, 9, pl. ivb). These finds collectively could indicate a date close to 800 bc
Until the complete publication of the Marlik material it is not easy to
argue strongly for an Iron II occupation here. Nevertheless, the excavator
(Negahban 1964, p. 38) and others who have seen the material (Dyson 1965,
240 chapter six
chart on p. 11; Young 1967, p. 22, note 69; Burney, Lang 1972, p. 118) agree that
some of the material from the tombs belongs in the early first millennium
bc (Compare Moorey 1971, pp. 23–24, who prefers a late second millennium
bc date.) I, too, think there is evidence for an Iron II occupation there on
archaeological and art-historical grounds (Muscarella 1972, pp. 42–43).24
Far away to the southeast at Tepe Yahya (III) a fragment of a gray bridged
vessel was found, and we are told that both gray and red wares occur in this
level (Lamberg-Karlovsky 1970, p. 27, pl. xiii). This information could indi-
cate that there was an Iron Age level at Yahya, but based on the published
material perhaps Iron II/III rather than Iron I/II, as suggested by the excava-
tor.
One final point will be presented here, a point already made by Young
(1967, p. 25), that practically all the sites that had Iron I material also had
Iron II material. Which is to say that from an archaeological view the Iron
Age I and II cultures lasted over a large area for a long time, and may
reflect the historical fact that there was a population continuum in much
of western Iran until the early eighth century bc. Of course, Sialk B is the
anomaly here because of its extraordinary painted-ware tradition, and here
alone one might be able to argue against stability (Dyson 1965, pp. 200–
201). The isolated Iron I burials at Dalma and Hajji Firuz, and at Godin to
the south, should be kept in mind, but they do not contradict a continued
distribution of the Iron I and II cultures.
24 At Klar Dasht a bridged spouted vessel with three small feet was found: H. Samadi, Les
découvertes fortuites Klardasht, Garmabak, Emam et Tomadjan (Teheran, 1950) pp. 8, 12, fig. 9.
Table I: Dinkha III Burial Data
Body Head Leg Arm No.
Orien- Posi- Posi- Head Posi- Posi- Ves- Weapons and Animal
Burial Type Sex/Age tation tion tion Faces tion tion sels Jewelry Miscellaneous Bones Comments Text Reference
B9aβ11 I C N-S B S F L:F 1 Bracelets, beads Poorly preserved
R:S
15 B F/MA N-S B N F F F 3 Bracelets, pins, X? P. 44, Figs. 9, 52
rings, needle, (385, 400)
headband, beads
16 I ?/MA N-S L S W F F 2 Pin, needle
17 I ?/MA N-S L N SE F R:? 3 Torque, bracelets, P. 44, Figs. 10, 11
L:S anklets, ring, pin,
beads, needle
18 I ?/MA N-S B S E F 1 Bracelet, pins, Poorly preserved
ring, beads
19 I C N-S R S E F L:F 4 Bracelet, ring Pp. 44–45, Fig. 12
R:S
the iron age at dinkha tepe, iran
Table I (cont.)
Body Head Leg Arm No.
Orien- Posi- Posi- Head Posi- Posi- Ves- Weapons and Animal
Burial Type Sex/Age tation tion tion Faces tion tion sels Jewelry Miscellaneous Bones Comments Text Reference
24 I ?/MA N-S L S F F 2 Bracelet, pins, P. 43, Fig. 6;
rings, needle, Muscarella, 1968,
beads fig. 19
25 I M/MA E-W B E NW F 2 Pin, beads Poorly preserved P. 40, Figs. 3, 5
26 I C N-S L N NE F F 3 Torque, bracelets, Bronze bowl P. 43, Fig. 7;
anklets, pins, Muscarella, 1968,
rings, earrings, fig. 16
needles, beads
27 I M/MA N-S B S NE F L:F 2 Bracelet, beads Spear X P. 43, Figs. 7, 8
R:S
chapter six
Table I (cont.)
Body Head Leg Arm No.
Orien- Posi- Posi- Head Posi- Posi- Ves- Weapons and Animal
Burial Type Sex/Age tation tion tion Faces tion tion sels Jewelry Miscellaneous Bones Comments Text Reference
TT IV β1 I F/YA N-S L N E F F 3 Torque, pins, Fig. 52 (509T)
earrings, ring,
beads
TT VII, β1 I F/YA N-S R N Ext. at 3 Torque, bracelets, Pp. 39–40, Fig. 3
sides pin, earrings
2 B M/YA E-W B E Sky Ext. F 2 Bracelet Dagger P. 40, Figs. 3, 4
TT IX β1 I F/A E-W B W Sky Ext. at 2
sides
5 I I N-S R S 2 Bracelets, anklets, Poorly preserved
pins, beads
chapter six
Table II (cont.)
Body Head Leg Arm No.
Orien- Posi- Posi- Head Posi- Posi- Ves- Weapons and Animal
Burial Type Sex/Age tation tion tion Faces tion tion sels Jewelry Miscellaneous Bones Comments Text Reference
8 B F/OA N-S B N E R:F 3 Pins, ring, needle, X Arthritic lipping Pp. 59, 67–68,
L:S beads of vertebrae; Fig. 43; Muscar-
partly in balk ella, 1968, fig. 2
9 B C E-W L W N F F 3 Pins, hairrings, Muscarella, 1968,
ring, earrings, figs. 2, 7
beads
11 B F/MA N-S B N E Ext. F 2 Pins, rings, beads Archer’s ring Pp. 61–62, Fig. 29
14 B ?/MA N-S L S W F 4 Partly destroyed Muscarella, 1968,
by β9 fig. 2
B10b, β16 B F/MA N-S L N E F F 7 Pins, rings, Fig. 52 (342T, 343P,
chapter six
Bibliography
Barnett 1967
R.D. Barnett, “The Urartian Cemetery at Igdyr,” Anatolian Studies 13 (1967) pp. 153–
198.
Boehmer 1965
R.M. Boehmer, “Die Grabungen auf den Zendan-i-Suleiman,” AA (1965) pp. 746–787.
Boehmer 1967
“Forschungen am Zendan-i-Suleiman,” AA (1967) pp. 573–585.
Burney 1962
Charles Burney, “Excavations at Yanik Tepe, Azerbaijan, 1961, Second Preliminary
Report,” Iraq 24, 2 (1962) pp. 134–149.
Burney 1964
“Excavations at Yanik Tepe, Azerbaijan, 1962, Third Preliminary Report,” Iraq 26, 1
(1964).
Burney 1970
“Excavations at Haftavan Tepe 1968: First Preliminary Report,” Iran VII (1970) pp.
157–171.
Burney 1972
“Excavations at Haftavan Tepe 1969: Second Preliminary Report,” Iran X (1972)
pp. 127–142.
Burney 1973
“Excavations at Haftavan Tepe 1971: Third Preliminary Report,” Iran XI (1973) pp. 153–
172.
Burney, Lang 1972
Charles Burney, Andrew Lang, Peoples of the Hills, (New York, 1972).
Burton-Brown 1951
T. Burton-Brown, Excavations in Azerbaijan, 1948 (London, 1951).
Carter 1965
T.H. Carter, “Excavations at Tell al-Rimah, 1964,” BASOR 178 (1965) pp. 40–69.
Contenau, Ghirshman 1935
G. Contenau, R. Ghirshman, Fouilles de Tepé Giyan, (Paris, 1935).
Dyson 1958
Robert H. Dyson, Jr., “Iran 1957: Iron Age Hasanlu,” Univ. Museum Bulletin 22, 2 (1958)
pp. 25–32.
Dyson 1960
“Hasanlu and Early Iran,” Archaeology 13, 2 (1960) pp. 118–129.
Dyson 1962
“The Hasanlu Project,” Science 135, no. 3504 (1962) pp. 637–647.
the iron age at dinkha tepe, iran 251
Dyson 1964a
“Notes on Weapons and Chronology in Northern Iran around 1000bc,” in Dark Ages
and Nomads, ed. M. Mellink (Istanbul, 1964) pp. 32–45.
Dyson 1964b
“Sciences Meet in Ancient Hasanlu,” Natural History Oct. 1964, pp. 16–25.
Dyson 1964c
“In the City of the Gold Bowl …,” ILN, Sept. 12, 1964, pp. 372–374.
Dyson 1965
“Problems of Protohistoric Iran as seen from Hasanlu,” JNES 24, 3 (1965) pp. 193–217.
Dyson 1967
“Early Cultures of Solduz, Azerbaijan,” Survey of Persian Art XIV (Tokyo, 1967) pp.
2951–2974.
Dyson 1968a
“The Archaeological Evidence of the Second Millennium BC on the Persian Plateau,”
CAH, fascicle 68 (Cambridge, 1968) pp. 3–36.
Dyson 1968b
“Hasanlu and the Solduz and Ushnu Valleys,” Archaeologia Viva 1, 1 (1968) pp. 82–101.
Ghirshman 1939
R. Ghirshman, Fouilles de Sialk (Paris, 1939).
Ghirshman 1954
Village perse-achéménide, MMA en Iran XXVI (Paris, 1954).
Ghirshman 1964
The Arts of Ancient Iran (New York, 1964).
Godard 1931
A. Godard, Les Bronzes du Luristan (Paris, 1931).
Goff Meade 1968
C. Goff Meade, “Luristan in the First Half of the First Millennium B.C.,” Iran VI (1968)
pp. 105–134.
Goff Meade 1969
“Excavations at Baba Jan 1967 …,” Iran VII (1969) pp. 115–130.
Hančar 1934
F. Hančar, “Kaukasus-Luristan,” ESA 9 (1934) pp. 47–112.
Herzfeld 1941
E. Herzfeld, Iran in the Ancient East (London, 1941).
Kleiss 1971
W. Kleiss, “Erkundungsfahrten in Iran 1970,” AMI 4 (1971) pp. 51–111.
Lamberg-Karlovsky 1970
C.C. Lamberg-Karlovsky, Excavations at Tepe Tahya, Iran, 1967–1969: Progress Report I
(Harvard University, 1970).
252 chapter six
Levine 1971
Louis D. Levine, “The Iron Age Revealed,” Expedition 13, 3–4 (1971) pp. 39–43.
Moorey 1971
P.R.S. Moorey, A Catalogue of the Persian Bronzes in the Ashmolean Museum (Oxford,
1971).
Muscarella 1966
Oscar White Muscarella, “Hasanlu 1964,” BMMA Nov. 1966, pp. 121–135.
Muscarella 1968
“Excavations at Dinkha Tepe, 1966,” BMMA Nov. 1968, pp. 187–196.
Muscarella 1971a
“Hasanlu in the Ninth Century BC and its Relations with other Cultural Centers in
the Near East,” AJA 75 (1971) pp. 263–266.
Muscarella 1971b
“Qalatgah: an Urartian Site in Northwestern Iran,” Expedition 13, 3–4 (1971) pp. 44–
49.
Muscarella 1972
“A Bronze Vase from Iran and Its Greek Connections,” MMA Journal 5 (1972) pp. 25–
50.
Muscarella 1973
“Excavations at Agrab Tepe, Iran,” MMA Journal 8 (1973) pp. 47–76.
Negahban 1964
E. Negahban, A Preliminary Report on Marlik Excavation (Teheran, 1964).
Rad, Hakemi 1950
M. Rad, A. Hakemi, The Description and Results of the Scientific Excavations at
Hasanlu, (Teheran, 1950) (in Persian).
Ralph et al. 1973
E.K. Ralph, H.N. Michael, M.C. Han, “Radiocarbon Dates and Reality,” MASCA News-
letter 9, 1 (Aug., 1973) pp. 1–18.
Schaeffer 1948
C.F.A. Schaeffer, Stratigraphie Comparée et Chronolgie de l’Asie Occidentale, (Oxford,
1948).
Schaeffer 1949
Ugaritica II (Paris, 1949).
Schmidt 1970
E.F. Schmidt, Persepolis III, The Royal Tombs and Other Monuments (Chicago, 1970).
Starr 1939
R.F.S. Starr, Nuzi (Cambridge, 1939).
Stein 1940
A. Stein, Old Routes of Western Iran (London, 1940).
the iron age at dinkha tepe, iran 253
Thrane 1964
H. Thrane, “Excavations at Tepe Guran, Luristan …,” Acta Archaeologia XXXIV (1964)
pp. 121–133.
Thrane 1965
“Archaeological Investigations in Western Iran,” Acta Archaeologia XXXV (1965)
pp. 153–169.
Thrane 1970
“Tepe Guran and the Luristan Bronzes,” Archaeology 23 (1970) pp. 27–36.
vanden Berghe 1959
L. vanden Berghe, Archéologie de l’Iran Ancien (Leiden, 1959).
vanden Berghe 1964
Le Nécropole de Khurvin (Istanbul, 1964).
Woolley 1955
C.L. Woolley, Alalakh (London, 1955).
Young 1962
T. Cuyler Young, Jr., “Taking the History of the Hasanlu Area Back … Another Five
Thousand Years,” ILN, Nov. 3, 1965, pp. 707–709.
Young 1965
“A Comparative Ceramic Chronology for Western Iran, 1500–500 bc,” Iran III (1965)
pp. 53–85.
Young 1966
“Survey in Western Iran, 1961,” JNES XXV (1965) pp. 228–239.
Young 1967
“The Iranian Migration into the Zagros,” Iran V (1967) pp. 11–34.
Young 1969
Excavation at Godin Tepe: First Progress Report, Royal Ontario Museum, Art and
Archaeology, Occasional Paper 17 (Toronto, 1969).
chapter seven
* This article originally appeared as “Warfare at Hasanlu in the Late 9th Century B.C.”
peoples here, which the Lord your God is giving you as a heritage, you must
not spare a living soul; but you must be sure to exterminate them, Hittites,
Amorites, Canaanites, Perizzites, Hivvites, and Jebusites, as the Lord your God
commanded you, so that they may not teach you to imitate all the abominable
practices that they have carried on for their gods, and so sin against the Lord
your God. (Deuteronomy 20:10–18)
And ten thousand others did the Judeans carry away alive, and they brought
them to the top of a crag, and cast them down from the top of the crag, so that
all of them were dashed to pieces. But the men of the band whom Amaziah
had sent back without allowing them to go with him to battle fell upon the
cities of Judah from Samaria to Bethhoron, and slew of them three thousand
and took a large amount of spoil. (2nd Chronicles 25:12–13)
(See also Josh 6:17, 24; Judges 1:6, 17:21; ISam 11:2, 30:17; II Kings 25:7.)
The inhabitants of Hasanlu apparently did not differ greatly from their
neighbors in the emphasis placed on military activities. Lacking written
documents from Hasanlu itself, we do not know whether its foreign policy
involved a defensive strategy or an offensive one, or both. Artifacts found at
the site, however, certainly indicate that the inhabitants were prepared for
war. Weapons were recovered in all areas of the Citadel, inside and outside
buildings. In some cases, their number and disposition indicate that they
were found as they had been stored—and never put to use in the final battle;
other isolated weapons were apparently abandoned or lost in the heat of the
battle.
Judging from the quantity excavated, the most common weapon em-
ployed at Hasanlu was the socketed spear, some made of bronze but the
majority of iron (Fig. 2); hundreds were found in the debris of the upper
floor of Burned Building II, where they apparently had been stored. Other
weapons include:
maceheads (Fig. 4)
48 stone
23 bronze
2 iron
3 iron/bronze
swords (Fig. 5)
2 bronze
28 iron
9 iron/bronze
1 with a gold cloisonné hilt
258 chapter seven
daggers (Fig. 5)
21 iron
1 bronze
3 iron/bronze
many blade fragments
bronze axes (Fig. 6)
iron pikes
Evidence for the use of the bow is provided by arrowheads (Fig. 7), some-
times found massed in a quiver. There were at least three bronze quivers, as
well as a unique iron quiver with bronze trim (Fig. 8).
Given this arsenal of weapons, it is not surprising that the Hasanlu finds
included protective metal armor. Among the bronze helmets were 2 crested
(Fig. 9) and 3 pointed (Fig. 11) examples, as well as 3 detached crests, their
(leather?) helmets having disintegrated; the crested helmets had separately
added ear guards. Three bronze shields, a pair of bronze shoulder guards and
a number of bronze and iron armor plates were also recovered. Even horses
sometimes wore armor on their heads (see de Schauensee, this issue).
Both military techniques and attitudes toward warfare can also be recon-
structed from representational art found at the site. In typical Near Eastern
fashion, the ruling elite of Hasanlu commissioned the illustration of bat-
tle scenes, which were displayed or stored in important buildings. These
scenes appear primarily on 72 small and fragmented ivory carvings that
were recovered (along with fragments bearing other motifs) from the second
story collapse of Burned Building II, identified as probably a religious struc-
ture or temple. (By contrast, only two battle scene fragments came from
Burned Building I, identified as probably the rulers’ residence.) Battle scenes
were also depicted on a silver and electrum beaker from Burned Building II,
and on five small rectangular bronze or iron plaques from Burned Building
I.
The ivory plaques had originally been attached with pegs to wooden
objects. Because the plaques had broken and scattered when they fell with
the collapse, the kinds of objects they decorated remain unknown, as does
the original grouping of the plaques and the way in which they were juxta-
posed. Lacking information on the relationship between plaques, we cannot
say to what extent they represent parts of a story or narrative. Both because
of their fragmentary nature and because none of the ivories has an inscrip-
tion (no local writing occurs at Hasanlu), we are also unable to identify or
explain a number of specific aspects of these scenes; for example, what is the
ethnic or political identity of the fighting forces, and what are the historical
event(s) represented?
warfare at hasanlu in the late 9th century bc 259
Figure 2a, b. Spears made of iron or bronze with sockets that fitted over a
wooden shaft were the most common weapons found in the ruins
of Hasanlu. (a) The iron weapons were often in fragments, but
a complete example was found in Burned Building VI. (b) This
bimetallic spearpoint has a bronze blade riveted to an iron socket and
mid-rib. This weak construction and the rarity of such weapons
suggest they may have had a ceremonial rather than a utilitarian
function (see Pigott, this issue). (a: HAS 74–249, L. 34.5cm, Musée Iran
Bastan, Tehran; b: UM 65–31–196, L. 57.0cm.
Photos courtesy of the Hasanlu Project.)
260 chapter seven
Figure 5a–d. Daggers and swords were made of bronze or iron, and often
had decorated hilts, inlaid with ivory or wood. (a: Metropolitan
Museum of Art 63.109.4, L. 10.3cm; b: UM 59–54–129, L. 29.5cm; c:
HAS 59–880, L. of hilt 23cm, Musée Iran Bastan, Tehran; d:
UM 61–5–112, L. 41.0cm. Drawings courtesy of the Hasanlu Project.)
warfare at hasanlu in the late 9th century bc 263
Figure 6a, b. Bronze axes such as these from Hasanlu were a traditional
weapon in Mesopotamia, depicted by the Sumerians as early as the 3rd
millennium bc. (a: HAS 62–1085, L. 18.0cm; b: HAS 58–194, L. 13.5cm. Both
in Musée Iran Bastan, Tehran. Drawings courtesy of the Hasanlu Project.)
Figure 7a–d. Arrowheads of iron and bronze were found throughout the
burned buildings. They vary in size and shape, so that a complete
study of their distribution within the site (including weapons in
storerooms) may allow us to distinguish between the arrows of
defenders and attackers. (a: HAS 74–269, L, 9.9cm; b: HAS 74–278,
L. 6.2cm; c: HAS 74–290, L. 9.5cm; d: HAS 74–276, L, 11.0cm. All from
Musée Iran Bastan, Tehran. Photos courtesy of the Hasanlu Project.)
264 chapter seven
Figure 12a. Scenes of battle are found in Near Eastern art as early as the
4th millenium bc. An impression made by a cylinder seal from the site of
Chogha Mish in southwestern Iran shows human figures on the walls of a
fortress, hands raised in distress or surrender. Two larger figures (left)
represent the conquerors, one of whom holds a small seated(?) captive by
the hair. (Drawing by Jon Snyder after Delougaz and Kantor 1972: Pl. 10:d.)
268 chapter seven
Figure 12b. A clay sealing from the site of Susa in southwestern Iran shows
a citadel, its walls ornamented with horns. Three nude figures
represent the defeated forces, falling or fleeing before a fully clothed
archer. One of the small figures appears to have his hands bound
behind his back. (Drawing by Jon Snyder after Amiet 1966: Fig. 11.)
Figure 13. During the 3rd millennium, Mesopotamian rulers recorded their
feats on sculpeted stone monuments. The “Stele of the Vultures” shows
the army of the city of Girsu trampling over the defeated army of Umma.
(The Louvre: pho courtesy of the Metropolitan Museum of Art.)
270 chapter seven
Figure 14. Some of the people who died in the columned hall of Burned
Building II were richly adorned with jewelry. The skeleton shown here
belonged to a young woman in her late teens. She had a necklace of
amber, carnelian, frit, and shell beads, and other strung beads lay below
her pelvis. On her upper body were three lion pins bearing cloth
impressions, one of them visible here (see Pigott, this issue). At her knee
was an iron dagger with a bronze hilt; traces of a scabbard adhered to
the blade. (HAS skel. 263. Photo courtesy of the Hasanlu Project.)
warfare at hasanlu in the late 9th century bc 271
Figure 16. Vigorous battle scenes depicting foot soldiers fighting with
spears and small shields are found on the Hasanlu ivory plaques.
The man on the left wears a crested helmet (see Fig. 9). Ht. upper
frag. 3.4cm. The Metropolitan Museum of Art, Rogers Fund, 1965
(65.163.8, .18) Image © The Metropolitan Museum of Art.
warfare at hasanlu in the late 9th century bc 273
Figure 19. Women are rarely depicted in the art of 9th century Hasanlu, in
which scenes of warfare and ceremony predominate. This ivory
plaque fragment shows a figure with braided hair; both the hairstyle
and the position of the hands suggest this is a woman, in
despair as she watches the enemy’s attack. An arrow juts out of
the tower behind her. Ht. 3.0cm. (Metropolitan Museum of Art
65.163.24. Photo courtesy of the Metropolitan Museum of Art.)
warfare at hasanlu in the late 9th century bc 275
archers shoot from platforms (like those of Fig. 18) seemingly set above forti-
fication walls, probably representing a city under siege. Other metal plaques
depict archers, cavalry, and chariots, but not in scenes where an enemy is
present. Only in one instance, on a silver and electrum beaker found at
Hasanlu (see Marcus, Fig. 1), is a prisoner of war apparently represented, for
here an armed soldier clasps the wrist of an unarmed figure.
As fragmented as they are, the ivory carvings enable us to present some
observations about battle arrangements. Cavalry and chariots are shown
fighting in action together side by side—depicted on the carvings by over-
lapping figures (Fig. 15a, lower right; note the hooves of a cavalry horse
directly in front of the chariot horses). Likewise, infantry are shown fight-
ing alongside chariot and cavalry horses (Fig. 15a, upper right; the hand of
an infantryman wielding a spear may be seen in front of the two chariot
horses’ heads). Several fragments depict soldiers attacking cavalry and char-
iot horses (Fig. 17), and others show infantry fighting hand to hand (Fig. 16).
The cavalry, albeit without stirrups for support, seem to have functioned as
lancers (Fig. 15b).
The distinct impression that one derives is that chariots, cavalry, and
infantry seem to have participated in battle together without functional
division- but it should be noted that the depictions may be merely an artistic
convention and not necessarily a literal recording of battle maneuvering. It
is thought by military historians that before the late 1st millennium bc (the
Achaemenid period) chariots functioned primarily to transport troops to
battle and to harass the enemy with arrows (see Xenophon Cyropaedia VI,i:
27–30). The chariots were not to get too close to their infantry, because of
the vulnerability of the horses to enemy weapons. The Hasanlu scenes do
not seem to conform to this concept, although in fact, we may be viewing a
representation of a mopping-up activity against a fleeing, broken enemy.
Light war chariots had been functioning in battle throughout the Near
East for about 1000 years before the destruction of Hasanlu, but the use of a
cavalry corps is not attested before the time of the Assyrian king Aššurnasir-
pal II in the early 9th century bc. During his reign and that of his succes-
sor Shalmaneser III, cavalry is often mentioned in battle descriptions and
depicted in art. In addition to its use in Assyria, a cavalry corps is recorded
as forming units in several other armies, including those of Urartu, Baby-
lon, Adini, Israel, and Syria. In the 9th century bc, then, both cavalry and
chariotry were normal constituents of a Near Eastern army. A large num-
ber of horse bits, protective breastplates, and harness equipment, as well as
the skeletons of horses have been excavated at Hasanlu. A fragmented mass
of wood together with a long pole from Burned Building IV–V may be the
warfare at hasanlu in the late 9th century bc 277
remains of a chariot; and some bronze and carved stone objects that may be
yoke saddle pommels for chariots were recovered (see de Schauensee, this
issue). Thus, the archaeological remains support the pictorial evidence that
chariots and cavalry functioned as elements of the Hasanlu army.
It need not be assumed that the representations of battle scenes on the
ivories were conceived by the local authority solely to emulate Assyrian
propaganda techniques projecting power and royal might. And it is prob-
able that the military forces of Hasanlu itself are represented fighting and
defeating an historical enemy. If so, the depictions of a cavalry and chariot
corps signify that there existed at Hasanlu an elite class that had the leisure
and skills to practice and perform the necessary complex maneuvering tac-
tics. Whether this class itself maintained and supplied the equipment and
horses or whether this was a state function is unknown. Equally unknown is
whether there was a conscript or a standing army, the latter being a standard
feature of the Assyrians by this time.
he holds a mace, while the other grasps a rope attached to a pair of seated
captives (Fig. 12c).
The earliest historical texts in cuneiform, written during the Early Dynas-
tic Period in Sumer (ca. 2900 to 2350 bc), contain records of war between
the independent city states of southern Mesopotamia and exaltation in the
slaughter of the enemy. Contemporary sculptures depict warfare in vivid
detail, and elucidate the cuneiform texts. One of the best-known monu-
ments of the Early Dynastic period, the “Stele of the Vultures” (Fig. 13), was
carved for King Eannatum of Girsu. On the stele are scenes showing dis-
ciplined uniformed troops from Girsu carrying spears and battle axes as
they trample over the nude bodies of the soldiers of Umma; nearby are
nude Umma bodies piled in a heap, while others are being buried under
a mound of earth. A mosaic panel from the Royal Cemetery at Ur (“the Ur
Standard”) shows four-wheel warwagons staffed by lancers that override the
nude bodies of an enemy, and a procession of prisoners, their arms pin-
ioned behind their backs. From this same period, shell plaques from Mari on
the Euphrates River depict both soldiers bearing axes and bound prisoners.
The portrayal of enemy dead in the form of nude bodies being overridden
by victorious forces—usually chariots or cavalry—and of bound prisoners
became standard victory motifs that continued to be represented in Near
Eastern art the the 1st millenium bc.
Beginning in the period of Semitic Akkadian control of Mesopotamia
(2334–2154bc) prisoners of war are mentioned for the first time in texts, and
Akkadian reliefs dramatically depict them with their hands or arms held
by rope behind their backs, and sometimes with their necks held in stocks.
Also during this periods texts mentions the slaughter of prisoners for the
first time; such an act is possibly portrayed on an Akkadian stele from Girsu
that shows soldiers killing unarmed nude men. Deportation of the captured
inhabitants of cities is first mentioned in the later texts of the Sumerian 3rd
Dynasty of Ur period (ca. 2112–2004bc).
Figure 20. Victims of the battle at Hasanlu were often found trapped
within buildings that had burned and collapsed. This group lay
at the northern end of the columned hall in Burned Building II,
near the outside door. Other items found on the floor included
charred beams (upper left) and a red deer skull with antlers (in
front of the workman). (Photo courtesy of the Hasanlu Project.)
crushed beneath fallen walls and roof (Fig. 20). The victims included men,
women, and many children. Some individuals were armed, while many of
the females and children were wearing jewelry, including relatively heavy
lionheaded pins (Fig. 14). Another group of about 89 people were found
where they fell in open areas, victims of slaughter. The cause of death is
graphically documented by head wounds, or by disarticulated limbs; in the
latter cases, the bodies could have been mutilated by animals and vultures,
but the head wounds patently tell another story. In each category, those who
perished from the collapse of the buildings, and those from slaughter, were
men, women, children, and infants.
Three clusters of skeletons revealed chilling episodes of death at Hasanlu.
In the first two cases, random slaughter of fleeing troops and inhabitants
seems to have occurred. South of Burned Building XI the skeletons of 11
280 chapter seven
Figure 21. A large group of people were slaughtered outside the entrance
to the settlement, to the south of Burned Building XI. Several in the group
had head wounds, visible on the skulls as bone damage. Skeleton 392, a
male of about 20, was killed by a mace that left a rounded depression on
the back of the skull (right). A healed lenticular (sword) wound is also
visible in the center of the photo; his age and the earlier wound suggest
that he may have been a soldier. (Photo courtesy of the Hasanlu Project.)
adults, 3 teenagers, 1 child, and 1 infant were uncovered close to one another;
none was armed. Six of the adults had head wounds, some multiple (Fig. 21).
The others were probably killed by wounds to other areas of the body, which
would not necessarily leave traces on the skeleton, a situation that no doubt
applies to other skeletons uncovered. And to the east of the Upper Court
Gate were uncovered the clustered skeletons of 6 adults, 2 teenagers, and 1
infant; one of the adults had a head wound.
The third group of skeletons almost certainly represents slaughtered pris-
oners. In Room 2 of Burned Building IV the skeletons of 27 people were
discovered—16 adults, the rest children and infants. Found overlapping or
warfare at hasanlu in the late 9th century bc 281
close together in a mass, the skeletons were lying in disarray over the burned
debris from the collapsed roof, and were covered by a layer of brick debris.
Four of the victims had head wounds, unambiguous evidence that they
had been killed by weapons; the form of bone damage indicates that the
weapon(s) used, perhaps to dispatch an already wounded and helpless per-
son, was a mace.
The fact that the victims were lying on burned debris indicates that they
were killed after the initial collapse of the building, but while the walls were
still standing. This evidence suggests that the victims seem to have been
purposely brought to Room 2 to be executed shortly after the fire had died
out and the destruction had been accomplished. At some time after their
death the walls collapsed and sealed them in the rubble.
Whether the killing of prisoners (if indeed such was the case here), as
opposed to the random slaughter of fleeing inhabitants, was an isolated
event at Hasanlu is unclear. Nothing is known of the dynamics that pre-
ceded the invasion of the Citadel by the enemy: whether the inhabitants
surrendered only to be killed nevertheless, or whether they did not surren-
der, and being defeated suffered the consequences. In this context we may
wonder if the massacre in Burned Building IV was a punitive political action
or a religious one, such as sacrifice?
Likewise we wonder about the presence or absence of the Hasanlu armed
forces: were they defeated in the neighboring plains, or were they absent on
a campaign? Wherever they were, they could not prevent the invasion, an
invasion that was accomplished by a strict following of the Assyrian method
of warfare.
The fire that destroyed the site at Hasanlu either was deliberately set or
resulted from an accident, but it must have started soon after the enemy
entered the city. Whatever the cause, the fire spread through the wood
and brick buildings of the Citadel quite quickly, apparently preventing the
enemy from acquiring much, if any, booty. There is no obvious evidence
within the debris that looting or post-destruction digging for treasure
occurred; rather the opposite is suggested. For example, the well-known
gold “bowl” (actually a beaker; see Winter, this issue) was found in the arms
of, apparently, a local inhabitant, who in the process of attempting to save
it died in the collapse of Burned Building I-West. And the vast quantity of
material recovered within the debris of all the buildings, some of it sump-
tuous (artifacts of silver, gold, ivory, Egyptian Blue), suggests that the City’s
contents remained essentially unplundered.
282 chapter seven
also by those Urartians represented on the bronze gates from Balawat com-
missioned by Shalmaneser III (858–824 bc).
The Urartian helmets on the Balawat gates, however, are depicted with
short, pointed ear guards made with the helmets, not added on as in the
Hasanlu examples, and thus it cannot be argued with force that the latter
belonged to the Urartian invaders. And if the ivory pictorial evidence indi-
cates that the Hasanlu infantry wore feathered/crested helmets, they are not
the same as the excavated examples. Who, then, wore the crested helmets
recovered at Hasanlu? Equally puzzling is the presence of the pointed hel-
mets (Fig. 11), which are very like the Assyrian standard form, and which after
the mid-9th century were represented in art as also worn by Urartians. Did
the Urartian troops wear both crested (for officers?) and pointed helmets
while at Hasanlu? Or, did the local forces themselves employ a variety of
helmet forms? More research and thought are needed to resolve this prob-
lem.
Another possible clue depends on the cultural attribution of a bronze
mace head (Fig. 4c) with star or rosette faces. Two very similar examples,
albeit found in a later context, derive from Altin Tepe, an Urartian site in
northeastern Turkey. Were the Altin Tepe examples booty from the Hasanlu
campaign kept as heirlooms, or were they locally made in Urartu under
northwestern Iranian influence; or is the Hasanlu example a weapon lost
there by an Urartian soldier? We do not know. Nor in fact do we know how
many (if any) of the other artifacts excavated at Hasanlu actually may have
been left by the enemy forces and are thus not to be documented as local
products. Here too, more research, not speculation, is required to resolve
this tantalizing issue.
From all the evidence made available by archaeology—the destruction,
the artifacts, the pictorial representations—it is attested that warfare was
not a casual or incidental activity for the people of Hasanlu IV. Nevertheless,
archaeology has also revealed that there was time, energy, and talent for
architects and workmen to construct monumental buildings, and for highly
skilled craftsmen to manufacture a large variety of objects, both luxury items
and objects for daily use. At Hasanlu there was a time for war and a time for
peace, but war was the ultimate event in its history.
284 chapter seven
Bibliography
Arniet, Pierre
1966 Elam. Paris: Centre National de la Recherche Scientifique.
Brentjes, Burchard
1986 “Kriegwesen im alten Orient.” Das Altertum 32:134–142.
Gelb, I.J.
1973 “Prisoners of War in Early Mesopotamia.” Journal of Near Eastern Studies
32(1–2):70–98.
Rolle, R.
1977 “Urartu und die Reiternomaden.” Saeculum 28(3):291–339.
Schulman, A.R
1979–1980 “Chariots, Chariotry, and the Hyksos.” Journal of the Society for the
Study of Egyptian Antiquities 10:105–153.
Yadin, Yigael
1963 The Art of War in Biblical Lands. 2 vols, New York: McGraw Hill.
chapter eight
* This chapter originally appeared as “The Hasanlu Lion Pins Again,” in A View from the
Highlands: Archaeological Studies in Honour of Charles Burney, ed. A. Sagona (Leuven: Peeters,
2004) 693–710.
1 Muscarella 1988, pp. 42–45 for discussion; Marcus 1993. Some years ago in a Van jew-
ellerís shop I saw a typical Hasanlu lion pin that the shopkeeper told me came from a Turkish
site north of Van. This ìsaid to be fromî site X was an attempt—either by the original seller
or the shopkeeper, or both—to cover the fact that the pin (and hundreds of other ancient
and modern objects sold in Turkey) was smuggled from Iran. For other plundered strays see
below and note 13.
2 Previously published in Muscarella 1989, p. 32, fig. 14. The upper figure in this photo-
graph is the present author; the other is Noruz from the village of Hasanlu, a superb excavator.
This photograph was taken shortly before Marcus 1993, p. 160, fig. 2 (Fig. 2 here): in the former
photo the lion pin near the elbow was only just emerging.
286 chapter eight
c) and, to document how lion pins were worn, a drawing was shown
depicting a female, arms akimbo, who wears bracelets, rings, several
necklaces, and two lion pins athwart one shoulder and one on the
other.3
In the discussion period, I confronted these assertions. Here I expand what
I stated in Van, but I do so by responding to a written paper, for the verbal
claims made in Van had appeared earlier and in more detail in Marcus 1993
(but I cannot avoid responding to brief discussions of lion pins in three other
articles published by Marcus, in 1994, 1995 and 1996). I arrive at different
conclusions and interpretations from Marcus. I argue here that a review of
the excavated evidence and data indicates that a number of statements and
cultural conclusions made in the various published papers reflect neither an
accurate archaeological history nor cultural explanation for the presence of
lion pins at Hasanlu.
It was at Hasanlu where I first met Charles and it is appropriate as well as
personally satisfying to present to him a work that addresses a special group
of artifacts from that extraordinary site.
Marcus’ argument can be summarised as follows: “60 adorned individu-
als” were recovered within the ruins of BB II, of which, “about half … were
wearing lion pins,” or “associated with lion pins;” that 69 of the approxi-
mately 100 examples recovered from the site at large derived from BB II;
and that “most [of the 69] were found on skeletons …” [italics added].4
Concerning the number worn, she states that children wore one lion pin,
claiming five children were recovered wearing lion pins, and “older children
and adults wore two or three pins”.5 With regard to their positioning on the
body—“Judging from the distribution of the lion pins on or near the upper
body [she cites her fig. 2, which is Fig. 2 in this paper],6 they were probably
originally pinned at the chest and shoulders …” and Marcus again records
“their placement on the upper body …”7
wore “three or more” and on p. 2502 that they “were worn, in groups of three …”
6 Her fig. 2 had been previously published in Dyson 1964, p. 374, fig. 12, and Muscarella
1988, p. 44, fig. 5, which is skeleton 263. In Dyson’s fig. 12 the caption mentions the two visible
lion pins; in his fig. 9 the caption mistakenly states that the drawing is of “the” pin in fig. 12.
The drawing is not from skl 263, but an isolated example from BB II (60–954): see Marcus
1993, p. 164, fig. 6. For the lion pins from skl 263, see Marcus 1993, p. 167, fig. 9.
7 Marcus 1993, pp. 164, 172; 1995, p. 2502: note the plural “shoulders”, and see infra.
the hasanlu lion pins again 287
8 I thank Robert Dyson for giving me much of his time to provide me with photographs,
field record sheets recording all the BB II skeletons, and answers to my many questions
promptly. Shannon White of the University Museum of the University of Pennsylvania also
generously helped me on numerous issues.
9 Skeletons 81, 82, 121–123, 125–130, 133, 136, 137, 142, 143, 145–149, 154, 155, 260, 262, 435,
436, 438.
10 Muscarella 1989, p. 29, fig. 9b.
11 For skeletons 123, 260, 262 in situ in BB II see ibid., fig. 20, lower right, lower left, top
right.
12 Muscarella 1988, p. 42.
288 chapter eight
from the outer town area; they are now in Teheran. [Unexcavated, purchased
examples include one example in the Louvre (acquired in 1958: AO 20472,
ex-collection Coiffard), two examples in Los Angeles, another in a Japanese
collection, one seen by me in Van (see note 1), and at least three that were
advertised in the bazaars.]13
In the mortuary furniture lists for 13 of the BB II skeletons (skeletons 119,
134, 135, 138–141, 144, 150, 152, 156–158), which include five adults and three
children (skeletons 119, 144, 156), are listed bracelets, necklaces, weapons,
etc., as well as lion pins. However, in not one instance did the excavator men-
tion precisely where, on or near (or how near) the skeleton, the objects, were
recovered (see also below). One may surmise that the recorded bracelets,
anklets, and rings were indeed found on their expected positions; for other
artifacts recorded, the situation is more difficult. In some cases it is clear that
artifacts recovered in the vicinity of a skeleton were recorded as belonging
to that particular skeleton’s personal adornment (however, specific loci on,
or measurements of distances from, a skeleton were not given). For exam-
ple, skeleton 158 has extant only its legs and part of its pelvis, with no body or
skull: where specifically was its furniture, listed as beads and one lion pin, in
fact recovered? In one case (skeleton 141) the excavator reported ‘9 lions’ in
the furniture inventory, surely an incorrect number with regard to on-body
placement. Although we remain ignorant whether any of these pins were
recovered actually on skeleton 141—for no loci positions were recorded—we
can claim based on the limited empirical evidence at Hasanlu (see below)
that at least six were not recovered on the skeleton, but recovered some-
where in its vicinity. The same excavator accurately recorded that six lion
pins were recovered in the vicinity of—not on—six skeletons (skeletons
128–133), and suggested in this instance that they “prob[ably] belong to”
them. This seems to be a viable interpretation but we cannot speculate
how many of the six lion pins belonged with each of the six skeletons.
13 The sum 103 with details comes from R.H. Dyson, personal communication, and Mus-
carella 1988, pp. 43, 45 and note 2 for published records of the unexcavated examples; also
A. Hakemi and M. Rad, The Description and Results of the Scientific Excavations of Hasanlu,
Solduz, Teheran 1950, figs. 31, 32 for four excavated examples. I do not know whether an exam-
ple once in the Brummer collection later surfaced in one of the mentioned collections, The
Los Angeles County Art Museum for example. For the Louvre example see Ghirshman 1964,
p. 27, fig. 28. Most probably these unexcavated examples derive from Hasanlu. Hence to the
103 excavated number we may add 7 or 8 others to the corpus. However, the problem remains
whether other Iron II sites in the area (viz. Nagadeh, Dinkha, and Geoy) may also have used
lion pins, and thus it would be too absolute to claim that all the orphans must have derived
from Hasanlu.
the hasanlu lion pins again 289
These examples, plus the lack of specific loci for the other listed lion pins,
compromise our knowing how many of these 13 skeletons actually wore
lion pins, and how many were worn per skeleton (see also skeleton 255,
below).
In fact, only three skeletons (skeletons 255, 256, 263)14 were explicitly
recorded as having lion pins on or immediately contiguous to their physical
remains. Skeleton 255 (a young adult) had a lion pin, pin towards body,
resting on its left humerus, and two pins found “adjacent to but not touching
skeleton”. However, a photograph (Fig. 3) and a sketch made in the field show
the former, but nothing of the other two is visible or drawn near the body
(this could signify that they were not found in the immediate vicinity, or
were removed before the photograph was taken; probably the latter).15 The
second skeleton, a child (skeleton 256), was recovered close to skeleton 255;
recorded in its inventory is a lion pin touching its only preserved humerus
and another “next to the body.” The field sketch shows one on the humerus
and another just above and parallel to it, close to the upper chest; clearly
both pins belonged to skeleton 256. The third, a young adult (skeleton 263,
Fig. 3) had three lion pins on its body: one touching the chest, pin up,
one recovered touching the inner side of the right arm, pin up, and a third
recovered “at neck/shoulder area”.16 From these field descriptions, skeleton
255’s in situ pin would have been attached to clothing on some part of the
upper body, probably close to or on the shoulder, or near the neck. Skeleton
256’s two pins were fitted one above the other at the chest. Skeleton 263 had
them in three superimposed rows, from the shoulder/neck to the waist area.
The field-report sheet for another skeleton (skeleton 157) records a
“bronze lion under leg”, a locus that could suggest either that it belonged to
that skeleton or had earlier fallen from another fleeing individual. Notwith-
standing the field record, however, a photograph of the skeleton in situ
shows a lion pin on the floor ca. 12–20cm. to the left of its left leg (Fig. 4).
14 I excavated these three skeletons in 1962—but except for skeleton 263 have no memory
of the excavation details. Several individuals and I excavated the others mentioned in this
paper in 1960 and 1962.
15 Note that a garment pin was also recorded with skeleton 255 as near the arm but it too
does not appear in Fig. 4. The problem of recognizing discrete skeleton inventories was quite
difficult given the mass of skeletons and scattered body bones (not to mention intruding
emotional feelings over the chaos we were uncovering and witnessing).
16 Muscarella 1989, p. 32, fig. 14.
290 chapter eight
positioned one over the other (probably horizontally) across the body, at
the stomach and chest areas, and the third either at the upper chest, or the
neck, or on one shoulder—but were displaced when the individual fell and
was crushed.
We are also informed that the excavated skeletons could not be sexed
… by methods of physical anthropology, their bones being too crushed by
fallen building debris.18
This is the understanding I recall from excavation at the site in 1960 and 1962,
and which is well-attested in field notes and photographs. Nevertheless,
aside from skeleton 135, it is reported that two (bona fide) lion pin-bearing
skeletons, skeleton 255 (see below) and 263,19 were sexed as females by a
physical anthropologist, Dr. Janet Monge, “on the basis of field photographs”
(i.e., here Fig. 2). Disregarding Monge’s warning (reported in her note 20)
that photo sexing is “highly speculative”, Marcus requests that we ‘tenta-
tively’ accept them as females. The information I have been provided is this:
a physical anthropologist has concluded “with 75 % confidence” that skele-
ton 263 is a female, “based on the skull and part of the pelvis”. Unfortunately
this report was followed by another, stating that in fact, and for reasons
unknown, there are two skeletons labeled skeleton 263, and it is impossi-
ble to determine which is the one with the lion pins—or if that skeleton is
in fact the one in the University Museum.20 Hence, the skeleton 263 pho-
tographed in situ with lion pins remains unsexed.
There is also a tendentious shift concerning interpretation and alleged
excavated data from the claim made in the 1993 paper that “children as
young as six or seven” wore lion pins, to that of the 1995 paper that “females,
including girls as young as six or seven” wore lion pins.21 Children transmo-
grified into girls, nothing about boys … (and skeleton 256, a child, has to my
knowledge not been sexed).
In 1993, figure 14, this notion was carried further. It is a drawing designed
to show the location of personal ornaments and alleges to demonstrate not
only that females wore lion pins but precisely where on the body—and why.
18 Marcus 1933, pp. 159, 170, and notes 19, 20, 23; also in 1994, p. 4, and 1995, p. 2502.
19 Marcus 1993, p. 170. I have no knowledge how my text in Muscarella 1989, p. 32 stated
that females and children wore lion pins, and citing fig. 14 as an example: but in fact, the
caption here was not written by me but by the editor. In any event, the present paper
confronts these issues.
20 These reports from Dr. Janet Monge reached me over several months from September
The caption says “Reconstruction of Skeleton 135 as female …”, but whose sex
we are simultaneously informed “is not based on any physical evidence”,22 for
it was very badly preserved. Listed in this skeleton’s field report sheet as its
furnishings were bracelets, a bowl and spoon, iron spears, a helmet, of which
more below,23 and three lion pins, whose positions whether on or near the
skeleton were not recorded. In the drawing three lion pins are depicted worn
on both shoulders and, nota bene, none are depicted worn on the upper body.
The shoulder positions are rationalised as a “possible [sic] arrangement …”
(some explanation for this ‘female’s’ preference for asymmetry would be
useful, or how she—and all the other BB II females?—choose two pins for
the right, one for the left shoulder). To document the drawing’s verisimili-
tude it is claimed (in note 23) that the execution is based on the “available
field photographs”—not one of which is published to support this alleged
fact: because none exist. Her own figure 2 shows precisely where lion pins
were worn, which is not three across the two shoulders. Skeleton 135 had no
finger rings or pins recorded in the field report, but there were two ‘snake’
headed bracelets, one with raised beads, and also a flat band form not shown
in figure 14 (several snake headed bracelets were also reported associated
with the skeleton of an infant, skeleton 148).
Among other adornments drawn on the akimbo female is a necklace
amulet composed of a fish on a chain.24 This amulet appears neither in the
inventory list in the furniture section of the skeleton sheet, nor on the field
object card as from skeleton 135. It was listed as from that skeleton in a
Register book, which information was passed to the Metropolitan Museum
of Art when the object came to reside there as its share of the excavation
division.25 The fact is, no one knows exactly where the amulet was recovered,
and it (along with the garment pins, and finger rings) should not have been
placed on the fleshed body of figure 14.
In wishing to demonstrate how females wore lion pins, the selection of
skeleton 135 was an unfortunate choice—it cannot be sexed, and the mor-
tuary furniture listing of the helmet and spears could indeed have allowed
it to be interpreted as a male. Indeed, where in figure 14 are (his?) helmet
and spears listed in the mortuary furniture, which listings are otherwise
accepted by Marcus? In fact, however examination of the field record reveals
helmet—ibid., p. 50, which also did not derive from skl 135 (see below and note 26).
the hasanlu lion pins again 293
that there was no helmet on or near skeleton 135. A sketch of their loci shows
that the helmet was recovered at some distance from the skeleton;26 no
other artifact, including the spears, is placed in the sketch—and their exact
loci too remain unknown. It follows that one may not use the ‘associated’
weapons to justify the claim that skeleton 135 is male, but neither may it be
declared a female: it remains anthropologically unsexed. Figure 14 is a ten-
dentious leap, deconstructing the ancient excavated and recorded evidence;
it is an artifact manufactured in the modern world.27
If sexing cannot be accomplished anthropologically, the BB II skeletons
can be sexed “however speculative” [sic] the exercise.28 by examining juxta-
posed, associated, or actually worn artifacts (which are differently defined
situations) from the skeletons and comparing them to those from the ceme-
tery, where it was generally possible to determine sex (but where no lion pins
were recovered). Fine; but the presented conclusions raise doubts in some
cases. For example, short garment pins worn on the shoulders to Marcus
signify females because she claims more females than males wore them in
the cemetery, and also “throughout western Iran”. But this equation obscures
the reality that males also wore them in the cemetery (at least one example),
and in BB II. An adult male, 30–40 years old (skeleton 260; anthropologically
sexed) from BB II had two straight pins recovered “at neck”, one of which
was 8cm in length—short—and a third “close by”. Further, a male is rep-
resented with a pin placed seemingly on one shoulder on the Hasanlu gold
beaker.29 Marcus genders four skeletons [135 (see above), 139, 140, 255] as
citations in her (in-part) overlapping articles because the majority of the Hasanlu artifacts
(those in the many burials for example) remains unpublished in a systematic manner 40
years after excavation.
29 Marcus 1994, pp. 4, 7, fig. 7, for garment pins from the cemetery. On the Hasanlu gold
beaker two females wear pins with the points protruding up from the garment; and one
seated/squatting male wears a pin—only one is shown, on the proper right shoulder (Porada
1965, pp. 98–99, figs. 633, 64); other males and females represented here do not wear pins.
The male’s pin has a large rounded top with the pin tucked down into the garment; in
size it appears to be large. In 1993, note 23, Marcus cites this gold beaker male as “visual
comparanda” for her creation of fig. 14—but how, whether for straight or lion pins is not
explained (I cannot see any parallels either way, and the garment pins are not in the same
positions). To explain why a male is a model for a female in pin adornment, in 1994, p. 12,
294 chapter eight
female because they have “short garment pins at the shoulders”.30 For skele-
ton 139 there is no mention of a garment pin in the mortuary furniture listing
(the skeleton, in poor condition, was not completely excavated); for skele-
ton 255 a pin is listed as “near arm”. She does not mention that male skeleton
260 (above) wore garment pins, at least one of which was short (a straight
pin was recovered juxtaposed to skeleton 263 but where was not recorded).
Further, in other venues31 it is large pins, “extremely long”, not short ones,
that she associates with females (see below).32
Negative evidence is introduced to support the claim that
… none of the skeletons with lion pins can be associated with male-gendered
artifacts …
which includes belts, armbands and weapons.33 But this conclusion ignores
her normal methodology that accepts (where convenient) all the listed mor-
tuary furniture, including lion pins, as ‘associated’ with, that is, belonging
to the assigned skeleton. For example, the spears and the helmet men-
tioned above in the mortuary furniture with skeleton 135 are overlooked,
whereas the lion pins listed in the same furniture inventory are emphasised
as manifestly belonging to a female skeleton. Overlooked is a dagger and a
knife recovered centimeters away from skeleton 263, which Marcus sexes
as female;34 in Fig. 2 the top of the dagger is seen at the lower right. That
this juxtaposition could have been fortuitous, the weapons having fallen
from another doomed individual, is surely possible—but we cannot know.
Nevertheless, here is an example of male-gendered artifacts that are indeed
‘associated’ with a skeleton wearing lion pins. Skeleton 152, an adult, had
she casually states that the male wearing the pin on the gold beaker “suggests [to Marcus]
that certain cult officials were gendered female” merely because he wears a shoulder pin!
(Whether to her the pin-wearing males recovered at Hasanlu are also to be understood as
cross-dressers is not revealed to us.) This is bald politically correct behavior.
30 Marcus 1993, p. 170; for skl 135’s pin see Marcus 1994, p. 7, fig. 7, E.
31 Marcus 1994, p. 11; 1966, pp. 46–47, but compare this to p. 49 where both males and
p. 50, fig. 24) is not short, as she claims, but fairly large; as is that worn by the male on the
gold beaker (see note 29). On the former, a female, the pin points up, on the latter, down; see
also an unexcavated but western Iranian, probably Luristan, disc pin where what is probably
a female wears two very large pins with the pins up (Porada 1965, p. 88, fig. 60). Thus, long
pins were worn by males and females in western Iran, and the pin (at least for females) could
be positioned up or down indicating that neither the size nor position of garment pins was
gendered; see also note 31.
33 Marcus 1993, pp. 170, 174, and note 19.
34 Visible in Muscarella 1989, p. 32, fig. 14.
the hasanlu lion pins again 295
three lion pins listed as part of its furniture, and an iron sword over its leg;
skeleton 157, also an adult, had a lion pin listed in its furniture inventory, and
an iron spear point near a leg. Marcus should have brought the ‘associated’
weapons into the gender discussion of these three skeletons, if only because
her methodology of sexing by means of associated material logically should
have indicated to her that they could be sexed as males.
Sexing artifacts is a growing research strategy of archaeologists; all are
concerned with investigating gender preferences and interpreting indige-
nous culture-controlled rules, which can be different from culture to cul-
ture.35 Warranting investigation are the cultural rules that determined
gender-specific forms of jewellery: anklets, bracelets, earrings, necklaces—
and lion pins. Evidence for gender-shared and gender-neutral categories
of artifacts at Hasanlu and elsewhere is presented by Marcus;36 anklets are
included in the latter category and she notes
… there is absolutely no reason to assume that the lion pins were gender-
specific,
a viable conclusion. But, then, did men and women equally wear lion pins,
and what of children?
Unsexed skeleton 255, a young adult, wore one dangle earring, anklets,
bracelets, a necklace and lion pins. Skeleton 256, a child, wore dangle ear-
rings, a bracelet, and at least one lion pin. Skeleton 263 a young adult
(female?), wore a necklace of amber beads, one garment pin, and three lion
pins. Now, Marcus claims that dangle earrings like those worn by skele-
tons 255 and 256 were recovered “exclusively with females” in the ceme-
tery burials. Since this evidence is not published, it cannot be verified,37
but if this earring-specific gendering is eventually demonstrated—dangle
earrings equal female (and also a number of earrings, still to be identified,
equal males)—then a child and a young person who wore lion pins may
have been females.38 Unless one accepts the weapons close to, “associated
35 Thus, in the Near East from the third millennium down through the Achaemenian
period, only royal males were represented with parasols, whereas in Classical Greece only
women and openly homosexual males were represented holding them. See my “Parasols in
the Ancient Near East”. 1999 SOURCE XVIII, 2: 1 ff.
36 Marcus 1993, p. 170; 1995, p. 2500, and fig. 14.
37 What is published are drawings of earrings that we are told come from female burials
strated nearly by the Ashurbanipal garden scene relief: the king and queen wear the very
same earring form; the females to the left of the queen and the males (eunuchs?) to the king’s
296 chapter eight
with”, skeleton 263 as evidence it is a male, one cannot otherwise yet iden-
tify a skeleton that manifestly wore lion pins as male (as noted the sex of
skeleton 263 remains ambiguous). It is probably correct to state that archae-
ologists have not demonstrated that lion pins were or were not gendered at
Hasanlu—only that it seems some females wore them. What appears cer-
tain is that whether gendered or not, they functioned as a marker exclusive
to Hasanlu.39 Oddly, especially given her figure 14, Marcus presents an anti-
gendering conclusion:
In the end however, in the absence of good physical criteria, the biological sex
of the skeletons with lion pins remains uncertain.40
In this context it would be valuable if one could determine not only if
straight pins were gendered by size, that is, whether or not males and
females wore different sized straight pins (above; more investigation is
needed), but relevant here, whether this situation equally existed for lion
pins. No one has ever introduced the subject of size differences among the
lion protomes and what this might suggest.41 An incomplete investigation
on the lengths of the lions yields a considerable variety: 3.6; 3.8; 4.2, 3; 4.8;
5.0; 5.2, 5.4, 5.6, 5.7, 5.8; 6.2; 6.4; 6.5; 7.1 and 9.3 cm. The longest is nearly
three times the length of the smallest. The two lions deriving from skele-
ton 256 (child) are 4.8 and 5.0cm in length; the three with skeleton 263
(young adult) are 5.4, 5.8, and 6.2cm in length. The three lions associated
with skeleton 255 (young adult) are 3.6, 4.3, and (the one from the humerus)
5.2cm in length. The measurements are not uniform, and the young adults’
pins do not closely conform in size, one being among the smallest in size
(3.6cm). Two proposals deserve consideration: that size differentials reflect
hierarchical differences among a select (still to be defined) selfreferential
group (religious devotees?), whether gendered or not; or, size relates to
right wear forms different from the royal couple as well as from each other. In the attendants’
case, gendering may inform the difference, in the royal situation high rank may inform the
wearing of the same earring form. If so, we may have before us two separate (and simulta-
neously depicted) functioning levels of artifact selection: gender and rank. See Barnett 1975,
pl. 170.
39 Marcus 1933, p. 165, plays with this idea; see below in my text. I find it of interest to
note the formal parallel of the lion pins to a metal foundation peg with a recumbent lion
terminal excavated at Bismya (Banks 1912, p. 237; I know this peg from a discussion with Jean
Evans): is there an underlying ideological issue here? And how does the Tishatal/Urkish lion
foundation peg, here with an upright lion (Muscarella 1988, No. 495), fit into this ideology?
40 Marcus 1993, p. 170, note 19.
41 Marcus 1995, pp. 2497, reference is made that they are “huge, heavy”, see also p. 2503. In
Muscarella 1988, I neglected to give measurements for the lions per se.
the hasanlu lion pins again 297
gender and age, large pins for males, medium ones for youths and women,
and small ones for children. This latter proposal is not mutually exclusive
from the first. But as we have seen, there is insufficient data to speak either
way about size differentials and age, but the limited evidence suggests that
size is not related to age. However, because of skeleton 256, we do know that
some children wore lion pins at Hasanlu.
The perception that women wore lion pins triggered Marcus to construct
a significant cultural interpretation, one founded under a canopy of theory:42
It is tempting [!] to think that competition over resources …
and Urartian and Assyrian expansion led to a situation where lion pins
… created a need to stress … unambiguous identities,
and that they
… may be seen as part of a broader cultural desire to create the illusion of an
armed and well-defended society … an actual military presence … as well as
artistic representations of military might …
This is especially the case, inasmuch as the lion pins
… resemble swords and daggers with lion hilts (cf. fig. 10) …
In fact, Marcus’s figure 10 shows only a non-human headed hilt that is not
a lion head (it looks like a snake), and, in any event, the weapon depicted
here derives from Sialk, not Hasanlu. Further, the lion pins
… carried a more overt [than straight pins!] message about military strength
…
and
… especially compelling about the Hasanlu situation is the notion [to the
author] that body ornaments and hence human bodies played an active role
…
in not only announcing social rank, but also
… promoting an ideology of military strength and power …43
And last but not least, women wore lion pins because their
… bodies complemented or reinforced traditional male signals of military
strength [and] adorned females would have been an essential part of the
42 Marcus 1993, p. 168; an attenuated version of the outside threat and lion pin reaction to
44 Marcus 1993, p. 172; note that in 1994, p. 10, she presented the same interpretation for
doned.
46 Marcus 1996, pp. 42, 46, 48, 49.
47 In Marcus 1993, p. 165 BB II is a palace. In 1995, pp. 2496, 2502, BB II’s identity is shifted
and it is called a temple. The function(s) of BB II is complex and is still not satisfactorily
explained.
48 Marcus 1995, p. 2503.
49 Marcus’ fig. 14 seems to illustrate the sense of the Western male labor song, here of
course transgendered: “When you see me coming, better step aside/a lot of men didn’t, a lot
of men died”. Whatever the case, women died alongside men at Hasanlu.
the hasanlu lion pins again 299
Bibliography
Banks, E.J.
1912 Bismya: or The Lost City of Adab. New York and London: G.P. Putnam’s Sons.
Barnett, R.D.
1975 Assyrian Sculpture in the British Museum. Toronto: McClelland and Stewart.
50 I cannot refrain from quoting Marcus (1995, p. 2488): “The Hasanlu material, however,
Dyson, R.H.
1964 “In the city of the golden bowl”. Illustrated London News 12 (September):
372–374.
Marcus, M.
1993 “Incorporating the body: Adornment, gender, and social identity in ancient
Iran”. Cambridge Archaeological Journal 3 (2): 157–178.
1994 “Dressed to kill: women and pins in early Iran”. Oxford Art Journal 17(2): 3–
15.
1995 “Art and ideology in ancient western Asia”, in Civilizations of the Ancient
Near East, edited by J.M. Sasson, J. Baines, G. Beckman and K.S. Rubinson,
pp. 2487–2505. New York: Scribners.
1996 “Sex and politics of female adornment in pre-Achaemenid Iran (1000–800
B.C)”, in Sexuality in Ancient Art: Near East, Egypt, Greece and Italy, edited
by N. Boymel Kampen, pp. 41–54. Cambridge: Cambridge University Press.
Muscarella, O.W.
1988 Bronze and Iron. New York: Metropolitan Museum of Art.
1989 “Warfare at Hasanlu in the late 9th Century B.C.”. Expedition 32 (2/3): 24–36.
Porada, E.
1965 The Art of Ancient Iran: The Art of Pre-Islamic Times, London: Methuen.
the hasanlu lion pins again 301
Introduction
Evaluation,” Bulletin of the American Schools of Oriental Research 342 (2006): 69–94.
† This article is the revised version of a lecture delivered at the Albright Institute of
Archaeological Research in Jerusalem in March 2006, when the author was the sixth Trude
Dothan Lecturer in Ancient Near Eastern Studies. This series, which also includes lectures at
Al-Quds and Hebrew Universities, was sponsored by the Albright Institute and endowed by
the Dorot Foundation.
306 chapter nine
venues; contradictory data, conclusions, and claims are often not addressed
or explained in subsequent publications. Hasanlu is a site about which it
can be said that too much has been published, and at the same time that
not enough has been published. To some extent, this state of affairs has
been on my mind for years, especially because of my long involvement with
the Hasanlu excavations and Iranian Iron Age archaeology. But only when I
first systematically reviewed and indexed the contents of the complete pub-
lication corpus in its published sequence did the full extent and range of
the problem fully surface. In addition to issues I already knew or sensed,
other significant ones—concerning the prevailing, characteristic manner
and method of reporting information and data—were exposed in the course
of this research.
The present study—an analysis and critique of the publications and how
the site was published—was generated precisely because of this review.
The questions I asked were these: What data and information have been
published and what have not? In what manner have they been published?
What do archaeologists know and not know about Hasanlu? And why?
Publication History
1 Two articles are essentially the same, Dyson 1962 and Dyson 1966; several are joint
articles; and I suggest that Dyson 1999a and 1999b should have been published as one article.
the excavation of hasanlu: an archaeological evaluation 307
organized in the traditional way. Text citations to works in this latter bibli-
ography are indicated with an asterisk (*).
Researchers must then organize a system to record data and information
systematically in the chronologically published sequence on specific or gen-
eral topics included in the many publications. It is frustrating work. Reports
on a specific year’s campaign are uneven, often brief, at times merely a para-
graph or a page (see the Addendum below). Sometimes it is unclear which
campaign season is being discussed. (For example, Dyson 1960c is in part a
report on the 1959 campaign, but this is not mentioned. I recognized this
only because the Bead, Artisan’s, and South Houses were excavated that
year.) And it soon becomes apparent that, along with new information scat-
tered across the publications, there is also a repetition of the same topics,
and photographs of the same artifacts are presented-some four to six times
in as many publications-while many others remain unpublished. Confusion
also arises from the character and minimal quantity of the information pre-
sented, from the many changes in factual statements and interpretations
presented across time with no back referencing or explanation, and from
the methods chosen to reveal the excavation results from each campaign.
To communicate and comment on these matters does not result in an
easy read, although this is not my design. Rather, I would argue, the difficulty
comes precisely from the very nature of the issues I address. And it is a major
aim of this paper to aid scholars to organize and facilitate future research on
this remarkable site.
Excavation History
The record of the early investigations at Hasanlu is given in Hakemi and Rad
1950 and Dyson 1967: 2951. In 1934–1935 Hasanlu was commercially dug by
licensed dealers. Stein (1940: 388) reported “pits dug here and there … for
Jew dealers from Urūmı̄yeh.” Some of the material recovered was published
by Ghirshman (*1939: 78–79, pl. C). The first manifest excavation there was
accomplished by Aurel Stein in 1936, published in 1940. In 1947 Hasanlu
was again commercially dug, by Mahmud Rad, some of which material
was published by Rad and Hakemi in 1950. The second excavation of the
site, a campaign of two months, was conducted in 1949 by Ali Hakemi
(Hakemi and Rad 1950). Finally, beginning with a 10-day survey in 1956
(Dyson 1957: 37–39), from 1957 to 1974 nine excavation campaigns were
directed by R.H. Dyson, Jr. The term “Hasanlu Project” was coined to identify
the strategy of excavating, sometimes simultaneously with Hasanlu, other
the excavation of hasanlu: an archaeological evaluation 309
2 Dinkha Tepe: a Bronze and Iron Age site (the former cultural area directed by Dyson, the
latter by Muscarella (Muscarella * 1968; * 1974)); Agrab Tepe: directed by T. C, Young, Jr. and
Muscarella (Muscarella * 1973); Qalatgah (Muscarella * 1971a; see n. 19 below): the latter two
are both Urartian sites (along with Hasanlu III B); Sé Girdan tumuli: directed by Muscarella
(* 1969; * 1971b; * 2003), a late fourth millennium b.c. Maikop site, The Neolithic sites of Pisdeli,
Haji Firuz, and Dalma Tepe were excavated in 1958, 1959, 1961, and 1968, directed by Charles
Burney, T.C. Young, Jr. and M. Voigt. Ziwiye was added to the Project’s goals; it was surveyed
and excavated by Dyson for several weeks in 1964; it remains incompletely published (Dyson
1965b: 205–206. and 1972: 50–51). See also Dyson in Dyson and Muscarella 1989: 22, n. 1. In
addition to Hasanlu, the sites of Dinkha Tepe, Sé Girdan, and Haji Firuz were first recorded
by Stein: he was the guide, and the Hasanlu Project followed in his footsteps.
310 chapter nine
Compare also Dyson 1997: 479, fig. 1, and 1989a: 10, fig. 11 (see fig. 2, this
article) with Dyson 1989b: 115, fig. 10 (see fig. 3, this article); the latter is more
accurate. And various pieces of information about specific structures are
presented in different publications-with rare reference to earlier statements
or interpretations.3
An extramural cemetery was excavated below and adjacent to the citadel
mound, in the area called the Outer Town (Dyson 1958a: 27; Young 1963:
22–23; Dyson 1989b: 107); and some burials were also recovered on the
citadel itself (Dyson 1965a: 158). Although the cemetery was a major source
of juxtaposed pottery and artifacts, no definitive information about the
precise number of burials recovered, their period-by-period breakdown, or
group contents have been published. In Muscarella 1966: 134, I claimed
there are “over 150 graves” from Periods II to V (see also n. 11). As late as
January 2005, in response to email enquiries, M. Danti told me that the
number of burials from all cultural periods recovered is not known (!), that
the figure remains to be tabulated (there are probably scores of burials).
Further, not a single burial from any period has been completely published
with all its burial goods and contexts discussed (see below for specific
details).
The following commentary is a synopsis and evaluation of what I submit
we know, and do not know, from the publication record regarding the
various cultural periods at the site, which are numbered consecutively in
Roman numerals from the top of the mound down, I to VII. Excavation
was controlled by laying out 11 × 11m squares, leaving a 1 m balk. These were
originally assigned Roman numerals, XL, XLI, etc. (Dyson 1959a: 8; 1966:
fig. 16-3; Young 1959a: 4), and had been laid out as a grid plan; these labels
were later converted to A, B, C, AA, BB, and so forth.
3 In rare examples, Dyson, in Dyson and Muscarella 1989: 22, n. 1. states that his publica-
tion 1965b is an “outdated article,” and Dyson 1999a: 136, says fig. 12, is a revised pottery chart
replacing an earlier one, but in neither case does the author give reasons to explain these
comments (see also Hasanlu III B and A issues, below).
the excavation of hasanlu: an archaeological evaluation 311
and the latter’s fig. 1027 provides drawings. Dyson and Pigott (1975: 182) noted
that there are three phases in this period and briefly discussed the pottery.
That Early Transcaucasian Culture wares were present here in VII C was
first reported in 2004 (Danti, Voigt, and Dyson 2004: 586,594–596, fig. 11:1).
The latter publication is the most extensive report of Period VII published;
included (pp. 588–593, 602–616) are descriptions and drawings of the pot-
tery wares from the three respective phases; and a section drawing is also
provided (2004: fig. 2). The paint-decorated vessel is mentioned, but not
illustrated (2004: 588), btit no burial is mentioned.
Period VI
Period VI is called the “Khabur Ware” or “button base” period, named after
very similar pottery from northern Mesopotamia, or “Late Bronze Age
Period VI” (Danti, Voigt, and Dyson 2004: 586). This pottery is mentioned
in Dyson 1958a: 27, 30; 1963a: 1–32; and 1965b: 193–195, where three ves-
sels are shown with comparanda (pl. 31). In Dyson * 1973: 703–706, the
Bronze Age period is discussed, but Hasanlu VI is barely mentioned; and
Dyson and Pigott 1975: 182 has one paragraph in which a fortification wall
is reported. In Dyson 1989b: 108–109, this period is again briefly mentioned,
and here for the first time we learn of a settlement, and that it is “consider-
ably larger than Iron Age Hasanlu.” This important information was left in
abeyance. To date, no plan of this settlement has been published. A burial
with one occupant is mentioned with its finds, but there are no photos
(Dyson 1958a: 30–31). Inadvertently, another Period VI burial containing a
male, female, and child, along with istikhan-shaped vessels, was reported
by Dyson (1964e: 36) but was considered by him to be from Period V (see
below).
Regarding chronology, Dyson 1963b: 33 gives a date “to around 1500bc”; in
1965b: 193, it is ca. 1600–1250 bc for the occupation range. Radiocarbon dates
are mentioned in Dyson 1968: 85 and 1977a: 165, as evidence that Period VI
terminated in, respectively, 1400 or 1650bc. Later, in Dyson 1989a: 6, this
same event is presented as having occurred ca. 1450 bc As correctly stated by
Dittmann (*1990: 112), the termination date for Hasanlu VI remains unclear.
We have very little information about Hasanlu VI, as Dittmann (* 1990: 108)
also indicated, noting that Dinkha Tepe, about 15 miles west of Hasanlu,
is presently the prime source for information about the Khabur culture in
northwest Iran.
312 chapter nine
Period V
In the Hasanlu literature, Period V has been labeled Iron Age I by Dyson
(1965b: 211, table 2;4 on this terminology, see the chronology discussion
below). Hasanlu V represents quite distinctly the appearance of a new
culture and population in the area (Dyson 1977a: 156, 166; Muscarella 1994:
140), with a new, characteristic monochrome, grey ware totally replacing the
previous painted wares; an extramural cemetery; and structures of different
form overlying the Period VI level, with “no visibly eroded surface or erosion
deposit” separating them (Dyson 1965b: 195). These buildings “lie for the
most part directly below” the following Period IV level (Dyson 1977a: 156).
However, no section drawings have ever been published to illustrate the
nature of the VI-V-IV transitions as described in print; but it is manifest.
Young (1963: 40) mentions an upper and a lower phase.
4 Young (1963: 133 and following) introduced the ceramic terminology Early and Late
Western Grey Ware Group, and Late Buff Ware Group, terms continued by him in 1965, to
define the three cultural levels. This tripartite terminology was relabeled Iron I, II, and III
by Dyson (in 1965a), who also employed there a variant of Young’s 1963 terminology (“Early
Monochrome Grey Wares”), but did not cite its original source; Piller, in Stollner, Siotta,
and Vatandoust * 2004: 315–316, recognized this. The terms Iron I, II, and III were used by
Dyson periodically (1973c; 1977a; 1999a: 115; Dyson * 1973: 705–707, 711), which we all followed.
For criticisms of the term Iron I employed for Hasanlu, see Muscarella *1974: 79. The most
vigorous criticism of the chronological problems associated with the Iron I and II terms is
Young 1985: 362, n. 1.
the excavation of hasanlu: an archaeological evaluation 313
the burial-was first reported in Muscarella *1968: 195; 1988: 378, n. 1; later in
1994: 141, fig. 12.2; also Young 1985: 373, n. 11.
Only one painted vessel derives from Hasanlu, from a Period V burial
excavated by Stein (1940: 401, fig. 110, pls. 24.3, 31.8). Remarkably, a very sim-
ilar vessel was excavated at Dinkha Tepe (Muscarella * 1974: 39–40, figs. 3,
5) from what I designated to be a Dinkha III (contemporary to Hasanlu V)
burial, recovered alongside other contemporary burials in the extramural
cemetery. Dittmann (*1990: 111, n. 33) discusses the Dinkha burial and con-
siders its contents to be from Dinkha III. Whether these vessels represent
strays or heirlooms from Period VI, or were imports, remains unknown.
Pottery remains essentially unpublished (noted by Medvedskaya * 1982:
33). Young (1963: 39–43; 1965: 70–72) presents a good review. But as Levine
(*1987: 233) stated, one looks to Dinkha Tepe for ceramic knowledge of this
period in northwest Iran, and Young (1985: 366) notes that Dinkha Tepe
“remains the type site for” this period (see also Piller in Stöllner, Slotta, and
Vatandoust * 2004: 315, 317, 325, n. 16; and Muscarella * 1968; 1974). Unpub-
lished also is a corpus of pottery recovered below Burned Building V of
Period IV, which is described as “typologically closer to that of period V,”
apparently representing transitional wares from Period V to IV (Dyson 1973a:
303). A decade later, Dyson (1983/1984: 303) also records that ceramics exca-
vated in 1972 and 1974 show a “gradual evolution of period V forms into those
of period IV with increasing elaboration.” Excavated over 30 years ago, these
important late Period V–early Period IV wares remain unpublished. Such is
the present condition at Hasanlu, where the Iron I grey ware introduction
and sequence were first charted.
The bones of 16 Period V skeletons from cemetery burials are recorded by
T. Rathbun (1972: 52–57), but the burials themselves remain unpublished
(as frustratingly noted by Medvedskaya *1982: 33).5 In some publications
(Dyson 1958a: 31; 1960c; 1960e; 1962: 639; 1964e: 36; 1965b: 196; 1966: 417; 1967:
2957), burials and their pottery, and beads, bracelets, and one iron ring are
sometimes cited, but no photos are published. The ring is the only iron
artifact recorded as deriving from Period V; no iron artifact was recovered
from the same cultural period at Dinkha Tepe (Dinkha III; Muscarella * 1974:
38). Inasmuch as Dyson reported that “the major remains of Period V come
from the cemetery” (1965b: 196), the non-publication of the burials is all
5 In 1966, I made enquiries regarding the number of Period V burials excavated. I was
informed by Maude de Schauensee that there were 10 from this period and that 5 others were
unclear, either from V or IV.
the excavation of hasanlu: an archaeological evaluation 315
the more unfortunate (Medvedskaya [*1982: 33, 103] understood this). Stein
(1940: 401–402, pls. 24, 31, fig. 110) published a grave photo from this period,
and Hakemi and Rad apparently published a few (1950: figs. 8b?, 10?, 14), in
each case together with their pottery. That some graves contained important
artifacts is indicated by a belt buckle reported by Winter (1980: 27) “from a
burial.”
An unresolved problem concerns the chronology of Period V as presented
in many publications across 30 years. In 1960, Dyson (1960e: 132) gives the
beginning date for the settlement as ca. 1500bc. Full-range estimates are
given in later publications: 1963b: 33, as “the end of the thirteenth or begin-
ning of the twelfth century bc”; 1965b: 195, ca. 1250–1000bc; 1967: 2957, as
1200–1000bc, 1968: 85, 1300–1000 bc; Dyson *1973: 705, 712–713, has it begin-
ning “around 1350+/-50 bc”; and in (the later written) Dyson 1973c (fig. 5
caption), it is 1200–1000; in de Schauensee 1988: 45, it is 1450–1200 bc. In the
latest recordings, Dyson 1989a: 6, it is ca. 1450–1250 bc, but in 1989b: 107, the
beginning date is “shortly after 1500bc,” repeated by Young 2002: 386. These
chronological dates for the incipience of V significantly differ and are in con-
flict with one another in their published sequence: 1500, 1250, 1300, 1200,
1200, 1350±50, 1200, 1450, after 1500–a back-and-forth difference of some 300
years (see also below)! And the terminating dates are given several times
as 1100, 1200, and then as 1250, a difference of 150 years. Equally troubling
is that not once are we offered a reflection or explanation regarding the
different chronologies presented over many decades. The same inconsis-
tencies obtain for the 14C dates also given in multiple venues, differing from
one another by hundreds of years, sometimes coinciding with the archaeo-
logical dates given above. These are reported in Dyson 1962: 641; 1964e: 39;
1965b: 197, n. 7; 1966: 420; 1967: 2967; 1968: 85; 1972: 57. In Dyson 1977a: 160,
161, 164, MASCA-corrected dates for several of the V buildings are given as
1300bc, 1300–1360, and 1220–1260bc, which suggests an “emphasis on the
1300–1350bc range,” but in the same year that dates for Period V were given
as 1450–1250 bc (1989a: 6, above), Dyson (in Dyson and Muscarella 1989: 8)
says the 14C range is 1350–1150 bc. Muscarella (1995: 989) places the termina-
tion in “the twelfth century.”
What are the beginning and end dates for Period V (and concomitantly
the incipient date of Period IV)? Although the published record is con-
fusing, we do know that the culture arrived at Hasanlu sometime after
the Khabur Ware period terminated. If this event occurred shortly there-
after, then Period V would more accurately be designated in archaeolog-
ical terminology as a Late Bronze Age phase (chronologically following
Hasanlu VI). And if such were the case, Hasanlu IV (below) would more
316 chapter nine
Period IV
This level was labeled Iron II by Dyson (see above and n. 4); it is the most
extensively published period at Hasanlu (figs. 2, 3). A major and complex
elite settlement was built directly over the ruins of the Period V buildings
(Dyson 1977a: 156); this early level is labeled Period IV C. It is often described
by Dyson as a “direct continuity” from V (1977a: 166), revealing “no apparent
stratigraphic hiatus” and “the transition [sic] to it from Period V (IV C) may,
in fact, have been less abrupt than presently appears” (1965b: 197–198); also,
there was a “major continuity between periods V and IV [written as VI, a
typographical error],” and the IV buildings were “superimposed on earlier
construction” (1983/1984: 303). The settlement experienced a destructive
fire that terminated Period IV C, dated variously to the mid-11th centurybc
(Dyson 1977b: 550), 1100 bc (Dyson 1989a: 6; Dyson and Voigt 2003: 222), or
the 12th century bc (Young 2002: 386). In some IV C buildings, BB II, IV East,
and V, it was recognized that a rebuilding occurred directly over the original
walls (Dyson 1973c: 2; Dyson and Muscarella 1989: 1); this second building
phase is labeled Period IV B (Dyson 1989b: 111, 114; Dyson and Voigt 2003: 222).
A final destruction ensuing from a violent attack and intense fire is secured
to the end of the ninth century bc, ca. 800 bc, determined by archaeology
and 14C data. That this event was caused by an Urartian penetration into the
Ushnu area seems certain (Dyson 1969: 44; Dyson and Muscarella 1989: 22;
Dyson 1989b: 109; see also Ivantchik *2001: 212, 240, 258, 261, for the ninth-
century chronology). A so-called squatters settlement, IV A, was uncovered
over the IV B ruins (Young 1963: 27; Dyson 1977b: 550; Dyson and Muscarella
1989: 2).
In the first decade of excavation (Dyson 1959a: 14; 1960e: 133; 1963b: 33;
1965b: 197; 1967: 2964; 1968: 85; 1969, 44), the beginning date for the period
was given as ca. 1000bc. That date shifted to 1100 bc in Dyson 1972: 42;
1973a: 303; and 1973c, 1; then earlier, to the mid-12th century bc in Dyson
1977b: 550; in 1989a: 7, it is a hundred years earlier, 1250bc, but in the
same publication (Dyson 1989b: 108), it is “the 12th century bc”; a few years
the excavation of hasanlu: an archaeological evaluation 317
later (Dyson and Voigt 2003: 221), it is “around 1250 bce.” Explanations for
these important revisions, a range of 250 years for the incipient date of
Period IV-and concomitantly for the termination of V (see above)—were
never presented or confronted. Radiocarbon dates are given in Dyson 1962:
642; 1965b: 197, n. 8; 1966: 420; 1972: 57; each date is different. When did
Period IV commence?
An intense fire that caused the collapse of the buildings and the violent
death of many individuals marks the final destruction (Muscarella 1989: 32,
35, figs. 14, 20). A total (?) of 246 skeletons of men, women, and children
were recovered in all areas of the IV B settlement; a good number had head
wounds from enemy violence, indicating that the fire and subsequent build-
ings’ collapse occurred in the course of the violent struggle within the site
(Muscarella 1989: 32–34, 36, fig. 21). Many skeletons were recovered in the
collapsed buildings, 62 in BB II alone (Dyson [1965b: 202] says “over 40 indi-
viduals”; Muscarella 1989: 32–35, figs. 14, 20; 2004).6 In all publications since
1965, the destroyers are assumed to have been the Urartians, which indeed
best fits the historical scenario (Muscarella *1971a; Dyson and Muscarella
1989: 19, 20–22; Dandamaev and Lukonin *1989: 19; Piller in Stöllner, Slotta,
and Vatandoust * 2004: 319).
Grey monochrome wares continue, but different forms are introduced.
Pottery is often mentioned, sometimes illustrated in drawings and a few
photographs, and contexts are rarely given (Dyson 1958a: figs. 23, 26, from
burials; 1960e: fig. 4, and from the settlement, figs. 6, 8; Crawford 1961: 87–
89; Dyson 1964e: 38, fig. 4; 1967: pl. 1485; 1989b: 113, fig. 7). For pottery from
the earlier excavations, see Hakemi and Rad 1950: figs. 15–28, and Stein 1940:
pls. 24, 30, 31 (all mixed with Period V vessels). The best publication of
the modern excavations of Period IV pottery is the analysis and drawings
in Young 1963: 31–39, pls. 8–12; also Young 1965: 55, 74–78, figs. 6, 7, and
the photographs in Dyson 1968: 84, 94–97. The bridge-spouted vessel, a
characteristic shape from this period, is common in burials, where it was
sometimes (but how often?) placed on a tripod (Ghirshman * 1939: pl. C; Rad
and Hakemi 1950: figs. 20, 23; Dyson 1960c: 121; Crawford 1961: 88, fig. 2). But
many other forms remain unpublished. Again, for this period’s pottery, in
burial contexts, see Dinkha Tepe (Muscarella *1968; * 1974; Piller in Stöllner,
Slotta, and Vatandoust *2004: 315, Abb. 7).
6 Because of the non-publication of the skeletons recovered in BB IT, I had to acquire the
2004 data from the Hasanlu skeleton (Burial) files (see n. 12 below), which were promptly
supplied to me by Dyson.
320 chapter nine
7 In an unsigned note, “Hasanlu, Iran,” in Expedition 1958 (Fall); 35, it is more accurately
described as a “large bucketshaped vessel.” It was the discovery of the “Gold Bowl” and the
public attention it received that projected the excavation and its director into prominence.
The first time it was published (spectacularly) was in Life magazine (“The Secrets of a Golden
Bowl,” January 12, 1959; 50–60). References to the vessel by the Hasanlu staff include Dyson
1960c; 1960e; 1960f; 2003; 45–46 (in a number of cases the “Golden Bowl” is included in the
title); also Muscarella 1987, and Winter 1989. Drawn depictions in several publications vary
in small details. It was Charles Burney who made the original drawing at Hasanlu in 1958
under difficult circumstances (a fact mentioned once, in Dyson 1959a; 14); M. de Schauensee
modified it in 1960 from photos, and in 1974 from photos and autopsy in Tehran. The “official,
corrected, drawing of the bowl, designated as MTMS 1974” (Dyson 2003: 46) is usually signed
by de Schauensee: see de Schauensee and Dyson 1983: fig. 5; Winter 1989 90, fig. 6; and Marcus
1991: 555, fig. 28. For a discussion of the various drawings and issues of their accuracy, see
Muscarella 1987: 137. That the vessel is in fact a beaker (see the drawing in Winter 1989: 88,
fig. 3) was noted in Muscarella 1989: 34; 1995: 991; and 1996: 211; also by Piller in Stöllner, Slotta,
and Vatandoust * 2004: 708. The bowl/beaker was recovered in BB I, Room 9—not 6, contra
Dyson 1989b: 124, left; see Winter 1989; 88: fig. 2; on p. 89 Winter says the silver beaker was
recovered in BB I East, but it was BB I West.
the excavation of hasanlu: an archaeological evaluation 321
and a fifth (field no. 64–611a) was recovered in a shop in the nearby town
of Nagadeh. All are clearly related in form, style, and iconography, clearly
reflecting an Assyrian background, either imported from there or closely
adapted (see Porada *1965: 120), and all the fragments derive from one
vessel; in fact, one of the 1964 fragments, the one purchased in Nagadeh,
neatly joins to the 1960 fragment. The 1960 fragment was published in
Crawford 1961: fig. 5; in Dyson 1962: fig. 5; in Porada * 1965: pl. 33; and again
by Dyson in 1966: fig. 16.5, and in 1968: 93—but in neither of the latter
two, nor in subsequent reports, were mentions made of the additional
fragments or the join made years earlier. Three of the 1964 fragments were
published for the first time in 2004 in a German museum catalog, courtesy
of the Tehran Museum (Piller in Stöllner, Slotta, and Vatandoust * 2004: 733,
no. 405). A number of other Egyptian Blue objects reflecting an Assyrian
background excavated in 1960, 1964, and 1970 also remain unpublished. For
several 1964 published examples of excavated Egyptian Blue objects from
BB II, see Piller in Stöllner, Slotta, and Vatandoust * 2004: 711, no. 362, also
710, no. 359 (the same object as Dyson 1972: 46–47, fig. 6, and Winter 1980:
22–23, fig. 58); another Egyptian object is shown in Winter 1980: fig. 59;
and Muscarella 1966: 133, figs. 30,31, and 33; for the probably Assyrian (or
North Syrian)-made lion bowl, see van Loon 1962 and Muscarella 1965b:
fig. 3.
The corpus of ivories, seals, and glass artifacts is fully published, as are the
imported lion bowls, ivories, and glass. Most of the artifacts are certainly
locally made, called “local style,” especially among the ivories, seals, lion
pins, and metal productions. For groups of artifacts from various Hasanlu IV
loci published together in one venue, see de Schauensee 1988; and Mus-
carella 1988: 15–79. Full or fairly extensive publications exist spread over
many venues:
– Ivories: Muscarella 1980.
– Glass, imported, including mosaic forms: von Saldern 1966; Marcus
1991; de Schauensee 2001.
– Seals: Dyson 1986; Marcus 1989; 1990a; 1990b; 1994a; 1996a (repetitive);
Dyson and Harris 1986.
– Metal vessels: de Schauensee 1988; Muscarella 1988: 24–32.
– Weapons and armor: Hakemi and Rad 1950: pls. 38, 39; Dyson 1964e;
Muscarella 1988: 53–63; 1989 (the drawings of alleged Hasanlu warriors
published here in figs. 3 and 10 were added without my knowledge and
are inaccurate. For an accurate, ancient, representation of a Hasanlu
warrior, see Muscarella 1980: nos. 54, 55—with earflaps, mustache, and
322 chapter nine
8 My 2004 article was written to confront claims by Marcus (1993; 1994b; *1995; and 1996b)
about alleged gendered function and cultural use of lion pins and other pins at Hasanlu.
the excavation of hasanlu: an archaeological evaluation 323
were Assyrian types and which plain glazed tiles (see 1989b: 121, fig. 18b). For
heirlooms, earlier artifacts recovered in IV B, see Dyson and Muscarella 1989:
1–2, and n. 8.
Relatively speaking, architecture is the most thoroughly published of all
the Hasanlu data, but this information is spread out in various publica-
tions (see figs. 2 and 3). For substantive architecture publications to date,
see Young 1966; 1994; 2002; Dyson 1977b; 1980; 1989b; and Muscarella 1988:
19, n. 4, 208–209, n. 3. Several smaller and separate buildings, labeled the
Bead House, South House (directly to the west of BB II), and the Artisan’s
House, are briefly mentioned. The latter is described as a “private house”
or “a small domestic structure,” situated outside the main citadel zone, in
the Outer Town; it contained molds, crucibles, stone vessels, and mortars
(Dyson 1960c: 121–122; 1966: 423; de Schauensee 1988: 46; Dyson 1989b: 109);
its locus does not appear on any published plan. The other two buildings are
located close to the northwest corner of BB II. The South House is merely
cited; of the Bead House, it is reported that it contained many beads, bone
handles, and stone bowls, the latter two items being shattered (Dyson 1960c:
122–123; Porada * 1965: 112–113). None of these structures’ contents were pub-
lished. The major architecture consists of large, manifestly elite structures,
with multicolumned halls, stairways to an upper level, a roof or second floor
(probably the latter), porticoes, and side storerooms; they are designated as
Burned Building (BB) I, II, and so forth. Throughout the publications, incom-
plete or inconsistent settlement plans are illustrated, creating a source of
confusion. Further, while it is stated in several publications that porticoes
were added to the fronts of several of the Burned Buildings in Period IV B
when they were reconstructed following the IV C fire (BB II, III, and IV–IV
E, and V: Young 1966: 55; 2002: 387; Dyson 1965b: 198; 1973a: 303; 1989b: 114;
Muscarella 1996: 210; Dyson and Muscarella 1989: 1), this is explicitly denied
for BB II in Dyson and Voigt 2003: 222, where it is claimed the portico was
present in IV C–but the earlier claims to the contrary were neither cited nor
explained away.9
9 The caption to fig. 20.1 (Dyson and Voigt 2003: 220, which “supercedes a preliminary
[sic] drawing” in Dyson 1989b: 112, fig. 6a) is confusing and in contrast to text references about
the portico existing in IV C (p. 222). In a personal communication (February 4, 2005), Dyson
informed me that recent research revealed that BB II’s portico had been added before the
fire terminating IV C, and thus belongs late in that period. However, the portico of BB IV was
added after the fire, constructed in IV B. (Note also that the caption for Dyson 1989b: 117, fig. 13,
an aerial view of the buildings, is completely erroneous.)
324 chapter nine
is being described within a section, and there are no figure citations in the text for the
illustrations.
326 chapter nine
positions are misleading in this respect, specifically, Dyson 1997: 479, fig. 1;
Winter 1980: fig. 2; Muscarella 1980: plan 1; Dyson and Muscarella 1989: 2,
fig. 1; and Dyson 2003: 42. For a correctly drawn plan, see Muscarella 1988:
17, fig. 2; 1995: 992, fig. 8; Dyson 1973c: fig. 8; 1989b: 115, fig. 10 (see fig. 3,
this article; but compare in the same journal, Marcus 1989: 58, and Dyson
1989a: 10, fig. 11; see fig. 2, this article; Dandamaev and Lukonin * 1989:’ 16
is completely wrong). Dyson and Pigott (1975: 183) and de Schauensee and
Dyson (1983) publish brief comments and some artifacts. Young (2002: 386)
says BB V was constructed “at the same time as” BB II, IV E, and III, but in
1983/1984: 303, Dyson says that V is older than BB II, offering no reference to
the previous claim.
BB IV–V, the corridor building between BB V and BB IVE, is mentioned
by Dyson and Pigott (1975: 183), Dyson and de Schauensee (1983: 71–75),
and later by Dyson (1983/1984: 303). Many pieces of horse gear/trappings are
listed, including one of the most striking artifacts recovered from Hasanlu:
a bronze breastplate decorated in relief with the figures of a powerful deity
holding two bulls. This magnificent object was fully published by Winter
(1980), along with the juxtaposed horse gear (see also Muscarella 1988: 66–
72,74–75, nos. 95–111, 120).
BB VI and BB VII are the two structures at the west of the mound,
adjacent to the “Gateway,” and are partially published (Dyson 1975: 181–182,
fig. 1; 1989b: 111–112, figs. 4, 6b). The “Gateway,” a major construction with
“enclosing walls” at the west of the site, was partially published with a plan
in 1975, and interpreted to be an access to the main structures of the citadel.
(This interpretation was challenged by Kroll [* 1992], who argued that the
structure was not a gateway but a stable.)
Replasterings were noticed and counted on a “block” on the hearth in
BB V. It is claimed that they were “burned and re-plastered perhaps 200
times” (Dyson 1977b: 551); but three years later the block is now a “column,”
and the number of replasterings is lowered to “over 100 times” (Dyson 1980:
150). No reason for the significant adjustment is presented (the latter figure
was chosen by Dandamaev and Lukonin [*1989: 17]).
Many of the contents of all the Burned Buildings remain unpublished
(BB II comes out best here, but often published artifacts from BB II are not
identified as such); no full description of the contents exists for any building
(pace Young’s comment of 40 years ago [1966: 48] about a forthcoming final
report).
Hasanlu IV seems to have had no fortification wall, although in many
publications up to 1975, it was accepted that the citadel was fortified; on this,
see the discussion below for Period III B.
the excavation of hasanlu: an archaeological evaluation 327
100 Period IV burials; in 1987 I was informed by Mary Virginia Harris of the University Museum
that there were 83 from this period. We thus have 200, 100, and 83 Period IV burials!
12 The cause of the disorder results from the recording system in effect throughout the
excavation campaigns: skeletons recovered at the site were registered on a so-called Burial
Sheet and assigned a beta-symbol and number provided in sequence to each skeleton as
encountered. Thus, all skeletons, whether excavated in burials in the cemetery or on the
citadel, or those who perished in the destruction, were all equally registered with a beta
number on Burial Sheets, No separate listing of the burials per se for any period exists (see
n. 11, above).
328 chapter nine
Crawford 1961: 88, 94; Ivantchik *2001: 97, n. 4); Porada (* 1965: 108, 110)
refers to the artifacts as “The Art of the Mannaeans.” Some have called them
Hurrian speakers (Dyson 1961b: 64; 1962: 642–644; 1966: 421); or an unknown,
unidentifiable polity (as in Dyson 1968: 89), where they may be Mannaeans,
but the “ethnic identity of periods V and IV … is at present inconclusive
…” (on this, see also Muscarella 1971: 264; 1987: 136). Early on Dyson also
argued for an Indo-European, in fact an incipient penetration of Persians,
at Hasanlu (1963b: 33). Winter (1989: 101–103) claimed that there is no solid
evidence to connect Hasanlu V and IV with Indo-Europeans, but allowed for
a Hurrian presence. Young (1985: 368, 374–375) also suggested that they were
possibly Indo-European-speaking Iranians.
Salvini, among others, has identified the site as Meshta, a state known
from Urartian texts as existing somewhere in northwest Iran (Salvini * 1995:
25, 41–42, 46; Dyson and Muscarella 1989: 19, n. 105). Reade (* 1979) iden-
tifies Hasanlu with Gilzanu (not Meshta, contra Winter 1989: 102), but the
identification essentially hinges on his geographical placements of this and
other polities recorded in Assyrian texts, positions that shift continuously
in modern interpretations; Salvini (*1995: 25) accepts Reade’s identification
as viable. Also, Reade’s argument regarding Assyrian material and influence
at Hasanlu, while certain, could be but one example where such manifesta-
tions occurred.
Period III
This period is first mentioned (in Dyson 1959a: 9) as the level recovered
directly above the ruins of Period IV. It was alleged here and in a number
of subsequent publications to be one architectural and cultural unit, often,
until 1999 (below), referred to as the “Triangle Ware Phase” or period, based
on a painted pottery motif claimed to be present throughout Period III
(Young 1959a: 5; Dyson 1961a: 534; 1962: 641; 1963b: 33; 1967: 2964, 2966,
and fig. 1036). In 1963, Dyson (1963a: 131–132) reported that in fact two
distinct levels are present within “Period III,” the later labeled III A, the
earlier III B (also 1964b: 372; the A, B distinction was ignored in Dyson
1967: 2966–2967). The recognition of the two levels was not supported or
explained by means of a section drawing (see further below). Young (1965:
53–55) saw the two levels as two phases of one cultural period, based on a
contemporary understanding of the painted pottery continuation; but he
did not distinguish between the phases, calling all the pottery (1965: figs. 1,
2) “Iron III” Thus, the pottery of III B and A was not kept discrete in the
excavation and recording.
the excavation of hasanlu: an archaeological evaluation 329
The architecture and pottery of these two major cultural periods, III A
and B, still remain sporadically, inadequately, or not published, a condition
noted by Dyson (1999a: 118) as the “unpublished state of Hasanlu III data”;13
for earlier comments on this, see Muscarella *1973: 71; also Levine * 1987: 234,
who notes the “unpublished material from Hasanlu III.”
Period III B
Even subsequent to the III B and A stratigraphical distinction, pottery with
triangle decoration continued to be recorded as coming from the III B level
(Dyson 1963a: 132; 1964b: 372; 1965b: 204–205, figs. 9, 10, 13; 1966: 419; 1967:
2958, 2960). In Dyson 1972: 46, it is flatly stated that painted wares were “the
most identifiable pottery of III B” (italics mine); the same was repeated in
1977b: 549 (about which see below). Later, however, in Dyson and Muscarella
1989: 4, the very opposite is claimed-namely, that only “a few sherds of …
Triangle Ware occur” here. Other references to III B include Dyson 1964b
(some discussion); 1965a; 1965b (with more extensive discussion); see also
Dyson 1968; 1977a; 1989a: 5–6. Aside from Dyson 1964b and 1965b, basically
only pottery is discussed (see discussion below).
The architecture, called Period III over time, is described as “simple stone
house foundations” (Young 1959b: 65); as “rows of rooms” (Dyson 1961a:
534); as “row houses set side by side” (Dyson 1963a: 131); as “small rooms”
(Dyson 1966: 417); and as “a series of single and double barracks-like rooms”
(Dyson and Muscarella 1989: 3). But these structures have never been pub-
lished except in a plan labeled Period III, and a fortification wall recorded
as from Period IV, reused in III (below): Dyson 1959a: 9; 1964b: 372, fig. 1. The
period assigned was corrected to III B in the plan published in Dyson 1989a:
7, fig. 6 (fig. 4, this article). Levine (*1987: 234) records (as a “personal com-
munication” from Dyson) that “a re-examination of Period III” revealed that
III B had two occupation phases; this was first reported by Dyson much later
(1999a: 132)—his section drawing there (1999a: fig. 10) is unclear on this; and
his publication does not mention when or how the phase modification was
recognized.
13 For a remarkable grave accompanied by four horses and arrows (probably socketed)
excavated in 1947 by Rad and Hakerni, see Dyson 1960c: 121; 1965b: 208–211; Muscarella 1988:
219–220, n. 3; Dyson and Muscarella 1989: 21; Derin and Muscarella *2001: 193; Ivantchik *2001:
185, 276–278. It is manifestly post-Period IV B, probably post–700 bce, but whether associated
with Period III B or A remains unknown at present.
330 chapter nine
14 This article presented many good points. But parts of the article are confused. Johnson’s
n. 45 (information given to her by Muscarella) states that the wall was indeed built in
Period III, not IV, but this note is a later addition to the manuscript shown to me; see
my comments on this issue in Muscarella 1980: 214, n. 39, and in Dyson and Muscarella
1989: 23 n. 11. It is possible that the alleged stratigraphy of the fortification wall lay behind
R.D. Barnett’s erroneous claim (* 1982: 321, 341) that Period IV was an “Urartian level”—but
his n. 60 is meaningless as a reference.
the excavation of hasanlu: an archaeological evaluation 331
given verbally), that led in 1974 (1972 was not cited here) to a reexamination
to test “the hypothesis” that the wall was constructed in III B. Finally, in the
fourth reference (Dyson 1989a), there is no mention of the background that
led to recognition of the accurate construction date. Rather, it is presented
as a normal ongoing result of excavations in 1972; included is a photograph
(1989a: fig. 7) of the trench that “clarified” the stratigraphy. But there is no
mention of the contributions of Kroll or Kleiss.
Also published in 1989 is a photograph taken in 1962 (Dyson 1989a: 5,
fig. 4; see also Muscarella 1966: 120) that clearly shows the III B wall cutting
into and destroying BB I’s rear and side walls-very clearly shown in the
plan published in Dyson 1959a: 6. Young (1966: 53) had concluded that
“foundation trenches dug for the structures of Period III” had destroyed
the rear wall of BB I. But the implication of these observations was missed,
because it was a given then that the wall was a Period IV construction.
332 chapter nine
15 See also Kroll * 1976: 164. Such was the confusion that P. Calmeyer’s entry “Hasanlu” in
the Reallexikon der Assyriologie (Calmeyer 1975: 129) states that III A contained only plain
ware and III B Piitnted Triangle Ware.
the excavation of hasanlu: an archaeological evaluation 333
“Period II/IIIA.” And the erosion surface was not recorded in any previous
publication, although, as noted above, he had earlier (in 1963a: 131–132 and
1964b: 372) called attention to the two, A and B, phases. The visual mes-
sage of fig. 10 also does not correlate with the section published years earlier
(see below), nor does it relate to the descriptions of the stratigraphy given
in Young 1965: 55 (and also in Dyson 1965b: 212) that the “shift from the
occupation of III B to III A is not clear” due to erosion, and that some III
B walls “continued in use” or served as “foundations” for “new walls of III A.”
These 1965 statements are in conflict with the description and the section
published 35 years later. If the newly published section and data accurately
reflect the excavated reality, the clear separation of III B from the succeed-
ing level, how could the excavator have asserted vigorously for decades that
III B contained painted wares-which were in fact not present there?
Only one other section drawing for the Hasanlu excavations had previ-
ously been published (probably drawn in 1960). It appeared three times
(Dyson 1962: 640; 1966: fig. 16-3; 1967: 2962, fig. 1032), with a thin layer labeled
“III.” Albeit not stated in Dyson 1999a (it took me some time to recognize
this), this earlier published section depicts the area directly adjacent to that
shown in 1999a: fig. 10. Note that the site locus trench labels given, “Op.
XLII, Op. XXIX,” were later changed to AA28 and BB28 (none of these trench
identifications were ever shown in any published plan). We are not given
expected information either about the locus on the site for the separately
published, contiguous sections, their position on the published plan of III
B (see fig. 4, this article), or why the Period III remains are labeled differ-
ently in the earlier (“III”) and later published sections (“II/IIIA”). I was able
to determine that the two sections depict the area immediately adjacent to
the south wall of BB II, extending to the fortification wall of III B. In this area,
only one isolated Period III B building was excavated (fig. 4, this article). Why
was not a section published of the stratigraphical sequences of Periods III B,
III A, and IV in the only areas on the site where Period III remains are con-
centrated, adjacent to the fortification wall, and here apparently above BB I
and III? (A number of section drawings were made over the years.)
Of the artifacts recovered, few are published. These include bone psalia
(Dyson 1964b: 372, figs. 2, 3): the former figure is labeled as from “Period III,”
and the latter is mislabeled as from IV in the caption, but from III B in the
text; in Dyson 1965b: 211, it is also assigned to IV, here with a major but incor-
rect historical conclusion regarding the presence of Scythians in northwest
Iran in the ninth century bc. This error (never corrected by Dyson) was inad-
vertently used by several scholars to arrive at the same wrong conclusion
about a Scythian presence outside their homeland (corrected in Muscarella
the excavation of hasanlu: an archaeological evaluation 335
*1974: 79, n. 16; 1988: 66, n. 2, and 220, n. 3; Derin and Muscarella * 2001: 199
and n. 83; see also Ivantchik *2001: 277, n. 66).16 In Dyson 1964b there is also
the fibula mentioned above; and another Urartian fibula was recovered in a
(still unpublished) tomb (Muscarella 1966: 135, fig. 37; 1988: 47–48, no. 53),
which I labeled “III A or B”; it is surely from III B. In Dyson and Muscarella
1989: 20, eight Urartian fibulae are cataloged as from III B. Dyson (1965b: 207)
also mentions a socketed trilobate (?) arrow, recovered from fill dumped
from the III B structures; a socketed bilobate arrow was incorrectly pub-
lished by me as from Period IV (Muscarella 1988: 63, no. 90, but see p. 107)—
in fact, it derived from Period III B (Derin and Muscarella * 2001: 193, and
n. 16). Two Urartian-style stamp seals were published by Marcus (1996a: 147–
149, figs. 117, 118) as deriving from two tombs “outside” the fortification walls.
Neither tomb is published, and hence other contents remain unknown-an
unfortunate omission inasmuch as (unplundered) Urartian burials are rare,
at Hasanlu and elsewhere.
For Period III B burials, we have no information other than that they were
excavated (above). Dyson (1965a: 158) refers to “a number of Period III burials
… found in the upper layer of the Citadel” (see Period II, below), but which
locations continue to remain unknown; nothing is mentioned here about
cemetery burials. Rathbun (1972: 12–15) refers to 13 skeletons from this period
that he examined, but he does not mention whether they came from the
cemetery or the citadel.
Period III B was dated in Dyson 1965b: 204, to 700–600 bc; in 1973c: 1, to
650–600bc; in 1972: 46, to “between (?) 750–600 bc”; in Dyson and Pigott
1975: 182, to ca. 750–600 bc; in Dyson 1989b: 110, to 8th and 7th century bc;
and in Dyson and Muscarella 1989: 8, 19, 20–21, to between 800 and 600bc,
and probably 7th century. The issue of its chronological relationship to the
earlier IV B destruction and IV C “squatters” phase was not discussed except
by Young (1963: 47–48), who ponders that a long gap may have existed.
The Dyson 1973c publication was the first to mention that Period III B
was an Urartian settlement (except for the fortification wall), which can
be considered a historical reality (Dyson and Muscarella 1989: 4, 19, 20;
Muscarella 1989: 34–35; Dyson 1989b: 104). The terminal date of ca. 600bc
+/- is approximately correct: III B (and Agrab Tepe) was probably destroyed
at the same time as many other Urartian citadels, during the reign of Rusa II,
or possibly later. But precisely when III B was built after 800 bc remains
unresolved.
Agrab Tepe, another Urartian site (a fort) close to Hasanlu, had been
destroyed twice in its history (Muscarella * 1973). Partly because of the pub-
lished reports that both Hasanlu III B and III A contained painted and plain
wares (Muscarella *1973: 65, 71, 74), I did not then recognize that the com-
plete absence of painted wares at Agrab and the presence of manifest Urar-
tian pottery there, as well as an Urartian sealing (Muscarella * 1973: 62–65;
see also pp. 69, 73–74), along with the fortification wall similarity, might indi-
cate that Agrab was Urartian, and contemporary solely with the III B period
at Hasanlu (Dyson and Muscarella 1989: 4, 19, 21). It was indeed an idea I
had considered, along with the possibility that the site may have contin-
ued into the post-III B period (Muscarella * 1973: 69, 74), but I could draw no
conclusions. The confusion is now obvious: I had used Hasanlu alleged III
A pottery (viz. as noted above, the pithos with triangles: a classic Urartian-
III B-vessel!) along with III B pottery parallels. In 1989 (Muscarella in Dyson
and Muscarella: 19, 21), I discussed Hasanlu III B’s Urartian history and its
incipient and terminal chronological problems.
A major omission in the Hasanlu publications regarding III B warrants
special mention, namely that, as recognized by P. Zimansky, there has been
a lack of recognition and hence relevant discussion of the “Urartians as a
significant presence there.”17 This omission occurs as well elsewhere in the
literature (viz. in Salvini *1995). I suggest that this is because the III B and
A division has not been articulated in the Hasanlu publications, nor has the
nature of the complete cultural differences between the two important and
distinct cultural periods, III B and A, been emphasized.
Period III A
No architecture is published for Period III A, and it is reported only (Dyson
1963a: 132) that this level is “less impressive [than III B], with more scattered
structures.” In Dyson 1964b: 372, it was claimed that the unique painted ves-
sel, recovered fragmented and “out of context” (see Dyson 1961a: 536, fig. 8),
“is now known to belong” to III A; also in 1965b: 212. However, a few years
later (Dyson 1967: 2960, fig. 1030),18 the vessel was dated to either Period IV or
17 P. Zimansky (* 1995: 112, n. 14). The Urartian sites of Agrab Tepe and Qalatgah (see n. 19
below) suffer the same fate: Zimansky ignores them in his discussion of Urartian sites in
northwest Iran. Based on the published record, he credited Dyson for recognizing that the
fortification wall was built in III B; he also oddly claims that no Urartian pottery was recorded
there.
18 Grace Freed Muscarella made the drawing, although this was not noted in the caption.
And the bone furniture tripod unit-possibly a drum-from BB II illustrated in Dyson and Voigt
the excavation of hasanlu: an archaeological evaluation 337
III, but without any reason given for the hesitancy and change (stratigraphy
confusion?). That uncertainty exists regarding III A pottery is exemplified by
the pithos and double-handled vessel (from Agrab Tepe) published as from
Hasanlu III A and III B in Dyson 1965b and 1999a (mentioned above), the
uncertainty is unrelieved by the fact that Young (1963: 30) said that painted
pottery continues into III A, while in 1972: 47, Dyson claimed that there is
no painted pottery in III A, and yet in 1977b: 549, the contrary was correctly
reported, but with no reference to or explanation of earlier information
provided. As observed above, this situation changed dramatically in Dyson
1999a, which states that III A alone yielded painted wares. In 1999b, Dyson
repeats much of 1999a (they are essentially one article split into two); but
1999b has more drawings of the painted pottery, Not mentioned in Dyson
1999a or 1999b is the fact that at the Urartian site of Qalatgah (Muscarella
*1971), ca. 15 miles west of Hasanlu, painted III A sherds were recovered,
demonstrating also that site’s post-Urartian history.19
Regarding chronology, we have shifts throughout the publication record,
as follows. Dyson 1965b: 212, gives 600–400bc; in 1972: 47, a provisional date
of 400 bc is suggested; Dyson and Pigott 1975: 182, 185, claim 600-?-300bc,
“Achaemenian period or shortly thereafter”; Dyson 1977b: 549, alleges “dated
to Achaemenian times sometime after 600bc”; likewise in 1999a: 137; and
Dyson in Dyson and Muscarella 1989: 8, suggests “around 400bc” In 1999a:
135, Dyson posits that III A commenced “long after the abandonment” of
III B, perhaps a century later, but correctly states that we do not know the
absolute date.
Period II
In Dyson 1959a: 9, and Young 1959a: 4, we are told that stone-lined tombs
(how many?) recovered in the debris of Period III could belong to Period II.
In Dyson 1957: 39; 1959a: 9; and 1961a: 534, there is reference to cist tombs that
could be from the “Parthian period,” or Period II; no photos were published.
In the latter publication-and thereafter-II is referred to as the “Mystery
2003: 231, fig. 20.7, was both excavated and drawn by me, rather than drawn by the individual
credited in the caption (he was not at Hasanlu). Compare the credit information given in
the same article, the caption on p. 220, fig. 20.1; see also Dyson 1989b: 115, fig. 10. The plan for
BB II was drawn by its various excavators, assisted in 1964 by Edward Keall, then serving as
architect. Each excavator drew a plan of the area s/he excavated in the field, which was then
joined to a master plan; the latter was thus enlarged and modified over the years. The plans
of BB I were drawn by both T. Cuyler Young and Dyson, and that of BB III by its excavator
T. Cuyler Young. (I do not know who drew the plans for the other buildings.)
19 For an account of the discovery of Qalatgah on July 31, 1968 (not July 30, as per
338 chapter nine
Muscarella * 1971a: 45), when I was directing the excavation of nearby Sé Girdan, see Mus-
carella * 1969: 5–6, n. 4, and * 1971a. A report on the 1968 campaigns at Haji Firuz, Dinkha
Tepe, Sé Girdan, and Qalatgah was published by Dyson with the names Muscarella and Mary
Voigt added (* 1969: 179–181). Voigt was director at Haji Firuz; I the director at Sé Girdan, who
on an exploration trip discovered Qalatgah; and Dyson was director of Bronze Age levels at
Dinkha Tepe. Another brief report is in Dyson * 1969: 19. After the discovery and several sur-
veys of Qalatgah, I invited the Dinkha team to the site; they came for the first time on August
9. My encounter and surveys of the site immediately revealed fortification walls (thought to
be a road by the locals), which, together with the surface finds of pottery, manifested the Urar-
tian nature of the site. On the survey with the Dinkha staff, an Urartian seal and an Urartian
stone inscription were discovered, the latter by a member of the Sé Girdan team, Christopher
Hamlin. For the inscription, see Muscarella * 1971a: 47–48, and M. van Loon (*1975).
Dyson’s brief statements in 1969: 44, Dyson * 1969: 19, and in Dyson, Muscarella, and Voigt
1969: 181, concerning Qalatgah do not reproduce the true nature of the events. However,
M. van Loon 1975: 201, n. * got it right, S. Kroll (in Stöllner, Slotta, and Vatandoust *2004: 363)
suggests that Qalatgah was the predominate military fortress in the area, not Hasanlu III B.
the excavation of hasanlu: an archaeological evaluation 339
Period I
The top level at Hasanlu, the “Islamic Period,” has been published by Danti
2004.
Conclusions
As a conclusion, I refer the reader back to the introduction above, the verac-
ity of which indictments I believe my text has, alas, amply demonstrated.
I add a couple of related comments: “[F]ailure to produce a full published
excavation report is simply tantamount to wanton destruction of an archae-
ological site. … [T]he excavation of a site is never warranted unless com-
prehensive publication is undertaken, nor is the excavation complete until
publication … is ended” (Renfrew *1980: 295). After this paper was com-
pleted, I came across Mousavi *2005. In pithy language and understandable
frustration, he writes of Hasanlu: “The final report of the excavations has
never been published, and the preliminary reports are either in the form
of news or general syntheses” (2005: 92 n. 11)—this in 2005, 50 years after
the beginning of excavations. Renfrew and Mousavi both communicate suc-
cinctly one of the failures of the Hasanlu publications analyzed above.
Acknowledgments
I wish to thank the two colleagues who carefully read a draft of this paper
or who communicated with me via email. They gave me what I expected—
valuable comments, criticisms, and suggestions, all of which helped me very
much. I, of course, am responsible for what advice I did not take.
Addendum:
Reports or Brief Mention of 1956–1964,
1970–1974 Hasanlu Campaigns
1940
Stein, A.: Old Routes of Western Iran. London: Macmillan.
1950
Hakemi, A., and Rad, M.: The Description and Results of the Scientific Excavations at
Hasanlu, “Solduz.” (Trans. P. Barzin, from Persian). Guzārishhā-yı̄ Bāstān Shināsi,
1329 A.H.: 87–103. Tehran: Vizārat-i Farhang, Idārah-i Kull-i Bāstānshināsi.
1957
Dyson, R.H., Jr.: Iran: 1956. University Museum Bulletin 21-1: 25–39.
1958
Dyson, R.H., Jr.:
1958a: Iran 1957: Iron Age Hasanlu. University Museum Bulletin 22/2: 25–32.
1958b: Pennsylvania Campaign in Iran. Archaeology 11: 128.
1959
Dyson, R.H., Jr.:
1959a: Digging in Iran: Hasanlu, 1958. Expedition 1/3: 4–17.
1959b: The Silver Cup of Hasanlu. Archaeology 12: 171.
1960
Dyson, R.H., Jr.:
1960a: The Death of a City. Expedition 2/3: 2–11.
1960b: Hasanlu-Azerbaijan Project. Explorers Journal 38/4: 10–11.
1960c: Hasanlu and Early Iran. Archaeology 13: 118–129.
1960d: Expedition News, Hasanlu, Iran. Expedition 3/1: 11.
the excavation of hasanlu: an archaeological evaluation 341
1960e: Where the Golden Bowl of Hasanlu Was Found: Excavations near Lake
Urmia Which Throw New Light on the Little-Known Mannaeans—Part 1.
Illustrated London News 236, January 23: 132–134.
1960f: The Golden Bowl and the Silver Cup-Treasures with a Dramatic History
and a Rich Significance: Excavations at Hasanlu, near Lake Urmia—Part II.
Illustrated London News 236, February 13: 250–251.
1961
Crawford, V.E.: Hasanlu 1960. Bulletin of the Metropolitan Museum of Art 20: 85–94.
Dyson, R.H., Jr.:
1961a: Excavating the Mannaean Citadel of Hasanlu; and New Light on Several
Millennia of Persian Azerbaijan. Illustrated London News 239, September 30:
534–537.
1961b: Hasanlu, 1960 Campaign. Archaeology 14: 63–64.
1962
Dyson, R.H., Jr.: The Hasanlu Project. Science 135/3504: 637–647.
van Loon, M.N.: A Lion Bowl from Hasanlu. Expedition 4/4: 14–19.
1963
Dyson, R.H., Jr.:
1963a: Hasanlu Discoveries, 1962. Archaeology 16: 131–133.
1963b: Expedition News. Expedition 5/2: 33.
Young, T.C., Jr.: Proto-Historic Western Iran. An Archaeological and Historical Re-
view: Problems and Possible Interpretations. Ph.D. dissertation, University of
Pennsylvania.
1964
Dyson, R.H., Jr.:
1964a: A Stranger from the East. Expedition 7/1: 32–33.
1964b: In the City of the Golden Bowl: New Excavations at Hasanlu in Persian
Azerbaijan. Illustrated London News 245, September 12: 372–374.
1964c: Ninth Century Men in Western Iran. Archaeology 17/1: 3–11.
1964d: Sciences Meet in Ancient Hasanlu. Natural History 73/8: 16–25.
1964e: Notes on Weapons and Chronology in Northern Iran around 1000BC, pp.
32–45 in Dark Ages and Nomads c. 1000bc: Studies in Iranian and Anatolian
Archaeology, ed. M. Mellink. Istanbul: Nederlands Historisch-Archaeolo-
gisch Instituut.
1965
Dyson, R.H., Jr.:
1965a: Hasanlu Excavations, 1964. Archaeology 18: 157–159.
1965b: Problems of Protohistoric Iran as Seen from Hasanlu. Journal of Near East-
ern Studies 24: 193–217.
342 chapter nine
Muscarella, O.W.:
1965a: A Fibula from Hasanlu. American Journal of Archaeology 69: 233–240.
1965b: Lion Bowls from Hasanlu. Archaeology 18: 41–46.
Young, T.C., Jr.: A Comparative Ceramic Chronology for Western Iran, 1500–500bc
Iran 3: 53–85.
1966
Dyson, R.H., Jr.: The Hasanlu Project. Pp. 413–435 in New Roads to Yesterday, ed.
J.R. Caldwell. New York: Basic.
Muscarella, O.W.: Hasanlu 1964. Bulletin of the Metropolitan Museum of Art 25: 120–
135.
von Saldern, A.: Mosaic Glass from Hasanlu, Marlik, and Tell al-Rimah. Journal of
Glass Studies 8: 9–25.
Young, T.C., Jr.: Thoughts on the Architecture of Hasanlu IV. Iranica Antiqua 6: 48–71.
1967
Dyson, R.H., Jr.: Early Cultures of Solduz, Azerbaijan. Pp. 2951–2970 in A Survey of
Persian Art 14, ed. A.U. Pope. London: Oxford University.
1968
Dyson, R.H., Jr.: Hasanlu and the Solduz and Ushnu Valleys: Twelve Years of Explo-
ration. Archaeologia Viva 1: 83–101.
1969
Dyson, R.H., Jr.: A Decade in Iran. Expedition 11/2: 39–47.
1971
Dyson, R.H., Jr.: Hasanlu. Iran 9: 170.
Muscarella, O.W: Hasanlu in the Ninth Century bc and Its Relations with Other
Cultural Centers of the Near East. American Journal of Archaeology 75: 263–266.
1972
Dyson R.H., Jr.: The Hasanlu Project, 1961–1967. Pp. 39–58 in The Memorial Volume
of the Vth International Congress of Iranian Art and Archaeology, Tehran-Isfahan-
Shiraz, 11th–18th April 1968, Vol. 1. Tehran: Ministry of Culture and Arts.
Rathbun, T.A.: A Study of the Physical Characteristics of the Ancient Inhabitants of
Hasanlu, Iran, Miami: Field Research Projects.
1973
Dyson R.H., Jr.:
1973a: Further Excavations at Tepe Hasanlu, Iran. Archaeology 26: 303–304.
1973b: Hasanlū. Iran 11: 195–196.
1973c: Hasanlu 1972. Proceedings of the 1st Annual Symposium on Archaeological
Research in Iran, ed. F. Bagherzadeh. Teheran: Iranian Centre for Archaeo-
logical Research (paper published as separate fascicle).
the excavation of hasanlu: an archaeological evaluation 343
1974
Muscarella, O.W.: The Third Lion Bowl from Hasanlu. Expedition 16/2: 25–29.
1975
Dyson, R.H., Jr.: Hasanlu 1974: The Ninth Century BC Gateway. Pp. 179–188 in Proceed-
ings of the IIIrd Annual Symposium on Archaeological Research in Iran, 2nd–7th
November 1974. Tehran: Iranian Centre for Archaeological Research.
Dyson, R.H., Jr., and Pigott, V.C.: Hasanlu. Iran 13: 182–185.
1977
Dyson R.H., Jr.:
1977a: Architecture of the Iron I Period at Hasanlu in Western Iran, and Its Impli-
cations for Theories of Migration on the Iranian Plateau. Pp. 155–169 in Le
plateau iranien et l’Asie centrale des origines à la conquete islamique, ed.
J. Deshayes. Paris: Éditions du Cente national de la recherche scientifique.
1977b: The Architecture of Hasanlu: Periods I to IV. American Journal of Archaeol-
ogy 81: 548–552.
Winter, I.: Perspective on the “Local Style” of Hasanlu IVB: A Study in Receptivity.
Pp. 371–386 in Mountains and Lowlands: Essays in the Archaeology of Greater
Mesopotamia, eds. L.D. Levine and T.C. Young, Jr. Bibliotheca Mesopotamica 7,
Malibu: Undena.
1980
Dyson, R.H., Jr.: The Question of Balconies at Hasanlu. Pp. 149–157 in From Athens to
Gordion: The Papers of a Memorial Symposium for Rodney S. Young, ed. K. DeVries.
University Museum Papers 1. Philadelphia: University Museum, University of
Pennsylvania.
Muscarella, 0. W.: The Catalogue of Ivories from Hasanlu, Iran. University Museum
Monograph 40; Hasanlu Special Studies 2. Philadelphia: University Museum,
University of Pennsylvania.
Winter, I.J.: A Decorated Breastplate from Hasanlu, Iran. University Museum Mono-
graph 39; Hasanlu. Special Studies 1. Philadelphia: University Museum, Univer-
sity of Pennsylvania.
1983
de Schauensee, M., and Dyson, R.H., Jr.: Hasanlu Horse Trappings and Assyrian
Reliefs, Pp. 59–77 in Essays on Near Eastern Art and Archaeology in Honor of
Charles Kyrle Wilkinson, eds. P. 0. Harper and H. Pitmann. New York: Metropoli-
tan Museum of Art.
1983/1984
Dyson, R.H., Jr.: Summary of Work at Hasanlu and Hissar, 1970–1980. Archiv für
Orientforschung 29–30: 302–304.
344 chapter nine
1985
Young, T.C., Jr.: Early Iron Age Iran Revisited: Preliminary Suggestions for the Re-
analysis of Old Constructs. Pp. 361–378 in De l’Indus aux Balkans: Recueil à la
mémoire de Jean Deshayes, eds. J.-L. Huot, M. Yon, and Y. Calvet. Paris: Recherche
sur les Civilisations.
1986
Dyson, R.H., Jr. and Harris, M.V.: The Archaeological Context of Cylinder Seals
Excavated on the Iranian Plateau. Pp. 79–110 in Insight through Images: Studies
in Honor of Edith Porada, ed. M. Kelly-Buccellati. Bibliotheca Mesopotamica 21.
Malibu: Undena.
1987
Levine, L.D.: The Iron Age. Pp. 229–250 in The Archaeology of Western Iran: Settle-
ment and Society from Prehistory to the Islamic Conquest, ed. F. Hole. Washington,
DC: Smithsonian Institution.
Muscarella, O. W, Review of Problèmes concernant les Hurrites, Vol. 2, ed. M.-T. Bar-
relet. Journal of the American Oriental Society 107: 135–137.
1988
de Schauensee, M.: Northwest Iran as a Bronze-Working Centre: The View from
Hasanlu. Pp. 45–62 in Bronzeworking Centres of Western Asia c. 1000–539bc, ed.
J.E. Curtis. London: Kegan Paul.
Muscarella, 0. W.: Northwest Iran, Excavated Objects, The Hasanlu Project. Pp. 15–81
in Bronze and Iron: Ancient Near Eastern Artifacts in The Metropolitan Museum of
Art, by 0. W. Muscarella. New York: Metropolitan Museum of Art.
1989
de Schauensee, M.: Horse Gear from Hasanlu. Expedition 31/2–3: 37–52.
Dyson, R.H., Jr.:
1989a: Rediscovering Hasanlu. Expedition 31/2–3: 3–11.
1989b: The Iron Age Architecture at Hasanlu: An Essay. Expedition 31/2–3: 107–127.
Dyson, R.H., Jr., and Muscarella, O.W.: Constructing the Chronology and Historical
Implications of Hasanlu IV. Iran 27: 1–27.
Marcus, M.I.: Emblems of Authority: The Seals and Sealings from Hasanlu IVB.
Expedition 31/2–3: 53–63.
Muscarella, O.W.: Warfare at Hasanlu in the Late 9th Century bc Expedition 31/2–3:
24–36.
Pigott, V.C.: The Emergence of Iron Use at Hasanlu. Expedition 31/2–3: 67–79.
Reese, D.S.: Treasures from the Sea: Shells and Shell Ornaments from Hasanlu.
Expedition 31/2–3: 80–86.
Winter, I.J.: The “Hasanlu Gold Bowl”: Thirty Years Later. Expedition 31/2–3: 87–
106.
the excavation of hasanlu: an archaeological evaluation 345
1990
Marcus, M.I.:
1990a: Centre, Province and Periphery: A New Paradigm from Iron-Age Iran. Art
History 13/2: 129–150.
1990b: Glyptic Style and Seal Function: The Hasanlu Connection. Aegeum 5: 175–
193.
1991
Marcus, M.I.: The Mosaic Glass Vessels from Hasanlu, Iran: A Study in Large-Scale
Stylistic Trait Distribution. Art Bulletin 73: 536–560.
1993
Marcus, M.I.: Incorporating the Body: Adornment, Gender, and Social Identity in
Ancient Iran. Cambridge Archaeological Journal 3: 157–178.
1994
Marcus, M.I.:
1994a: In His Lips He Held a Spell. Source 13/4: 9–14.
1994b: Dressed to Kill: Women and Pins in Early Iran. Oxford Art Journal 17/2: 3–15.
Muscarella, O.W.: North-western Iran: Bronze Age to Iron Age. Pp. 139–155 in Anato-
lian Iron Ages 3: The Proceedings of the Third Anatolian Iron Ages Colloquium Held
at Van, 6–12 August 1990, eds. A. Çilingiroğlu and D.H. French, London: British
Institute of Archaeology at Ankara.
Young, T.C., Jr.: Architectural Developments in Iron Age Western Iran. Bulletin of the
Canadian Society for Mesopotamian Studies 27: 25–32.
1995
Muscarella, 0. W.: Art and Archaeology of Western Iran in Prehistory. Pp. 981–999 in
Civilizations of the Ancient Near East, Vol. 2, ed. J.M. Sasson. New York: Scribner’s.
1996
Marcus, M.I.:
1996a: Emblems of Identity and Prestige: The Seals and Seatings from Hasanlu, Iran:
Commentary and Catalog. University Museum Monograph 84; Hasanlu
Special Studies 3. Philadelphia: University Museum, University of Pennsyl-
vania.
1996b: Sex and the Politics of Female Adornment in Pre-Achaemenid Iran (1000–
800B.C.E.), Pp. 41–54 in Sexuality in Ancient Art: Near East, Egypt, Greece,
and Italy, ed. N.B. Kampen. Cambridge: Cambridge University.
Muscarella, O.W.: Hasanlu. Pp. 209–211 in The Dictionary of Art, Vol. 14, ed. J. Turner.
New York: Grove.
346 chapter nine
1997
Dyson, R.H., Jr.: Hasanlu. Pp. 478–481 in The Oxford Encyclopedia of Archaeology in
the Near East, Vol. 2, ed. E.M. Meyers. New York: Oxford University.
1999
Dyson, R.H., Jr.:
1999a: Triangle-Festoon Ware Reconsidered. Iranica Antiqua 34: 115–144.
1999b: The Achaemenid Painted Pottery of Hasanlu IIIA. Anatolian Studies 49: 101–
110.
2001
de Schauensee, M.: A Note on Three Glass Plaques from Hasanlu. Iraq 63: 99–106.
2002
Young, T.C., Jr.: Syria and Iran: Further Thoughts on the Architecture of Hasanlu.
Pp. 386–398 in Of Pots and Pans: Papers on the Archaeology and History of Meso-
potamia and Syria Presented to David Oates in Honour of his 75th Birthday, eds. L.
al-Gailani Werr, J. Curtis, H. Martin, A. McMahon, J. Oates, and J. Reade. London:
NABU.
2003
Dyson, R.H., Jr.: Hasanlu Teppe. Pp. 41–46 in Encyclopaedia Iranica, Vol. 12, ed.
E. Yarshater, Boston: Routledge and Kegan Paul.
Dyson, R.H., Jr., and Voigt, M.M.: A Temple at Hasanlu. Pp. 219–236 in Yek bud,
yeki nabud: Essays on the Archaeology of Iran in Honor on William H. Sumner,
eds. N.F. Miller and K. Abdi. Monograph 48. Los Angeles: Cotsen Institute of
Archaeology, University of California.
2004
Danti, M.D.: The Ilkhanid Heartland: Hasanlu Tepe (Iran) Period I. University Muse-
um Monograph 120; Hasanlu Excavation Reports 2. Philadelphia: University of
Pennsylvania Museum of Archaeology and Anthropology.
Danti, M.D.; Voigt, M.M.; and Dyson, R.H., Jr.: The Search for the Late Chalcol-
ithic/Early Bronze Age Transition in the Ushnu-Solduz Valley, Iran. Pp. 584–
616 in A View from the Highlands: Archaeological Studies in Honour of Charles
Burney, ed. A. Sagona. Ancient Near Eastern Studies, Supplement 12. Leuven:
Peeters.
Muscarella, O.W.: The Hasanlu Lion Pins Again. Pp. 693–710 in A View From the
Highlands: Archaeological Studies in Honour of Charles Burney, ed. A. Sagona.
Ancient Near Eastern Studies, Supplement 12. Leuven: Peeters.
the excavation of hasanlu: an archaeological evaluation 347
*Supplementary References
Barnett, R.D.
1982 Urartu. Pp. 314–371 in The Cambridge Ancient History, Vol. 3, Pt. 1: The
Prehistory of the Balkans; and the Middle East and the Aegean World, Tenth
to Eighth Centuries bc, eds. J. Boardman, I.E.S. Edwards, N.G.L. Hammond,
and E. Sollberger. 2nd ed. Cambridge: Cambridge University.
Calmeyer, P.
1975 Hasanlū, Reallexikon der Assyriologie und Vorderasiatischen Archäologie 4:
128–131.
Dittmann, R.
1990 Eisenzeit I und II in West- und Nordwest-Iran zeitgleich zur Karum-Zeit
Anatoliens? Archäologische Mitteilungen aus Iran 23: 105–138.
Fehr, B.
1971–1972 Zur Geschichte des Apollonheiligtums von Didyma. Marburger Win-
ckelmann-Programm 1971/72: 14–59.
Ghirshman, R.
1939 Fouilles de Sialk près de Kashan 1933, 1934, 1937, Vol. 2. Série archéologique 5.
Paris: Geuthner.
348 chapter nine
Ivantchik, A.I.
2001 Kimmerier und Skythen. Steppenvölker Eurasiens 2. Moscow: Eurasien-
Abteiling. des Deutschen Archaologischen Instituts.
Johnson, E.C.
1976 Urartian Influence upon Iranian Architecture in the Early First Millennium
bc Marsyas 17: 21–38.
Kleiss, W.
1973 Planaufnahmen Urartäischer Burgen in Iranisch-Azerbaidjan im Jahre 1972.
Archaeologische Mitteilungen aus Iran 6: 81–89.
Kroll, S.
1975 Ein Schüssel der Triangle Ware aus Azarbaidschan. Archaeologische Mit-
teilungen aus Iran 8: 71–74.
1976 Keramik urartäischer Festungen in Iran. Archäologische Mitteilungen aus
Iran, Ergänzungsband 2. Berlin: Reimer.
1977 [Urartäische Plätze in Iran]. B: die Oberflächenfunde des Uranu-Surveys
1976. Archaeologische Mitteilungen aus Iran 10: 83–118.
1992 Ein “triple road system” oder “Stallbauten” in Hasanlu. Archäologische Mit-
teilungen aus Iran 25: 65–72.
Levine, L.
1987 The Iron Age. Pp. 229–250 in The Archaeology of Western Iran: Settlement
and Society from Prehistory to the Islamic Conquest, ed. F. Hole. Washington,
DC: Smithsonian Institution.
Marcus, M.I.
1995 Art and Ideology in Ancient Western Asia. Pp. 2487–505 in Civilizations of
the Ancient Near East, Vol. 4, ed. J.M. Sasson. New York: Scribner’s.
Medvedskaya, I.
1982 Iran: Iron Age I. Trans. S. Pavlovich, from Russian. BAR International Series
126. Oxford: B.A.R.
Mousavi, A.
2005 Comments on the Early Iron Age in Iran. Iranica Antiqua 40: 87–99.
Muscarella, O.W.
1968 Excavations at Dinkha Tepe, 1966. Metropolitan Museum of Art Bulletin 27/3:
187–196.
1969 The Tumuli at Sé Girdan: A Preliminary Report. Metropolitan Museum
Journal 2: 5–25.
the excavation of hasanlu: an archaeological evaluation 349
Porada, E.
1965 The Art of Ancient Iran: Pre-Islamic Cultures. New York: Crown.
Reade, J.
1979 Hasanlu, Gilzanu, and Related Considerations. Archaeologische Mitteilun-
gen aus Iran 12: 175–181.
Renfrew, C.
1980 The Great Tradition versus the Great Divide: Archaeology as Anthropology?
American Journal of Archaeology 84: 287–298.
Salvini, M.
1995 Geschichte und Kultur der Urartäer. Darmstadt: Wissenschaftliche Buchge-
sellschaft.
van Loon, M.
1975 The Inscription of Ishpuini and Meinua at Qalatgah, Iran. Journal of Near
Eastern Studies 34: 201–207.
Zimansky, P.
1995 Urartian Material Culture as State Assemblage: An Anomaly in the Archae-
ology of Empire. Bulletin of the American Schools of Oriental Research 299/
300: 103–115.
chapter ten
Abstract: The site of Muweilah in Sharjah, United Arab Emirates, has been pub-
lished by its excavator Peter Magee over a number of years as having flourished
during the Iron II period of the UAE/Oman, Arabian chronological system, ca. 1100/
1000–600bc. He has further asserted that within this long period, Muweilah’s exis-
tence can be dated to the time of the north-western Iranian Iron II period, which
terminated ca. 800bc, dating his site specifically to ca. 920–770 bc. Evidence used
to affirm the Iranian Iron II chronology includes Iranian and local architecture
and pottery parallels, and C 14 data. I rejected the viability and relevance of these
parallels in print in 2003, to which Magee responded, reaffirming his 10th-early 8th-
century bc chronology. Here I respond to the excavator’s ongoing defence, and argue
for a considerably later north-western Iranian Iron III date for Muweilah.
The Background
Reacting to two of Peter Magee’s articles (of 1997 and 2001) about his site
Muweilah in Sharjah, United Arab Emirates, I challenged the chronology he
presented, claiming it was too high, and disagreed with his cultural/histori-
cal conclusions relating to the nature and date of the pottery and architec-
ture recovered.1 Magee responded to my challenge,2 vigorously defending
(appropriately) his position. Here I offer my response to his 2005 defence:
I have not changed my mind regarding his dating of Muweilah and present
* This article originally appeared as “The Iranian Iron III Chronology at Muweilah in the
bibliography, and sharing their views with me. And Peter Magee for graciously allowing me
to publish photographs from his publications.
1 Muscarella 2003, 249–250, n. 102. My critique was presented briefly in a footnote in an
article a new C 14 date recently proclaimed by the excavators of Gordion in Anatolia, arguing
that it was too high. The footnote was presented to give another example of a C 14 date that I
believed to be incorrect. I first encountered Muweilah at a Bryn Mawr lecture by Peter Magee
in October 2002, where I first expressed (verbally) my disagreements about the chronology
he assigned to Muweilah’s and architecture.
2 Magee 2005a.
352 chapter ten
my arguments for rejection here. Beginning with his first reports Magee
has continuously reported that Muweilah came into existence during the
UAE/Oman, Arabian Iron II period, which he dates from ca. 1100/1000 to
600bc.3 The reader must understand from the beginning of the discussion
that this chronological period, Iron II, is alleged to have lasted for 500 years;
and that in Iran this same 500-year time-span encompasses two separate
and distinct cultural and chronological periods, Iron II and Iron III—crucial
issues not articulated by Magee. Within that broad Arabian Iron II time
frame he specifically situates Muweilah’s construction and existence con-
temporary with the chronology of the Iranian Iron II period, with Hasanlu
Period IV, which terminated ca. 800 bc (i.e. 200 years earlier than the appar-
ent termination of the Iron II period in Arabian terminology).4
The Ceramics
3 Magee 1996a, 208; 1996b, 246, 249; 1997, 96; 1999, 44; 2002, 161; 2004, 32; 2005a, 161; 96.
4 I am not a scholar of Arabian archaeology and I found it confusing that the very same
terminology used in Iranian archaeology, Iron I, II, III, is employed in Arabia. For a discussion
of the Iranian/Hasanlu Iron Age terminology and problems, see Muscarella 2006.
5 Magee 1996a, 203, 205–206, figs. 16–17; 1999, 45, figs. 5–6; 2001, 121, 123, fig. 12; 2002, 164–
165, fig. 2; 2005a, 165, fig. 1, right (compare the Hasanlu vessel at the right); 2005b, 99, 112, figs. 5
(Fig. 1 in this paper), 20; Magee et al. 2002, 141, fig. 13.
6 Magee 1996a, 208; 1996b, 246–248, fig. 7.B, C, E, G; 1997, 93–95, fig. 2; 2005a, 165, fig. 1,
centre; see also Boucharlat and Lombard 2001, 218, fig. 11.
7 Magee 1999, 45, fig. 5; 2005b, 100, fig. 6.
the iranian iron iii chronology at muweilah 353
8 Overlaet 2005, pls. 8.2, 9, 11.3–4. Further, the drawing of B (in n. 5 above) seems to be a
restoration.
9 Magee 1997, 94, 96, fig. 2; 1999,45; 2005a, 162; 2005b, 93–94.
10 Dyson 1965, 208. On this also, see Young 1965, 76–80, figs. 13–14.
11 This reality has been known for decades: viz. Haerinck 1978, 85. For details and bibliog-
I too would agree with Magee that ‘it is too early to make definitive
statements concerning the date of Iron Age Sialk’,12 which is a complicated
subject. But I would add, that based on the Sialk pottery, one cannot date
Rumeilah (or, indeed, Muweilah) within the Iranian Iron II chronology as
established in north-western Iran. Magee himself admits (but then ignores
it) that the Arabian ‘decorative patterns’ are ‘unlike the examples found in
Iran during this period’, i.e. in the context here, Iron II.13
Magee continuously uses the terms ‘bridged’ or ‘bridged-spouted’ to
describe the Muweilah spout form, but inexplicably does not inform us
whether the bridged spouts are positioned vertically or horizontally. The
position of the spout determined how the liquid was poured out into
another container, and although Magee mentions this process, he does not
inform us how far the Muweilah people had to tip their spouts.14 Spouted
vessel forms from the Iranian Iron III period exist, viz. at Hasanlu, Ziwiye
and Yanik Tepe.15
In response to my 2003 brief comments on this significant matter, Magee
replies16 it is ‘difficult to know’ how Muscarella can be certain that they
are ‘horizontally or vertically spouted’. My answer is that lacking textual
information, I looked at his published photographs, which do not portray
horizontal bridged spouts; and how and why does Magee know that they
are horizontally spouted: which must be the case to qualify for his Iron II
attribution? But nowhere has he addressed this, recognised its significance
in his discussions of parallels and chronology. He subtly adds two ambigu-
ous modifications to his previously published claims.17 One is ‘that not all
the Muweilah examples are comparable to the more elongated horizontal
bridge-spouted examples from north-western Iran’. If by ‘more elongated’
he is now stating that the (some?) spouts at Muweilah are bridged horizon-
tally, but are short, why not say it straightforward? Note that in this 2005
article the ‘horizontal’ word for the Muweilah vessels is mentioned for the
first time, again ignoring its chronological importance.18 He also claims that
recognised by Haerinck 1987, 87); Ziwiye: Young 1965 60, fig. 4.5; Yanik Tepe: Haerinck 1978,
84, 87, fig. 7.
16 Magee 2005a, 163–164.
17 Magee 2005a, 164.
18 Magee 2005a, 164.
the iranian iron iii chronology at muweilah 357
he is not able to see how any of the Muweilah examples are ‘significantly
[his emphasis] different from Iranian Iron Age II bridge-spouted vessels
in general [my emphasis]’. In another venue published the same year19 he
also stated, obliquely and casually, that ‘there are many differences between
the Arabian … and west and northwest Iranian examples. The east Arabian
examples contain a more open spout that differs from the very elongated
horizontal [sic] spout found on some [sic] northwest and central Iranian
examples’ (no profile drawings are published). The vertical spout position
remains unmentioned here, concealed from us, although the vertical word
is used when informing us that vertical spouts of a different form exist at
Sialk. Here also he generalises about a long life for bridge-spouted vessels
down to the 8th century bc.20
I suggest that Magee’s 2005 defence contra Muscarella 2003 is an attempt
to modify all his previously published, and strongly stated, pottery parallel
claims. Here is the core of the issue under review: I see only bridged vertical
spouted vessels among those published from Muweilah (Figs. 1–2), a form in
Iran that is stratigraphically distinguished as occurring later than the Iron II
horizontal spout examples. In 1996 Magee asserted an indefinite generalisa-
tion, ‘Bridge-vessels are leitfossils for the Iron II period in western Iran’,21 but
omitted the qualifying ‘horizontal’, a term essential in any characterisation
of Iranian Iron II leitfossil spouted vessels. For excellent published Hasanlu
Period IV examples see Stein; for the record, there are examples of bridged
horizontal spout vessels in post-Iron II Iran.22 Magee23 also cites examples
of ‘bridged spouted vessels’ at Godin and Baba Jan, and which (again) is
meaningless, and in the context a misleading use of ‘bridged’ (why omit
the crucial position of the spout?). In fact, all the vessels from these sites
are dated to the Iranian Iron III period—and all the cited vessels have verti-
cal spouts; also, Baba Jan has much painted pottery:24 which as such should
drawing in pl. XXVIII, 17. An excavated example from Ziwiye has hanging triangles and incised
chequerboard decoration (Dyson 1965, 206, fig. 11): it has a parallel at Muweilah (Magee
2005b, 113, n. 28); Nush-I Jan (Stronach 1969, 18, fig. 7); War Kabud in Lurisran (Overlaet 2005,
pl. 11.1), which may possibly be dated in Luristan chronology to Iron IIB, 8th century bc or
later.
23 Magee 2005b, 94.
24 Magee (2005a, 94, n. 9) cites Goff 1985 as a Baba Jan reference. This is an error, repeated
358 chapter ten
in the bibliography: the journal reference should be Goff 1978, and the issue is Iran 16, pp. 29–
66 (for a painted vessel with vertical spout, see pl. 1Ia). Nn. 6 and 7 also have wrong references.
The Manor house at Baba Jan does nor seem to have a columned hall; it has three columns,
probably for a colonnade. For the post-9th centuty dating of Baba Jan, see Muscarella 1988,
140, n. 1, 209, n. 4 (contra Boucharlat and Lombard 2001, 222, n. 9).
25 Magee 2005a, 163.
26 See Magee 1996a, 208; 1997, 93–95 (Rumeilah); 2001, 123; 2002, 164.
27 Magee 1996b, 247.
28 Magee 1996b, 247.
29 Magee 2005b, fig. 7, centre.
30 Kroll 1976, 115; see also Young 1965, 58, fig. 3.6, 11, fig. 4.6.
31 Magee 2004,27–28, fig. 4 for one example.
the iranian iron iii chronology at muweilah 359
it is not perforated. The next rounded cover example known is from the 7th
century bc. Ashurbanipal garden relief (Goldman—LL), also unperforated.
Only those from the Achaemenid period have all the characteristics of the
Muweilah censor cover: form, multi-perforations and, sometimes, figured
handles (Goldman—A, B, F, G, I).34 Some are domed, others pyramid shaped
and stepped, for which also see the Achaemenid examples from Uşak.35 One
can argue that the Achaemenid-form censor derives from a late phase at
Muweilah, the time prior to its destruction, but then indeed this could also
obtain for the vertically spouted vessels as well: which were recovered in
the destruction level there.36 The archaeological evidence indicates that the
Muweilah censor cover cannot be employed to date the site’s destruction to
‘a fire sometime after 770 bc but before c. 600 bc’.37 Its presence contradicts
this asserted chronology: and brings into the discussion an Achaemenid
period at Muweilah (whether it can be argued that such censors existed
earlier in Arabia, I leave to the specialists); Magee plays down involvement
with Fars.38
The Architecture
Carbon-14 Data
ca. 920bc … destroyed by fire 770bc but before c. 600 bc’;54 in 2002 the initial
construction date is lowered to ca. 900 bc;55 in 2003 it is after 920 bc and with
‘an upper limit of 800 bc’ for the site’s destruction (wood and date seeds).56
In 2004 the construction date for the site is now the ‘end of the ninth cen-
tury bc’, based on dates and date seeds samples.57 The following year Magee
emphasised the C 14 evidence from dates and burnt beams for the chronol-
ogy, repeating that the ‘main buildings’ were ‘constructed after ca. 920bc
and destroyed sometime after ca. 800 bc but before ca. 600bc’.58 And ‘bridge’
spouted vessels were ‘in use, therefore, sometime after 920bc’, perhaps even
‘from c. 1000 bc onwards’, and ‘exhibit some form of influence from Iran’ [my
emphasis].59 In the final analysis, C 14 evidence is the core of the issue at
hand. Surely it was the C 14 dates that firmed his chronological claims, his
associated dating of the pottery and architecture. All three are presented as
if they naturally form a closed, self-created triangle of mutually supporting
evidence.
The C 14 dates given in Magee 2004, along with data given in table 1 there,
do not manifestly prove the Iranian Iron II chronology argued. Here I stick
my neck out, for I see a problem with the C 14 analysis; and I must leave a
close analysis to C 14 statistical experts to debate (as they will also with my
2003 article that disagrees with another C 14 determination).
One more issue, an important conclusion of Magee’s, should be ad-
dressed. Magee believes that Arabian letters inscribed on a vessel fragment
excavated at Muweilah manifest that the South Arabian script existed in
south-eastern Arabia, and thus, based on his chronology, indicates that ‘seri-
ous cultural interaction with Yemen began 300 to 400 years earlier than
previously thought’.60 If I am correct in my disagreements with Magee’s
Muweilah’s chronology, this cultural assertion is erroneous, but warrants
independent discussion by the experts.
Conclusions
Based then on the pottery, the censor and formal architectural parallels, and
a challenge of the C 14 date, I argue that Muweilah came into existence in the
post-Iranian Iron II period, probably in the 7th century bc (or later?), what-
ever UAE/Arabian terminology is employed (late Iron II?). When Magee
states that there is an ‘absence of iron III material culture’ at Muweilah,61 he
can only be referring to Arabian terminology. And Muscarella never commit-
ted the ‘fundamental error’ of only equating ‘this pottery and thus this site’.62
(One may aptly use the same phrase to describe Magee’s equating Hasanlu
with Muweilah.) What I said and repeat, is that no ceramic or architectural
evidence present at Muweilah, or Rumeilah, supports an Iranian Iron II date
there. The issue is not, and never was, whether there was Iranian influence
and trade with Arabia,63 but when it commenced.
Addendum
In the latest issue of BASOR (No. 347 [2007], 83–105; it appeared at the
Metropolitan Museum of Art on 20th September 2007) Peter Magee has a
long article on Arabian sites (‘Beyond the Desert and the Sown …’), including
Muweilah. Here in a long discussion on chronology he dates Muweilah
(along with other sites) generally within the Arabian Iron II period, with a
broad 500 year range date of 1100–600 bc. But not once here does he mention
that in his many (and repetitious) earlier writings he has consistently dated
Muweilah to the 9th century bc, vigorously asserting alleged Hasanlu IV B,
9th-century bc ceramic and architectural parallels. These articles, eleven in
all—and of which a total of five are omitted/ignored in his argument and
from the bibliography—are discussed and critiqued by me in the above,
in the present article. There is more. Nowhere in the BASOR article does
he mention or include in his bibliography my original challenge to his 9th-
century bc dating of Muweilah that appeared in AWE 2.2 (2003), n. 102. And
nowhere does he include a reference to his long article in AWE 4.1 (2005)
contra my 2003 note, wherein he challenged my late dating (hundreds of
years later than his dating), and vigorously defended the 9th-century date
for Muweilah that he has touted for a long time. Furthermore, he does not
mention that he had read with my permission (upon the request of AWE’s
Editor-in-Chief) in August 2006 the then unpublished article submitted to
AWE—the present article—rebutting his 2005 AWE paper!
Bibliography
Boucharlat, R. and Lombard, P. 2001: ‘Le Bâtiment G de Rumeilah (Oasis d’al Ain).
Remarques sur les Salles á Poteaux de l’ Age du fer en Péninsule d’Oman’. IranAnt
36, 213–238.
Curtis, J. and Razmijou, S. 2005: ‘The Palace’. In Curtis, J. and Tallis, N. (eds.), Forgot-
ten Empire The World of Ancient Persia (London), 50–55.
Dyson, R.H. jr 1965: ‘Problems of Prehistoric Iran as seen from Hasanlu’. JNES 24,
193–217.
Goldman B. 1991: ‘Persian Domed Turibula’. StudIran 20, 179–188.
Haerinck, H. 1978: ‘Painted Pottery of the Ardebil Style in Azerbaidjan (Iran)’.
IranAnt XIII, 75–91
Kroll, S. 1976: Keramik Urartäischer Festungen in Iran (Berlin).
Magee, P. 1996a: ‘Excavations at Muweilah. Preliminary Reports on the First Two
Seasons’. Arabian Archaeology and Epigraphy 7, 195–213.
——— 1996b: ‘The Chronology of the Southeast Arabian Iron Age’. Arabian Archae-
ology and Epigraphy 7, 240–252.
——— 1997: ‘The Iranian Iron Age and the Chronology of Settlement in Southeastern
Arabia’. IranAnt32, 91–108.
——— 1999: ‘Writing in the Iron Age: the earliest South Arabian inscription from
southeastern Arabia’. Arabian Archeology and Epigraphy 10, 43–50.
——— 2001: ‘Excavations at the Iron Age Settlement of Muweilah 1997–2000’. Pro-
ceedings of the Seminar for Arabian Studies 31, 115–130.
——— 2002: ‘The Indigenous Context of Foreign Exchange between South-eastern
Arabia and Iran in the Iron Age.’ The Journal of Oman Studies 12, 161–168.
——— 2003: ‘New chronometric data defining the Iron Age period in southeastern
Arabia’. Proceedings of the Seminar for Arabian Studies 33, 1–10.
——— 2004: ‘The impact of southeast Arabian intra-regional trade on settlement
location and organization during the Iron Age II period’. Arab Archaeology and
Epigraphy 15, 24–42.
——— 2005a: ‘Columned Halls, Bridge-Spouted Vessels, C 14 Dates and the Chronol-
ogy of the East Arabian Iron Age: A Response to Some Recent Comments by
O. Muscarella in Ancient West & East’. In AWE 4.1, 160–169.
——— 2005b: ‘The production, Distribution and Function or Iron Age Bridge-
Spouted Vessels in Iran and Arabia: Results from Recent Excavations and Geo-
chemical Analysis’. Iran 43, 93–115.
Magee, P., Thompson, E., Mackay, A., Kottaras, P. and Weeks, L. 2002: ‘Further
evidence of desert settlement complexity: report on the 2001 excavations at the
Iron Age site of Muweilah, Emirate of Sharjah, United Arab Emirates’. Arabian
Archaeology and Epigraphy 13, 133–156.
Muscarella, O.W. 1988: Bronze and Iron (New York).
the iranian iron iii chronology at muweilah 367
——— 2003: ‘The date of the Destruction of the Early Phrygian Period at Gordion’.
AWE 2.2, 225–252.
——— 2006: ‘The Excavation of Hasanlu: an Archaeological Evaluation’. BASOR 342,
69–94.
Overlaet, B. 2005: ‘The Chronology of the Iron Age in the Pusht-i Kuh, Luristan’.
IranAnt 40, 1–33.
Özgen, T. and Öztürk J. 1966: The Lydian Treasure (Istanbul).
Stein, A. 1940: Old Routes of Western Iran (London).
Stronach, D. 1969: ‘Excavations at Tepe Nush-i Jan, 1967’. Iran 7, 1–20.
Young, T.C. jr 1965: ‘A Comparative Ceramic Chronology for Western Iran. 1500–
500BC’. Iran III, 53–85.
chapter eleven
Abstract: In 714 bc, Sargon II, king of Assyria, conducted a major military campaign
across the Zagros Mountains into western Iran. It was the eighth campaign of his
reign, and the fourth into Iran. This time, Sargon’s main goal was to contain the
state of Urartu within its territory and to subdue its allies. The itinerary of the
Assyrian army and the events of the campaign were written in the form of a long
letter from the king to the god Assur, and it is preserved almost in its entirety. The
text is of great importance because it gives one of the most extensive itineraries of
an Assyrian campaign and, unlike other Assyrian reports, it mentions by name, and
often describes, a large number of cities and fortresses as well as many geographical
features.
For over 70 years scholars have attempted to reconstruct the route of Sargon from
Assyria to Iran and to identify by epigraphical and archaeological research and
survey the cities and features he mentions. Unfortunately, there has been little
agreement among the modern researchers with regard to the specific directions
taken by Sargon and equally so with attempts to link a site on the archaeological
map with one mentioned in the ancient text. The present article gives a summary
of the various solutions offered over the years and argues for the identification of
two archaeologically known sites with two of those mentioned by Sargon.
Introduction
The study of the historical geography of ancient Iran is laden with difficulties
and lack of verification. Although Mesopotamian cuneiform texts often
refer to specific districts, states, cities, mountains, and rivers in western
Iran, it is not possible by some easy methodology to equate with certainty
the toponyms with actual geographical loci. The main reason for this gap
between text and place is the absence from excavated sites in western Iran
of textual material that might furnish information about their names. In
the neo-Assyrian period of Mesopotamian history (10th–7th centuries bc)
many texts exist that refer to and cite scores of features and places in western
* This article originally appeared as “The Location of Ulhu and Uiše in Sargon II’s Eighth
Iran, and many valiant attempts have been made by scholars to identify and
locate them. The major text of Sargon II’s eighth campaign gives an account
of the king’s conquests and collection of booty in areas of western Iran
and contains quite specific citations and descriptions of both geographical
features and inhabited cities and fortresses. Attempts to link up or equate
the toponyms of the text with archaeological sites and geographical features
are based on the examination of local topography and site placement in
western Iran. But the various conclusions and identifications presented by
different scholars, using the same text and examining the same topography,
manifestly demonstrate that the problems of identification are not readily
resolved.
If one is able, therefore, to identify with some degree of certainty and
viability a single known Iranian site that is mentioned in Sargon’s text, it
becomes possible to renew the investigation with a fresh approach. Replac-
ing a modern mound-name with its ancient name would not only give some
indication of direction for Sargon’s journey from Assyria, but it might also, by
reference both to the sequence of movement supplied in the text and to local
archaeological features, allow further site identifications to be obtained. It
is the conclusion of this paper that, by utilizing epigraphical and geographi-
cal data and the results of recent archaeological excavation and survey, two
of the ancient fortress cities mentioned in Sargon’s text may with some cer-
tainty be identified with two known sites in nw Iran.
Earlier Interpretations
In the course of reviewing Paul Zimansky’s book Ecology and Empire: The
Structure of the Urartian State (1985) I was obliged to consult in some detail
the text recording the eighth campaign of the Assyrian king Sargon II in
western Iran and in Urartian territory in 714 bc (Thureau-Dangin 1912; Meiss-
ner 1922; Luckenbill 1927:73–79; Weidner 1937–1939). While Zimansky is not
primarily concerned with the labyrinthine problems associated with mod-
ern reconstructions of Sargon’s route, he engages these issues when they
touch upon the terminology employed in the Assyrian text to describe the
physical characteristics of Urartian settlements. Thus, when discussing the
great fortress (birtu rabû) of Uaiais (line 299 of the Assyrian text), Ziman-
sky (1985: 42, 112, note 64) forcibly states that Sargon’s description of the site
fits well with Qalatgah. This equation, linking the important city of Uaiais
mentioned by Sargon with Qalatgah, an Urartian site in the Ushnu Valley
just sw of Lake Urmia in nw Iran (Azerbaijan), intrigued me, and the vigor
the location of ulhu and uiše in sargon ii’s eighth campaign 371
of the claim compelled me to review the evidence that would allow such an
identification. This entailed a close reading of Sargon’s text and an exam-
ination of the archaeological data available from Qalatgah itself and from
the neighboring terrain. The results of this review, which led me to disagree
with Zimansky’s identification, generated the writing of this paper.
The longest account of the eighth campaign of Sargon is written in cunei-
form on a large clay tablet (37.5 × 24.5cm) that derived from the Assyrian
capital city of Assur in Iraq. It was pilfered from that site and presumably
placed in the antiquities market; it eventually was acquired by the Louvre
“zu Unrecht” (Meissner 1922:113). The text was first published with commen-
tary by F. Thureau-Dangin in 1912, at a time when very little was known
about the archaeology and ancient cultural geography of Anatolia and Iran.
What is certain from the context of the text is the general direction of Sar-
gon’s route: that he crossed the Lower Zab River in Mesopotamia and passed
over Mt. Kullar into the Zagros Mountains and Iran, where he campaigned
before turning north to attack Urartian-controlled territory. After at least
one major battle was fought and a number of cities subdued, some near
a “sea,” Sargon returned home to Assyria. His army returned by one route,
while Sargon with 1000 picked troops went by another, and difficult, route
to attack Musasir, a holy city in ne Iraq (Boehmer 1973). Inasmuch as it was
assumed in 1912 (and continued to be assumed for decades) that Urartian
territory encompassed only the area around Lake Van in eastern Turkey,
Thureau-Dangin (1912:ix and map; here fig. 1) reconstructed Sargon’s route
from Iran to Urartu as involving a passage around the east and north shores
of Lake Urmia in nw Iran, a turning west across the Zagros to Lake Van, and
a circling of that lake before the homeward journey.
The discovery in 1968 of the Urartian site of Qalatgah with its late 9th-
century Urartian inscription (Muscarella 1971) and the extensive explora-
tions and surveys in Azerbaijan conducted by W. Kleiss of the German
Archaeological Institute1 made manifest the assumption that Urartian ter-
ritory from the late 9th to the 7th century bc in fact also included the area
around the western and southern shores of Lake Urmia. This archaeologi-
cally-documented evidence of the extent of Urartian-controlled land has
forced the conclusion that Thureau-Dangin’s reconstruction itself required
1 The results of these surveys and explorations have been published by W. Kleiss and his
colleagues in AMIran beginning in 1970 and continuing into the 1980s. Kleiss’s surveys are a
significant contribution to Urartian and Iranian archaeology. See also Kleiss (1976) and Kroll
(1976).
372 chapter eleven
revision.2 Sargon’s Urartian campaign and his reference to the “sea” could
only signify that the Assyrian text was concerned with the Urartian area
around Lake Urmia, that Sargon never campaigned in Urartu proper, and
that he never went to Lake Van. What is more, there is no indication from
the text that could allow one to conclude that Sargon circled around Lake
Urmia from the east. Nor is there any archaeological evidence for an eastern
route; in fact, there are no Urartian settlements along the eastern side of
the lake (Kleiss 1976: pp. 1, 2, with map).3 Sargon traveled to Lake Urmia and
2 For the relevant literature offering interpretations of the eighth campaign, see Ziman-
sky (1985: 112, note 55); Lehmann-Haupt (1926: 309); Oppenheim (1960); Kleiss (1969–1970);
Muscarella (1971); and Reade (1978: 141).
3 Doubts about Sargon’s alleged circling of Lake Urmia first surfaced in my mind in 1960
when I traveled by truck along its eastern shore from Tabriz to Hasanlu: the lake was too large
the location of ulhu and uiše in sargon ii’s eighth campaign 373
moved some distance up the western shore before turning west and south
for the journey to Musasir.
As this investigation is primarily concerned with the identification of
Qalatgah, it will be limited to an examination of the areas cited by Sargon as
being in or bordered by Urartian-controlled territory. The text records that
after leaving Uishdish, apparently a Mannaean-controlled region, Sargon
passed through Subi, Sangibutu (with many cities), Armarili, Aiadi on the
“sea” (i.e. Lake Urmia), then on to Uaiais, and finally to Nairi-Hubushkia,
from which area he began the journey home.4 Attention will be focused first
on Uaiais, appropriately in the context of its identification with Qalatgah by
Zimansky.
Thureau-Dangin (1912: x, note 1) was the first to suggest that Sargon’s
Uaiais (Assyrian text, line 298) is a variant of Uisi, Uiše, etc., known to
cuneiformists from a number of Assyrian documents, and his observation
has been accepted by all scholars. Believing that Sargon circled Lake Van,
Thureau-Dangin placed Uiše (as I shall continue to cite it) at Bitlis, near
the sw shore of that lake; others who accepted the encirclement of Lake
Van agreed with Thureau-Dangin (e.g., Piotrovskii 1959: 105, with map; Bur-
ney 1972: 156). Lehmann-Haupt (1916: 143–144; 1926: 310, 317–319, 322), who
believed Sargon traveled close to but did not encircle Lake Van, situated Uiše
further east, at the modern town of Bashkale; König (1955: 68, note 1, 207)
placed it somewhere sw of the lake. The placement of Uiše in se Turkey
continued to be accepted in more recent times. Çilingiroğlu (1976–1977: 265,
fig. 1) accepted the fact that Uiše is near Bashkale; while Levine (1977: 145,
147) and Mayer (1980: 15, fig. 1, 29) situated it further south between the
Upper Zab River, near the modern town of Hakkiri-Çölemerik, and Lake
Urmia.
(some 80 miles in length) not to have been mentioned if circled. These thoughts reached
fruition in 1968 when I discovered Qalatgah and its inscription.
At least three scholars in recent times continue to believe that Sargon circled the lake from
the east, even though they reject the long route to Lake Van (i.e. a “medium” route): Wäfler
(1976: 20); Çilinğiroǧlu (1976–1977: 259–265); Reade (1978: 140, fig. 2, 141). And, to mention
only post-1970 writers, Burney (1972: 138–139, 155–156), van Loon (1975: 206–207), and Kleiss
(1977a: 138–140) continue to support Thureau-Dangin’s long route around the east shore of
Lake Urmia to Van. The short route, essentially to the sw shore of Lake Urmia, first proposed
by the writer in 1971, has been accepted by Levine (1977: 144–148, fig. 1), Mayer (1980: 224–
225), and Salvini (1982: 387; 1984: 15, 46–51). Zimansky (1985: 40, 93, 113, note 81) is ambiguous;
on page 93, however, he summarizes his important research to conclude that the “lengthier
routes that have been proposed” may be ruled out.
4 For the sequence, see Levine (1977: 137–148, and fig. 1 [here fig. 2]), and Zimansky (1985:
41, fig. 7) for a schematic drawing of the route in the northern areas.
374 chapter eleven
I was tempted to accept the equation between Uiše and Qalatgah, because
for some time I was uncertain whether Sargon had traveled far up the
western shore of Lake Urmia. If he did not travel up the shore Uiše would
have to lie, on the basis of the campaign’s sequence, somewhere by or below
its south shore. But a careful reading of Sargon’s reference to Uiše left me
unsatisfied with Salvini’s and Zimansky’s opinion that Qalatgah was being
discussed. True, Qalatgah may be a “great fortress” and is situated in a fertile
area, necessary features for the identification. But they are not sufficient
features alone, and no obvious element or characteristic compels one to
5 The Qalatgah inscription is written on a stone block, 0.88 ×0.36 ×0.44cm, which is
about half its original length. Fortunately, the opening lines are among those surviving. They
mention that the inscription was written by Ishpuini and his son Menua, which places it in
the last decade of the 9th century bc. For photographs and a drawing of the stone see van
Loon (1975: figs. 1–3).
the location of ulhu and uiše in sargon ii’s eighth campaign 375
connect the site with the textual reference. For example, it is by no means
clear what part of the steep site would qualify as its rear. And regarding the
Uiše citation in the inscription, it is unclear from the incomplete context
(see note 5) whether or not it actually refers to Qalatgah itself.
Further study of Sargon’s text in a search for other possible candidates
that might relate to Qalatgah brought me first to the passages concerned
with the cities of Sangibutu, and particularly the city of Ulhu. And here I
immediately sensed that I was reading a description of Qalatgah.
Ulhu, like Uiše and many other cities and districts mentioned by Sargon,
has been moved by modern scholars back and forth across the ancient
landscape, and for the same reasons. Ulhu is mentioned in Sargon’s text
as a city in Sangibutu, at a distance from other local cities and therefore
apparently in its northern section. Thureau-Dangin (1912: viii), Lehmann-
Haupt (1916: 145, note 1; 1926: 319), Piotrovskii (1959: 105, with map), and, in
more recent times, van Loon (1966: 18) and Reade (1978: 140, fig. 2) situated
Ulhu in the fertile Marand plain, ne of Lake Urmia. Burney (1972: 140, 154)
and Kleiss (1969–1970: 131–132, fig. 3; 1977a: 140) believe it is to be situated
either near Marand or further sw, near the modern town of Shapur, where
earlier Wright (1943: 185) and Laessøe (1951: 21, note 2) had placed it. Kleiss
and Burney also suggest that if Ulhu was in the Shapur plain, it is to be
identified with the ancient site of Haftevan Tepe (Kleiss 1976: maps 1 and
2, site no. 32).
As all these identifications were based on the assumption that Sargon
passed around Lake Urmia from the east, thence either west to Van or
southward, they can readily be dismissed. And even if one could assume
that Sargon marched up the west shore of Lake Urmia as far north as
the Shapur Plain, Haftevan Tepe cannot by any archaeological reasoning
qualify as a major Urartian site, as Zimansky (1985: 39, 45, 113, note 100) has
demonstrated.
Ulhu is described by Sargon as a strong city (āl dannūti, line 200) situ-
ated at the foot of Mt. Kishpal,6 adjacent to a fertile area with orchards and
gardens. These features do not conflict with the topography at Qalatgah, a
walled city built on the slopes of a mountain and adjacent to a higher one,
and situated in one of the most fertile valleys in Iran.7 What is significant,
6 In the text, the line mentioning the name of the mountain is broken away, but in a
parallel passage in Sargon’s annals it is mentioned (Thureau-Dangin 1912: 71, line 114).
7 For a description, with plans and photographs of the site, which was extensively sur-
veyed but never excavated, see Kleiss (1971: 63–64, figs. 13, 14; 1977b: 70–72); and Muscarella
(1971: 42).
376 chapter eleven
8 Qanats are underground tunnels used to collect ground water; the tunnels eventually
reach open air where the collected water flows into irrigation channels. In line 211 of the text,
hirı̄tu is used (Meissner 1922: 116), alongside of which Rusa built a palace. But would Rusa
˘have built a palace by a qanat, rather than by a canal, whose flowing water could be utilized?
For a lack of qanats in Urartian territory, see Zimansky (1985: 119, note 128).
9 See also the Chicago Assyrian Dictionary 16 (1962: 259): surru, “heart, center (of an
˙
object),” but where the author also supplies canal: “he had countless irrigation ditches flow
out of it (the canal).” Note also that Laessøe (1951: 27, note 46) and Oppenheim (1960: 142)
both interpret the broken line 201 (“They [the people of Ulhu] did [not] drink, they did not
satiate themselves”) to signify that water was in short supply at Ulhu until Rusa discovered
new water sources. Line 201 is far from clear, but one notes in the previous preserved phrase in
the same line, “… his people like fish …”, which phrase could hardly suggest that Ulhu had no
water until Rusa arrived. What Sargon is clearly noting is the irrigation works built by Rusa,
and not that there was no water at Ulhu until Rusa discovered it.
the location of ulhu and uiše in sargon ii’s eighth campaign 377
For where indeed did the water that flowed as copiously as the Euphrates
into the canal have its source but from the interior—from the interior of the
city of Ulhu? This meaning surfaced in my original reading of the text and
it was reinforced by the observations given above. Further reinforcement
came when I learned that the word for city (alu) is feminine (personal
communication from L.D. Levine). This fact might allow one to suggest
that the feminine “its” refers to the city (line 200), and there is no other
antecedent feminine reference in the surviving text.
This vivid description of water gushing from the city of Ulhu precisely
records the nature of the water source at Qalatgah. And, as written in Sar-
gon’s text, it is a fine example of Oppenheim’s (1960: 134) astute comment
that the Assyrian writer had in mind a literary concept that sought to cre-
ate a visible perception or imagery, a “representation of reality as in the
[carved stone] reliefs.” Qalatgah has within its perimeters two separate water
sources: springs that gush forth from the mountain rock; one is at the south,
the other to the west. To this day, its waters are directed into channels used
for irrigation, and there is a pool with a dam that collects this water before
it flows further. It was among the stones of the dam that the inscription was
discovered.10 Nowhere else in the text does one encounter the enthusiasm
expended on the water source of a city, and it is worthy of note that no other
site in nw Iran but Qalatgah qualifies for such enthusiasm.
There is yet another specific observation in the text that catches our
attention, for it too relates to a feature at Qalatgah. As Sargon departs from
Ulhu, and before he reaches Armarili, he mentions Urartian watchtowers
on the summit of mountains (lines 249–250). A tower-like structure that
has been interpreted as a watchtower or observation post (or possibly, of
course, a fire-signal tower) was discovered by S. Kroll and W. Kleiss on a
height separated by a saddle from Qalatgah’s peak.11 The correspondence of
10 See Kleiss (1971: 64, fig. 14) for the springs and dam (Teich), and Muscarella (1971: 47). No
attempt has been made to date the pool beyond its present usage. The water from the springs
flows in quantity and is most delicious.
11 Kleiss (1977b: 70–72, figs. 22, 23), but with no mention of the Sargon reference. Zimansky
(1985: 45) mentions another watchtower near Verachram, south of Lake Sevan.
Sargon also mentions plane trees at Ulhu. Wright (1943: 185, note 60) says they grow
“only” in Salmas and Urmia (Rezaiyeh), not further north, but he says nothing of the south.
H. Bobeck, however (1968: 286), says they occur in the Zagros from Turkey south to Fars,
growing only in ravines. None occur today at Qalatgah, but (although I do not remember
myself) I am informed by a colleague that she saw them growing in the town of Ushnu. They
are a favorite tree in modern Iran and are planted wherever there is sufficient water, as in
Teheran.
378 chapter eleven
12 The chronology is known from the results of many surveys (Muscarella 1971: 46–47;
collectively make up Meshta, mentioned only in late 9th- and early 8th-century bc Urartian
texts at Tashtepe and Karagünduz (1979: 177; 1982: 391–392; also 1984: 19–21, 32–33, 47, fig. 2).
He further suggests that the latter inscription actually records the destruction of Hasanlu IVB.
He does not suggest an historical name for Hasanlu-Solduz in 714bc, except to note that the
Assyrians could have employed a different name than Meshta, i.e. Gilzanu (Salvini 1982: 392;
also 1984: 21), as Reade (1978: 140, fig. 2; 1979: 175–179) suggested.
the location of ulhu and uiše in sargon ii’s eighth campaign 379
Levine in part). That Sargon does not mention Lake Urmia in connection
with Armarili may cause concern, inasmuch as Solduz at its north reaches
the lake. The omission, however, may merely signify that his troops did not
journey there through the marshes while in Armarili, whereas when in Aiadi
(see below) they could not avoid mentioning the lake. (The western sector
of Armarili would be where Hasanlu is in fig. 3.)
Yet another problem in connection with the ancient nomenclature and
relationship of the Ushnu and Solduz valleys is raised by the fact that after
departing from Ulhu and before arriving at Armarili, Sargon attacked 21
strongholds in a mountainous and fertile region. If the Solduz valley was
Armarili, then those strongholds must have been in the Ushnu valley. Such
an interpretation is not impossible, as the valley is bordered on three sides
by mountains; we may also recall that it was at this stage of his campaign that
Sargon mentioned the watchtowers (see above). What is more, Zimansky’s
(1985: 42) reading of this event is that the 21 cities “were clearly grouped
together” on a single mountain, a view that does not conflict with the
suggestion that the action described occurred near Ulhu, and in the Ushnu
the location of ulhu and uiše in sargon ii’s eighth campaign 381
14 On his 1977 map (here fig. 2), Levine clearly does not include the Tashtepe area in
Armarili. He also (1977: 150) considers the fortress city of Hasanlu (Period III) in western
Solduz to be a city within Armarili. But was it one of the 21 cities of Sangibutu captured by
Sargon—or one of the cities of Armarili?
15 Note that in the Ushnu valley there is but one river, the Godar Çay, ca. 2.5km south of
Qalatgah.
382 chapter eleven
is to be counted as the first of the three crossed by Sargon. The first river
above Rasakan is the Baranduz Çay, running e-w across the plain; north of
it is the Berdesur Çay, also running e-w; above it is the Sahr Çay and the
Ruzeh Çay, both meeting at Urmia-Rezaiyeh, the modern city; still further
north is the Nazlu Çay. Whichever river is counted as the first mentioned
by Sargon, however, it appears likely that Uiše lay north of the Berdesur
Çay.16 Here we leave the modern map and turn to the archaeological map
filled in by the important work of W. Kleiss. From this map it is immediately
clear that only one major site exists in the Urmia plain, and only one
qualifies for consideration as reliably reflecting Sargon’s description of Uiše,
the strongest of all the fortresses of Rusa. That site is Qaleh Ismael Aga on
the south bank of the Nazlu Çay.
Sargon’s text does not expend much space describing Uiše except to
record that it was “a great fortress, which was stronger than any other of his
[Rusa’s] fortresses,” that it had orchards and forests, and that he captured
it from the rear. Kleiss’ survey and plan (1977b: 64, figs. 14–16) show the site
to be situated in a “strategisch sehr günstiger Position auf einem isoliert vor
dem Gebirgsrand sich erhebenden, doppelgipfligen Felsmassif … [which]
liegt an einem alten und auch von dem Urartäern benutzten Weg …” Striking
is the cliff face below the site at its north, directly above the Nazlu Çay,
which could be perceived as the rear or back of the site. Qaleh Ismael Aga
is the third largest Urartian site in Iran (Qalatgah has yet to be measured),
“a complex site [that] dwarfs all other ruins of the Urartian period in the
plain of Urmiyeh” (Zimansky 1985: 39, table 2). The pottery collected in
surveys (Kroll 1977: 98–106) consists of Iron Age grey wares (probably all
of Iron Age II) and Iron Age III types. Qaleh Ismael Aga was a powerful
fortress in the late 8th century bc (Zimansky 1985: 10);17 and its geographical
16 The point is, we do not know whether Aiadi extended from, say, the area only between
Heyderabad and the Rasakan plain, consisting of a narrow shore and the adjacent mountains,
or whether it extended further north to the Urmia plain. In the text it is not made clear
whether the three rivers were considered to belong to Aiadi or to Uiše.
17 Kroll (1977: 107–108) assumes from the non-diagnostic greyware asemblage that the site
was founded by Sarduri or Ishpuini and that the site is older than Qalatgah, even as old as
Tushpa; he also believes that Qaleh Ismael Aga is located in Gilzanu (see note 13 above). A city
named Uišini is mentioned in a text of Ishpuini (Meher Kapisi) and also in a text of Menua
(Akdaman) (König 1955: 52, 55, 67–68 and notes 10, 31). If Uišini is the Urartian for Uiše, as
König posits (and also Kinnier-Wilson 1962: 109–110), then there was a city Uiše existing in
the 9th century bc Sargon (line 285), however, mentions a city in Aiadi called “the old Uiše”
on the shore of Lake Urmia, which is clearly distinct from the city of Uiše mentioned after
leaving Aiadi and crossing three rivers. Therefore, it is not clear whether the Urartian texts
the location of ulhu and uiše in sargon ii’s eighth campaign 383
position coincides with that of Uiše on “the lower border of Urartu and on
the Nairi frontier” (line 298).
If Qaleh Ismael Aga is Uiše, then Levine’s (fig. 2) and Mayer’s maps must
be considerably adjusted: Aiadi should be situated along the southern, lower
third, of Lake Urmia’s west shore and the district of Uiše must be shifted
east (i.e. in general, to a position on their maps where Aiadi is written),
presumably to include a large part of the Urmia plain. Aiadi and Uiše were
provinces or districts within the borders of modern Iran (fig. 3).
At this stage in the study I learned that an Italian team had excavated
at the site of Qaleh Ismael Aga and published the results in Tra los Zagros e
l’Urmia.18 But more than publishing the excavation results (Salvini 1984: 215–
239), the identification of the site is discussed by M. Salvini, who concludes
that it was ancient Uiše. Before commenting on this identification, however,
he returns (1984: 12, 24) after more study (see Salvini 1979) to the presence
of U-i-še in the inscription from Qalatgah, which site he had previously
tentatively identified as Uiše. He is now able to report that the reading of the
name is in fact far from certain, that a lacuna between the allegedly written
i and sē could allow for a totally different reading, and his conclusion is that
a new collation is warranted. He further notes that even if the originally
proposed reading holds up, we still do not possess the missing part of the
text that would explain why and in what context the city is mentioned. In
short, the Qalatgah inscription itself cannot at present be used to identify
that site.
With regard to Qaleh Ismael Aga’s identification with Uiše, Salvini (1984:
46–51) notes that on the basis of Assyrian documents other than the eighth
campaign, Uiše must lie se of Urartu proper and near Hubushkia, and that
the three rivers Sargon crossed fit the local topography of the Urmia plain.
refer to old Uiše or to (new) Uiše, which leaves Kroll’s early dating of Qaleh Ismael Aga far
from certain. Old Uiše and Uiše were not far apart geographically, as Zimansky (1985: 119,
note 130), contra van Loon, correctly noted.
18 I learned this from L.D. Levine, whom I called to share my views on the Ulhu and Uiše
identifications. The volume was in my library for some months waiting to be read. It is a rich
book, filled with valuable information and insights on “le province orientali dell’Urartu.” The
section on Qaleh Ismael Aga is only one of many topics discussed. Further, I then wrote to
M. Salvini (January 10, 1986) to share my views on Ulhu and Uiše and to call attention to the
fact that on independent grounds we had both identified Uiše. On February 3, 1986, by which
time this manuscript was finished and about to be sent off for publication, I received from
Salvini the article he co-authored with P.E. Pecorella (1982) which I did not possess. Here, the
authors already concluded that Qaleh Ismael Aga was to be identified with Uiše. In addition,
they cover much of the ground dealt with in Salvini (1979, 1982, and 1984).
384 chapter eleven
He calls attention to the large size of the site and the fact that no other large
one exists in the area, that it protects a main road, and that it has a back—
the lower castle over the cliff (see Salvini 1984: 219, 228, figs. 40, 49). And as
a result of his identification of Uiše, Salvini (1984: 47, fig. 2) makes the same
adjustments to Levine’s map as I suggested above for Aiadi and Uiše.19
Two independent approaches, one starting from the identification of
Ulhu and tracking Sargon’s route from that posited position, the other work-
ing from local geographical and topographical observations, have reached
the same conclusion that Qaleh Ismael Aga was ancient Uiše. Ulhu and Uiše,
it is suggested, are to be considered (and reviewed) by scholars as points
fixed on the ancient map of nw Iran, and they join the two others hitherto
recognized in the general area, Musasir and Tashtepe-Meshta.
Having identified and reunited Ulhu and Uiše with their ancient mounds,
the aim of this paper is accomplished. Yet, having tracked Sargon to Uiše-
Qaleh Ismael Aga, one hesitates to abandon him. This is especially so,
because my understanding of Sargon’s text is that his journey from Uiše
may have been shorter than has previously been recognized. Leaving Uiše,
Sargon went to Nairi where he was visited by the king of Nairi-Hubushkia,
Ianzu, who arrived from his royal city, also called Hubushkia, a distance of
four bêru (double-hour). Abruptly, in the next line, and with no topograph-
ical commentary, Sargon is in the royal city receiving tribute. The state of
Hubushkia is considered to be situated in se Turkey, near the headwaters of
the Upper Zab, in the vicinity of Çölemerik-Hakkiri and Yüksekova (Levine
1977: 144; Salvini 1982: 386; 1984: 40, 49). But did Sargon journey far west,
to or near the headwaters of the Upper Zab, some 50–60 miles as the crow
flies, without comment, to reach the royal city? I doubt it. Further, after leav-
ing Hubushkia for Musasir, the only recognizable geographical reference
mentioned is the Upper Zab, a river locally called Elamunia, which Sargon
crossed shortly after he started on his commando raid. This reference cannot
be interpreted with certainty to signify that Sargon traveled the long dis-
tance to the Upper Zab headwaters from Uiše.
19 Salvini does not refer to Ulhu and has not attempted to identify Qalatgah with an
ancient name, except to place it within Armarili. On his map, figure 2, Armarili seems to
occupy more territory than suggested in his text (Godar Çay valley). On this map, Salvini
places Sangibutu south of Mahabad. I do not claim to know how far south on the map
this province is to be placed, but inasmuch as I have argued that the Ushnu valley was the
northern part of Sangibutu, then it follows that I would prefer Subi to be placed in or near
the position where Salvini has written Sangibutu, and that the latter be shifted north.
the location of ulhu and uiše in sargon ii’s eighth campaign 385
20 Lehmann-Haupt (1916: 140; 1926: 313) suggests the Topzawa Çay; Boehmer (1979: 124)
suggests the Barasgird River, ca. 20 km north of Musasir. But if the Upper Zab/Elamunia was
crossed early in the journey, as it is presented in the text, then the river would have to be
further north. Reade (1978: 141) claims Sargon crossed the Rowanduz Çay, but this river lies
to the south of Musasir (Boehmer 1973: fig. 1); see the following note.
21 The routes I propose here are of course speculative, although, I insist, viable and
attractive; no one knows from a map how Sargon got to Musasir. One wonders whether
Sargon traversed the Kel-i Shin Pass, which is best entered from the Ushnu valley.
The problem is that we do not know whether Sargon returned to the Ushnu valley, or
whether earlier Assyrian kings used this pass, and Sargon claims that his route to Musasir was
unique. Further, the pass was controlled by the Urartians (Salvini 1982: 389–390; 1984: 43), and
perhaps Sargon would therefore not have used it for this reason and because doing so would
have compromised his surprise attack. Lehmann-Haupt (1916: 139–140; 1926: 309–310, 313,
326–327) van Loon (1975: 207), and Zimansky (1985: 112, note 64) believe Sargon took the Kel-i
Shin Pass; Boehmer (1973: 39, note 24; 1979: 124) and Reade (1978: 140–141, fig. 2) do not. The
different opinions on this issue are based on views concerning whether Sargon circled Lake
Urmia and concomitantly the locations of Uiše and Hubushkia. Reade, for example, believes
with Kinnier-Wilson (1962: 110) in the encirclement of Lake Urmia and that Hubushkia lay sw
of Lake Urmia near Kaneh. In this view, Sargon approached Musasir from the south.
386 chapter eleven
References
Bobek, H.
1968 “Vegetation,” in The Cambridge History of Iran. Cambridge: The University
Press: 280–293.
Burney, Charles A.
1972 The Peoples of the Hills. New York: Praeger.
Çilingiroğlu, Altan A.
1976–1977 “The Eighth Campaign of Sargon II,” Jahrbuch für kleinasiatische For-
schung 4/5: 252–269.
Kinnier-Wilson, J.V.
1962 “The Kurba’il Statue of Shalmaneser III,” Iraq 24: 90–115.
Kleiss, Wolfram
1969–1970 “Zur Ausbreitung Urartus nach Osten,” Istanbuler Mitteilungen 19/20:
125–136.
1971 “Bericht über Erkundungsfahrten in Iran im Jahre 1970,” Archaeologische
Mitteilungen aus Iran 4: 51–111.
1976 Topographische Karte von Urartu. Archaeologische Mitteilungen aus Iran,
Ergänzungsband 3.
1977a “Alte Wege in West-Iran,” Archaeologische Mitteilungen aus Iran 10: 137–151.
1977b “Urartäische Plätze in Iran,” Archaeologische Mitteilungen aus Iran 10: 53–
83.
Kroll, Stephan
1976 Keramik urartäischen Festungen in Iran. Archaeologische Mitteilungen aus
Iran, Ergänzungsband 2.
1977 “Die Oberflächenfunde des Urartu-Surveys 1976,” Archaeologische Mittei-
lungen aus Iran 10: 83–118.
Laessøe, Jørgen
1951 “The Irrigation System at Ulhu, 8th Century B.C.,” Journal of Cuneiform
Studies 5: 21–32.
the location of ulhu and uiše in sargon ii’s eighth campaign 387
Lehmann-Haupt, C.F.
1916 “Musasir und der achte Feldzug Sargons II (714 v. Chr.),” Mitteilungen der
vorderasiatisch-aegyptischen Gesellschaft 21: 119–151.
1926 Armenien Einst und Jetz, 2, 1. Berlin and Leipzig: B. Behr/F. Feddersen.
Levine, Louis D.
1977 “Sargon’s Eighth Campaign,” in L.D. Levine and T. Cuyler Young, Jr., eds.,
Mountains and Lowlands. Malibu, CA: Undena Publications, 135–151.
Mayer, Walter
1980 “Sargons Feldzug gegen Urartu—714 v. Chr. Eine militärhistorische Würdi-
gung,” Mitteilungen der Deutschen Orient-Gesellschaft zu Berlin 117: 13–32.
Meissner, Bruno
1922 “Der Eroberung der Stadt Ulhu auf Sargons 8. Feldzug,” Zeitschrift für Assyri-
ologie 34: 113–122.
Oppenheim, A. Leo
1960 “The City of Assur in 714 bc,” Journal of Near Eastern Studies 19: 133–147.
Piotrovskii, Boris B.
1959 Vanskoe Tsarstvo. Moscow: Vostochnoi Literaturi.
Reade, Julian
1978 “Kassites and Assyrians in Iran,” Iran 16: 137–143.
1979 “Hasanlu, Gilzanu, and Related Considerations,” Archaeologische Mitteilun-
gen aus Iran 12: 175–181.
Salvini, Mirjo
1979 “Die urartäischen Schriftlichen Quellen aus Iranisch-Azerbaidjan …,” Ar-
chaeologische Mitteilungen aus Iran, Ergänzungsband 6: 170–177.
1982 “Forschungen in Azerbaidjan: ein Beitrag zur Geschichte Urartus,” Archiv
für Orientforschung, Beihaft 19: 384–394.
1984 “La Storia della Regione in Epoca Urartea,” 9–51 and “I Documenti,” 53–134
in Paolo Emilio Pecorella and Mirjo Salvini, Tra lo Zagros e l’Urmia. Rome:
Edizioni dell’Ateneo.
388 chapter eleven
Thureau-Dangin, François
1912 Une relation de la huitième campagne de Sargon (714 av. J.-C.). Paris: Paul
Geuthner.
Wäfler, Markus
1976 “Urartu und Assyrien,” in Hans-Jörg Kellner, ed., Urartu: ein wiederentdeck-
ter Rivale Assyriens. Munich: Museum für Vor und Frühgeschichte, 19–21.
Weidner, Ernst F.
1937–1939 “Neue Bruchstücke des Berichtes über Sargons achten Feldzug,” Archiv
für Orientforschung 12: 144–148.
Wright, Edwin M.
1943 “The Eighth Campaign of Sargon II of Assyria (714BC),” Journal of Near
Eastern Studies 2: 173–186.
Zimansky, Paul
1985 Ecology and Empire: The Structure of the Urartian State. Chicago: The Ori-
ental Institute of the University of Chicago.
chapter twelve
Abstract: In 1938 Erich Schmidt, taking time out from his major work at Persepolis,
excavated for three weeks the site of Surkh Dum in eastern Luristan, in western
Iran. Although very little has been published on the finds and architecture, aside
from two brief and summary reports by Schmidt and Maurits van Loon, Surkh Dum
is recognized by Iranian archaeologists to be one of the most important sites in
Luristan, and in Iran in general. Not only was Surkh Dum a settlement site, rather
than a cemetery—which is the typical circumstance in the archaeological history
of Luristan—but many hundreds of objects of bronze, ivory, bone, faience, and
terracotta, as well as about 200 cylinder and stamp seals, were recovered. To date,
only seven of the objects have been published, and nothing has been published
about the two buildings partially uncovered. In 1943 The Metropolitan Museum of
Art acquired 41 objects excavated at Surkh Dum, only five of which had previously
been published. Because of the importance of the material for modern knowledge
of the art and archaeology of Luristan, an area plundered since the late 1920s, and
the source of countless thousands of unexcavated objects, the presentation of even
a small group of excavated artifacts from Luristan is considered to be of great value.
The present paper offers a history of our present knowledge of the site, a tentative
discussion of its chronology, and a catalogue discussion of the Surkh Dum material
in The Metropolitan Museum of Art.
The Background
Following his air reconnaisance trip in 1937 organized to survey sites for
future excavations, Erich Schmidt in the summer of 1938 began his arduous
“epic” overland journey into Luristan, in western Iran, through the “ragged
crests of the Zagros” in search of the sites plotted by air. After “weeks of
failure” he struck his “archaeological Bonanza”: the site of Surkh Dum in
the Kuh-i Dasht region of eastern Luristan (fig. 1).1 At the time, Schmidt’s
* This article originally appeared as “Surkh Dum at the Metropolitan Museum of Art: A
Institute for Iranian Art and Archaeology V, 3 (1938) 205. The name of the site has been
published as Surkh Dum, Dum Surkh, Surkah Dum, Surkh-i Dum, Surkh Dum-i Luri: see Oscar
White Muscarella, Unexcavated Objects and Ancient Near Eastern Art: Addenda (Undena 1979)
12, note 8; Helene Kantor, “Embossed Plaques with Animal Designs,” JNES V (1946) 234, note 3,
235, 237; Peter Calmeyer, Datierbare Bronzen aus Luristan und Kirmanshah (Berlin 1969) 87,
143–145, 150, 159, 169, 188; Pierre Amiet, Les Antiquites du Luristan (Paris 1976) I wish to thank
Trudy S. Kawami for reading my manuscript and for sharing with me her notes and thoughts
on Surkh Dum.
2 The nature of the sponsorship is not made clear in Schmidt’s report: on page 204 it is
stated that “the staff of the Persepolis Expedition was put at the disposition of the American
Institute. …”
3 Schmidt’s report appeared very shortly after his Luristan campaign was terminated, and
as stated on page 208: “the individual objects have not yet been photographically recorded.
…” The excavations at Surkh Dum were obviously rushed, taking only three weeks time.
V.E. Crawford informs me that Richard Carl Haines, the architect, told him that the finds
appeared in such profusion that excavation had to cease in order to register the objects; this
may explain why some of the objects do not have a SOR field number painted on them. (SOR
derives from Sorkh Dum, which is a variant spelling of Surkh Dum.) The campaign was also
hampered by the terrain; among other problems, fodder for the donkeys used in transport
had to be imported.
surkh dum at the metropolitan museum of art 391
4 Phyllis Ackerman, Guide to the Exhibition of Persian Art, 2nd ed. (New York 1940) 541.
5 Ibid. 130, KK, 532, 534–536, 540, 542–548.
6 Ibid. 547, XX. Both Ackerman, 532, F, and Maurits van Loon in his review of Dark Ages
and Nomads c. 1000 B.C., Machteld Mellink, ed. (Istanbul 1964), in BibO XXIV, 1/2 (1967) 24,
mention a spiked axe with a lion-head juncture. This axe is now in the University Museum
(SOR 1633) and has a blade and spike formation of the same type as P.R.S. Moorey, Catalogue
of the Persian Bronzes in the Ashmolean Museum (Oxford 1971) 53–55, no. 20: but with the lion’s
head facing the spikes, not the blade. Thus, it is not the same form as suggested by Moorey,
ibid, 51, nos. 14, 15, and in his Ancient Persian Bronzes in the Adam Collection (London 1974)
43, nos. 7, 8.
7 “Prehistoric Nature Worship in Western Iran,” ILN (March I, 1941) 292.
8 See Muscarella 1979, op. cit. (in note 1) 13.
surkh dum at the metropolitan museum of art 393
9 Phyllis Ackerman, “The Gemini are Born,” Archaeology 8, 1 (1955) 26; the date 1937 is
repeated.
10 Phyllis Ackerman, “A Luristan Illustration of a Sunrise Ceremony,” Cincinnati Art Muse-
um Bulletin 5, 2 (1957) 4. Trudy Kawami has shown me a sketch plan of the site: two buildings
were incompletely cleared. The larger has 17 rooms or areas, the less excavated has five; no
circular walls exist. In the larger building there is a big room, not centered, that has a central
unit, the altar, no doubt.
11 A.U. Pope, Masterpieces of Persian Art (New York 1945) 16; Mehdi Bahrami, “Some
Objects Recently Discovered in Iran,” Bulletin of the Iranian Institute VI, 1–4 (1946) 71. The
sketch plan mentioned in my note 10 does show six entrances to the large room. Bahrami
may have had access to the same plan in Teheran. We will have to await a final report to get
information about the tiles and to learn if the central unit in the large room was in fact a fire
altar.
12 Van Loon, op. cit. (in note 6) 23–24. Van Loon himself has the responsibility for the final
publication.
13 Schmidt, op. cit. (in note 1) 208, note 3.
14 In BibO 29 (1972) 69, note 22.
394 chapter twelve
mentioned by Schmidt, e.g., disc-headed pins, and added others not specifi-
cally mentioned either by Schmidt or Ackerman: a square-framed openwork
plaque (a “wand”?), lobed bronze rings, bone lion pins on an iron shank,
pottery, incised frit vessels, and, surely of some importance, “dedicatory
inscriptions.” He also enumerated the types of pins recovered, straight ones
with a variety of decorated heads, and disc-headed, decorated in repoussé.
Van Loon specifically noted that the typical Luristan standards were non-
existent at Surkh Dum; and it is relevant to mention at this point that neither
Schmidt not Van Loon mentioned any quivers or large decorated plaques,15
nor objects of precious materials. With regard to the pottery cited, it would
be rash to reach firm conclusions on the basis of the brief descriptions given,
but mention is made both of triangles used as decoration and of the absence
of red and grey wares, indications that we are dealing with a post (or late)
Iron II/Iron III context.
To date, the above cited reports and obiter dicta contain all the basic infor-
mation from first- and second-hand sources available about the excavations
of Schmidt at Surkh Dum. Whether all the types of objects recovered have
been mentioned, or there are still others unreported, will not be known until
a final publication appears (in the meantime, see below).
Up to the present time only six of the many hundreds of objects excavated at
Surkh Dum and mentioned by Schmidt, Ackerman, and Van Loon have been
recorded and published with photographs:16 four cylinder seals (Nos. 33–
35, 40, below), a frog-headed pin (No. 11 below), and a glazed faience ves-
sel (cf. No. 31, below). The objects listed by Ackerman in the 1940 cata-
logue of the Persian Exhibition were not illustrated. Trudy S. Kawami has
recently informed me, however, that yet another object from Surkh Dum has
been published, albeit without a provenience reference, and it is an impor-
tant object indeed. In his work on decorated bronze nipple beakers, Peter
Calmeyer17 published a drawing of a small fragment of one of these beakers
15 It is not clear what Van Loon means on page 25 by “hammered bronzes;” and it seems
misleading to claim that they “all seem to come from the top level,” inasmuch as he placed the
hammered disc-headed pins in the earlier levels. What is more, his dating of these hammered
bronzes to the 7th century is based on an incorrect, low dating of the nipple beakers (see
below).
16 Muscarella, op. cit. (in note 1) 12.
17 Peter Calmeyer, Reliefbronzen in babylonischem Stil (Munich 1973) 32–33, A24.
surkh dum at the metropolitan museum of art 395
in the Teheran Museum (No. 1124), where it is listed merely as from Luris-
tan. According to Kawami, who saw the original drawing in the Oriental
Institute in 1974, this fragment was found in the refuse dump at Surkh Dum
and has the field number SOR 1712. No doubt the fragment was given to the
Teheran Museum as part of its division of the excavated finds and presum-
ably the records were lost or misplaced so that Calmeyer was not informed
of the provenience. The discovery that a decorated nipple beaker was exca-
vated at Surkh Dum is significant, for, to my knowledge, among the many
scores known to exist, it remains the only one to derive from a controlled
excavation.18 Furthermore, the provenience is a site in Iran, which reinforces
the generally held view that the beakers as a group derive from that area.19
Equally important is the occurrence of a decorated nipple beaker in a sanc-
tuary, a fact that refutes Calmeyer’s suggestion that all the beakers must have
derived from tombs,20 and which also neatly demonstrates that excavated
material is the only evidence that can be used to form hypotheses regarding
alleged proveniences and the functional value of an artifact.
At the same time, a number of objects, all of which derived from the
antiquities market, that is, from clandestine plundering, have casually been
attributed to Surkh Dum by a number of scholars. In his 1938 report Schmidt
mentioned that digging by local villagers had occurred at Surkh Dum before
he began his campaign and a number of scholars have either called atten-
tion to this fact or have claimed that after Schmidt left more digging took
place.21 Thus, after 1938, when a dealer attributed an object to Surkh Dum
it was found to be convenient to accept the claim, given the references to
the clandestine activity—but which information, let it be recognized, came
from the same dealers.22 It should also be noted that some scholars were mis-
led by information from individuals who supposedly were in a position to
18 Oscar White Muscarella, “Decorated Bronze Beakers from Iran,” AJA 78 (1974) 243–244;
idem, review of Calmeyer 1973, op. cit. (in note 17) in JAOS 97, 1 (1977) 77.
19 For discussion see Muscarella 1974, ibid. 243–245, 248–249; 1977, ibid. 77.
20 Calmeyer 1973, op. cit. (in note 17) 123, 151, 231, 233, and discussed in Muscarella 1977, op.
cit. (in note 6) 19–20; idem, “Some Elaborately Decorated Bronze Quiver Plaques Made in
Luristan, c. 750–650 BC,” Iran XIII (1975) 20; Amiet, op. cit. (in note 1) 1; Roman Ghirshman,
Bichapour II (Paris 1956) 120, note 1; idem, The Arts of Ancient Iran (New York 1964) 48; André
Godard, L’ Art de l’Lran (Paris 1962) 52; Jorgen Meldgaard, et al., “Excavations at Tepe Guran,
Luristan,” ActaA XXXIV (1963) 98, note 5; D. de Clercq-Fobe, Epingles votives du Luristan
(Teheran 1978) Introduction, page 2.
22 More pins seem to have been offered for sale after 1938 than before.
396 chapter twelve
know what in fact came out of the site.23 Others made what they considered
to be intelligent guesses concerning attribution, on the assumption that cer-
tain types of objects, disc-headed pins for example, derive only from Surkh
Dum. We will now examine a few of these misattributed objects.
The well-known bronze quiver plaque in the Metropolitan Museum of
Art, with its rich and important evidence of style and iconography for west-
ern Iran,24 was purchased from A.U. Pope in 1941, two years before he sold
the Museum the material from Surkh Dum to be discussed below. The site
of Surkh Dum was not mentioned in the transaction and the only informa-
tion given was that the plaque came from the Kuh-i Dasht region in Luristan.
Of significance for the issue of origin is that a year before the purchase Ack-
erman, in the aforementioned Persian Exhibition catalogue,25 published the
plaque, naming as its owner the antiquities dealer R. Rabenou, thus reveal-
ing candidly that the object was available on the art market, and that, con-
sequently, its final ancient resting place cannot be known.26 Yet within a few
years, by 1945, Pope published the quiver as coming from Surkh Dum27 and
from this time onward most scholars who cited the plaque also assigned it
to that site.28 A puzzling exception was Ackerman, who in 1955 repeated her
23 See Muscarella 1979, op. cit. (in note 1) 5–6, 13; cf. also statements by René Dussaud,
“Anciens Bronzes du Luristan et Cultes iraniens,” Syria XXVI (1949) 213, or Pope apud Schmidt,
op. cit. (in note 1) 210–211, note 5.
24 MMA number 41.156; Moorey, 1975 op. cit. (in note 21) 19, note 1, 24–26, pl. 1; Edith
Porada, “Iranische Kunst,” in Der alte Orient, W. Orthmann, ed. (Propyläen Kunstgeschichte:
Berlin 1975) 397–398, no. 317. From what information I have been able to gather, the quiver
was on the market at least by 1939. Stylistically judged, the quiver is definitely Iranian in
manufacture.
25 Ackerman, op. cit. (in note 4) 115: not 199as in Muscarella 1979, op. cit. (in note 1) 13,
and 1957, op. cit. (in note 10) 4. There are no references to quivers in the Surkh Dum field
records: cf. Moorey 1975, op. cit. (in note 21) 19, note 1, where the Surkh Dum archives are
cited and where the distinct impression is given that the quiver was excavated by Schmidt.
V.E. Crawford told me that Carl Haines claimed that no works of art were found at Surkh Dum.
Presumably it could be argued by the interested parties that the quiver, and others, were
found after Schmidt left, by the local peasants: but this type of argument involves gratuitous
guessing, not archaeological reasoning, and begs many questions.
27 Pope 1945, op. cit. (in note 11) pl. 15.
28 Dussaud, op. cit. (in note 23) 213, citing Mme. Godard as an authority; Calmeyer 1969,
op. cit. (in note 1) 87, note 289, 290, D, 145, 159; Amiet, op. cit. (in note 1) 85; Helene Kantor,
“The Shoulder Ornament of Near Eastern Lions,” JNES VI (1947) 258; Edith Porada, “Nomads
and Luristan Bronzes,” in Mellink, op. cit. (in note 6) 27, note 61; idem, The Art of Ancient
Iran (New York 1965) 87–89, 236, note 10; in her 1975 paper, op. cit. (in note 24) she modified
her position, omitting the site attribution; Henrik Thrane, “Archaeological Investigations in
surkh dum at the metropolitan museum of art 397
original claim that it was known through the antiquities market.29 At least
three scholars further believed that not only this quiver, but still others, were
found in the Surkh Dum sanctuary, or belonged to a “Surkh Dum group.”30
Disc-headed pins are perhaps the main type of object (aside from quiv-
ers) most often associated with Surkh Dum, because indeed many were
recovered there. To date, however, not a single example of those excavated
by Schmidt has been published (now see No. 3, below). What has been
published, on the other hand, are scores of examples, all of which derived
from the antiquities market, and the majority of which have been casually
attributed to Surkh Dum. Disc-headed pins began to surface during the time
of the earliest appearance of the Luristan bronzes,31 but whether they were
plundered from one or several sites is of course not known. This latter point
notwithstanding, some scholars in the past have claimed to know the mul-
tiple sources. Thus, Pope, in an editorial comment about the Surkh Dum
wands, stated that “closely related pins have been found at other sites …,”
while Amiet believes that disc-headed pins have been “découvertes dans
des tombes.”32 Presumably, the pins discussed were those known before
Schmidt went to Surkh Dum, for after 1938 only Surkh Dum is mentioned.
And it was Pope again33 who, forgetting or ignoring his 1938 comment about
“other sites”, began to cite stray disc-headed pins as coming from Surkh Dum.
As with the quiver, various scholars over the years followed Pope’s lead—
probably because of his institution’s role in the excavations—and accepted
as received knowledge that any stray disc-headed pin, including these in the
Coiffard and Graeffe collections, came from one site, from Surkh Dum.34
Western Luristan,” ActaA XXXV (1964) 159, note 7 (not 59, as Muscarella 1979, op. cit. [in note 1]
13); Irene Winter, A Decorated Breastplate from Hasanlu, Iran (Philadelphia 1980) 18, note 89.
29 Ackerman 1955, op. cit. (in note 9) 27, 29, fig. 4.
30 Ghirshman 1964, op. cit. (in note 21) 70; Moorey 1975, op. cit. (in note 21) 19–26; E.D. Phil-
lips, “The People of the Highland,” in Vanished Civilizations, E. Bacon, ed. (New York 1963)
227; Amiet, op. cit. (in note 1) 85, claims that the David-Weill quiver came from Surkh Dum
“certainement.”
31 Viz. A.U. Pope in ILN (September 6, 1931) 390, fig. 16; André Godard, Les Bronzes du
Luristan (Paris 1931) pl. XXXIV; Leon Legrain, Luristan Bronzes in the University Museum
(Philadelphia 1934) pl. VI; Ernst Herzfeld, “Das Ornament nach der Mitte des II. Jahrtausends,”
AMIran VIII (1937) 156–157, Abb. 118; J.A.H. Potratz, “Scheibenkopfnadeln aus Luristan,” AfO
XV (1941–1945) 39, note 5; cf. Ghirshman 1956, op. cit. (in note 21) 120, note 1, and Clercq-Fobe,
1978 op. cit. (in note 21) Introduction, page 1.
32 Pope apud Schmidt, op. cit. (in note 1) 210, note 5; Amiet, op. cit. (in note 1) 75.
33 Pope 1945, op. cit. (in note 11) 16, pl. 16A, B.
34 Viz. Kantor 1946, op. cit. (in note 1) 234 and note 3; Bahrami, op. cit. (in note 11) 71–73,
figs. 1, 2 (fig. 1 = Ghirshman 1964, op. cit. [in note 21] fig. 490, listed as from the Teheran art
398 chapter twelve
Chronology
market); Dussaud, op. cit. (in note 23) 196–205, figs. 1–7, pl. IX, X; Ghirshman 1956, op. cit.
(in note 21) 120–122; Calmeyer 1969, op. cit. (in note 1) 87, note 288, 145, note 463: but cf. 143,
note 458; Amiet, op. cit. (in note 1) 1, 75; Clercq-Fobc, op. cit. (in note 21) Introduction, p. 2,
115; Louis vanden Berghe, Archéologie de l’ Iran Ancien (Leiden 1959) 93, 276, pl. 123e–h; Pierre
Amandry, “Un Motif ‘scythe’ en Iran et en Grèce,” JNES XXIV, 3 (1965) 151, with Surkh Dum in
quote marks.
35 Pope 1945, op. cit. (in note 11) 16, pl. 16A.
36 See Muscarella 1979, op. cit. (in note 1) 8, Iranian no. 6.
37 Cited by me as forgeries in “Unexcavated Objects and Ancient Near Eastern Art,” in
Mountains and Lowlands, eds. L.D. Levine and T.C. Young, Jr. (Undena 1977) 171, 172, nos. 1, 2,
5, 173, note 75; see also Muscarella 1979, op. cit. (in note 1) 3, nos. 3 and 5, also page 6. From
private sources I know that Pope was the vendor of the silver plaque and that he attributed it
to Surkh Dum. I recently examined the pin illustrated in Dussaud, op. cit. (in note 23) fig. 6,
pl. x; cf. Muscarella 1977, ibid. 173, note 75: I have no doubt that it is not ancient.
38 Dussaud, op. cit. (in note 23) 210, fig. 10; Muscarella 1977, op. cit. (in note 37) 175, no. 56.
surkh dum at the metropolitan museum of art 399
the one hand, the alleged lack of stratigraphy, and on the other, the existence
of several phases, has already been mentioned. Inasmuch as Van Loon had
access to Schmidt’s records, we must assume his observations to be correct
and that the existence of floor levels and rebuildings are implied in his state-
ment. Defining these phases, Van Loon claimed that there was a 10th, a 9th,
an 8th, and a 7th century level (four phases?), and that the sanctuary was
in existence from the 10th through the 7th centuries bc, a period of over
300 years.39 Five years later the dates of some of the levels were revised: the
9th century level is actually 8th century; the 8th century level is actually late
8th–early 7th century; nothing was said of the old 10th or 7th century levels.40
Nowhere, in either of the two statements on chronology, is an explanation
given for the determination of dating, although one may infer that it is based
on art historical analyses of the artifacts.
Actually, an important clue concerning dating exists in the presence of
iron, which Van Loon noted occurs in all levels, including the earliest. This
fact alone surely suggests that there is no 10th century level at Surkh Dum,
for there is no evidence to date that the presence of iron is documented
anywhere in Iran before the late 9th century bc.41 It therefore follows that
the earliest level cannot predate the 9th century bc at the earliest, and may
actually not pre-date the 8th.
Confusing the issue of chronology at Surkh Dum is the material of pre-
sumably earlier periods (i.e., earlier than the posited late 9th/8th century
date for the incipient level) recovered there: cylinder seals, a spiked axe with
a lion’s mask juncture (see note 6), duck-headed pins (see No. 13, below), and
the nipple beaker fragment mentioned above.42 Spiked axes with crescent-
shaped blades are generally dated to the last centuries of the 2nd millen-
nium bc on the basis of related inscribed pieces, none of which, however, has
the zoomorphic juncture,43 and the recent invaluable discoveries of Vanden
Berghe at Bard-i Bal and Kutal-i Gulgul in Luristan have not only furnished
additional Luristan locations for the type (one from the latter site has a
Museum 6 (1969) 34; Louis vanden Berghe, “La Chronologie de la Civilisation des Bronzes du
Pusht-i Kuh, Luristan,” Proceedings of the 1st Annual Symposium of Archaeological Research
in Iran (Teheran 1973) 4; Vincent Pigott, “The Question of the Presence of Iron in the Iron I
Period in Iran,” in Mountains and Lowlands, op. cit. (in note 37) 218, 223, 226, 231.
42 Schmidt, op. cit. (in note 1) 208, note 3, 210; see also Elizabeth Williams-Forte’s discus-
Archeologia 36 (1970) 10, 13; idem, “Recherches Archéologiques dans le Pusht-i Kuh Luris-
tan,” Bastan Chenasi va Honar-i Iran 6 (1971) 20–21, 26, figs. 11, 13, 28; idem, “Recherches
Archéologiques dans le Luristan,” Iranica Antiqua X (1973) 35, fig. 20, pls. XVI, XVIII, 1; idem,
“La Nécropole de Kutal-i Gulgul,” Archeologia 65 (1973) 18, 22, 24, 25; for revised dating, see
Vanden Berghe 1973, op. cit. (in note 41) 4; Edith Porada, “Ancient Persian Bronzes,” Apollo
(February 1979) 142.
45 See here notes 88, 94, 106.
46 Calmeyer 1969, op. cit. (in note 1) 224–228; Muscarella, 1974 op. cit. (in note 18) 243–249;
Carquois du Luristan,” Syria LI (1974) 244; idem 1976, op. cit. (in note 1) 30, 75, 103. Cf. Clercq-
Fobe, op. cit. (in note 21) Introduction, page 1, who dates the site as beginning in the second
half of the 2nd millennium bc, based on a misunderstanding of Schmidt’s statements and
Pope’s comments about early seals at Surkh Dum.
surkh dum at the metropolitan museum of art 401
Iran Bastan Museum, and the remainder subsequently in the United States
among the Oriental Institute of the University of Chicago, the University
Museum of the University of Pennsylvania, Mrs. Boyce Thompson, and
Pope. It was the latter collection that eventually came to the Metropolitan
Museum, but it is not known to me whether some pieces in Pope’s share
were sold or given to other institutions or individuals, or whether his whole
share arrived intact.48
The objects have been in the Metropolitan Museum for almost 40 years,
during which time only four cylinder seals and a frog-headed pin have
been summarily published. Inasmuch as many hundreds of objects were
excavated at Surkh Dum, the Museum’s collection represents only a fraction
of the total. Yet it may be stated without elaboration that the value for
Iranian archaeology in presenting even this fraction should be obvious.
For, given the recognized significance of the finds from one of the most
important sites in Iran, and one of the few excavated settlements in Luristan,
as well as the attribution to the site of the many dealer-derived objects
mentioned above, a publication is surely long overdue. With these thoughts
in mind, the present paper is to be considered not so much a preliminary,
but rather a mini-report of the Surkh Dum excavations.
In the lists that accompanied the objects sent to the Metropolitan Muse-
um in 1943, the only information supplied was a Surkh Dum field number for
each object, sometimes a plot number and a “pr” (presumably a plot record)
number, and rarely a R (room) reference; only some of the objects have a
field number painted on it (see Note 3). In a few cases a reference to the
Persian Art Exhibition Catalogue of 1940 was furnished. But there was no
information regarding stratigraphy, so that until a final publication appears
with full details the incomplete information presented here remains tem-
porarily without significant [stratigraphic] meaning.
The aim of the following catalogue is to make available for study and
discussion that group of artifacts existing in one of the several reposito-
ries of the Surkh Dum corpus. And it will be clear that the catalogue does
not impinge upon the final report, it only anticipates it. The bronzes are
presented first, followed by the bone and ivory objects, the faience vessel,
and, finally, the cylinder seals, which are described and discussed by Eliz-
abeth Williams-Forte. The headings contain the Museum and Surkh Dum
record numbers. For each object a description is given along with a brief
48 The Museum’s records do not reveal in what capacity Pope functioned when he sold
the objects.
402 chapter twelve
Catalogue
Metal Artifacts
No. 1. Plaque. 43.102.11; Surkh Dum 1721. Bronze; P.H. 9.5 cm.
Two stylized leonine creatures face each other in heraldic position. Their
mouths are open and tongues and fangs protrude; manes are depicted as
thick curling tufts. The upward curling tail, the neck and the thigh on the left
figure are accentuated with punched dots. The paws apparently originally
touched and both creatures seem to be rampant, standing on rear legs. The
border consists of raised dots framed by narrow bands; the top and part of
the right border are extant.
surkh dum at the metropolitan museum of art 403
There is at least one other fragment of a similar plaque from Surkh Dum,
preserving only the eye and mane of a creature exactly like the present
example. I know of only one parallel to these two plaques, one more com-
plete, and formerly in the David-Weill collection.49 From this example we
are able to restore our fragments as a rectangle, with perhaps a dead ani-
mal under the creatures, indicating perhaps that the latter are fighting over
their prey. The David-Weill plaque has subsidiary motifs around the crea-
tures lacking on ours; what is more, those lions are winged and the necks
and bodies are rendered in a more baroque fashion than ours, indicating a
separate workshop. What function the plaques had is not known, but, inas-
much as there are no holes along the borders, we may presume that they
were not meant to be attached to leather or another backing. The creatures
are not typical Luristan types and thus add another dimension to the reper-
tory.
No. 2. Plaque. 43.102.12; Surkh Dum 1269; Plot JH R2. Bronze; L. 5 cm.; H. 2.5 cm.
This small, thin rectangular plaque has no holes for attachment to another
object or material and is complete as is. A recumbent, horned animal in low
relief and with no body decoration faces right. Its legs are tucked under its
body, and its hooves touch. The animal is framed by raised dots.
Whatever function plaques like this and No. 1 had at Surkh Dum is as yet
unknown, pending the publication of their original find spots. Recumbent
49 Pope 1941, op. cit. (in note 7) 293, fig. 7; Amiet 1976, op. cit. (in note 1) no. 196: which may
animals with their feet and hooves in the same position are to be seen on
disc-headed pins and other Iranian objects attributed to Luristan, as well as
from Marlik and on objects from Elarn;50 the image is thus well established
within Iran. A plaque published by Godard51 that depicts recumbent birds
may have had a similar use and meaning: votive plaque?
50 Luristan: Amandry, op. cit. (in note 34) plate XXVI, 3; Clercq-Fobe, op. cit. (in note 21)
nos. 27, 40, 41. Marlik; Ezat Negahban, Preliminary Report on Marlik (Teheran 1964) pl. VII A;
Elam: Amandry, ibid., pl. XVII, XXVIII; Pierre Amiet, “Appliques iraniennes,” Revue du Louvre
(1977) 64–65, fig. 4.
51 Godard 1962, op. cit. (in note 21) fig. 35.
surkh dum at the metropolitan museum of art 405
52 Deity: Dussaud, op. cit. (in note 23) 200; Ghirshman 1956, op. cit. (in note 21) 196; Moorey
1971, op. cit. (in note 6) 214–215. Deities and humans: Godard 1962, op. cit. (in note 21) 63–64;
Clercq-Fobe, op. cit. (in note 21) 22, 40.
406 chapter twelve
concerning whether all the heads are those of females, or some are females,
others males.
Aside from the pins from Surkh Dum, a large number with a variety
of motifs, many formally matched among the excavated examples, have
surfaced on the antiquity market since the first time Luristan bronzes began
to appear about 1930 (see note 31). Parallels for the pin under discussion,
where the human face occupies the whole disc, may be found in addition to
those from Surkh Dum itself among several stray examples.53 On the basis of
the little information published by Van Loon about the site of Surkh Dum, it
appears that the disc-headed pins do not pre-date the late 8th century and
may actually continue into the 7th century bc.
53 Ghirshman 1956, op. cit. (in note 21) pl. XXIV; Godard 1962, op. cit. (in note 21) figs. 68–71,
73, 74; Clercq-Fobe, op. cit. (in note 21) nos. 29–32.
54 A.U. Pope, “Prehistoric Bronzes of a Hitherto Unknown Type …,” ILN (May 6, 1939) 790,
fig. 7; Godard 1962, op. cit. (in note 21) figs. 34, 36; Clercq-Fobe, op. cit. (in note 21) no. 50; Peirre
Amiet, “Les Bronzes du Luristan,” Revue du Louvre (1963) 16, fig. 8; idem, “Notes d’Archéologie
iranienne,” Revue du Louvre (1969) 328, fig. 5. Because of the extraordinary variety of forms and
motives, and different workshops involved, it is always a difficult and complex problem to
separate the genuine from the forgeries among the many unexcavated works of art attributed
to Iran. Thus, in Muscarella 1977, op. cit. (in note 37) 173, no. 25 I listed as suspicious a disc-
headed pin in Geneva depicting a kneeling demon holding snakes. After restudying this pin
in context with others depicting kneeling figures I must retain my doubts: note that the
surkh dum at the metropolitan museum of art 407
Geneva disc is thick and may not be a pin head (see Bernard Goldman, “The Asiatic Ancestry
of the Greek Gorgon,” Berytus XIV [1961) 1–9, plate I, 1, who thinks it is a mirror). Note the
sharp outlines of the demon’s body and the sharp demarcation line of the projecting right
leg separating it from the body, and note the drawing of the feet, features not present on the
Surkh Dum figure, nor on other examples. The head of the demon also bothers me, especially
when seen from the rear. I also do not cite a disc-headed pin published in Mostra d’Arte
Iranica (Rome 1956) plate XVI, left (Muscarella 1977, ibid., no. 20). Here I am not convinced by
the drawing of the demon’s hair, eyes, face in general, and the presence of the pubic triangle,
as well as that of the two animals; it is at least suspicious.
55 Clercq-Fobe, op. cit. (in note 21) 31, 120. Goldman, op. cit. (in note 54) 1, 3, 5, 6–9, sees
No. 5. Open-work cast pin. 43.102.1; Surkh Dum 1573; Plot J1 pr 178/45.59 Bronze;
L. 14.3cm.
Cast in one piece, this open-work pin depicts a squatting female en face,
her legs spread and touching the frame. Small pellet breasts and exposed
pudenda identify the figure as a female. The face and head are corroded but
one may see facial features: small eyes, lips, as well as the horns that identify
her as a deity. Earrings or spiral hair locks exist on either side of the face, and
a grooved area above the face may represent hair. The female is nude but
there is a grooved rectangular area above the pudenda that may represent
a girdle. Held at bay by her thin, unnaturally curved arms are two horned
animals—antelopes/goats—stylistically rendered only by their heads and
long, thin necks that join in a continuous curve, and which enclose the
deity in a frame; unidentifiable curved units connect the animal heads to
the deity. A non-descript thin unit joins the frame to the pudenda, but it is
not clear whether it is a strut or had a more significant meaning, namely
representing the process of birth.
Open, cast pins depicting either a mistress or master of animals are one of
the most characteristic forms among the Luristan bronzes. The iconography
occurs on many pins but is not limited to them, for, among other items, it is
who is a Great Mother-Lamashtu-Gorgon figure, an interpretation also held for the other
kneeling figures on the pins. There is indeed a formal parallel between the Geneva demon
and the Larnashtu depictions, but if the former is not ancient, the idea collapses. Moreover,
the conclusions regarding the sex and attributions of the kneeling figure are to my mind not
so secure as presented.
56 Amiet 1976, op. cit. (in note 1) no. 186.
57 Cf. a pin formerly in the Bach collection, Bronzes de la Perse (Paris, Hotel Drouot
12/12/1973) no. 28, and Godard 1962, op. cit. (in note 21) fig. 81.
58 Moorey 1971, op. cit. (in note 6) 208; Clercq-Fobe, op. cit. (in note 21) 223, no. 26.
59 Listed in Ackerman 1940, op. cit. (in note 4) 543, G. She called the unit joining the
represented standing, but the horns and spiral hair curls, and sometimes the
presence of the grooved girdle, identify her as the same or a related deity, or
essence, as the squatting female;62 and sometimes the female has no legs rep-
resented, or merely a head.63 The variety is large and pins of the type under
discussion are but one type of a large group reported from Luristan that is
clearly related iconographically: defined by the primary motif, a mistress or
master of animals framed either in a crescent, or within a square or circular
unit.
The squatting female motif is not limited to open-work pins, for at least
two examples are known depicted in relief on disc-headed pins, in one
case where the female is actually giving birth.64 The squatting position in
general, a birthing position for women, as well as the specific birth scene,
surely supports the interpretation that these pins, open work or in relief,
had a votive value associated with fertility rather than female sexuality per
se. And probably the standing females, including those with animal heads,
likewise were involved with fertility matters. As for the standing males,
linked with the females iconographically, they, no doubt, had a different
charged function.
Moorey65 has suggested that the open-work pins, some of which are large
and heavy with frames, may have served as icons rather than as garment
pins. He has rightly noted their stylistic and iconographical relationship
with the finials, which were probably icons. Perhaps he is right in essentials,
especially given their presence in the sanctuary of Surkh Dum, but it need
not follow that they were not also worn on the body as a charm or protective
amulet, while also functioning as clothing fasteners.
note 31) pl. V, 14; Moorey 1971, op. cit. (in note 6) no. 348, a variant with a square frame; Amiet
1976, op. cit. (in note 1) nos. 178, 180, the former with a mouflon’s head over the female, and
exactly matched by a pin in the Khoramabad Museum, Mina Sadegh-Behnam, Anita Koh,
Lorestan Bronzes and Islamic Metalwork (N.D.) no. 4.
62 L. Speleers, “Nos Nouveaux Bronzes perses,” Bulletin des Musées Royaux d’Art et d’His-
toire (1932) 102, fig. 27. Other bronzes in the Brussels Museum’s collections are published in
the same journal and under the same title for the years 1932 and 1933: (1932) 56–71, 93–104,
115–119; (1933) 62–69, 86–95. Cf. similar pieces in Moorey 1971, op. cit. (in note 6) no. 346, and
J.A.H. Potratz, Luristanbronzen (Istanbul 1968) no. 101.
63 Amiet 1976, op. cit. (in note 1) no. 178; Moorey 1971, op. cit. (in note 6) no. 347; Godard
1931, op. cit. (in note 31) pl. XXXV, 150; H. Potratz, “Das ‘Kampfmotif’ in der Luristankunst,”
Orientalia 21:1 (1952) 26–28, figs. 21–38, 41, illustrates a number of these pins of different types
and considers them to represent a Luristan moon goddess.
64 Amiet 1976, op. cit. (in note 1) no. 189; Godard 1962, op. cit. (in note 21) figs. 77, 78: the
former pin was once in the Godard collection, not David-Weill as thought by Moorey 1971, op.
cit. (in note 6) 204.
65 Moorey 1971, op. cit. (in note 6) 200; idem, 1974, op. cit. (in note 6) 124.
surkh dum at the metropolitan museum of art 411
No. 6. Pin. 43.102.7; Surkh Dum 1539; Plot JI pr 178/11.66 Bronze, iron; P.H.
3.8cm.; W. 7cm.
A central motif is flanked by the heads and long necks of two stylized
Luristan type felines. The necks are joined and continuous and half-way
enclose the central unit in a frame, similar in form to some open-work pins
(see No. 5). Seen from the front, it is difficult to recognize what the central
motif represents; seen from the side, it is easily recognized as a duck’s head
turned back to recline on its wings (cf. the bracelet from Bard-i-Bal).67 The
shank of the pin, now missing, was made separately of iron, judging from
the color at the join. A number of examples of this pin type, some framed
with feline, others with antelope, heads, occur at Surkh Dum.
I know of only two published parallels to our pin.68 Surkh Dum also
yielded many examples of straight pins with the very same central motif as
on ours but lacking the flanking heads and necks; other examples exist in
various collections.69 These pins are now confirmed as being from Luristan,
and the occurrence of iron surely attests to a date not earlier than the late
9th century bc, and probably later.
66 Listed in Ackerman 1940, op. cit. (in note 4) 544, P, and incorrectly described.
67 Louis vanden Berghe, “La Necropole de Bard-i Bal,” Archeologia 43 (1971) 21, fig. 15.
68 Speleers 1932, op. cit. (in note 62) 102, fig. 28, and Hotel Drouot Catalogue, 5/22/80,
no. 273 bis; cf. L. Speleers, “Antiquités iraniennes,” Bulletin des Musées Royaux d’Art et d’His-
toire (1938) 42, fig. 16: is this a pastiche?
69 Godard 1931, op. cit. (in note 31) pls. XXXIII, LVI, 119, 125, 205; F. Basmachi, “The Luristan
Bronze Objects in the Iraq Museum,” Sumer XIX (1963) pl. 6; Moorey 1971, op. cit. (in note 6)
no. 317.
412 chapter twelve
The solid cast head of this pin is rendered in a stylized manner, suggesting
as an impression that a large head alone is depicted, when in fact a whole
figure is sculpted. The head clearly is meant to predominate and apparently
represents a dominant figure. This interpretation is suggested by the promi-
nent, sharp nose, thick brows encircling the eyes, and either a large lantern
jaw or a beard clearly offset from the mouth area; no ears are depicted. A
sloping, flat beret-like cap does not completely cover the hair, for a band of
wavy locks or curls runs across the forehead. The head joins a tubular sec-
tion that functions as both neck and body, albeit that arms are not depicted.
At the base of the “body” are thighs and legs in a squatting position, and
the feet grasp the top of the shank on either side. The impression is that the
figure sits on a pinnacle, holding on by the feet. Breasts are not indicated,
but between the knees is a raised oval area with a central depression, which
suggests that it is a vulva. If, however, the figure has a beard and not a long
jaw, then we have something else here. Note that laboratory analysis has
determined that the figure is made of copper, not bronze.
I can find no parallels for this figure, for its position, or face and hat.
It is further unique in that it is at present the only published example of
a figure in the round to have been excavated in Luristan. The literature is
filled with examples of stray bronze (copper?) figurines claimed to derive
from Luristan, but they are all standing figures and none has the armless
neck-body arrangement as ours. This latter feature, however, does seem to
exist on the demon figure on the many finials reported from Luristan, and
in one small but significant detail they present another parallel to our piece:
on a few finial examples it seems—one is not certain—that the neck-body
has feet that grip the base below the neck section.71 If this observation is
correct then we have another formal connection between our figurine and
the typical Luristan finial demons.
No. 8. Human-headed pin. 43.102.17; Surkh Dum 44; Plot J1 N.E. Bronze; H.
3.5cm.
71 Viz. Potratz 1968, op. cit. (in note 62) pl. XXXVI, XXXVII, nos. 228–231 (note that on
nos. 232, 234, 235, 238, 242, the visible feet may belong to the heraldic creatures).
414 chapter twelve
That this is a pin is indicated both by the presence of a shank, now mostly
missing, and because there is another, more complete example from Surkh
Dum (SOR 201) now in the University Museum, Philadelphia. Our example
is topped by a small, beardless human head with a prominent nose and thick
lips, centered on a curved unit that projects on either side and which may
represent either arms, otherwise not represented, or more probably, wings.
On the Philadelphia example the wings are longer and thinner, and curve
up in a pronounced manner at the tips. At Surkh Dum there are examples
of pins exactly like ours in form except that instead of the head at the center
there is a short, raised scalloped unit.
Outside of Surkh Dum no other examples of this pin type with a human
head are known. On the other hand, an example with the raised central
scalloped unit, exists in Brussels.72 And clearly related examples, with a plain
small central unit, exist in the Ashmolean Museum and among the bronzes
claimed—without verification—for Khurvin;73 another related type, one
with a swelling at the central part of the wings, is also claimed for Khurvin.74
A more developed type, perhaps also related to our pin in concept, depicts
the torso of a male centered on the wings, or a male torso centered on ram’s
horns.75
If the projections on our pin, and the others, are indeed wings, one would
be right to assume that we have a representation of a deity, unnamed to be
sure, but appropriate for dedication at a sanctuary.
No. 9. Animal- and demon-headed pin. 43.102.3; Surkh Dum 209; Plot JI pr
2597.90.76 Bronze; P.H. 3.5cm.
The head of this pin consists of a stylized head of a demon or deity sur-
mounted by a recumbent animal; it is meant to be viewed from the front
only. The face is formed by pellet eyes encircled by a thick line that is the
brows and that continues to create the nose; no ears or mouth are depicted.
Two loops placed on top of the head may represent hair curls and curved
72 Speleers 1933, op. cit. (in note 62) 89, fig. 26.
73 Moorey 1971, op. cit. (in note 6) no. 289; Louis vanden Berghe, La Nécropole de Khurvin
(Istanbul 1964) pl. XLIII, no. 316. Note that many of the objects published as from Khurvin
derive from a Teheran private collection.
74 VandenBerghe, ibid., no. 314; see also Godard 1931, op. cit. (in note 31) pl. XXXIII,
129; cf. A.U. Pope, ed., A Survey of Persian Art (Tokyo 1964–1965) pl. 59 D; Wolfram Nagel,
Altorientalisches Kunsthandwerk (Berlin 1963) pl. VIII, 18.
75 Ernst Herzfeld, Iran in the Ancient East (London and New York 1941) 155, fig. 275, center
units encircling the face may be horns. An animal, whose head is now
missing, rests on the head of the demon, its feet touching the sides. Under
the grooved moulding is a hole that once held a separately made pin; it is no
longer possible to know if it was bronze or iron.
To my knowledge, no parallels for this specific pin exist, but there are
at least two examples that in form closely relate to ours. These pins have
what Porada has called “two profile lion heads (which also combine to form
a single frontal mask).”77 On both these pins the masks, if that is what they
77 Porada 1979, op. cit. (in note 44) 142–143, note 17, fig. 9; A.U. Pope, “Mute, Yet Eloquent:
The Significant Luristan Bronzes …,” ILN (September 13, 1930) color plate, lower left.
416 chapter twelve
actually are, are surmounted by a mouflon with large, majestic horns. While
the legs on the Surkh Dum example are placed on either side of the central
mask, however, on the other pins they rest directly on top. It may be that our
animal also originally had a head with the same majestic horns.
An exact parallel for the stylized mask with top loops on our pin exists
in relief on a disc pin in Brussels, and in the round on a handle (?) in the
Erlenmeyer collection.78
No. 10. Animal-headed pin. 43.102.8; Surkh Dum 1078.79 Bronze; P.H. 5.7 cm.
Cast in the round, the head is in the form of a standing goat whose feet are
drawn together to rest on a plinth. Below the plinth is a tubular moulding
into which a separately made pin, of unknown material, was originally
inserted.
78 Clercq-Fobe, op. cit. (in note 21) 208, no. 19, pl. 19; M.L. and H. Erlenmeyer, “Frühi-
ranische Stempelsiegel, II,” Iranica Antiqua 5 (1965) 8, pl. V, 24, pl. XI; cf. similar heads in
J.A.H. Potratz, “Die Luristanbronzen des Museums für Kunst und Gewerbe in Hamburg,”
ZAssyr 17 (1955) pl. I, 4.
79 Listed in Ackerman 1940, op. cit. (in note 4) 545, W.
surkh dum at the metropolitan museum of art 417
Goats and mouflons in the round were very commonly depicted on the
art of Luristan; on finials, horse harnesses, pins, and so forth. With regard to
goats on pins, there is a group that often represents the animal either stand-
ing or recumbent with the head turned toward the viewer, as on horse cheek-
pieces;80 our example is distinct in that the goat faces forward. Characteristic
of most examples is the position of the feet, which are drawn together as if
balanced on a point, a mountain peak.81
Pins of this type may have been icons rather than, or in addition to, being
used as clothing fasteners and charms.
No. 11, Frog-headed pin. 43.102.4; Surkh Dum 400; Plot II pr 24.82 Bronze; L.
4.8cm.; W. 2.5cm.
The strange creature seems to be a frog depicted in the round and as seen
from the top. Its eyes bulge, and all four legs project in the same direction;
the body is simply rendered except for the back ridge that connects the legs.
The frog is clearly the head of a pin, the shank of which is cast with it and
part of which is still extant.
A similar pin, complete, ending in a frog’s body was at one time in the
David-Weill collection;83 it was made by a different hand than the one that
made ours and has a loop at the shank’s base to hold a chain or cord to
facilitate securing the pin to a garment. In addition, the Boston Museum
of Fine Arts has an amulet in the form of a frog that was purchased in 1930;84
a frog amulet also is in Baghdad.85 Although apparently rare, the frog is thus
attested as a decorative element for pins and amulets in Luristan. Aside from
our Surkh Dum example one or two more were found at Surkh Dum; no
others, either pins or amulets, have been excavated.
No. 12. Animal-headed pin. 43.102.20; Surkh Dum 1432; Plot J1 R3 pr 175/63.
Bronze; L. 6.6cm.
The head is formed with two individual units: a rounded band, decorated
in low relief with studs or cast granulations framed by beaded mouldings,
which is joined to the head of an animal, whose head, horns, and ears
are rendered naturally in the same plane as the shank. Under the head is
a groove running its whole length and with its side arms grooved at the
nose end. This feature, seen in side view, could indicate the animal’s legs
and the swelling behind the eyes could be its shoulders. Thus we have the
forepart of an animal, not just its head; this feature distinguishes this pin
from the antelope-headed pins, Nos. 14–17. From a formal point of view,
one thinks of course of the zoomorphic straight-headed pin from Baba
Jan,86 where the whole body of a typical Luristan feline creature forms the
head.
No. 13. Duck-headed pin. 43.102.19; Surkh Dum 423; Plot II pr 24–25 98.00.87
Bronze; L. 20.6cm.
Cast in one piece, the head is in the form of a reclining duck that is separated
from the shank section by a series of grooves. In addition to this pin and
others from Surkh Dum, the only other site that has yielded this type pin is
Kutal-i-Gulgal, also in Luristan.88 A large number of stray examples exist, all
exactly the same as those from Surkh Dum and Kutal-i-Gulgal,89 and which
now have a confirmed Luristan provenience.
The presence of the same type of object both at Surkh Dum and at an
excavated cemetery site in Iran is archaeologically significant. First of all,
aside from gaining knowledge about distribution, it demonstrates that, as
in the present case with regard to pins, the object could have a votive and
86 Clare Goff Meade. “Luristan in the First Half of the First Millennium B.C.,” Iran VI (1968)
Basmachi, op. cit. (in note 69) pl. 5; Pope 1964–1965, op. cit. (in note 74) pl. 60 K; Potratz
1968, op. cit. (in note 62) 36, note 4, pl. XXIV, 140; Moorey 1971, op. cit. (in note 6) 193–194,
nos. 314–315; Anton Moortgat, Bronzegerät aus Luristan (Berlin 1932) pl. VII, 19; W.D. van
Wijngaarden, “De Loeristanbronzen in het Rijksmuseum van Oudheden,” Oudheidkundige
Mededelingen XXXV (1954) pl. XlI, 75, 76.
420 chapter twelve
perhaps a secular function. And second, it serves as a warning that all stray
objects of a type usually related to Surkh Dum may not in fact have derived
from there; some of these pins cited above have been known since 1930.
The date of the Kutal-i-Gulgal tomb containing the pins was originally
dated by Vanden Berghe to a time around 1100–1000bc, but later the date
was modified and lowered by a century, to ca. 1000–900bc: this is still earlier
by a century or more than the time it is suggested the Surkh Dum sanctuary
flourished (supra), and it creates a paradox. Either the Kultal-i-Gulgal tomb
is even later than suggested (see also No. 19), or the Surkh Dum sanctuary
begins earlier than perceived, or, a third possibility, the pins at the latter site
are heirlooms or long lived. In any event, all that can be stated at present
is that there seems to be a considerable difference in time between the
occurrence of the pins at two excavated sites.
Four other animal-headed pins (Nos. 14–17) are exactly the same in all
details, differing only in size and horn positions (indicating that they were
made by the lost-wax process). All terminate with the head of a horned
animal—an antelope? The horns are free from the head and pass between
the upright ears. The pin shanks vary in length and thickness; No. 17 is bent
and No. 16 is broken.
A very large number of antelope-headed pins were excavated at Surkh
Dum and an equally large number of strays have been recorded from the
time of the earliest appearance of Luristan bronzes on the antiquities mar-
ket.90 The former examples thus neatly confirm a Luristan provenience for
the class. Note that No. 17 is specifically mentioned as having been found in
a wall.
90 Moorey 1971, op. cit. (in note 6) 193 for references, nos. 312, 313; see also Godard 1931,
op. cit. (in note 31) pl. XXXIII, 123, 132, 133; Herzfeld 1941, op. cit. (in note 75) 155, fig. 275;
Wijngaarden, op. cit. (in note 89) pl. XII, 77–80; Calmeyer 1964, op. cit. (in note 81) no. 67;
Jean Paul Barbier, Bronzes iraniens (Geneva 1970) no. 41.
surkh dum at the metropolitan museum of art 421
No. 15. 43.102.21; Surkh Dum 279; Plot J1 pr 1997.45. Bronze; L. 13.9 cm.
No. 16. 43.102.22; Surkh Dum 197; Plot II. Bronze; L. 8.3 cm.
No. 17. 43.102.23; Surkh Dum 1585 or 1685; Plot KH, in the wall. Bronze; L.
9.2cm.
No. 18. Pin. 43.102.24; Surkh Dum 578; Plot KI pr 11.91 Bronze; L. 12 cm.
No. 19. Animal-terminal bracelet. 43.102.2; Surkh Dum 1632. Bronze; D. 8.2 cm.
The terminals on this cast penannular bracelet are in the form of the forepart
of a stylized animal of indistinct species (lion?); on each back is a loop,
probably used to hold a cord or chain to keep the bracelet from slipping
off the wrist; the arms are round in section and plain.
One of the most characteristic and numerous of the objects reported
from the plundered tombs of Luristan is the bracelet with zoomorphic
terminals, hundreds of which have surfaced through the antiquities market.
Bracelets of this type are rare from excavations in Luristan, but it is known
now that they exist not only at Surkh Dum (i.e., the present example and
others reported), but also at Bard-i-Bal,94 so the Luristan provenience for
the class in general is documented. Moorey95 has discussed the importance
of these objects, both with respect to their earlier occurrence in Iran at
Hasanlu and Marlik, and equally significant, their flourishing continuity
in Achaemenian times. Concerning the latter occurrence, animal-terminal
bracelets represent one of the most clearly documented examples of earlier
Iranian art forms taken over and developed by the Achaemenians.
94 Vanden Berghe 1971, op. cit. (in note 67) 21, fig. 15; idem 1973, op. cit. (in note 44: Iranica
Aside from the Surkh Dum examples, no other bracelets like our example
with terminals in the form of the whole or the forepart of an animal (as
opposed to the many with only the animal head) have been excavated, but
strays exist.96
No. 20. Lobe-shaped ring. 43.102.13; Surkh Dum 1601.97 Bronze; D. 2.5 cm.
No. 21, Lobe-shaped ring. 43.102.14; Surkh Dum 10298 Bronze; D. 2.3 cm.
Ring No. 20 was made from a sheet of bronze wider in front than at the
back where the slightly narrower ends touch. The design, which may have
been cast with the sheet, consists of two different horned animals, one
with a curved horn shown in profile, the other with two horns shown
frontally, flanking a stylized tree; a multi-petalled rosette or star is behind
each animal. In the field is an inscription in cuneiform which reads Dinger.
Mesh, part of a prayer invoking the gods. The second ring is similar in shape
96 Godard 1931, op. cit. (in note 31) pl. XXVIII, 94; Pope 1964–1965, op. cit. (in note 74)
pl. 57 C; Ghirshman 1964, op. cit. (in note 21) fig. 94; Louis vanden Berghe, et al., Bronzes
Iran-Luristan Caucasus (Paris 1973) pl. XXXVI, center.
97 Listed in Ackerman 1940, op. cit. (in note 4) 547, QQ.
98 Listed in ibid. 547, RR.
424 chapter twelve
to the first, although cast closed and with a more pronounced lobed front.
The design, while neat, is executed in eneven lines and seems to have been
added after casting. A bird facing right is placed above a horned animal
striding left with its head down: on the sealing the positions are reversed.
Along the outside borders is an incised line. Both rings may have been used
as seals.
Edith Porada99 has made a study of these rings, seeing them as stylistic
indicators for establishing a chronology for certain Luristan bronzes. She has
categorized rings of our first type as “sheet” rings, those of the second type as
“lobed” rings. She has also perceived a chronological distinction between the
two types, exhibited by the fact that the engravings on the sheet examples
are usually carefully rendered and often show heraldic animals flanking a
stylized tree. To Porada, the scenes on the lobed rings are usually cruder in
execution and were added after casting. While not explicit, it seems that
she considers those rings cast with the arms closed to be lobed, and the
penannular ones to be the sheet rings.
Inasmuch, however, as some “lobed” rings have penannular ends and
finely executed scenes, and because the “sheet” rings are all lobed in shape,
the division is not so clear as assumed. It may be that the type and style of
the scenes themselves should be the criteria for the division among the lobe-
shaped seals: the neatly rendered heraldic animals and tree, all of which
seem to have penannular ends, and the others, some crude others neatly
rendered, with animals or demons not in a heraldic position and usually in
a different style than the first grouping.
The scenes on the “sheet” rings have been compared by Porada to Elamite
and Babylonian art and dated accordingly to 1200–1000bc; the “lobed” rings
are dated ca. 1000–800bc.100 Van Loon, while noting that the relative se-
quence suggested by Porada, that sheet rings precede lobed rings, is sup-
ported by the Surkh Dum stratigraphy, has nevertheless claimed that the
former occur there in the 8th century level101 (= 8th–7th centuries bc), the
latter in the 7th century bc level. Given the fact that early seals occur at
Surkh Dum, it is not impossible that Porada’s date of the sheet rings to the
99 Porada 1964, op. cit. (in note 28) 16–19; idem 1965, op. cit. (in note 28) 75–78.
100 In Porada 1964, ibid. 16 the sheet rings are dated ±1100 bc; later, 28, “about the twelfth
or eleventh century bc;” the lobed ones are dated on page 17; see 1965, ibid. 76, figs. 47, 48, 78.
Cf. Erlenmeyer, op. cit. (in note 78) 2–5 for a different arrangement and dating, but based on
gratuitous comparisons and to my mind not convincing.
101 Van Loon 1967, op. cit. (in note 6) 24.
surkh dum at the metropolitan museum of art 425
late 2nd millennium bc is correct: yet one has to accept a 200–400 year
difference between the two ring types, which in shape at least are not so
different. How many lobe-shaped rings were excavated at Surkh Dum is still
not revealed, but it would be of value to know if in fact sheet rings occur only
in one level, lobed rings only in another.
Thus, the issue of dating remains to my mind still unresolved, especially
with regard to the so-called sheet examples: we have a situation where
style suggests an early dating and stratigraphy a later one. Porada has also
noted that penannular rings of sheet metal and with lobed faces, some of
gold, occur in other areas of the Near East in the late 2nd millennium bc;
very distinctly lobed rings of iron and bronze have also been excavated at
Hasanlu of 9th century bc date.102
No. 22. Pendant. 43.102.15; Surkh Dum 617. Bronze; L. 2.9 cm.; H. 2.2 cm.
This small cast pendant seems to depict a dog. Its raised tail curves up and
forward above a flat rear end; eyes are small raised pellets and the ears are
small; a suspension loop connects the neck and back.
The number and variety of pendants at Surkh Dum (see also No. 23)
is still not known but few are known elsewhere from excavations. Aside
from a bird pendant from Bard-i-Bal (see No. 23), pendants in the form of
102 Porada 1964, op. cit. (in note 28), 16; for Hasanlu see Aurel Stein, Old Routes of Western
Iran (London 1940) 398, pl. XXV, 2; others, unpublished, are known from the recent exca-
vations. Cf. similar, but less pronounced, lobed rings from neighboring Dinkha Tepe, Oscar
White Muscarella, “The Iron Age at Dinkha Tepe, Iran,” MMJ 9 (1974) figs. 43, 52, nos. 133, 342,
620.
426 chapter twelve
No. 23. Pendant. 43.102.16; Surkh Dum 1013. Bronze; L. 0.6 cm.; H. 1.9 cm.
103 A. Hakemi and M. Rad, The Description and Results of the Scientific Excavations at
202; W. Heimpel, “Hund,” Reallexikon der Assyriologie IV (1922–1975) 494–497; Betty Schloss-
man in Ladders to Heaven, Oscar White Muscarella, ed. (Toronto 1981) 114–116; Daphne
Achilles. in ibid. 201. For recent discussions see I. Fuhr in B. Hrouda, Isin-Išan Bahrı̄yāt I
(Munich 1977) 135–145.
105 Speleers 1932, op. cit. (in note 62) 115, fig. 10; Pope 1964–1965, op. cit. (in note 74) pl. 59 J;
center; Moorey 1971, op. cit. (in note 6) 231–232, nos. 416–418.
surkh dum at the metropolitan museum of art 427
No. 24. Miniature axe. 43.102.9; Surkh Dum 1500.108 Bronze; L. 4.8 cm.
In miniature size, this axe duplicates full-sized examples known both from
excavations and the antiquities market. Characteristic of this particular type
is both the cut away, slanted lower part of the socket, and the flange-butt
with a horizontal ridge; the socket and the flange have a thick outline. Exam-
ples of this type have different blade shapes that define them as chisels,
picks, or axes, but the slanting socket and flange interrelate the group as
belonging to the same class or type. Our example is an axe, the upper edge
horizontal, the lower curving up to the socket.
The type was studied by Maxwell-Hyslop, Deshayes, and Calmeyer,109 all
of whom isolate the characteristics as well as give evidence for geographi-
cal distribution in the Near East. Full-sized examples have been excavated
at Til Barsip in north Syria, and in Iran at Susa and Kalleh Nisar in Luris-
tan.110 Other examples have been attributed to Nimrud and Tepe Giyan, but
without verification; and Calmeyer mentions two from Mari.111 Counting the
108 Listed in Ackerman 1940, op. cit. (in note 4) 547, XX.
109 K.R. Maxwell-Hyslop, “Western Asiatic Shaft-Hole Axes,” Iraq XI (1949) 99–100, Type 9;
Jean Deshayes, Les Outils de Bronze de l’Indus au Danube, I, II (Paris 1960) I, 166, II, 70, Type
A5c; Ca1meyer 1969, op. cit. (in note 1) 32–34.
110 Calmeyer 1969, op. cit. (in note I) 33, fig. 32; Amiet 1976, op. cit. (in note 1) 9, fig. 5; Louis
note 1) 34–35.
428 chapter twelve
present example, three axes of our type have thus been excavated in Iran.
All these examples are dated to the last centuries of the 3rd millennium bc.
There is no evidence at Surkh Dum that the site pre-dates the 1st millen-
nium bc, so that the presence of a 3rd millennium axe type is an anomaly.
One might assume that it is an heirloom, or a stray found by locals and dedi-
cated at the sanctuary, or an indication that some examples of the type were
still being made at a later date; we do not know which possibility obtained.
Miniature weapons, daggers and axes, while not rare in the ancient Near
East are not common either.112 The occurrence of a miniature axe at a sanc-
tuary indicates that it was dedicated as a model of a functioning weapon.
Note that another axe of a different type (see note 6) was also excavated at
Surkh Dum.
In fragments when found, parts of all four edges of this rectangular plaque
are preserved. On side A there is no defined upper border while the lower is
a band of vertical lines; on side B both upper and lower borders consist of
heavy four-petalled rosettes. On both sides the right and left borders have
an irregular but neat guilloche pattern, the centers and curves of which are
drawn with a compass. The upper edge has two drilled holes, the lower has
three.
112 Hans Bonnet, Die Waffen der Völker des Alien Orients (Leipzig 1926) 71, fig. 9, from a
temple at Assur; Maxwell-Hyslop, op. cit. (in note 109) 119, 120; E.A. Speiser, Excavations at
Tepe Gawra (Philadelphia 1935) pl. XLIX, 3; Nagel, op. cit. (in note 74) nos. 41, 42, 100; Vanden
Berghe 1964, op. cit. (in note 73) pis. XLIV, XLV, 332, 335–337; Pierre Amiet, “Bactraine Proto-
historique,” Syria LIV (1977) 106–107, fig. 14; Calmeyer 1969, op. cit. (in note 1) 27, D is over
twice the size as ours and may not be a true miniature.
surkh dum at the metropolitan museum of art 429
Side A is decorated with two identical bull men (the horns are clearly
part of their heads), en face and touching hands standing side by side; each
holds a creature at bay with his outside hand. The bull men have a triangular
face with a rectangular nose, slit mouth, thick round eyes that look compass
drawn with a central dot, and a beard that frames the whole face; vertical
lines between the horns may be hair and no ears are depicted. They are
dressed in a belted, calflength one-piece (?), short-sleeved jumper that is
fringed throughout its length and at the lower border; a vertical band runs
from the neck to the belt. Feet are visible below the skirt and seem to point
to right and left. The creature on the viewer’s right is an upright lion whose
body is outlined in bordered lines, as are the bull men’s dress. The creature
to the left is larger, but because of a break is not readily identified. He seems
to rest on his haunch and his feet and paws, one with claws pointed up, the
other down, touch the bull men; this creature may be a bear, not a lion. The
decoration, except for the guilloche and eyes of the bull men, is rendered in
a crudely incised manner, especially the feet and hands of the bull men and
the bodies of the creatures.
Side B is equally rendered in a crude fashion. Here a lion at the left attacks
a horned animal fleeing to the right. One paw of the lion touches its prey,
both of whose front feet are off the ground, perhaps to show that it is falling.
The body outlines are like those on side A.
430 chapter twelve
The scenes on both sides are complete and the plaque is an individual
unit. Inasmuch as both sides are decorated, both sides were obviously meant
to be viewed, which makes it difficult to understand how the plaque was
used. That it was set into some type of frame is indicated by the upper and
lower holes.
The en face figures with triangular outlined face are of course the same as
that of the figure on the faience concave vessel also from Surkh Dum (No. 31).
Equally matched on these two objects from the same site is the crudeness
of the incised rendering of the decoration. And while no exact parallels
to match the two juxtaposed figures functioning as masters of animals are
available, there can be no hesitation in recognizing the Luristan style of the
figures. Bull men (or figures with bull horns) are common there, as is the
en face position and the flaring skirt on other figures.113 The plaque, then, is
clearly a product of a Luristan atelier (at least the incised scene, if not the
original cutting and carving of the guilloche: cf. No. 31), and thus documents
ivory working in Luristan in the early centuries of the 1st millennium bc.
No. 26. Animal-headed pin. 43.102.26; Surkh Dum 1254. Bone; L. 5.5 cm.
113 Godard 1962, op. cit. (in note 21) figs. 34–36, 38, pls. 16, 18, 21, 23; Moorey 1975, op. cit. (in
note 21) 21, 24, figs. 1, 5, pls. I, IIb, IVd, the latter with an outlined face; Clercq-Fobe, op. cit. (in
note 21) nos. 50, 56.
surkh dum at the metropolitan museum of art 431
114 Van Loon 1967, op. cit. (in note 6); Schmidt, op. cit. (in note 1) 211 for reference to bone
“wands.”
115 Moorey 1971, op. cit. (in note 6) 196–197, nos. 324–327; Porada 1965, op. cit. (in note 28)
166, fig. 67, pl. 29; idem, 1975, op. cit. (in note 24) 393–394, no. 310.
432 chapter twelve
No. 27. Handle (7).43.102.27; Surkh Dum 824. Ivory (7); L. 7.6 cm.
This badly corroded and pitted object is pierced at one end to a depth of
about 1.7cm. One can barely make out the incised decoration which consists
of rows of lines framing bands of zigzag patterns at each end; a clear space
of about 1.3cm. separates the decoration.
The hole at one end suggests that the object was a handle. What it held
may only be conjectured, but it was probably a perishable material, since
nothing was found in the hole.
Nos. 28–30. Lion figurines. 43.102.28, 29, 30; Surkh Dum 666d. Ivory; L. 2.5, 2.5,
2.9cm.
surkh dum at the metropolitan museum of art 433
Although No. 30 is slightly longer than the others, all are the same in all
details. Each figurine represents a recumbent lion with its head resting on
its paws. The thighs, head, and ears are in relief while the mouth is formed
by two grooves. All have flat bases pierced with two holes; No. 30 has the
holes piercing the whole figure. Few distinct stylistic features are present.
At Nirnrud and Hasanlu116 small recumbent ivory calves pierced at their
bases were used as handles or grips on the lids of ivory pyxides and it is
probable that our three lions had the same function.
No. 31. Faience vessel(s). 43.102.45a, b, c; Surkh Dum 14. Faience; greatest H. of
preserved area: 12.6 cm.; D. of rim: ca. 10.6cm.
116 Max Mallowan, Nimrud and its Remains I (London 1966) 219–220, figs. 173, 174; Oscar
White Muscarella, The Catalogue of Ivories from Hasanlu, Iran (Philadelphia 1980) 195–196,
nos. 242–245.
434 chapter twelve
Only part of this vessel is preserved, including about one-half of the rim
and a section both below and to the viewer’s left of the protome. The upper,
rim area is decorated with a wide border consisting of a central outlined
band of guilloches—a center enclosed by an S-shaped motif—framed by
two bands of vertical lines. Below is the main decorative scene depicting
a central figure flanked by large birds moving away. The central figure, en
face, has a triangular, outlined face, oval eyes lidded at the bottom only, a
rectangular nose anda slit mouth; ears are unnaturally placed at the top of
the head and frame vertical hair lines. No mustache is indicated but random
incisions on the face suggest a beard. Nothing else remains of this figure,
but he is clearly represented in the master-of-animal position. To his right
is a creature identified as a large bird by his beak and one wing, awkwardly
placed before him. The bird has a crest and feathers at its back rendered
by short incisions. Only the back feathers remain of the bird at left. All these
figures are incised in a crude and cursory manner compared to the relatively
neat rim decoration and protome.
The protome projects from the decorated area of the vessel and consists of
a bearded male wearing a bulbous hat decorated with vertical lines and two
small upright horns at the front center. His face is human with a distinctly
surkh dum at the metropolitan museum of art 435
large nose in a straight line from the forehead, a thick-lipped mouth, human
ears, and eyes that appear to be closed, the lids meeting at the center; if
there is meant to be a mustache it is not evident. Two small legs ending in
cloven hooves join just under the beard. The figure is clearly a demon, a bull
man. Below the protome is an incised curved border (?) and another incised
object that may be the wings of a bird, now missing. In a few areas one may
still see the remains of a glaze.
The curvature of the vessel as restored may be too flared as there is very
little evidence for a flare on the preserved part; the restoration is apparently
based on comparative vessels.
Two glazed fragments, one incised with the winged area of a bird or crea-
ture (b), the other a rim fragment (c), came with the other more complete
fragments (a). While (b) might belong to the vessel—its incisions are not
noticably different in execution, (c) clearly does not. Although it has the
same rim decoration pattern and band sizes, the guilloche is not the same,
neither in the size of the central circle nor in the S-curve pattern: we there-
fore must consider that fragment (c) at least represents another similar
shaped vessel.
436 chapter twelve
117 Amiet 1976, op. cit. (in note 1) 60, fig. 39 (in Teheran).
118 Pierre Arniet, Elam (Auvers-sur-Oise 1966) 500–501, no. 376; H. 20.5cm.
119 Vanden Berghe 1973, op. cit. (in note 44: Archeologia) 28 and color plate.
120 Amiet 1966, op. cit. (in note 118) 498–499, no. 375; Porada 1965, op. cit. (in note 28) 72,
fig. 46.
surkh dum at the metropolitan museum of art 437
121 Amiet 1966, ibid., 495, no. 372; Porada 1965, ibid., 72, fig. 45.
122 Porada 1965, ibid., 70.
123 Pierre Amiet, “Éléments Émaillés du Décor architectural Néo-Élamite,” Syria XLIV
Cylinder Seals
by Elizabeth Williams-Forte
No. 32. Akkadian cylinder seal. 43.102.34; Surkh Dum 1124. Shell C?; H. 2.8 cm.;
D. 1.45cm.
This seal shows two groups of battling gods. To the left appear two deities
grasping both a mace and the top of the other god’s crown. The second group
of two deities flank a god with arms held down with palms up. The attacking
gods grasp the central deity’s crown while the god to the right smites him
with a mace.
surkh dum at the metropolitan museum of art 439
125 Mesopotamia: R.M. Boehmer, Die Entwicklung der Glyptik während der Akkad-Zeit (Ber-
lin 1965) Abb., 318, 324 (Kish), Abb., 321, 348 (Tell Asmar). Susa: Pierre Amiet, Glypticue
susienne, Memoire de la Delégation archéologique en Iran XLIII (Paris 1972) 188–192, pl. 146,
no. 1550.
126 Amiet 1966, op. cit. (in note 118) no. 157.
127 See Amiet 1972, op. cit. (in note 125) 190.
128 Ibid.; also Pierre Arniet, “Pour une interpretation nouvelle du Répertoire iconogra-
129 Pope, apud Schmidt 1938, op. cit. (in note 1) 208, note 3; Porada 1964, op. cit. (in note 28)
tamia: Henri Frankfort, Stratified Cylinder Seals from the Divala Region (Chicago 1955), esp.
pp. 7–11; temples: ibid. 7, Table 1. For a discussion and rejection of the possibility that the
large number of Jamdat Nasr seals discovered in later layers indicate continued production
of this style, see p. 3.
131 Porada 1970, op. cit. (in note 130) 89–91, no. 107 (late Akkadian to Old Babylonian
period). For the only other cylinders older than the majority of the seals discovered in the
chapels see the Mitannian seals, nos. 110–111, 113.
132 Ibid. 7–105, 127–131.
133 Ibid.; see Roman Ghirshrnari’s quote on p. 4.
134 Pierre Amiet, “Glyptique élamite à propos de nouveaux documents,” Arts Asiatiques
XXVI (1973) 3–65, especially p. 22. Another example of votive seals for which no impressions
are known is cited by Frankfort 1955, op. cit. (in note 130) 16–17.
135 For this Akkadian seal, No. 32, see Amiet 1966, op. cit. (in note 118) no. 157; for the Old
Babylonian seal No. 33 see Amiet 1972, op. cit. (in note 125) no. 1692; and for the Middle
Elamite seal No. 34 see ibid, nos. 2026–2027.
136 For No. 32 see Boehmer 1965, op. cit. (in note 125) no. 347 (Nippur); for No. 33 see
surkh dum at the metropolitan museum of art 441
clearly were not created originally as votive objects. Perhaps the 1st millen-
nium bc inhabitants of Surkh Dum considered these seals suitable temple
offerings because of their great antiquity. The later seals (Nos. 36–40) may
have been fashioned as votive objects since no impressions of seals of sim-
ilar style and iconography are known to me. This implies that they date to
the same period as the Surkh Dum sanctuary (ca. 800–650 bc; see above) for
which they were intended, as was the case at Tchoga Zanbil. Indeed, as will
be discussed below, seals Nos. 36–40 do exhibit characteristics of a seem-
ingly regional, Luristan, origin. Their exact dates, however, are difficult to
determine because of the paucity of comparative archaeological material
from this “dark” age in Iranian cultural history.
No. 33. Old Babylonian cylinder seal.137 43.102.35; Surkh Dum 786. Hematite: H.
2.3cm.; D. 1.28cm.
Louis Delaporte, Catalogue des Cylindres, Cachets, et Pierres gravées de Style oriental I (Paris
1920) pl. 12:7 (Tello); for No. 34 see Edith Porada, Seal Impressions of Nuzi AASOR XXIV (New
Haven 1947) nos. 613–614; and for No. 35 see Anton Moortgat, “Assyrische Glyptik des 13
Jahrhunderts,” ZAssyr 47 (NF 13 1941) Abb. 57, 59–60.
137 Previously published by Vaughn Crawford, et al., The Metropolitan Museum of Art: Guide
to the Collection of Ancient Near Eastern Art (New York 1966) 16, fig. 26.
442 chapter twelve
Inscription
dUTU-KU5.DI Šamaš-dajjān(!?)
ÌR i-la-ni servant of the gods
(uninscribed)
(translated by Dr. John Huehnergard)
No. 34. Middle Elamite cylinder seal.140 43.102.39; Surkh Dum 1317. Serpentine;
H. 2.56cm.; D. 1.15cm.
A deity sits upon an animal-headed throne and holds a rod and ring. Before
him is a worshipper with an animal offering. A star is in the field between
them. To the left is a secondary scene divided into two registers: above, a
lion stalks a horned animal; below, a worshipper stands before a deity who
holds a staff. In the field surrounding them are a bird above a fish, a fly, and
a fox (?).
Worship scenes similar to the present example have been found im-
pressed on tablets at Nuzi in northern Mesopotamia and on less securely
stratified examples at Susa in Iran.141 The typically Elamite characteristics of
these scenes such as the worshipper’s vizor-like hair-style, the deity’s crown
with outward curving horns, and his animal-headed throne have been dis-
cussed by Edith Porada.142 All seals with these stylistic features, including
138
Dr. Huehnergard points out that the empty inscription case is unusual.
A. Spycket, “Un Elément de la Parure féminine a la Ire Dynastie babylonienne,” RAssyr
139
XLII (1948) 89–96, especially pages 93–94; idem, “La Déesee Lama,” RAssyr 54 (1960) 73–84.
140 Previously published by Edith Porada, “The Origin of Winnirkes Cylinder Seal,” JNES 5
(1946) 257–259, fig. 4; idem, 1965, op. cit. (in note 28) 47, fig. 22; idem, 1975, op. cit. (in note 24)
386, no. 297e; Amiet 1973, op. cit. (in note 134) 19, pl. 12. P.
141 Porada 1946, op. cit. (in note 140) 257–259, figs. 1–3; Amiet 1972, op. cit. (in note 125)
nos. 2022–2027.
142 Porada, ibid., 257–259; idem, “Aspects of Elamite Art and Archaeology,” Expedition 13
the present example from Luristan, have been dated to the mid-2nd mil-
lennium bc by Porada on the basis of one of the Nuzi seal impressions
which bears the inscription of Winnirke, the mother of the Mitannian ruler
Tehiptilla (ca. 15th–14th centuries bc).143 Amiet however, in his discussion of
the related Susa examples cites numerous similarities between this group
of seals and late Old Babylonian examples. He suggests that Winnirke’s
seal might have been inscribed after its importation to Nuzi perhaps from
Elam, or, alternately, that seals of this style were popular over several cen-
turies.144
Until securely stratified examples of similar seals are uncovered by
controlled excavations, the exact date of these seals, including the present
“heirloom”, will remain a point of discussion. The appearance of two char-
acteristic features of these scenes on earlier Syrian seals of Middle Bronze
Age date (ca. 1800–1600bc), however, might support the slightly higher date
suggested by Amiet. First, the goddesses with stream-like lower bodies sup-
porting the god’s animal throne on two of the Susa seals are analogous
143 Porada 1946, op. cit. (in note 140) 258; for a date ca. 1400/1300bc for this seal now see
Porada 1975, op. cit. (in note 24) 386, no. 297e.
144 Amiet 1973, op. cit. (in note 134) 19–20, pl. XII.
444 chapter twelve
No. 35. Middle Assyrian cylinder seal.148 43.102.37; Surkh Dum 528. Greyish
chalcedony (?); H. 2.72cm.; D. 1.18cm.
On this beautifully carved seal, a hero holds two long-horned animals each
suspended by one rear leg. He wears a kilt with pendant tassels and a
helmet surmounted by a lobe-like projection. In the field to the right are
an eightpointed star and a moon crescent.
Similar heroes conquering two animals appear on seal impressions on
tablets excavated at Assur in North Mesopotamia. Three impressions dated
to the reigns of the Assyrian kings Shalmaneser I and Tukulti-Ninurta I
(ca. 1274–1208 bc) show a similar hero wearing a kilt with tassels and holding
145 Susa: Pierre Amiet 1972, op. cit. (in note 125) nos. 2023, 2026. Mari: idem, “Notes sur
Nationale (Paris 1920) no. 464; H.H. von der Osten, Ancient Oriental Seals in the Collection of
Mr. Edward T. Newell (Chicago 1934) no. 303; idem, Altorientalische Siegelsteine der Sammlung
Hans Silvius von Aulock (Uppsala 1957) no. 293. See also the comments of Amiet 1973, op. cit.
(in note 134) 19 and note 1.
147 Rene Labat, “Elam c. 1600–1200 BC,” CAH II, part 2, chapter XXIX (Cambridge 1975) 379–
Glubok, The Art of the Lands of the Bible (New York 1963) 45.
surkh dum at the metropolitan museum of art 445
two horned animals each suspended by one hind leg. Other impressions of
similar date provide exact parallels for the eight-pointed star and the moon
crescent that appear on the Surkh Dum example.149
The enormous curving horns of the animals, and the hero’s helmet sur-
mounted by a lobe-like projection, however, are unparalleled on seals pro-
duced in Assur, or on artifacts from any other region of Mesopotamia as
far as I know. Reversed animals, probably gazelles characterized by their
long lyre-shaped horns, appear on a cup in the Louvre, which is not from
stratified context, but is of probable Iranian origin on the basis of its style
and bitumen material. Dated to the early 2nd millennium bc on stylistic
grounds by Amiet, these creatures are held aloft by a bull man rather than
a human hero.150 Later Neo-Elamite artifacts including a seal and a stone
relief from Susa show similar horned animals.151 Thus, whereas no repre-
sentations of a gazelle with these distinctive undulating horns appear in
Mesopotamian art, examples are found on Iranian artifacts of varying date.
Therefore, these animals may belong to a genus of gazelle indigenous to
Iran.
149 Moortgat 1941. op. cit. (in note 136) 50–88, esp. pp. 77–79. Abb. 59, 60, 61.
150 Amiet 1966, op. cit. (in note 118) no. 200 A, B.
151 Pierre Amiet, “La Glyptique de la fin d’ Elam,” Arts Asiatiques XXVIII (1975) 3–45, no. 62
(Susa), also no. 55; Amiet 1966, op. cit. (in note 118) no. 432.
446 chapter twelve
152 Porada 1971, op. cit. (in note 142) 28–34, fig. 7 (Assur) and fig. 9 (Foroughi). For a possible
example of this knobbed turban see also the badly effaced seal from Marlik, Ezat Negahban,
“The Seals of Marlik Tepe,” JNES 36 (1977) 81–102, esp. p. 92, fig. 8.
153 Porada 1975, op. cit. (in note 24) 386, no. 297g; also see B. Parker, “Cylinder Seals from
Tell al Rimah,” Iraq XXXVI (1975) 21–28, esp. p. 35, no. 48; and idem, “Middle Assyrian Seal
Impressions from Tell al Rimah,” Iraq XXXIX (1977) 257–268, esp. p. 260, pl. XXVII, no. 12.
154 J. Munn-Rankin, “Assyrian Military Power 1300–1200BC,” CAH II, part 2 chapter XXV
(Cambridge 1975) 274–298, esp. pp. 284–285; Labat 1975, op. cit. (in note 147) 386–387.
155 For a discussion of Assyrian “influence” see Irene Winter “Perspective on the ‘Local
Style’ of Hasanlu,” in Mountains and Lowlands, op. cit. (in note 37) 371–386, and Muscarella
1980, op. cit. (in note 116) 170, 200–202, 210–217, 222.
156 Helene J. Kantor, “The Glyptik,” in C.W. McEwan, Soundings at Tell Fakhariyah (Chicago
1958) 69–85. For possible indigenous Syrian elements see p. 82, no. LIII and note also Kantor’s
views concerning the earliest occurrence of the winged human-headed sphinx in Assyrian art
on illiustration XI. Compare the earlier winged male sphinxes on 18th–17th century bc Syrian
seals like E. Williams-Forte, Ancient Near Eastern Seals. A Selection of Stamp and Cylinder Seals
in the Collection of Mrs. Williiam H. Moore (Metropolitan Museum of Art 1976) no. 3.
surkh dum at the metropolitan museum of art 447
No. 36. Late Middle Elamite (?) cylinder seal. 43.102.36.; Surkh Dum 1461.
Unglazed yellowish faience (?) with apparent metal stain; H. 3.3 cm.; D.
0.92cm.
157 Porada 1970, op. cit. (in note 130) figs. 74–76, 79–80; Amiet 1972, op. cit. (in note 125) 265,
nos. 2055–2063.
448 chapter twelve
158 For development of opinions concerning the date of the border see Amiet ibid. 273–
274, nos. 2131, 2134, 2091, who dates the border primarily to the Neo-Elamite period; Porada,
ibid. 98, no. 117, 50, no. 51, states that the border might last a very long time on the basis of
the Susa seals like Amiet, ibid. nos. 2131–2134; Amiet 1973, op. cit. (in note 134) 24–25, note 1
redates the seals published in Amiet 1972, nos. 2131–2134 to Middle Elamite on the basis of
the Tchoga Zanbil material.
159 Delaporte 1920, op. cit. (in note 136) pl. 33:4 (Susa).
160 G. Contenau and R. Ghirshman, Fouilles de Tepé Givan (Paris 1935) pl. 38:36.
161 R.D. Barnett, “Homme masqué ou dieu-ibex,” Syria XLIII (1966) 259–276, pls. XXIV, 1, 2,
XIX, 1, 2; for early stamp seals see pl. XX, fig. 1, etc.
162 Van Loon 1967, op. cit. (in note 6) 24.
163 R. Ghirshman, Fouilles de Sialk II (Paris 1939) pl. XXX, 2. For chronology see Oscar White
Muscarella, “Excavations at Agrab Tepe, Iran,” MMJ 8 (1973) 70–71, note 14.
surkh dum at the metropolitan museum of art 449
No. 37. Late Middle Elamite (?) cylinder seal. 43.102.32.; Surkh Dum 131.
Chalcedony with iron rust; H. 3.38cm.; D. 1.34cm.
On this seal, a pair of crosses, one placed above the other, appears in the field
alongside three figures with stick-like limbs, beak noses, and horns (?). The
first figure to the right seems to hold a weapon in its upraised hand. With
elbows jutting, the two remaining figures hold a spear with one hand while
placing the other hand on a hip.
The three figures shown in procession on this seal are unparalleled. No
exact analogies appear to exist for these flat, shallowly engraved figures,
arms akimbo, with hands defined by small drillings. Their beak-like profiles
surmounted by a bent form ending in a drilling is similar to, but less clearly
rendered than, the heads of the attendants on No. 36. The relationship
between the figures on the latter seal and a particularly Iranian ibex-headed
demon or masked being was cited above and may be applicable as well to
the present more crudely carved figures. Whether the figures on this seal
are human beings wearing ibex-horned masks or composite creatures is
impossible to determine, but both clearly are related to an Iranian cultural
164 Porada 1965, op. cit. (in note 28) 78, fig. 49 and 235, chapter VI, note 6 for the pin
reference.
165 Pierre Arniet, La Glyptique mesopotamienne archaique (Paris 1961) 42, 158, pis. 37, 38; for
animals participating in banquets on earlier seals see pl. 99, no. 1308 (Ur) and no. 1313 (Tell
Asmar).
450 chapter twelve
tradition involving the hunt or worship of the ibex. The composite ibex-
headed creature is a manifestation of myth or fable, as may be the case
on No. 36, while the masked individuals perhaps reenact the myth through
ritual. The latter possibility may be applicable to the present scene for the
absence of animal prey near the armed figures and their strange gestures
may suggest that they are involved in a ritual, perhaps a dance, prior to the
hunt.166
A date for our seal in the late 2nd millennium bc is suggested solely on
the basis of the six crosses, motifs most commonly found on Kassite and
post-Kassite artifacts (ca. 1400–1000bc).167
The metal-like accretions on the stone may suggest that this seal was
deposited in the Surkh Dum sanctuary alongside metal artifacts. Interest-
ingly, the only other Surkh Dum seal in The Metropolitan Museum of Art
that bears similar metallic stains is the faience seal No. 36, which also shows
166 Barnett 1966, op. cit. (in note 161) 259–276 for discussion and modern ethnographical
parallels. See also Porada 1964, op. cit. (in note 28) 15, and Amiet, loco cit. (in note 165).
I neglected to discuss similar connections for a seal of Proto-Elamite date in Ladders to
Heaven, op. cit. (in note 104) 191, no. 155.
167 Thomas Beran, “Die Babylonische Glyptik der Kassiten-zeit.” AfO XVIII (1958) 255–278,
No. 38. Early Neo-Elamite (?) cylinder seal. 43.102.40.; Surkh Dum 103. Burned
chlorite (?); H. 2.58 cm.; D. 11.7cm.
A rampant griffin attacks a couchant winged bull on this seal. In the field
surrounding these animals are a fly (?), a fish, a star, and the lower body of a
monkey (?).
Similar griffins with inward curving wings occur on two seals from Susa
and one from Tchoga Zangil dated by Amiet to the early Neo-Elamite pe-
riod.168 A seal from Tepe Sialk, Necropole B (ca. late 9th–7th centuries bc)
shows an animal in analogous rampant pose menacing winged horned
animals. And a vase from Sialk dated to the same period shows griffins that
also can be compared to the creatures on this Surkh Dum seal.169
As noted by Amiet, creatures having crescent-shaped wings with feathers
indicated on their outer edge are characteristic of artifacts produced in the
early centuries of the 1st millennium bc170 An unprovenanced quiver dated
168 Amiet 1972, op. cit. (in note 125) 273, nos. 2126–2127; idem, 1973, op. cit. (in note 134)
No. 39. Early Neo-Elarnite (?) cylinder seal. 43.102.30; Surkh Dum 807. Burned
chlorite (?); H. 4cm.; D. 1.13cm.
An animal flanks a sacred tree on this tall, slender cylinder seal. Above the
animal’s head is a Maltese cross and behind it are a monkey above a bird.
Although iconographically related to late 2nd millennium bc post-Kassite
artifacts of Mesopotamian and Iranian origin, the scene on this seal is
stylistically similar to later 1st millennium bc artifacts produced in Iran. A
171 Porada 1965, op. cit. (in note 28) 51, fig. 30; for Proto-Elamite seals see Amiet 1961, op.
73, fig. 55; Amiet 1975, op. cit. (in note 151) 3–45, esp. p. 17, pl. IX. no. 71. For trees, birds, and
monkeys, see also Amiet 1972, op. cit. (in note 125) nos. 2121–2122 (Neo-Elamite).
176 Porada 1965, op. cit. (in note 28) 78, fig. 49.
177 For bulls with long tails see Beran 1958, op. cit. (in note 167) Abb. 16, 28, 31; caprids with
short tails, Abb. 22–24; caprid with long tail hanging down, Abb. 30.
454 chapter twelve
The Surkh Dum creature’s extremely long, only slightly arched neck with
spikey forms marking its outer curve also is unusual. Linear striations fre-
quently decorate the interior volume of animal’s necks on post-Kassite arti-
facts but never protrude beyond the neck’s outline.178 On our seal, however,
the linear strokes issue from the edge of the neck and thus may indicate the
creature’s mane. Analogous short linear details, some slightly up-curving
as on the present example, decorate the neck of a short horned, leonine-
clawed, birdfooted creature on a seal of probable 1st millennium bc date in
the Yale Babylonian Collection. Identified as the horned dragon of Marduk,
two of these creatures with more elaborate curling manes appear on the
seal of the son of Shutur-Nahunte II, a Neo-Elamite ruler (ca. 7th–6th cen-
turies bc).179 Thus, the short strokes stretching from our creature’s horn to
the base of its neck may represent a similar mane and identify it as a horned
dragon. Moreover, the curving, rounded forms of our creature’s body are
more closely related to these horned dragons than to the linearly patterned
bodies of 2nd millennium bc animals.
Horned dragons never appear alongside sacred trees on 2nd millennium
bc artifacts. Seals of the 1st millennium bc, such as the Yale seal and the
seal of Hupan-Kitin, the son of Shutur-Nahunte II discussed above, however,
show the horned dragon rampant alongside a spade, another emblem of
Marduk. On the latter seal, the spade is shown as a stylized sacred tree. Since
the creature on our seal appears to be Marduk’s animal attribute, the horned
dragon, perhaps the sacred tree with its spade-like branch is in some way
related to that god’s triangular emblem.
Thus, this seal finds its closest stylistic analogies in artifacts from early
1st millennium bc Iran. A 1st millennium bc date and Iranian origin may
be supported by the seal’s material, which is the same burned chlorite of
Nos. 38 and 40, seals of probable early Neo-Elamite date. Because of the
unique nature of the scene, however, and the analogies mentioned above
to post-Kassite artifacts, the possibility cannot be excluded that this scene
represents a previously unknown regional style of the late 2nd millennium
bc that perhaps is indigenous to Luristan.
178 Ibid. Abb. 28; for later material see Porada 1965, op. cit. (in note 28) pl. 33; also see
Madeline Noveck, The Mark of Ancient Man … (Brooklyn Museum 1975) no. 36.
179 Amiet 1975, op. cit. (in note 151) 18–19, pl. IX, no. 67 (Yale), pl. VI, no. 34 (Hupan-Kitin).
surkh dum at the metropolitan museum of art 455
No. 40. Early Neo-Elarnite (?) cylinder seal.180 43.102.33; Surkh Dum 1299.
Burnt chlorite; H. 4.31cm.; D. 1.28cm.
The central scene shows two rampant horned animals flanking a tree with
spikey branches and tendrils. To the left an archer kneels above a horizon-
tally flying bird with denticulated wings. Above the archer is a star.
This tall, slender seal, published and discussed in depth by Edith Porada,
has been compared to several seals of different style but similar iconography
from Tchoga Zanbil.181 Of faience and late Middle Elamite date (ca. 12th–11th
centuries bc), the Tchoga Zanbil seals also show kneeling archers along-
side enormous horned animals flanking vegetation. The size differential
between man and beast suggested to Porada that the caprids might be con-
sidered supernatural by the carver of the seal.182
This seal has been placed chronologically by Porada between the late
Middle Elamite Tchoga Zanbil examples (ca. 1200–1000bc) and the later 9th
century bc Neo-Assyrian linear style seals which more closely parallel the
180 Previously published in Porada 1964, op. cit. (in note 28) 15, pl. 1, fig. 1.
181 Ibid. 15, text fig. 1; also idem 1970, op. cit. (in note 130) pl. IV, nos. 35–36.
182 Ibid. 15.
456 chapter twelve
Conclusion (O.W.M.)
In 1977 and 1979 I presented a list of Luristan objects that have been exca-
vated both in Iran and elsewhere.184 Several more excavated pieces may
now be added, although, unfortunately, none has been published with pho-
tographs: a bronze tube with Janus heads at the top and a “screw” base, from
Baba Jan in eastern Luristan;185 a bronze standard finial consisting of con-
fronting felines from Xatunban in the Ilam area of western Luristan, and
from the same site five bronze horse bits, at least one of which is in the form
of winged goats trampling a gazelle.186 Collectively they add up to a total of
28 excavated Luristan objects (25 from Luristan itself and three from else-
where),187 a pathetically small number when compared to the thousands of
was furnished: from the description the tube seems to be similar to those illustrated in
Moorey 1971, op. cit. (in note 6) pls. 37, 38. We thus have two Luristan bronzes from the
settlement site of Baba Jan.
186 Published too late for my 1979 paper: Exposition des denières Découvertes Archaeolo-
giques 1976–1977 (Musée Iran Bastan 1977) 42, nos. 384, 385; no. 386 mentions “1 des 4 mors en
bronze”, but it is not clear if they are plain or decorated with figures. The confronting felines
are apparently like those illustrated in Moorey 1971, op. cit. (in note 6) pls. 31, 32, but we do
not know if they are of the naturalistic or the stylized type. Note also that a bronze bucket
excavated by Louis vanden Berghe, “La Necropole de Charnzhi-Mumah.” Archaeologia 108
(1977) 60, 61, has a scene of a city under siege and a chariot battle, known to me from a drawing
kindly sent to me by the excavator. This bucket may be an Assyrian import.
187 When the final publications of Vanden Berghe appear more objects may be added;
I have not counted. except for one example (above) the weapons excavated by Vanden
Berghe. Note also that the objects allegedly found at Maku in Iranian Azerbaijan can in no
archaeological sense be accepted as excavated objects; cf. Moorey 1971, op. cit. (in note 6) 16,
143–144.
surkh dum at the metropolitan museum of art 457
objects claimed to derive from Luristan and given much prominence in pub-
lications for the last 50 years. It is against this background that the present
report should be viewed, for it increases by over 100 % the number of exca-
vated Luristan objects available for study and discussion. Equally important,
a variety of objects not recognized among the recorded excavated types
may now be included within the bona fide repertory of Luristan artifacts,
e.g., sheet-metal work, anthropomorphic-headed pins, pendants, and ivory
and bone material. And, finally, the cylinder seals significantly amplify our
knowledge concerning the foreign relations of the Luristan culture, knowl-
edge based not on dealer-derived material, but rather on the only evidence
that can lead to meaningful archaeological conclusions-excavated objects.
chapter thirteen
* This chapter originally appeared as “North-western Iran: Bronze Age to Iron Age,”
in Anatolian Iron Ages III, eds. A. Çilingiroğlu and D. French (Ankara: British Institute of
Archaeology at Ankara, 1994), 139–155.
460 chapter thirteen
some cases very similar to the painted ‘Urmia’ wares from the Urmia region
(Edwards 1986: 69). But when one encounters painted vessels with the very
same decoration that were purchased as far apart as Adana and Van (respec-
tively, Çilingiroğlu 1984: 130, 137, Figs. 1, 13 and 132, Figs. 5–8), one must be
hesitant about assigning provenances.1 These vessels most probably derived
from somewhere in Turkey and, as such, inform us of cultural relations, even
a common ceramic tradition, with North-western Iran sometime prior to the
15th–14th centuries bc.2
In North-western Iran, excavations at Hasanlu and Dinkha Tepe in the
Solduz and Ushnu valleys respectively in Azerbaijan have revealed a strati-
graphic sequence from the early-Second through most of the First Millen-
nium bc. At both sites there exists the same Bronze Age culture (Hasanlu VI,
Dinkha IV) defined by painted, incised and plain buff wares that are related
to the Syrian and Mesopotamian Khabur wares (Fig. 2, nos. 3, 6, 11–14 and
Fig. 3). This culture terminated sometime in the 17th–16th century bc (Dys-
on and Muscarella 1989: 23 n. 25; Stein 1984: 29). At Dinkha polychrome ware
(Pl.1.1) occurs primarily in the Bronze Age erosion levels and in the over-
lying basal Iron Age fill, with no firmly associated architecture. Although
not yet fully studied, this ware seems to represent a post-Khabur period
occurrence that briefly existed just before and at the very beginning of the
new Iron Age period (see also Edwards 1986: 63). Further evidence for this
brief overlap are two polychrome vessels, one each from Early Iron Age I
burials at Hasanlu (Stein 1940: Fig. 110, Pls 24:3, 31:8) and Dinkha (Fig. 4.1;
Muscarella 1974: 39, 48, Figs. 3, 5); these could be imports from the north or
have been locally acquired.3 A similar overlap is also evident from painted
vessels present with Iron Age pottery at Kordlar on the western shore of
Lake Urmia (Lippert 1979: Figs. 9–11), at Sialk (Ghirshman 1939: Pls 37, 40, 41,
S444, S476, S495; Young 1965: 70, 72)4 and, intriguingly, at Kizilvank across
1 Modern trade has surely passed the vessels far from their plundered homes. Edwards
(1986: 70) says archaeologists believe they came from north of Van.
2 The ante quem date would be determined by this beginning date for the EWGW/Iron I
culture. Edwards (1986: 69) seems to accept the vessels in Adana as locally acquired, a prove-
nance which is not demonstrated. He also suggests (p. 70) that plundered painted vessels in
Munich are “possibly from North-western Iran”, a conclusion without archaeological value.
3 Çilingiroğlu (1984: 130, Fig. 2) compares a vessel in Adana with the Hasanlu example.
While close, it is not exact. Edwards (1986: 63 n. 61) incorrectly assigns the Dinkha vessel to a
stone tomb. It actually came from an inhumation cut into the Bronze Age level. He dates the
burial to the Bronze Age, with no documentation or reference to the published text.
4 Medvedskaya does not discuss these painted vessels, which may represent a degree of
continuity. In her illustrated pottery chart they are not shown in the Sialk row.
north-western iran: bronze age to iron age 461
the Araxes River (Dyson 1973: 705), where polychrome painting occurred on
Iron Age (in Iranian terms) pottery shapes.
Stratified, pre-Iron Age polychrome wares occur elsewhere in the Urmia
area—abundantly at Haftavan, at Kordlar, in the Trans-Caucasus and, as
noted, in Anatolia (Çilingiroğlu 1984; Edwards 1986: 64, 67, 70, 72),5 all of
which attests to a cultural tradition of painted wares linking these areas in
the Bronze Age, sometime about the middle of the Second Millennium bc.
Commencing with Hasanlu period V and Dinkha period III, the Bronze
Age cuLture was succeeded in the 15th or 14th century bc by a distinctly
new culture characterized by burnished monochrome grey and orange-red
wares (Fig. 4.2 and Pl. 1.2). The culture was first called the Early Western Grey
Ware (EWGW) horizon by Young (1965: 70; 1985: 362 n. 1), then simplified to
Iron I by Dyson (1965: 211; Young 1967: 22 n. 68), although still technologi-
cally a Bronze Age culture (on terminology, see Muscarella 1974: 79; Lippert
1979: 134: “die bronzeitlichen Tradition”; Haerinck 1988: 64 n. 1). The follow-
ing period, Hasanlu IV and Dinkha II, clearly a continuation of the same
culture, was called the Late Western Grey Ware (LWGW) horizon or Iron II
(Pl. 2.1; Young 1967, 1985; Dyson 1965). At Hasanlu this period was terminated
in a major destruction about 800 bc, most probably accomplished by the
Urartians, who then occupied the site (Dyson and Muscarella 1989: 3, 20;
Muscarella 1989: 34).
When the Urartians destroyed Hasanlu IV, and most probably Dinkha II,6
they destroyed a culture that had existed in the area for about 600 years
(Dyson 1965; idem 1977; Muscarella 1974; Dyson and Muscarella 1989). But
the culture that had existed in the Van area during this period remains
archaeologically unknown.
It was clear from the early excavations at Hasanlu that the Iron Age cul-
ture was distinct in a number of aspects, most obviously in pottery forms
and surface treatment from that of the earlier Bronze Age culture. This dis-
tinction was made manifest by the subsequent excavations at Dinkha, where
5 Edward’s map (1968: 71, Fig. 4) excludes Dinkha from the polychrome zone. For painted
and polychrome wares from pre-Iron Age Kordlar (not discussed by Edwards 1986), see
Kromer and Lippert (1976: 65–82, Pls 4–6) and Lippert (1976: Pls 1, 2). Also compare Kromer
and Lippert (1976: Pl. 4.2) with Çilingiroğlu (1984: 130, 137, Figs. 1, 13 and Pl. 12.1) here with
Çilingiroğlu (1984: 131, 135, 136, Figs. 2, 9, 12).
6 We have no evidence at Dinkha for or against a destruction, as the Iron II buildings
preserve only one course of foundation stone, but no post—LWGW/Iron II material exists
there. Dinkha lies just across the valley and in sight of the Urartian site of Qalatgah built by
Ishpuini and Menua, and surely would not have been left free (Muscarella 1974: 58, 82; 1989:
34).
462 chapter thirteen
stratified architecture, burials and pottery of the Bronze Age was available
for study alongside the later EWGW/Iron I finds there (similar but quanti-
tatively less evidence is also now known from Hasanlu). That this distinc-
tion indicates a major break and a new beginning in the history of the area
has been recognized by all scholars who have reviewed the evidence (e.g.
Dyson 1965: 197; 1977: 156, 166; Young 1965: 57; 1967: 22, 24, 31; 1985: 361, 364,
373; Burney 1972: 115, 117; Muscarella 1974: 52, 81; Levine 1987: 233). Further
research demonstrated that the same EWGW/Iron I horizon is present at
other sites surveyed and excavated in the Solduz valley (e.g. Hajji Firuz,
Nagadeh); on the western shore of Lake Urmia, at Kordlar and Haftavan;
in the Elburz Mountains near Teheran, at Khurvin and Gheytaryeh (which
have not yielded earlier material); at Sialk; and in limited form at Godin and
Giyan in Central-western Iran but not in the Mahidasht area or in Luristan
(Young 1985: 364; Levine 1987: 241, 247). The EWGW/Iron I horizon was char-
acteristic in Iran only in the north-west and along the southern face of the
Elburz mountains (Young 1985: 367, 374): an essential understanding neces-
sary to keep in mind in discussions of the Iron Age(s) in Iran at large (Young
1985: 363 n. 1).
Furthermore, that this new culture, appearing in the area a century or
more after the termination of the Bronze Age Khabur culture,7 indicates
the presence of a new ethnic entity, seemed to most scholars a reasonable,
albeit not provable, conclusion. However, to go beyond the archaeological
evidence and attempt to identify the ethnic affiliation of the newcomers
is another matter. Thus the assertion that the newcomers were Iranian
speakers (e.g. vanden Berghe 1964: 46; Young 1867, 1985: 372, 374—“some
Iranians”; Burney 1972: 115; Ghirshman 1977: 47, 50, 54, 62) is speculative,
being incapable of documentation, since written records have yet to be
recovered.8 And opinions regarding the direction and loci from which this
Iron I culture arrived are still under review and are not easy to resolve
(compare Young 1967 and 1985 with Dyson 1977).
Nevertheless, taking a position that may accept or reject a specific ethnic
identification or a preference for one route over another, one may admit that
7 At Dinkha in the B9 and 10 areas, along the north, there is evidence of a chronological
gap between the end (by burning) of the Bronze Age and the incipience of the trash deposits
into which was created the Iron Age cemetery (Muscarella 1974: 52). The excavations at
Hasanlu were too limited for this period to allow for a firm conclusion about a gap (Dyson
1965: 195).
8 Young (1967: 31) also thinks that Marlik (which has metal spouted vessels similar to
Iron I shapes) was an Iranian (language) site. Medvedskaya (1982: 11) also has the same view,
which seems a contradiction to her general conclusions (infra).
north-western iran: bronze age to iron age 463
there are issues separate from that which accepts the conclusion that a new
culture (most probably brought by a new people) entered North-western
Iran, which is strictly an archaeological issue.9 And this latter conclusion is
a basic fact of the archaeological record.
In a paper (1977) and monograph (1982) I. Medvedskaya is convinced that
she has succeeded in demonstrating that there was no cultural break from
the Bronze to the Iron Age in North-western Iran, that there was a continuity
in the culture. Furthermore, in as much as there was continuity, there can be
no argument that Iranians arrived there in the Iron I period; indeed, there
can be no claim that any new ethnic group arrived there at this time (1977:
104; 1982: 38, 96, 99, passim).10
To demonstrate these historically significant conclusions, Medvedskaya
claims that a careful reexamination of the finds and their contexts indicates
that 1) there is a lack of cultural unity in the Iron I period since it is not stable
(1977: 94, 99; 1982: 13, 34, 39, 50, 96) in pottery distribution, that 2) Iron I
pottery forms are “not new for Iran in the period Iron I” for they contain
Bronze Age forms (1977: 103; 1982: 34, 38, 98) and consequently the origin
of the pottery is to be “sought in situ in the roots of the local (Bronze Age)
culture or the cultures of adjacent areas”, and that 3) burial forms are not
distinct in the two periods.
In all these assertions Medvedskaya is utterly wrong. An analysis of her
presentation of the evidence reveals that in almost every instance there is
misunderstanding of the published material, omission of relevant and sig-
nificant information and errors of fact. What is unfortunate is that her con-
clusions, unexamined, are being cited by scholars either as a new and correct
interpretation of the early history of Iran (Genito 1986: 50; Dandamayev and
Lukonin 1989: 12) or as “useful” (Curvers 1985: 198) or casually, as a “different
view” (Henrickson 1983–1984: 216 n. 51). Lest still others uncritically cite the
subjective revisionism of Medvedskaya, that is, cite her without first going
back to the data, I will here review her claim.11
9 Note that neither Dyson nor the author ever stated that the EWGW/Iron I culture
represented Iranian speakers. Dyson at one time (e.g. 1960: 119, 121; 1964b: 3) thought the
entity in Hasanlu IV (LWGW/Iron II) was that of the Manneans, a position later dropped;
see Boehmer (1964: 19 n. 64), who rejects Hasanlu IV as a Mannean site.
10 Since Medvedskaya (1982: 1) believes that the Urmia area was occupied by Hurrians in
the Second Millennium bc, it would follow that she must also believe (a belief avoided in the
text) that the inhabitants of Hasanlu and Dinkha were Hurrian speakers in the Iron I and II
periods.
11 The text of the monograph is often difficult to understand. Information on specific sites
is not given in one place but scattered. The same scattering exists in the pottery and burial
464 chapter thirteen
discussions. Sometimes the meaning of a sentence is unclear: this may be a problem with the
translation.
12 Burney (1973: 164) dated istikhans to the Iron I period solely because he thought
(wrongly) that the Godin examples were from this period; but note that Burney’s text here is
unclear: were true istikhans found at Haftavan?
13 These corrections are not reported by Medvedskaya although the works are cited in her
bibliography. Medvedskaya also omits the istikhan found by Stein (1940: Pl. 30: 11) in an Iron II
burial. There are two metal examples with handles from Tepe Guran (Meldgeard, Mortensen
and Thrane 1964: 128, Fig. 29). The published Hasanlu istikhans derived from a period VI
(Bronze Age) burial, OP 10, burial 2 (Dyson 1959: 9). Other istikhans occur in the period VI
level at Hasanlu.
14 Medvedskaya (1982: 19, 1(b)2) (she gives no reference and refers to grey and cream
colour) may be referring to the one example from Dinkha III but she does not mention the
one from Hasanlu (supra). Medvedskaya (1977: 94) mentions painted wares “in significant
numbers” in Hasanlu IV: there are no painted wares from this period.
north-western iran: bronze age to iron age 465
Medvedskaya (1977: 96; 1982: 13, 39) claims that pattern-burnished pottery
occurs often and is a decoration characteristic of Iron I pottery but (Dyson
1964a: 36) specifically stated that it was rare in that period. Furthermore, she
finds it possible to cite a unique pattern burnished sherd from Dinkha IV
(Hamlin 1974: 129) as an indication of continuity, even though there is no
example known from Dinkha III (Muscarella 1974: 48).
Medyedskaya (1982: 38,98) notes that “grey wares” are ubiquitous in the
Near East in the Second Millennium bc, in Mesopotamia, the Caucasus,
Greece and Iran, including Dinkha IV, all of which attests “to the continuous
existence of grey and also grey-black ware throughout the Second Millen-
nium bc. Therefore, the sharp increase in its quantity in the Iron Age I can-
not be considered sudden and, what is more important, it was not brought
from outside”. What is not realized in this argument is that grey ware is here
treated generically, grey ware equals grey ware. Notwithstanding this facile
generalization, there is a difference between the grey wares of the Bronze
Age and that of the EWGW/Iron I horizon in fabric, technique of manufac-
ture, surface treatment and shape.15 No one who has closely worked with
the pottery in Iran would confuse the two (Young 1985: 373: 373, n. 11; Dyson,
personal communication). Indeed, surveys in the Solduz valley have allowed
several Iron Age sites to be recognized specifically because of the character-
istic grey wares present. The grey wares of the EWGW/Iron I horizon were
brought from outside as a coherent assemblage. There is no evidence for any
in situ evolution.
Medvedskaya further posits that there was no ceramic horizon in the
corpus of the Iron I period in Western Iran (1977: 95; 1982: 50, 96). In the
north-west the lack of unity is indicated because all the diagnostic shapes
do not occur together in the same tomb or because they do not always
occur together at the same time. Consequently if every tomb does not
contain all the diagnostic shapes or if they do not occur together at each
site, cultural unity is lacking, And the pottery of the other areas in Western
Iran—Luristan, Southern Caspian—is “different”.
Medvedskaya does not include a definition or explanation concerning
the reasons which determined the utilization of other geographical and
cultural areas in her study of North-western Iran and she omits significant
information from the north-west that witnesses against her views. Thus, in
the first place, to document diversity she gives equal weight to the finds from
Hasanlu V and Dinkha III in the north-west, both classical EWGW/Iron I
15 Hamlin (1974: 128) notes that the grey wares from Dinkha IV consist mainly of bowls.
466 chapter thirteen
sites and to those of Giyan 116 and Godin to the south-east (1982: 33, 39,
41, 54, 64). Now, while participating in some manner, and being somewhat
contemporary, with the EWGW/Iron I culture (Young 1965: 62; 1985: 366),
the southern sites cannot be brought forth to document a lack of unity in
the north. Godin preserves only three graves. The culture of Giyan both in
the Bronze II and Iron I Age periods—which exhibit a lack of continuity—
was essentially peripheral to that in the north-west during both periods
(Henrickson 1983–1984: 210 n. 29, 215). The neighbours valleys (including
Luristan) also lack evidence of extensive occupation of the EWGW tradition
(Young 1985: 367; Medvedskaya 1982: 96!).
Of course there was diversity in the cultures of Western Iran in the gen-
eral, chronological Iron I period, a fact recognized by all scholars who have
studied that period (e.g. Levine 1987: 232, 243; Young 1965: 362 n. 1, 364;
Henrickson 1983–1984: 216 n. 51; Haerinck 1988: 64, 67). None of these schol-
ars has challenged unity in a specific area. The unity of the EWGW/Iron I
horizon in the northwest cannot be refuted either by reference to other, dis-
similar, Iron I cultures or by omitting information from the relevant area
that indeed attests to unity. This is the methodology of Medvedskaya.
Medvedskaya’s analysis and statistics are significantly compromised by
omitting from the analyses the pottery from the important northern site of
Kordlar, which with Hasanlu and Dinkha is a typical EWGW/Iron I site, and
that from the burial at Hajji Firuz (Pl. 2.2; infra), which is most probably a
deposition from one of the several Iron Age sites surveyed in the Solduz val-
ley. When the finds from these sites are placed back (because removed by
Medvedskaya) into the corpus and into the discussion, they play a signifi-
cant role in the collapse of Medvedskaya’s conclusions.
As has long been recognized, there are three basic diagnostic shapes
for the EWGW/Iron I horizon: 1) the bridge less spouted vessel, 2) the
pedestal-based, goblet with handle and 3) a bowl with a small modelled
ridge in crescent form on the inside, the so-called ‘worm bowl’ (Fig. 4.2 and
Pl. 1.2); also illustrated by Muscarella (1968: 193, Fig. 17).17 In her eagerness
16 It may be stated that she considers Giyan to be more important in her argument than
Dinkha, for inexplicably in her Fig. 4, an illustrated chart showing “ceramic links between
the Iron I sites of Iran”, there is a Giyan column but none for Dinkha, which has the largest
and most comprehensive corpus of EWGW/Iron I pottery published to date. This is but one
example of the distortion of her analysis.
17 Medvedskaya (1982: 37, 40, 42) gives much attention to cups, which she claims are
diagnostic. For cups from this period, add to her listing of Sialk and Khurvin (= Young 1965:
Fig. 9.2), examples from Hasanlu, Haftavan and Gheytaryeh, all in North-western Iran. For
Fig. 12.4.2, the B 10b, burial 13, see Muscarella (1974: 85, the catalogue).
north-western iran: bronze age to iron age 467
to deny cultural unity, Medvedskaya does not mention (or is unaware of)
a fact essential to the discussions (and it is one that readily contradicts her
conception), namely, that all three diagnostic shapes are associated together
at six sites in North-western Iran: Geoy Tepe (Dyson 1965: Fig. 2; here Fig. 5),
Haftavan (Burney 1970: 165), Hajji Firuz (Muscarella 1974: 49; here Pl. 2.2),
Kordlar (Lippert 1979: Pls 5, 6), as well as at Hasanlu and Dinkha Tepe
(Fig. 4.1, 2 and Pl. 2; Muscarella 1974: 48). Furthermore, at Geoy Tepe and
Hajji Firuz18 the three diagnostic shapes occur together in one tomb and at
Kordlar they were all found together in one room.
It should also be noted in the discussion that Medvedskaya is unaware of
the importance of the ‘worm bowl’ as a diagnostic shape in the EWGW/Iron I
corpus; it is found at six sites, not three as recorded in scattered references
by Medvedskaya, and it occurs only in the core area of the EWGW/Iron I
culture.19
Medvedskaya believes that she has demonstrated the lack of unity and
stability in the EWGW/Iron I horizon by asserting that its diagnostic pottery
shapes never existed side by side in a horizon. To her, they were manufac-
tured at different times and accordingly arrived in the north at different
times and from different areas.20 Aside from ignoring the evidence of the
Geoy Tepe and Hajji Firuz finds, these assertions are presented without anal-
ysis of stratigraphical or archaeologically derived chronological evidence
that might shed light on the major and still unresolved problem of the rel-
ative dating of the various Iron I sites in Iran. For example, she notes the
predominance of the diagnostic goblet in the north (1982: 37) but asserts
without analysis or evidence that it arrived there from “Luristan [sic]
(Giyan)”, where it “appeared earlier”. In fact, our knowledge of the relative
chronology of the appearance of the goblet at Giyan and the northern sites
is not sufficient to argue that it “could have begun [at Giyan] either ear-
lier or later than in North-western Iran” (Henrickson 1983–1984: 212).21 And
when Medvedskaya (1982: 50) states that the EWGW/Iron I goblet must have
“come from the pottery of the previous period”, she in fact cannot cite a sin-
gle earlier example.
18 Hajji Firuz is mentioned in her catalogue of sites (p. 102) but its vessel inventory is
east, east to west, south to north, but never from north to south or east.
21 Furthermore the goblet form at Giyan is the only shape that links that site to the north
22 Medvedskaya never notes that the EWGW/Iron I shapes (1982: Fig. 7: 110–113) isolate
themselves from the forms of other spouted vessels in the Near East (1982: Fig. 7).
23 Medvedskaya (1977: 99) incorrectly states that all Dinkha III burials contained goblets;
the goblet (1982: Fig. 3) is omitted from the Dinkha III row.
north-western iran: bronze age to iron age 469
24 A vessel from this burial is paralleled by another from Dinkha IV (Hamlin 1974: 133,
Fig. 2; idem 1974: 37, Fig. 2) is not an extramural cemetery (p. 65) because no
settlement of that period was recovered! The fact that at Dinkha the period
III cemetery has the same “topography … as the preceding Dinkha IV period”,
i.e. it was deposited over Bronze Age burials, is a further indication to her
that the cemetery was not extramural.
Although Medvedskaya is correct that no period III settlement was recov-
ered at Dinkha (the whole mound was not cleared: Muscarella 1974: 36.
Fig. 1), manifestly a settlement did not exist in the cemetery area, a fact
which signifies in normal archaeological reasoning that the settlement and
cemetery were discrete and that the latter was extramural. Bronze Age buri-
als did exist below the Iron Age burials but Medvedskaya neglects to report
that the Bronze Age burials were associated with architecture. They were
intramural burials and thereby markedly of a different nature from the over-
lying burials in a cemetery dug into trash debris (whose depositors probably
knew nothing of the underlying architecture and burials).
In Medvedskaya’s unique system of classification, not only is Hasanlu’s
cemetery (which is situated in an area outside the settlement) not extra-
mural, it is not even a cemetery (p. 61). It is a “pseudocemetery”, which
unfortunate term she defines as a cemetery “on the edge of a settlement”.
That this “edge” is outside the settlement, and therefore extramural, is not
appreciated.
Medvedskaya (1982: 55, 65) also takes much time to discuss burials at
Giyan, Bronze and Iron Age Godin26 and in Luristan proper, all of which are
irrelevant to the archaeological history of the Bronze and the EWGW/Iron I
cultures of North-western Iran (supra), allegedly the topic of the paper.
Finally, there is a long, and equally irrelevant, chapter (Medvedskaya
1982: 68) on daggers, axes and arrow heads, the purpose of which is to
demonstrate that in the Iron I and II periods of Western Iran various forms
existed in different regions. There is absolutely nothing in this conclusion
that is new, nor in the claim that the “Iranian”27 weapons do not reveal
evidence of the presence of “the Iranian nomads” (p. 94). The chapter is but
another attempt to focus on the cultural diversity throughout Western Iran
during the Iron Age in order to document an alleged lack of cultural unity
in the northwest.28
26 The Godin excavations had nothing to do with the Hasanlu Project: it was an indepen-
dent venture. Its three Iron I graves were also extramural (Henrickson 1983–1984: 205).
27 Medvedskaya does not clarify her usage of the term Iranian for the cultures of the area
known as Iran and for its linguistic meaning. These are two different concepts!
28 Medvedskaya says nothing of the continuity of toggle pins and torques from the Bronze
north-western iran: bronze age to iron age 471
Acknowledgements
All photographs (except Plates 1.2 and 3) and all figures (except 4.1 and 2)
are courtesy of the University Museum, University of Pennsylvania.
Age in general to the Iron Age (Muscarella 1974: 51, 52, 54, 80) nor of the polychrome wares
supra. Torques are a new element at Dinkha III. They do not appear in the Bronze Age tombs
and toggle pins all but disappear.
29 Relevant bibliography is missing, for example Boehmer and Cleuziou on arrowheads.
Medvedskaya is unaware of the literature on fibulae in the Near East and Iran: on p. 75
she dates one to the 9th century bc, when it first appears a century or more later; see also
Haerinck (1988: 65).
30 Medvedskaya uses words such as “accidental” or “chance” finds to describe what was,
and is in fact, organized plunder. Her usage distorts the magnitude of the destruction of
Iranian cultures (Muscarella 1988: 34).
31 Young (1985: 374) suggests that the people of the EWGW/Iron I culture may have
included Iranians and non-Iranians. It begs the question, How do we know. This position
seems to be an artificial compromise with his original 1967 paper.
472 chapter thirteen
Bibliography
vanden Berghe, L.
1964 La nécropole de Khurvin. Istanbul.
Boehmer, R.M.
1964 Volkstum und Stadte der Mannäer. BaM 3: 11–24.
Burney, C.
1970 Excavations at Haftavan Tepe: 1968. Iran 8: 157–171.
1972 Peoples of the hills. New York with D.V. Lang.
1973 Excavations at Haftavan Tepe 1971: third preliminary report. Iran 11: 153–172.
Çilingiroglu, A.
1984 The Second Millennium pottery tradition of the Van Lake basin. AnatSt 34:
129–139.
Curvers, H.H.
1985 Review of Medvedskaya 1982. BibO 42.1/2: 197–198.
Edwards, M.
1986 “Urmia Ware” and its contribution in North-western Iran in the Second
Millennium BC. Iran 24: 57–77.
north-western iran: bronze age to iron age 473
Genito, B.
1986 The Medes: a reassessment of the archaeological evidence. EW 36.1–3; 11–
81.
Ghirshman, R.
1939 Fouilles de Sialk. Paris.
1977 L’Iran et la migration des Indoaryene et des iraniens. Leiden.
Haerinck, E.
1988 The Iron Age in Gilan—proposal for a chronology. In Curtis, J. (ed.) Bronze-
working centres of Western Asia c. 1000–539BC: 63–82. London.
Hamlin, C.
1974 The early-Second Millennium ceramic assemblage of Dinkha Tepe. Iran 12:
125–153.
Henrickson, R.C.
1983–1984 Giyan I and II reconsidered. Mesopotamia 18–19: 195–220.
Levine, L.
1987 The Iron Age. In Hole, F. (ed.) The Archaeology of Western Iran: 229–250.
Washington, DC.
Lippert, A.
1976 Vorbericht der österreichischen Ausgrabungen am Kordlar-Tepe in Per-
sisch-Aserbaidschan: Kampagne 1974. MAGW 106: 83–111.
1979 Die österreichischen Ausgrabungen am Kordlar-Tepe in Persisch-Westaser-
baidschan (1971–1978). AMIran 12: 103–137.
Medvedskaya, T.
1977 On the ‘Iranian’ association of the grey ceramics of the early Iron Age.
Vestnik Drevnei Istorii: 93–104. (In Russian).
1982 Iran: Iron Age I. Oxford (BAR International Series 126).
Muscarella O.W.
1968 Excavations at Dinkha Tepe, 1966. BMMA. November 187–196.
1974 The Iron Age at Dinkha Tepe. MMAJ 9: 35–90.
1988 The background to the Luristan bronzes. In Curtis, J, (ed.) Bronzeworking
centres of Western Asia c. 1000–539BC: 33–44. London.
474 chapter thirteen
1989 Warfare at Hasanlu in the late-9th century BC. Expedition 31.2/3: 24–36.
Stein, A.
1940 Old routes of Western Iran. London.
Stein, D.
1984 Khabur ware and Nuzi ware: their origin, relationship and significance. Assur
4.1. Malibu.
Figure 2. Hasanlu Period VI: nos. 3, 6, 11–14; all others are Period V
(scale 2:5). Courtesy of the Penn Museum, image #73867.
north-western iran: bronze age to iron age 477
Rudyard Kipling’s poem (from The Elephant’s Child) has been with me for
years—and I have previously cited it.1 In researching the Jiroft story it resur-
faced in my mind because the epistemological questions it poses are per-
tinent to the following discussion, and thus I (appropriately) cite it again
here:
I keep six honest serving-men
(They taught me all I knew);
Their names are What and Why and When
And How and Where and Who.
I send them over land and sea,
I send them east and west;
… For they are hungry men,
But different folk have different views …
Indeed; and it is the different views of the issues of asking or not asking,
and not answering, the questions raised by the honest serving men that
generates this review.
1. Introduction
Yousef Madjidzadeh, Jiroft: The Earliest Oriental Civilization” Bulletin of the Asia Institute 15
(2001): 173–198.
1 See Bronze and Iron (New York: Metropolitan Museum of Art, 1988), p. 9.
486 chapter fourteen
2 On the book spine and its cover the author’s name is spelt Madjidzadeh, but in the three
pre-text publisher pages it is spelt Majidzadeh. I use the former spelling because libraries
will use it. Fred Hiebert kindly sent me copies of photographs previous to my acquiring the
volume—as a gift from from Ali Vahdati in Tehran.
jiroft and “jiroft-aratta” 487
record here in one venue is valuable and important for continuous exami-
nation and research.
Up to the time of completing this review (August 2004) no excavated
artifacts have been reported from Madjidzadeh’s excavation (it is not a
cemetery site) that relate or can be compared to the dissimilar, unexcavated
488 chapter fourteen
objects published in his Catalogue. Therefore, only these latter objects per
se are capable of being analyzed—and which objects, I argue, must in
archaeological discourse be cited and labeled “Jiroft,” not Jiroft, artifacts.
An evaluation of information and assertions about “Jiroft” artifacts and the
nature of their ancient culture explained in the 2003 Catalogue and the
other published reports forces one to confront (again!) basic archaeolog-
ical methodology, here concerning cultural and historical interpretations
and evaluations involving unexcavated artifacts known solely from mul-
tiple confiscations, and with no records preserved of these acquisitions.
Revealed is this, at present there is no knowledge of Jiroft (i.e., from excava-
tions in the Jiroft area), and very little available of “Jiroft” except a number
of objects. Between the two conceptions there is a profound epistemolog-
ical and archaeological difference (but one would never know this from
the “Jiroft” literature: alleged archaeological attempts at conclusion-forming
procedures have altered very little in the last fifty years).
2. History of “Jiroft”
and national police and archaeological authorities over two years, we have
the article by Hamid-Reza Husseini in the Persian Morning Daily online,
9/8/02.
As if the same script writer were involved, the history of the discovery and
confiscation of the “Jiroft” material uncannily parallels in all formal details,
including local professional inaction, the discovery and partial confiscation
of the plundered Kalmakarra cave material in western Iran: see Henkelman
2003, 214ff. To his credit, Henkelman also distinguishes between Kalmakarra
and “Kalmakarra objects”; see below.
In the Bonham’s antiquity sale of 9/22/88, at least nos. 171, and 172, a
sculpted snake vessel and another vessel with cut-out rosette decoration
are neatly paralleled in the Madjidzadeh Catalogue: pp. 109 and 115. This
may indicate that plundering in the Jiroft area began much earlier than
reported, at least sporadically, or that these objects have a wide geographical
distribution.
comes to about 260 objects. Lawler (2003, 974) records that “many hundreds
of vessels” were confiscated—how many were plain (a number of which
were confiscated, see Madjidzadeh 2003, 159, 163), and how many decorated
with motifs, is not revealed. We await an inventory of the confiscated objects.
Excavated Tombs
Tombs have been excavated by an archaeologist (H. Choubak) at Riganbar
(one of the five named plundered sites). Although Madjidzadeh publishes a
photograph of one with its burial goods intact (in 2003b, 25), nothing of its
contents is mentioned. Ali Vahdati informs me from Teheran that this tomb
was the only Bronze Age tomb excavated, and that the others are Islamic,
and he confirmed my suggestion that no decorated vessels were recovered
here. Worth noting is that Madjidzadeh is engaged in an excavation strategy
of digging only a settlement site, not searching or excavating the plundered
cemeteries (Lawler 2004a, 46). Such unilateral action ignores the precious
work and model of the great Belgian archaeologist Louis Vanden Berghe,
who spent fifteen years surveying cemeteries in Luristan that had been
plundered for decades—and thereby recovered hundreds of intact burials.
His model should be followed, not ignored—even one tomb excavated with
Catalogue material would be a very significant archaeological discovery for
Jiroft!
From information gathered from local plunderers regarding the number
of objects recovered from the plundered tombs Madjidzadeh (2003,6) says,
“in many cases a single grave contained up to sixty objects.” And from Lawler
(2004a, 46) we are further informed from the same local sources that “each
grave contained at least one stone vessel; the largest one contained 30.” I
think it not unfair to say that these inventory records are hearsay, and may
not be cited as historical reality.
jiroft and “jiroft-aratta” 493
“Jiroft” Style
There are several different styles, depictions, and a variety of iconographies
represented in the confiscated corpus—especially articulated on bowls,
pedestal footed vessels, beakers, and “weights.” Many of these styles and
iconographies have not hitherto been encountered (“strange and different,”
having “many entirely new items” as one archaeologist—who defends the
corpus’ integrity—accurately reported to me). Perrot (in 2003b, 111) uses the
phrase “style de Jiroft,” but styles in the plural is a more accurate term, given
that for many “entirely new items” parallels do not exist, except in “Jiroft.”
Madjidzadeh (in 2003b, 26) reported that clandestine diggers reported to
him that relief decorated vessels were recovered, a claim again, which (here
parti pris) has no archaeological value. And no one to my knowledge has
reported finding a decorated vessel in surveys.
One archaeologist believed at one time (email message to me) that there
is a “virtual absence of the classic intercultural style …” but later shifted
entirely, claiming that “… the vast [sic] majority are indeed of the IS!”
Cleuziou (in 2003b, 116C 122) notes differences between “Jiroft” and Tarut
artifacts, but stresses “stylistic” parallels (viz. entwined snakes); and Mad-
jidzadeh (2003, 7) places all the chlorite objects in the série anciennes, i.e.,
early Intercultural Style (IS).3 A good number of the confiscated objects are
canonical members of the IS corpus, or readily relate to them: see Mad-
jidzadeh 2003 for imbricate, whorl, spiral, palm tree, hut, and animal and
scorpion patterns: pp. 44,67–75, 110–111, 117–118, 125, 127–129, 142; also the
vessels illustrated in Pittman (in 2003b, middle and bottom of p. 85; also
Cleuziou 2003b, 117, fig. 4, and 122). For convenient excavated examples for
these and other motifs in the IS corpus see of course Burkholder 1971, Kohl
1975 and 2001, Zarins 1978, and Lamberg-Karlovsky 1988.
According to Madjidzadeh the corpus reveals a “high quality of work-
manship,” (2003a, 37, also in 2003, 8), but he is also aware (2003, 10) that
it “is not always equal,” that (here no doubt reflecting the multiple styles
and iconographies) the corpus was “produced in different workshops, and
by different stone-cutters, having different levels of skill and talent. But, in
comparison with the Mesopotamian reliefs in stone, they appear in general
to be of superior skill, talent and capability.” He does not explain or develop
any of these important cultural and aesthetic issues further, give specific
examples, or discuss whether the different workshops recognized could be
3 For a defense of the term Intercultural Style, see Kohl 2001, 209, 215f.
494 chapter fourteen
correlated with different confiscation loci. But no one can dispute his obser-
vations on the skill and workshop issues.
(In a discussion with Philip Kohl an idea occurred: stone analyses should
be undertaken of both Yahya and “Jiroft” stone material to determine
sources—we know there was one near Yahya. I see no evidence that one
can claim that Yahya craftsmen manufactured the latter’s stone artifacts.)
Chronology
Since most archaeologists involved in “Jiroft” accept that the published cor-
pus derived from (somewhere around) ancient Jiroft, chronological ranges
may be estimated and have indeed been proposed. Madjidzadeh (2003a,
37, 44; 2003, pp. 7, 12) asserted an early date for chlorite vessel production,
3000bc, or late 4th to first half of the 3rd millennium bc; this chronology
is accepted by Perrot—“il y a 5000 ans,” 3100–2900bc (in 2003b, 97, Ill).
Cleuziou (ibid., 116) raises the question whether this early chronology is cor-
rect or a later date, Early Dynastic II–III, even into the Akkadian period, is
better, but seems to favor the later dating; Pittman (ibid., 81, 87) accepts a
general 3rd millennium date; Lawler (2004b, 50) reports (from an anony-
mous source) an “around 2500 bc” date. Recently P. Amiet (2002/2004, 95 f.)
rejected Madjidzadeh’s early dating, arguing for the late 3rd millennium.
He is correct; and one would expect that the genuine IS period artifacts in
the corpus be invoked for chronological determinations, comparing them to
excavated material from Mesopotamia and Yahya, and that recovered from
Tarut.
Madjidzadeh, however, knows (and therefore need not explain) that the
Mesopotamian artifacts post-date those from “Jiroft” and Yahya. But dis-
interested analysis indicates that his and Perrot’s beginning and flourit
chronology of “Jiroft” is fundamentally wrong, too high by more than a half
millennium. Concerning at least the IS material, the second half of the 3rd
millennium bc, which includes the late Early Dynastic and at least part of
the Akkadian period, is an accurate general chronology. Lamberg-Karlovsky
(1988, 54) sees IS material continuing into post-Akkadian times, as does
Kohl (1975, 30; idem 2001, 215, 220f., 222, 224, 226 f.), where it is claimed
that at Yahya IS material was made late in the style, in Akkadian or even
post-Akkadian times. One of the issues here is recovery of IS objects in post-
Early Dynastic, Akkadian contexts: see Martin in Aruz and Wallenfels 2003,
no. 233, an IS feline-snake combat scene in Berlin bearing an inscription
signed by Rimush, Lugal Kish. This object, albeit not excavated, is quite
important for establishing the chronological range for these scenes.
jiroft and “jiroft-aratta” 495
What must be faced unblinkingly here is that the early unanchored chro-
nology is generated not by archaeological reasoning, but by a priori tenden-
tious, self-serving conclusion formations about “Jiroft” and “civilization.”
2004a, 42) rejects Madjidzadeh’s “civilization” claims. I too have argued that
the ecstatic claims under review do not reflect reality, that even if all the
“Jiroft” material were ancient artifacts, it “is not world-shaking” (in Lawler
2004a, 48—with my original phrase restored).
Are all “Jiroft” objects manifestly of ancient manufacture? When I first en-
countered the “Jiroft” material in a lecture by H. Pittman in April 2003 (I
heard another in March 2004), aside from typical IS present, I was surprised
by the variety of styles, iconographies, and forms other than what is known
either in Mesopotamia or Iran, including the Kerman area. Locally, Yahya
shares the typical geometric IS non-representative motifs—the snake-lion
and snake-snake combat, whorl, and scorpion representations (see Kohl
1975 for patterns with locus map, p. 24, and p. 26, nos. 2, 3, 4; Kohl 2001, 222,
226, figs. 9.8, 9.13; Lamberg-Karlovsky 1988, 55 ff., figs. 1, 3, 4, pls. IV–X; omit
from both articles references to Azerbaijan and Palmyra as excavated sites
yielding IS objects). My sense, however, was that a number of objects could
not automatically be accepted as ancient productions merely because they
had been confiscated in Iran (of which action little is known). These views
were reinforced after further reading, but I readily admit that with ongoing
study over time I have changed my opinions several times regarding the ages
of specific objects.4
Judging from the publications and personal discussions, raising the issue
of possible forgeries in the “Jiroft” corpus is not a popular opinion. The fervor
aroused in some who accept the whole corpus as genuine may be under-
stood from email comments by an archaeologist with whom I exchanged
views regarding ancient or modern. Initially believing many objects to be
“impossible,” my correspondent noted that this position was soon reversed:
4 I experienced similar views in the past regarding then strange iconography’s surfacing
on the antiquities market that related them to IS objects (viz. snake-lion confrontations),
and raised the issue of “whether they are ancient or not” (Muscarella 2000a, 171). In some
instances an abeyant cautious view was recommended ibid., 171, nos. 16, 18a–f. For the
record, of the complex no. 16, I am now more at ease; I also believe in the antiquity of
no. 18-e—primarily because of the execution of the guilloche/whorls, for no. 18-c, p. 488, the
iconography is not an issue, and aside from the snake’s ears and body line, cannot condemn it;
pp. 169, 486, no. 7, still puzzles me but I cannot condemn it outright. Nota bene, that changing
one’s mind over time is a correct and necessary activity, it results from what I call the gift of
the bazaar, the chaos caused by the antiquities market.
jiroft and “jiroft-aratta” 497
and lifetime jobs), but because I cannot claim to know, to perceive all the
possible scenarios in their ancient/modern histories. Hesitancy is formal, an
attempt to keep all options open, one of which is the possible presence of
forgeries. An archaeologist experienced in IS and Iranian scholarship shared
this view with me: “I would argue that it would be almost impossible to
decide whether most [“Jiroft”] objects were genuine or fakes.” Which neatly
defines the problem I am articulating here: absent a Jiroft against which
to compare “Jiroft,” one is compelled to focus on problematic orphaned
objects. If we do not attempt to know which “Jiroft” objects are ancient and
which modern, how can we begin to discuss Jiroft?
A number of “Jiroft” objects stand out from the IS corpus, suggesting (at
least) that archaeologists not automatically accept and introduce them into
discourse on ancient artifacts; these are listed below as what at best are
called probable forgeries (below, “b”).
In addition, my eyes recognize a number of complex and ambitiously
made pieces with which I continuously wrestle (see below, “c”). They may
not be ancient—but I have changed my mind more than once with regard to
the age of several of them. Given this unclear view—a mon avis of course—I
propose that they be kept in abeyance, subject to physical and continuous
stylistic analysis. And cited within parentheses, with a caveat. The objects
themselves are not numbered and can be identified only by page references
in Madjidzadeh’s Catalogue (which I employ here). I have not autopsied a
single object from the “Jiroft” corpus, but note that those who accepted the
corpus objects originally did so also from photographs.
In the discussions I list some examples of possible “Jiroft” objects that
have surfaced outside of Iran and appearing for sale in auction and dealer
catalogues. If I am correct about attribution they should be considered as
additions to the corpus. A number of objects offered for sale—both ancient
and forgeries—that may have derived from “Jiroft” are not cited here, as I
remain uncertain about their source.
b. Probable Forgeries
1. bowl, pp. 45–46 (fig. 2), Hero mastering bulls: I suggest that the very poor
execution and inconsistencies of the bulls’ heads—eyes, noses, beards, bod-
ies, and the very insecurely and badly planned and executed wavy water
lines over the bulls (compare pp. 53, 125), the position of the leg tufts jut-
ting from the hooves themselves, and the man’s head condemn this piece.
Compare all the details of the bulls’ physical forms, and especially the flow-
ing water lines above the heads, with those on a conical vessel recovered
from Tarut (recorded there by G. Burkholder; eventually it was sold to a col-
lector), and an unexcavated but genuine vessel in the British Museum (Aruz
in Aruz and Wallenfels 2003, nos. 226, 227); the Tarut vessel must have been
the model for the “Jiroft” object (see also below, no. 3).
See also the bull depicted in Bonhams 11/7/02, no. 196, a conical vessel
most likely from “Jiroft.” The vessel shape is the same as that of pp. 54–
56 (see below, no. 3), and which P. Kohl informs me is not a classic IS
form.
2. pyxis, pp. 47–48 (fig. 3): mermen (not intended) holding something in
each hand apparently feeding bulls—for which compare those represented
on pp. 53 and 125. Also, the uniform gross workmanship, style, and iconog-
raphy force us to reject this piece as an ancient artifact.
cannot exclude the possibility. A stone fragment from Uruk depicts a figure
identified as a male with the same chignon form situated to the right of
an unidentifiable motif—not a snake (as Lindemeyer and Martin 1993, 128,
pl. 61, no. 690). These two examples document the IS chignon hairstyle, but
was the chignon gendered for men or women? (Lindenmeyer and Martin’s
alleged male parallels are not parallels.)
The “Jiroft” bowl has a neat motif-mate; it was purchased (of course “in
good faith”) by the Louvre (A. Benoit, Revue du Louvre, October 2003, 13 ff.,
fig. 1) after being offered for sale in Hotel Drouot, Sept 30, 2002, no. 212
(see also Cultural Heritage News online, 10/29/03). Depicted are two human
heads facing each other within a centrally placed vase that is framed by
snakes. The motif is precisely that of pp. 54–56; both remain unique. The
Louvre human heads have feminine-appearing faces and a rear chignon,
all of which features seem to be correctly executed—far better, than that
of pp. 54–56. Is the latter ancient?—uncertain, not impossible; it is much
better executed than the bowl.
The Louvre curator Benoit (ibid., 14) does cite the parallel with pp. 54–56,
as does Perrot (in Lawler 2003, 975), but in the latter publication A. Caubet
(Benoit’s senior curatorial colleague) is (transparently) indignant, “dis-
mayed by the accusation” that her purchased vessel derived from the “Jiroft”
plunder—even though aware that the only known parallel derives from that
corpus. In fluent museum-speak both curators dissimulate, disingenuously
disassociating the (if genuine, then manifestly plundered) piece they pur-
chased from the vulgar plundering of Jiroft. Benoit notes a different color
of stone, which is not relevant to stone composition, and Cleuziou correctly
observes (2003b, 120) that its shape does not occur at “Jiroft.”5
Perhaps we will also be instructed that other “Jiroft”-like vessels recently
purchased (in good faith) by the Louvre most absolutely did not derive from
the “Jiroft” (or of course, as any curator knows, from any other) plundering
activity (see A. Benoit, Revue du Louvre, June 2003, 86 f., figs. 1, 2; Hotel
Drouot 2/26/03, no. 221); rather—as everyone in the Louvre knows—they
were found by poor peasants when tilling their gardens.
5 If evidence exists, as Benoit and Caubet claim in the Louvre and Lawler articles, that
it was in a private European collection since 1968 (which I go on record as doubting: it was
not mentioned in the Hotel Drouot sales catalogue), it should be presented to archaeologists
(who do not work for the Louvre) for documentation examination.
jiroft and “jiroft-aratta” 503
4. p. 106 (fig. 5): vase with entwined snakes: ambitious but very crowded and
not good workmanship or execution of the heads and ears, and especially
the uncanonical presence together of both round and oval body markings.
Certainly it remains at least suspicious (as Amiet 2000, 96). Perhaps p. 103,
which has a similar scene, is ancient as here the snake heads are more
securely executed.
5. p. 118 (fig. 6): top two conical beakers: both decorated with a whorl
pattern that displays incompetent workmanship—a lack of understand-
ing of the pattern, and insecure execution of each whorl. Such qualities
I do not find among the ancient corpus (viz. de Miroschedi 1977, pl. II-
a; Zarins 1978, nos. 114, 308; Lamberg-Karlovsky 1988, 78, pl. X-b). P. 117
is a vessel with a superbly executed whorl pattern, one matched exactly
except in size by a vessel offered for sale by the Safani Gallery (Ancient
Form, 2004, 11). Two other vessels decorated with whorl patterns are in
the “Jiroft” corpus: Pittman in 2003b, 85, lower left and right. Both are
not made by the same hand; I am more comfortable with the latter
example.
Some other vessels with overall whorl patterns were offered for sale at
Christies, London, 5/15/02, no. 254, and the same venue, 5/13/03, no. 13. The
Louvre purchased a very obvious, stupid forgery of this motif (P. Amiet in
Revue du Louvre 1, 1987, 16, no. 7; Muscarella 2000a, 170, no. 9).
6. p. 126, “weight” (fig. 7): an ambitious work depicting raptor and snakes on
one side, floating scorpion men on the other. No such decorated weights
have been excavated, but the raptor motif is of course well known. Here
I see poor workmanship and inconsistencies and a lack of uniformity in
the execution and positioning of all details of the depicted figures, which
makes me reluctant to accept it without question. Note the scorpion-men’s
thumbs, fingers and arms, ears, eye positions; the whorl constructions and
positions, the tails; also the raptor’s body, talons, beak, and beard (compare
also pp. 15–17, 112–113, nos. 3 and 15 in section “c,” below). For excavated
raptors flanked by snakes see the examples from Nippur and Tarut (Aruz in
Aruz and Wallenfels 2003, no. 234; Kohl 2001, 223, fig. 9.9; Zarins 1978, pl. 68,
no. 159), as well as “Jiroft” (Madjidzadeh 2003, 92–94). For the snakes’ ear
forms, noses, mouths, compare the British Museum vessel (Aruz in Aruz and
Wallenfels 2003, no. 227), and again the Nippur and “Jiroft” vessels. See also
Muscarella 1993, 144, no. 1, fig. 5-a.
504 chapter fourteen
a. a weight offered for sale by Christies, London 12/7/1994, no. 181 (Muscarella
2000a, 169, no. 5; see also n. 58). That it is not ancient is manifest by the
snakes’ teeth—in what looks actually like felines’ heads; the legs and talons
of the raptor are misrepresented; and the execution of the snakes’ incised
body oval decoration are not those of an ancient craftsman.
c. the weight decorated on both sides offered for sale in Hotel Drouot
2/27/03, no. 30, probably from “Jiroft,” and decorated on both sides—a Hero
with a tail mastering snakes, on the other a raptor-snake scene; in both cases
the snakes’ bodies continue onto the handle itself. This is a complex and
very skillfully made work, better than that of p. 126 and the Barakat example
(“d,” below)—but the snake’s body crossing over the raptor’s body puzzles
me, and the Hero has an oval inlay in his hair. Nevertheless, a fragment of a
weight handle from Yahya seems to preserve the body of a snake (it has oval
markings), but not in the same manner as the present example (Lamberg-
Karlovsky 1988, fig. 3-F, pl. IV). It may be genuine, but warrants more study.
d. the weight for sale in the Barakat Gallery catalogue vol. 32 is surely from
the “Jiroft” corpus, primarily because of the lion-raptor combat motifs. It
has very complex and amazingly ambitious decorations on both sides, lion-
snake combats and central figures. But a close reading of all the details
reveals many poor and inconsistent carvings: on the obverse, the central
double-headed monster’s hands, shoulders, ears, necks, feet, and his stand-
ing on air; the flanking snakes’ awkward bodies and positioning, nose mark-
ings, mouths, tail, vertical body markings; and the lions’ crudely executed
not-uniform claw structures, mouths, feet, tail terminals—one is sculpted
as a unit of the animal’s back: compare Madjidzadeh’s pp. 87–90—models
508 chapter fourteen
for this weight? The same execution problems exist for the other side: the
lions’ feet, head, and eyes, and tails; the squatting man holds water flowing
from the addorsed bulls’ heads; his face, body, his leg lacking a foot, and his
kilt should be compared to an example from Tarut (Muscarella in Aruz and
Wallenfels 2003, no. 224-d), one in the British Museum (Aruz in Aruz and
Wallenfels 2003, no. 227, fig. 85), one in Japan (ibid., no. 235); and to section
“c” below, no. 2. This ambitious piece generates reservations, and at present
cannot automatically be accepted as ancient.
e. the weight offered for sale in Hotel Drouot, 2/23/02, no. 393, seems also to
be a member of the “Jiroft” corpus, but whether from this area or elsewhere
I suggest it is not an ancient artifact: examine the raptor’s body decoration
and beak, and the single animal in one talon; the mountain motif (compare
Aruz and Wallenfels 2003, no. 226), the isolated and upside-down moon
crescent (compare ibid., fig. 85). (In the same auction catalogue, again
whether from “Jiroft” or not, nos. 397 and 398 are also not ancient.)
7. p. 139, small idol (fig. 8): is there any reason that allows one to assert it is
ancient? Note the mouth open (caught, in surprise?), the unique hairstyle
(“Jiroft” style), and the presence of scorpions as the figure’s arms.
jiroft and “jiroft-aratta” 509
c. Problem Pieces
As noted, there may be more forgeries in “Jiroft” than listed above. A number
of objects presented in this section may in reality belong in section “b,”
above—or indeed, perhaps in “a.” This category exists because I find it
difficult to react positively to unparalleled or inconsistently executed styles
and iconographies (merely because they were confiscated), and think more
stylistic and technical analysis is required; here too I have changed my mind
on several objects. What is on trial here—the jury is still out—are the very
objects that in fact define the “Jiroft” Style.
1. pp. 11–12, pedestal goblet (fig. 1, lower left): two Heroes master two upside-
down lions: I see the human head positions, their faces, eyes, ears, hair to be
frozen, not alive; equally so the legs, feet, clothing. I am uncomfortable with
the articulation of the felines’ rear leg positions, and they stand on their front
legs; also, the scorpion uncharacteristically lacks its sting. I do not know any
specific parallels.
2. pp. 13–14, beaker (fig. 1, right): motif of a Hero with bulls, while another
is in the mountainous sky holding water or a rainbow, or a jumping rope?
All is seemingly satisfactory until we compare the men’s heads, faces, the
nose mouth curves, hair, clothing, different belts, the articulation of their
shoulders, forearms and fingers, and the leg and feet of the seated Hero—
with excavated male figures in the same kneeling position: Zarins 1978,
pls. 70, 72, no. 546; also Muscarella in Aruz and Wallenfels 2003, no. 224-d;
Aruz in ibid., no. 227. The stylistic and detail differences are apparent, and
probably instructive. The iconography of the lower bulls below a mountain
motif is closely paralleled to a genuine vessel recovered at Tarut (see also
above, “b,” no. 1), here lacking the male figure (Aruz in Aruz and Wallenfels
2003, no. 226, for mountain motif and bulls), but the execution skills are
quite different—when was the one modeled from the other? I also think
the star and crescent are poorly executed (compare Burkholder 1971, pl. IV,
no. 11). And contrast the bulls’ heads, double peaks at the top, water pattern,
and hump to p. 53 in the Catalogue.
Cleuziou (2003b, 122) accepts the authenticity of this vessel, comparing it
favorably to the Tarut vessel (above). I tend toward the negative, but remain
indecisive—which is again an example of the “Jiroft” problem!
510 chapter fourteen
3. pp. 15–17, bowl (fig. 9): Hero mastering scorpion-men; for the latter com-
pare p. 126 (no. 6 in section “b,” above), and no. 15, below. I find the human
and scorpion-men’s heads difficult to judge—but I am more than not com-
pelled to accept it as probably ancient.
4. pp. 18 through 33, a difficult group to work with given their unexcavated
status and stylized elements: six pedestal goblets with grazing animals and
stylized trees. Some of these trees I find too stylized, with outlined leaf
borders: pp. 21–23 (fig. 10), 27–28 (fig. 1, upper left), 30–31 (fig. 11)—although
the trees of pp. 18–20 (fig. 12), 24–26, perhaps 32–33, do seem more natural
with their isolated leaves. Compare the similar trees from Mari (Wilson in
Aruz and Wallenfels 2003, no. 231), and the ancient trees of the Catalogue’s
pp. 40–41. Pp. 18–20 has at least four separate animals, with their young,
in no regular order or position, depicted eating at trees. Pp. 24–26 has two
separate levels of different animal groups; the trees are different. Pp. 32–
33 seem well executed. These three vessels warrant more attention; they
should be examined and compared to the others I list here. They seem
easier to accept as ancient, more confidently executed and the motifs seem
natural—was one or more the model for the others? Note that a number
of plain pedestal vessels were also confiscated (p. 122): how many of the
original were plain and later embellished with scenes—copied from which
vessel?
There are similarly formed excavated tree representations in naturalistic
depictions from Mari (Aruz in Aruz and Wallenfels 2003, no. 231); see also
naturally executed palm trees on pp. 37–30, 118, 127, 128 of the Catalogue;
also Muscarella in Aruz and Wallenfels 2003, nos. 225-a, b.
The Barakat Gallery has offered for sale a conical vessel with the same
basic animal and tree decoration; see also Hotel Drouot 9/30, 2002, no. 213.
5. pp. 34–35, pyxis (fig. 13): the Christmas tree ball effect—compare a sim-
ilar motif on pp. 24–26, the positions and executions of the raptors—one
trampled by a caprid, others hitting the ground—and three different trees,
collectively bother me. But from what I can see of the execution, it looks
good: the work is possibly ancient.
7. pp. 42–43, pyxis: two lions each with a totally unique head and body-form
design. Were they added to a damaged ancient vessel to increase value? Two
conical vessels on pp. 37–39 seem also unique, but their execution looks
fine and I am not compelled to indict them. Compare the executions of
the trees on all these vessels, also the date tree to pp. 110 and 111. A problem
piece.
8. pp. 49–50, pedestal goblet, recumbent ibexes below trees: the ibexes’ eyes,
tails, horn tips, and the trees are poorly executed. A companion decorated
goblet made by a different craftsman is in the Barbier-Mueller Museum
(Amiet in Arts and Culture, fig. 15).
9. pp. 58–59 (fig. 14): back-to-Back lions gored by bulls, a young bushy-tailed
animal and a raptor rests on the back the bulls, another is between them.
The inconsistency of the drawing and execution of the lions’ underbelly and
their claws, and the raptors wings creates doubts.
10. pp. 65–66, conical vessel with a bull-leaping scene (fig. 15): I do not call
attention to this piece here because I doubt that this scene could appear
in central Iran, for the motif existed from the Aegean to the Indus Valley
(see Aruz in Aruz and Wallenfels 2003, 409 and fig. 100-c). But again—
and of course at the very core of the “Jiroft” problem—I am held by the
crudeness of the execution and pose of the figure standing on a bull while
holding a tree at the same time, his slightly raised right foot, and his bald,
speckled head. I find, on the other hand, that the leaping figure situated
between the bull’s horns with his foot wrapped around the tree not difficult
to accept. The bull is tied to the tree, an original and intriguing scene, worthy
of further research. If the scene is anciently executed—not impossible—
then the fact that the human’s contact with the trees has cultural signifi-
cance.
11. pp. 78–90, and Pittman 2003b, 85, upper left: there are eight bowls each
with feline-snake encounters. I do not find it difficult to accept as ancient
the conical vessel on pp. 76–77, but am less certain with the execution
on two other bowls: pp. 78–79 (fig. 16), 84–85, viz. the felines’ heads, their
aberrant claw constructions, the snakes’ heads and ears, the form of the
oval body decorations: compare these with the same forms or scenes on
excavated pieces from Yahya, Mari, Nippur, and Tarut (viz. Kohl 1975, 26,
no. 2, idem 2001, 222, fig. 9.8; Lamberg-Karlovsky 1988, 78, fig. 3-G, pl. IV;
Muscarella in Aruz and Wallenfels 2003, nos. 224-a, 224-e, 234; Aruz in ibid.,
512 chapter fourteen
no. 232 and fig. 87; Godarzi in ibid., no. 242-a; Zarins 1978, nos. 58, 135,
157, 542, 545; Kohl 1979, fig. 5). Compare also the superb snake and lion
combat scene on a fragment in Berlin, (Martin in Aruz and Wallenfels 2003,
no. 233) bearing an Akkadian inscription on its rear—Rimush Lugal Kish
(see above). Some of the excavated examples depict snake combats or snake
and lion combats; on the latter examples note especially the lions’ claw
construction.
The six vessels on pp. 80–81, 82–83, 86, 87–88, 89–90, and Pittman, above,
appear to be ancient (see section “a,” above; and compare the entwined
snake heads of pp. 89–90, 93–94, 99–100 to the Khafajeh vessel cited by
Madjidzadeh 2003, p. 10, n. 8; see in addition Kohl 2001, 215, fig. 9.5). The
fragment on p. 91 is most probably ancient (are the claws unfinished?).
Representations of the same scene and style occur in the antiquities
market: Hotel Drouot 2/13/02 nos. 399, and 405—purchased by the Barbier-
Mueller Museum (Amiet in Arts and Culture, fig. 1) but here the lion bites
into the snake, which to my knowledge is unique (?). A very bad forgery of
the scene is in a Japanese collection (Ishiguro?): The Ancient Orient Museum,
Tokyo 1978, no. 48, Muscarella 2000a, 171, no. 17.
12. pp. 95–96, bowl: raptor and snakes (fig. 17); the placement of the serpents
across the raptor’s wings is I believe unique, but not impossible; and we
should expect to see a beard on the raptor’s chin, and is his beak too elon-
gated. But not insignificant, the snake has round, not oval, body patterns,
not present on excavated snakes, which feature catches our attention.
13. pp. 101–102, bowl: what is the scene? Note the floating snakes’ heads and
an unidentifiable unit—a head, fire? Perhaps genuine, but to be kept in
abeyance.
14. p. 105, pedestal goblet (fig. 18): the bodies of the entwined combating
snakes seem to get lost in the entwining; other examples in the corpus are
better executed—an example from Tarut (Burkholder 1971, pl. VII, no. 20).
Also how do we explain the line that divides the decoration on each snake’s
body? A problem piece to be further investigated.
for sale (Amiet in Arts and Culture, fig. 11; Christie’s, London, 5/15/02, no. 265
have the very same swimming scorpion-man motif), but all were made by
different craftsmen.
16. p. 120, lower right, vase: the upper row of attempted guilloches is very
badly executed in form and spacing, and differ in each case, and is not
paralleled from the excavated corpus.
17. pp. 123 (fig. 20), 124: two weights in the form of openwork entwined
snakes. Both are entirely different in shape, body decoration, and sculptural
symmetry; no. 124 seems less finished. Models for the motif exist: a weight
from Soch in Uzbekistan, and an example in the Louvre (Muscarella 1993,
144, fig. 4, 149, no. 9; Kohl 2001, 227, fig. 9.14). Are both the “Jiroft” weights
ancient? Or is one ancient and the other a modern copy?
These two snake weights call our attention to a handle offered for sale in
Hotel Drouot 6/26/03, no. 113, a handle sculpted in open work that depicts
a male figure sculpted in the round from his kilt to his head. No feet are
depicted, but below the kilt there is an unparalleled unit of three triangles
at the front and a unit of four curved forms at the rear, so he cannot be
said to be kneeling conventionally; he masters two snakes, one of which is
connected to him by a strut. Of interest is that the male’s eyes seem to depict
a blind person—consciously, or unintended? Perhaps this weight doesn’t
belong to the “Jiroft” corpus. It is a complex, unique, well-made object—but
its uniqueness demands more investigation before secure acceptance.
18. p. 133, double-headed raptor plaque (fig. 21): aside from not first-rate
workmanship and five and six claws terminating in different lengths, it is
an unparalleled work—why accept it unconditionally?
19. p. 135, scorpion plaque: doubts raise themselves with regard to this plaque
unparalleled elsewhere; and the excellent execution of the whole, especially
the face, confuses me. Why are its “wings,” body and tail forms and their
decoration different than the other scorpion plaque on p. 136 (fig. 22)? I am
not so secure with this scorpion plaque either but cannot outright condemn
either one.
20. p. 147, two odd, seemingly unfinished stone figurines, one a human, one
an animal head (?). Amiet (2002, 96) considered these to be forgeries. He
may be correct as they are quite formless, but who knows. In any event, they
can have no archaeological value.
514 chapter fourteen
21. p. 166, two resting felines, both said to be made from lapis lazuli. They
are close but not the same and quite simple, and it is difficult to form a
conclusion about their ages. Note the intentionally scarred right eye of one
lion, and the different front feet constructions, one open, the other closed.
The figures on p. 167, however, do not cause concern.
Conclusion
Acknowledgements
I want to thank Philip Kohl, Jean Evans, and Paul Collins for a close reading
of a manuscript of this paper and giving me intelligent insights and opinions.
jiroft and “jiroft-aratta” 521
Bibliography
Background
In 714 bc, the Assyrian king Sargon II (721–705bc) documented his military
campaign east across the Zagros Mountains in western Iran, then north
into Urartian territory. His record survives on a now incomplete terracotta
tablet (37.5 by 24.5cm.) purchased by the Louvre Museum in 1910 from the
antiquities dealer J.E. Gejou. The tablet was published in a superb French
translation by F. Thureau-Dangin in 1912. Years later, a missing fragment of
the tablet that had been excavated at Assur in northern Mesopotamia by
German archaeologists was recognized in Berlin (lines 195–236; nine lines of
the original tablet are still missing). This remarkable recovery and German
scholarly research ascertained that the Louvre tablet had in fact derived
from Assur (Meissner 1922). Most probably a workman had plundered the
tablet at Assur and sold it locally from where it was smuggled abroad to Paris.
The tablet was written as a report taking the form of a letter to the God Assur;
and its date, 714bc, is known from the eponymy name recorded on it. It is
one of the three known Assyrian Gottesbriefes, with the other two belonging
to Shalmaneser IV and Esarhaddon (Zaccagnini 1981: 264, note 3). Aside
from Thureau-Dangin’s translation, there is another by D.D. Luckenbill in
English (Ancient Records of Assyria and Babylon II, London 1989 73–99),
and a German translation by W. Mayer (MDOG 115, 1983: 65–132). For an
historical, ideological, and literary analysis of the text see Zaccagnini 1981.
In the 1940’s, the letter engaged the attention of a few scholars involved in
the history and culture of Assyria, Iran and Urartu during the early first mil-
lennium bc (Rigg 1942; Wright 1943). Renewed interest and study began in
the 1970’s and 1980’s, generated both by extensive archaeological excavations
* This chapter originally appeared as “Sargon II’s Eighth Campaign: An Introduction and
Overview,” in The Eighth Campaign of Sargon II: Historical, Geographical, Literary, and Ideo-
logical Aspects, trans. Samad Elliyoun and Oscar White Muscarella (The Hasanlu Translation
Project, Tehern, 2012), 5–9.
524 chapter fifteen
Sargon’s journey was long, lasting for months. Most scholars situate his
Urartian campaign, his terminal target, in the Lake Urmia region, com-
mencing either at its western or eastern shores (see Zimansky 1990:19, and
note 78); some others believe (with Thureau-Dangin) that Sargon continued
his campaign from Lake Urmia to Lake Van, the Urartian heartland. How-
ever, identifications of site and region locations presented by scholars vary,
sometimes considerably, all equally based on modern archaeological and
survey data. Scholars who have discussed these issues extensively include
Mayer, Levine, Muscarella, Salvini, Kroll, Kleiss, Burney, Zimansky, Reade,
Çilingiroǧlu, and Liebig. Some basically agree on the identification of the
geographical locations, others strongly disagree, positing distinctly differ-
ent geographical locations for places in the text (viz. Levine 1977: 141–142,
145, map; Muscarella 1986, and maps figs. 2, 3; Piotrovskii 1959: 105; Zimansky
1990: 14–16: who “tentatively suggest(s)” his geographical attributions). Spe-
cific archaeologically known Urartian sites have been identified with their
sargon ii’s 8th campaign: an introduction and overview 525
Before his venture against Urartu in the region of Lake Urmia, Sargon had
campaigned extensively south of the lake. Scholarly interpretations regard-
ing the precise directions he undertook here, how far and where in west
ern Iran, east and north, and the approximate locations of the many
526 chapter fifteen
A. Short Route
1. Encirclement of Lake Urmia from the East: The scholars arguing this
route include: Piotrovskii 1959: 105, map; Burney 1972: 138–140, 155;
Van Loon 1975: 206–207; Çilingiroǧlu 1976–1977; V. Selim and O. Belli
(Anadolu Araşstirmalari 1976–1977): 89–90; W. Mayer 1980: 22–23;
Zimansky 1990; Liebig 1989: 33; Reade 1994: 186; 1995: 38: but to him,
the campaign extended as far as the Caspian Sea; Kroll 1997: 206.
2. No encirclement of Lake Urmia: This view suggests that the campaign
was limited to its western shore area then returned south: Rigg 1942:
132, 135, 138. Rigg was the first scholar to suggest this route. I originally
advanced this view based on observations at Lake Urmia, first in 1960
when travelling down its eastern shore, and reinforced by the discovery
of the Qalatgah inscription (Muscarella 1971:49; 1986. See also Salvini
1984: 46–51; 1995: 28–29, 92). Zimansky (1990: 5, note 22) states that I
“even claim[ed]” I was the “first” to make this observation. Indeed, I did
suggest this, not then knowing the ideas of Rigg 1942, along with those
of others, including Levine 1977. Here (page 150) Levine gives a buried
reference to my 1971 revision of Sargon’s route, which he dismisses by
describing it as tentative.
B. Long Route
1. Reaching Lake Van: Thureau-Dangin 1912; Wright 1943; Van Loon 1975:
205–207; 1987: 260; Çilingiroǧlu 1976/77: 264–265; Kleiss 1969–1970: 131–
132, Abb. 3.
528 chapter fifteen
Bibliography
Burney, C. (& Lang, D.M.), 1972, The Peoples of the Hills, New York. Çilingiroǧlu,
A., 1976–1977, “The Eighth Campaign of Sargon II,” Jahrbuch für keinasiatische
Forschung (Anadolu Araştirmalari) 4–5: 252–269.
Fales, F.M., 2003, “Evidence for West-East Contacts in the 8th Century bc: The Bukan
Stele,” in Continuity of Empire (?) Assyria, Media, Persia, G.F. Lanfranchi et al eds.,
Padova. Pp. 131–147.
Kleiss, W., 1969/70. “Zur Ausbreitung Urartus nach Osten,” AMI 19/20, pp. 125–
136.
———, 1977, “Alte Wege in West-Iran” AMI 10, pp. 137–151.
Kroll, S., 1997, Review of Salvini 1995, in Die Welt des Orients XVIII, pp. 203–208.
Lanfranchi, G.B., 2003, “The Assyrian Expansion in the Zagros and the Local Ruling
Elites,” in Continuity of Empire (?) Assyria, Media, Persia, G.F. Lanfranchi et al eds.,
Padova. Pp. 79–118.
Levine, L., 1977, “Sargon’s Eighth Campaign,” in Mountains and Lowlands: Essays in
the Archaeology of Greater Mesopotamia, Bibliotheca Mesopotamia Vol. 7, pp. 135–
151.
Liebig, M., 1989, “Zur Lage einiger im Bericht über den 8. Feldzug Sagons II. von
Assyrien gennanter Gebiete,” ZfA 81, 1, pp. 31–36.
Mayer, W., 1980, “Sargons Feldzug gegen Urartu—714 v. Chr.,” MDOG 112, pp. 13–
32.
Meissner, B., 1922, “Die Erorberung der Stadt Ulhu and Sargons 8. Feldzug,” ZA,
pp. 113–122.
Muscarella, O.W., 1971, “Qalatgah: An Urartian Site in Northwestern Iran,” Expedition
13, 3/4, pp. 44–49.
———, 1980, The Catalogue of Ivories from Hasanlu, Iran, The University Museum,
Philadelphia.
———, 1986, “The Location of Ulhu and Uiše in Sargon II’s Eighth Campaign,” JFA 13,
pp. 465–475.
———, 2010, “Hasanlu and Urartu,” in Bianili-Urartu, J. Reade and S. Kroll eds.,
Munich. Pp. 265–279.
Pecorella, P. & Salvini, M., 1984, Tra lo Zagros e L’Urmia, Rome.
Piotrovskii, B.B., 1959, Vanstoe Tsarstvo, Moscow.
Reade, J., 1995, “Iran in the Neo-Assyrian Period,” in Neo-Assyrian Geography, M. Liv-
erani ed., Rome. Pp. 31–42.
———, 1994, “Campaigning Around Musasir,” in Anatolian Iron Ages 3, A. Çilingiroǧlu
and D. French eds., Ankara. Pp. 185–188.
Rigg, H.A., 1942, “Sargon’s Military Campaign,” JAOS 62, pp. 130–138.
Salvini, M., 1984, in P. Pecorella and M. Salvini, Tra lo Zagros e L’Urmia, Rome.
Salvini, M., 1995, Geschichte und Kultur der Urartäer, Darmstadt.
Thureau-Dangin, F., 1912, Une Relation de la Huitième Campagne de Sargon, Paris.
van Loon, M., 1975, “The inscription of Ishpuini and Menua at Qalatgah, Iran,” JNES
34, 3, pp. 201–207.
———, 1987, “Review of Pecorella and Salvini 1984,” BibOr XLIV, 1/2: 252–262.
Wright, E.M., 1943, “The Eighth Campaign of Sargon II of Assyria (714BC),” JNES 2,
pp. 175–186.
530 chapter fifteen
Zaccagnini, C., 1981, “An Urartian Royal Inscription in the Report of Sargon’s Eighth
Campaign,” in Assyrian Royal Inscriptions: New Horizons, Orientis Antiqui XVII,
F.M. Fales, ed. Pp. 259–295.
Zimansky, P., 1985, Ecology and Empire: The Structure of the Uraratian State, Chicago.
———, 1990, “Urartian Geography and Sargon’s Eighth Campaign,” JNES 49, no. 1,
pp. 1–21.
Section Two
Anatolia
chapter sixteen
Three Great Early Tumuli. The Gordion Excavations, Final Reports, Volume I.
(1981) R.S. Young, with contributions by K. DeVries, E.L. Kohler, J.F.
McClellan, M.J. Mellink, and G.K. Sams. E.L. Kohler, Editor. University
Museum, Philadelphia. 310 pp., with Appendices, illustrated with figures and
plates. $75.
Long awaited, and after several missed deadlines (the last being Decem-
ber, 1981) the publication of the architecture and contents of the three
largest and richest tumuli excavated by Rodney S. Young at Gordion has
now appeared (July, 1982). It will be immediately noted here that the vol-
ume is one of the most important and significant publications of excavated
material from a major Near Eastern site to have been published in years. The
vast amount of material systematically presented in numerous photographs,
plans and drawings, and in interpretative textual commentaries, revealed in
a detailed manner never before attempted, will surely generate among stu-
dents a renewed interest in the culture and history of the Phrygians. Young
died (October 25, 1974) before he was able to organize for publication his
manuscript with photographs, write the essays and conclusions he planned,
and express all the ideas he kept in his fertile mind. A committee was there-
fore organized in 1975 by F. Rainey, then Director of the University Museum
to collate the material in hand, to supply technical data and to contribute
and solicit appropriate essays. The committee was also charged with orga-
nizing the topics and schedules for future publications of the 17 Gordion
campaigns (1950–1973); this work is now in progress and soon (one cau-
tiously hopes) we shall see a volume on the “lesser” tumuli (E.L. Kohler) as
well as a detailed study of Phrygian pottery (G.K. Sams).
In this first volume the committee has accomplished the difficult task
of gathering and organizing the scattered and unpaginated notes of Young,
* This article originally appeared as “King Midas’ Tumulus at Gordion,” The Quarterly
printing them in the sequence he desired, and (so we are informed) leav-
ing them intact and unchanged with proper concern to preserve his ideas
and style. Aside from assembling the necessary photographs, plans, and
drawings and adding appendices on technical subjects, some of the com-
mittee members, along with two invited scholars (gks, jfmc), have writ-
ten entries and essays on various subjects, including the final Conclusions
chapter. Thus, the volume is the work of several scholars, indeed primar-
ily of Young, but with important contributions by the others, all intimately
familiar both with the excavations at Gordion and with the mind of its
excavator.
Following the wishes of Young, the three tumuli (p, mm, w) are published
in the order of their excavation, not in the chronological order perceived
by the committee. Narration of the excavation problems and strategy, tomb
description, and a catalogue of most of the contents of Tumulus p and
mm are by Young, with further discussion of the pottery by gks, the wood
remains and bronze quadriga by elk, the iron by jfmc, and the Egyptian
Blue and paste by mjm. At the time of Young’s death he had completed
a catalogue of the finds from Tumulus w, but not a description of the
excavation of the tomb and its contents in situ. This latter work has now been
accomplished by kdv, based on Young’s field notes and published writings;
there is also a discussion of the pottery of w by gks. A separate chapter
(IV Commentary) includes discussion of selected groups of objects found
in three tumuli, bull cauldrons (kdv), omphaloi (mjm), belts (elk), pottery
(cks), wood (elk), and bronzes by Young. A final chapter, Conclusions,
is written by mjm. Analyses of wood, food, bronzes, and textiles from the
three tumuli, as well as the latest C-14 dates and a discussion of Phrygian
inscriptions are given in the appendices.
Given the situation that there is a total of six authors in the main body
of the text, it would have been helpful to the reader if each non-Young
entry had been identified by initials alongside the section heading, rather
than with an asterisk matched at the bottom of the page with the author’s
name. Because these entries are interspersed within Young’s in the Tumulus
p and mm chapters, too often I discovered that I was unaware that a shift
had occurred from one author to another (for example pages 219 to 233).
In the appendices the names of the contributors are properly supplied at
the heading, as are also the dates when their manuscripts were submitted
(from 1959 to 1979). But dates are not supplied for the time of completion or
submission either of Young’s or the committee’s contributions.
This multiplicity of contributors has also resulted in the inevitable sit-
uation that there is at times a lack of uniformity in discussions of the
near east invited review: king midas’ tumulus at gordion 535
continue to clarify this very important issue and to determine if other Gor-
dion “bronzes” are actually brass?
A few more minor contradictions could also have been noted. On page 156
Young refers to 37 fibulae on the Tumulus mm bier, of which ten had disin-
tegrated, while on page 169 the numbers 32 and 5 are given respectively. On
page 114 Young mentions ten studded belts in mm, where there are in fact
nine (147ff.), the tenth being of a different type and all bronze. And while
kdv says (221) correctly to my mind—that there is no Urartian material at
Gordion, mjm (268) speaks, with no evidence given, of Urartian influence
there (see also Mellink 1979, 256).
Although it is maintained (xxix) that an “updating of the notes” of Young
was supplied, the documentation presented does not support this claim,
and one is disappointed by the lack of both relevant bibliographical ref-
erences and discussions of comparanda, not only in Young’s entries but in
some of the contributors’ as well. For example, there is not a single com-
ment concerning the techniques of tumuli construction elsewhere in Ana-
tolia, nor in the Near East and Europe. Indeed, a separate essay devoted to
this subject in a volume on Phrygian tumuli is not too much to expect, all
the more so as Young basically neglected this aspect of Phrygian archae-
ology. That a number of excavations have recorded the various methods
of tumulus filling and the placement of the underlying tomb off center is
ignored (e.g. on Cyprus, at Sardis, Kerkeneş Daǧ, Ankara, and in Iran at Sé
Girdan and Takht-i-Suleiman: for a summary see MMA Jour 2, 1969, 22 f.,
MMA Jour 4, 1971, 25f.). In the wood catalogue for Tumulus p we miss a
discussion of the ivory carvings from the city mound at Gordion, as well
as reference to those at Nimrud, Hasanlu, and elsewhere, with pertinent
comments on the relationship of ivory and wood carvings. Nor is there a
mention of the preservation of wood sculpture at other Near Eastern sites—
at Ebla, Nimrud, Hasanlu, Samos, Karmir Blur, the Altai—nor to the pos-
sible reasons for their preservation in each instance. Formal parallels are
given for one wood sculpture (54f.) “related to Iranian work,” but the cited
examples are all Achaemenian gold, and about 200 years later in date (also
note that although attributed to Hamadan, they are not excavated pieces).
That Egyptian Blue artifacts (268) and animals headed vessels (121 ff.) were
excavated at Nimrud and Hasanlu, that vessel attachments in the form of
birds occur at Hasanlu, is nowhere mentioned. Further, that parasols are
represented, aside from Ehnali, on Assyrian and Achaemenian reliefs in
association with royalty, that fans are commonplace in Near Eastern rep-
resentations of banquet scenes (74ff., n. 145, 146), and that there is recent
work published on Phrygian fibulae in the West (K. Kilian, 1975, 151 ff., and
near east invited review: king midas’ tumulus at gordion 537
ca. 300 meters in diameter). After the drilling period it took a further two
full months of digging, half of this time in 24-hour-a-day shifts, involving
first opening a trench, and then digging a tunnel for a combined length of
135.20 meters, to reach the tomb. It was discovered to be surrounded imme-
diately by a revetment log wall and a square stone wall three meters in height
outside that, all covered with many tons of stone rubble. A gabled structure
made of neatly cut pine wood timbers, the tomb was intact and uncollapsed,
in itself an historic discovery. Inasmuch as there were no windows or doors,
an opening was cut through the west wall, and upon entering one found one-
self directly before the bier holding its occupant. For the record, Machteld
Mellink was given the honor of being the first non-Phrygian to enter the
tomb (although I am informed another staff member improperly made that
claim).
Young’s text is, typically, clear and concise, yet at times difficult to follow
primarily because of the problems inherent in describing such a complex
tomb structure. And often there is inadequate cross-referencing and coor-
dination between the text and an appropriate figure or plan (presumably all
references to figures were added after 1974). Thus, for example, figures 61,
63, and 64 better illustrate mortising techniques than do the cited figures 54
and 59 (88). The important figures 52–62 (executed by C.K. Williams ii)
require internal labelling in order to identify details mentioned in the text—
e.g. the arrows in figure 62—and they are not sufficiently cited in the text.
(Figure 63 is a masterful isometric drawing of the whole tomb complex
by C.K. Williams ii, but we selfishly miss an isometric drawing of internal
structure of the tomb itself.) Certain features of the initial building of the
tumulus described by Young (80), that there was “a successive line of stones”
laid to determine the center and that “there was a laying down of a ring of
filling [clay] around a predetermined perimeter for the tumulus”, are not
indicated in the plans and sections (figures 50–52). Moreover, the section
drawing of figure 52 is merely a sketch, not an empirical drawing; it was
copied in Philadelphia from Young’s notebook. No sections are supplied for
Tumulus p and w. Further, a section drawing of Tumulus p (figure 3), shows
a pit although it is not mentioned in the text, and the description of the
excavation suggests that such a feature could not have been noted even if
present (note that in AJA 61, 1957, 325f., there is mention of a pit). Tumulus
w had a tomb pit, but here too the section drawing does not show it (fig-
ure 114).
On page 263 Mellink casually notes that Young’s “description of the enter-
prise [the excavation of Tumulus mm] allows the reader to read between the
lines that the discovery … was a rare confrontation with an intact part of the
540 chapter sixteen
past …” “Between the lines” presents the problem precisely. For although
the description takes the reader step by step in the long and arduous—and
courageous—journey that led to the discovery of the tomb, one is aware that
it consciously lacks any hint that the excavator and his team may have expe-
rienced personal excitement. While the description is correct, it is written
with an intent to avoid any expression or pride and joy in a major discov-
ery. “A rare confrontation with … the past.” Alas, it is an attitude that also
obtained during the excavation itself.
The most important conclusions expressed in the volume concern the
relative and absolute chronology of the tumuli, in particular the absolute
dates assigned to Tumulus w and mm. These conclusions are issues of con-
siderable importance and will keep scholars’ pens occupied for some time.
Indeed, they crucially affect not only the internal history of Gordion itself,
but of Phrygia at large, and equally the dating of the Phrygian alphabet, the
dates for the transmission of Phrygian material abroad, and, not least, the
nature of archaeological interpretation. With these issues in mind, the fol-
lowing comments are offered to initiate discussion.
Young expressed the belief in all his previous writing, and never changed
his mind, that the three tumuli, along with Tumulus iii, excavated in 1900
by the Körte brothers, were pre-destruction (ca. 696 bc) in date. To him,
w was the earliest, and he dated it to the end of the 9th century bc or
“slightly later” (199). mm was next in date, about six or seven decades later,
ca. 725/720–717 bc, or the “latter half of the eighth century” (102, 109, 232),
followed by iii and p in time: thus, the sequence w, mm, iii, p, with a range
from ca. 800 to 700bc or a full century. The contributors hold a different
view, with regard to both the relative and the absolute dates. While they
all share Young’s view that w is the earliest, they believe that mm is the
latest in the sequence, and it seems that some believe further that mm was
not built until after the Kimmerian destruction, which we must presume
dates it ca. 696/5 bc. In this sequence we have w, iii/p, mm (176, 198 f., 215 f.,
233ff., 236ff., 239, 254, 263f., 266, 269ff.); the relative sequence of iii and p
is unclear, and they are considered to be too close in time to separate them
chronologically.
Turning first to Tumulus w, we note that kdv (199) dates it “decades if not
generations before the catastrophe,” which I interpret to mean (assuming
that “if not” means “even”) that w is either ca. 760 or 790 bc (accepting
the conventional 30 years for a generation). Equally ambiguous is Mellink’s
suggested chronology, which implicitly reveals uncertainty. On the one hand
(264, 270) she seems to accept a ca. 800bc date, in part based on the dating
of a bronze vessel, w9, that Young originally associated with one excavated
near east invited review: king midas’ tumulus at gordion 541
through its bronzes than through its pottery” (215). Parallels in form and
type among the finds from tumuli include (and I give them at length to
document my position; the letters and numbers stand for the tumulus and
catalogued object): bronze spouted vessels (w5, p72–78, iii, mm14); pottery
spouted vessels with stepped interior spouts (w61, p75); geometric painted
vessels (w61 [is this not the most developed of all?]; p49–53, iii); painted
panel vessels (p55–57, iii: on page 49 ks says the panels of an example from
the city mound [pl. 96, b] are small and relates them to iii; on page 270
mjm implies it has larger panels than iii); monochrome spouted vessels
(w63, p72–78, iii); leather belts (w25, 26, iii, mm170–179); bronze belts (p34–
36, mm180); type xii, 7a fibulae (w iii, also city mound destruction level);
bull cauldrons (wi, 2, mmi, 12, 13); oinochoe base with a pin embedded in
wax (w6, mm27); animal headed finials (w8, p36); bronze ladles (w, iii, p,
mm); omphaloi (w10, p11; and compare w11, p12, mm125–130; w12–14, p14–
28, mm131–167; compare also the plain bowls w23, p27–30, mm168–169);
pottery amphorae (w65–72, p91–104, mm372–377); wood screens (w80, p151,
mm378, 379); wood utensils (w75–79, p121–154, mm380–387). w and mm
share ten of the above parallels, p and mm share seven, p and w share
thirteen. The main differences between w and mm are the lack of type xii, 7a
fibulae and painted vessels in mm, and the lack of dinoi in w—all negative
evidence.
External evidence does not help in the dating of w. Although the often
cited bowl from Assur bearing the name Assurtaklak, the same name as
the Assyrian eponym of 806bc, has been brought forth in discussions of
w’s chronology (198, 234, 270), it may not in fact be so employed. In the
first place, as Mellink has noted, its shape is not the same as w9 to which
it has been compared. In the second place, it is quite possible that the
vessel is almost a century later in date than 806 bc. Tomb 30 at Assur also
contained a fibula, the date of which cannot be earlier than the late 8th
century and could be 7th century, and a lamp of a type excavated in a late
8th early—7th century bc burial at Baba Jan in Iran (Iran vii, 1969, 125,
fig. 7:1, 2). These artifacts date the deposition of the objects in Tomb 30 a
century or more after 806bc: which suggests either that the bowl was an
heirloom (but visible and in use for a long time), or that another man named
Assurtaklak inscribed it with his name (see also P.H.G. Smith in BaBesch 56,
1981, 10).
Because Urpallu, the king of Tyana, is represented at Ivriz wearing Phry-
gian clothing and a Phrygian fibula of type xii, 9, the exact date when the
relief was carved is clearly of great importance for the dating of Phrygian
fibulae, type xii, 9 in particular, and by extension back in time, xii, 7a.
near east invited review: king midas’ tumulus at gordion 543
Akurgal (Die Kunst Anatoliens, 118) dated the relief after 730bc (see also
Orthmann in PKG 1975, 432, “um 730 v. Chr.”). Young (245, n. 115) placed the
relief earlier “in the years around 738bc, more probably before than after”
(a position taken by me in 1967, 20). This chronology is based on a reference
to Urpallu in the Assyrian annals of Tiglath-Pileser iii; but other, albeit
complex, evidence for the date of Urpallu now exists. A tablet excavated
at Nimrud written by an unnamed Assyrian king mentions both Midas
(Mita) and Urpallu as contemporaries (J.N. Postgate, Iraq xxxv, 1973, 21 ff.).
Postgate firmly dated the letter to the time of Sargon, to 709bc specifically
(agreeing with Saggs, Iraq xx, 1958, 205ff., who first published the letter).
Mellink (371; see also 1979, 249ff.) is inclined to date the letter to the time
of Tiglath-Pileser iii, to ca. 735 bc (following H. ten Cate 1967, 122), and she
believes that the relief at Ivriz was carved at or close to the same time.
Inasmuch as the dating of the relief would supply independent evidence
about the time when type xii, 9 fibulae were in use, and if, as plausibly
argued by kdv and Young, this type is a development of type xii, 7, itself
a development of type xii, 7a, then the latter two types would have been
created sometime earlier than the relief. Thus, if the Ivriz relief was carved
in the 730’s, we would have a date in that decade for the floruit of all three
types, and equally a projected date back in time of some undetermined years
for the first appearance of types xii, 7 and 7a. But if the Assyrian letter, and
concomitantly the relief, is from the late Sargonid period, from ca. 709bc,
we would then have a contemporary date for the three types in the previous
decade, and probably in the 720’s, even the 730’s, for types xii, 7 and 7a.
In short, the Urpallu relief is of potential, but at present limited, value for
determining the earliest chronology of the three fibulae types discussed, and
consequently for the objective dating of the xii, 7a fibulae found in quantity
in Tumulus w.
I do not think I am wrong to suggest that the attempts to date Tumulus w
close to 800 bc are ultimately based on an underlying assumption (but one
never mentioned) that the archaeological record indicates that Gordion was
an established city by this time. In AJA 1966, 273 ff., and in AJA 1968, 239 f.,
Young presented evidence for the existence beneath the destroyed level at
Gordion of an earlier level with seven phases. One may speculate about
the incipience, duration, and chronology of this level and its phases, very
little of which was actually exposed, but there is nothing in the published
record that compels acceptance of a pre-8th century bc date for the time of
the construction. Nor does the published evidence from the deep sounding
(AJA 1966, 276ff.), which revealed no architecture or discrete levels as such,
compel us to accept the existence of a 9th century settlement at the site.
544 chapter sixteen
Nevertheless, I am aware that all the data from the sounding have not
yet been published and that disinterested scholarship requires an open
mind on this fundamental problem. Therefore, in the context of the present
discussion one may only conclude (patiently awaiting enlightenment from
future publications) that even if there was a Phrygian (and n.b., not a pre- or
proto-Phrygian) settlement at Gordion ca. 800bc, it cannot be assumed that
Tumulus w dates to that time; and if there was no such settlement before the
8th century bc (whatever incipient date), then Tumulus w must have been
built subsequent to that time.
Evidence from other Phrygian sites in the east seem to be less ambigu-
ous on the matter of chronology and yields no evidence for pre-8th century
occupation. At Boǧazköy (Bossert, MDOG 89, 1957, 58 ff.; MDOG 94, 1963,
53ff., 68ff.; Opificius MDOG 95, 1965, 81ff.) excavations indicate an 8th cen-
tury date for the earliest post-Hittite occupation there. And I suggest that
Opificius’ attempt (ibid., 87) to push the Phrygian pottery sequence back to
ca. 800 bc, based on an incorrect 9th century date assigned to a vessel from
Carchemish, and an equally incorrect date for the fibulae there, is based on
speculation, not empirically determined evidence (see also G.K. Sams, AS
xxiv, 1974, 181). The same situation with regard to the post-Hittite settlement
and chronology obtains equally for Alishar (Akurgal 1955, 37, 155; also K. Bit-
tel, Hallusha, NY, 1970, 138). Indeed, the earliest post-Hittite level at Alishar,
4c, contained both Near Eastern fibulae, of late 8th/7th century types, as
well as Phrygian examples, surely indications of an 8th century occupation,
not earlier.
Except for the possible, and still to be demonstrated, exception of Gor-
dion, no Phrygian site can be convincingly dated before the 8th century bc.
In this respect Akurgal’s (1955, 24, 112ff.; 1965, 468 ff., 472) short chronology
for Phrygian culture (but not his suggested low date of 725 bc for the found-
ing of Gordion) still holds up; Phrygia had a “glorieuse mai très courte vie.”
In addition to archaeological—internal evidence—it is against this back-
ground that Tumulus w should also be studied and its chronology evaluated.
We do not know whether the Phrygians constructed tumuli burials imme-
diately after they built their earliest cities, but on the evidence from Tumu-
lus w as interpreted here the earliest Phrygian tumulus recorded up to the
present time does not pre-date the mid-8th century bc.
Until the very last page (272) only one of the contributors offers his
views on the absolute date for the construction of mm. gks (176) sees it as
“roughly coeval with the city’s destruction,” but he also cautions “against
too low a dating” and suggests that “mm is still close to the date of p and
the time of the early city’s end.” I read this to mean that mm was sealed a
near east invited review: king midas’ tumulus at gordion 545
relatively short time before the Kimmerian raid, an opinion I believed all
the contributors held. It therefore came as a nonsequitur that one/some of
the contributors (see page xxix for confusion concerning who is responsible
for the Conclusions attributed to mjm) consider mm to have been built and
sealed after the destruction.
Aside from being a nonsequitur within the context of the volume itself,
the announcement that mm is a post-destruction construction is more sig-
nificantly an archaeological nonsequitur, one that bends the evidence avail-
able and takes it beyond its natural boundaries. However late the tomb
may be, cognizance of and fidelity to the full archaeological record compels
comprehension of a context within predestruction Gordion. The contrary
statement offered with no set of detailed evidence, demands of others a pre-
sentation of proof that it does not have, and that cannot be accomplished.
The impulse behind the chronology as presented is informed by a fiat, and
it floats in the air (272): “Preliminary analysis supports the interpretation of
the occupant of mm as Midas.” Which announcement automatically raises
the question, what analysis, and why Midas? Indeed, if the buried man is
Midas it follows that the tomb is post-destruction, inasmuch as it is accepted
in the text that Midas killed himself—as reported by ancient historians—at
the time of the Kimmerian attack; the tomb therefore will have been built
after the suicide/Kimmerian attack, not, obviously, before. However, if one
takes a disinterested position with regard to the identity of the entombed
individual, Midas ceases to be a viable candidate. For the chronology of
the tomb alone, determined by both the internal evidence and historical
and economic analyses, will inform us whether Midas is buried therein, not
the assumption that because the deceased is Midas we therefore know the
chronology. (Of course, if the historians are in error and Midas died some-
time before the destruction, Midas could be buried in mm: but since we
cannot know this, we cannot discuss it.)
Mellink herself (270; assuming her to be the sole author of the Conclu-
sions) provides some of the pertinent archaeological evidence for dating
mm in the late 8th century bc (evidence discussed by others, including
Young): the situlae and type xii, 7 fibulae, both represented in Sargon’s city
at Khorsabad, and the xii, 9 fibulae, represented at Ivriz. The city mound
parallels with Tumulus mm might suggest (with gks) that the latter was
built close to the time of the destruction, but they do not necessarily indi-
cate a post-destruction date. More study is required for a full understand-
ing of the chronological relationship between certain objects found on the
city mound and in the turnuli, But one surely cannot cite only those from
the former that parallel finds in mm while ignoring the occurrences of
546 chapter sixteen
geometric painted pottery, amphorae, and type xii, 7a fibulae found both in
w and on the city mound.
Then there is also to be considered, and not parenthetically, the matter of
the great quantity of the deposited material, including wood and cloth, that
could not have survived the fire recorded almost throughout the remains
on the city mound. Many questions are begged when it is claimed that
the material could have derived from other Phrygian cities because Midas’
“citadels were not all looted at the time when Gordion suffered destruction.”
Not unexpectedly, no specific cities are mentioned, for the truth or error of
this claim defies precise analysis. Most, probably all, Phrygian cities, west
and east, are archaeologically attested to have been destroyed, although it
is not possible to know whether one was destroyed at the same time as, or
much later (months, years?) or earlier than, another. No other Phrygian site,
not even the Ankara tumuli, has yielded evidence to indicate that anyone, or
all sharing the responsibility, could have sent hundreds of precious objects
to Gordion after destruction, let alone even before.
It is further suggested that the actual work involved in constructing the
tumulus and tomb, if not provisioning it, was locally accomplished: “The
effect of the Kimmerian raids was not such as to destroy the city of Gordion
or its dynasty. Tumuli continued to be erected in the 7th century, and the
rebuilding of the citadel started gradually under the protection of the mud-
brick fortification of Küçük Hüyük.” But the citadel was destroyed and aban-
doned, with no clear evidence of reoccupation, and the alleged “rebuilding”
is by no means certain or fully understood, nor did it gradually or other-
wise reach fruition. Further, while the Küçük Hüyük (the excavation and
history of which is not discussed) was built sometime after the destruction
of the city mound, we do not know how much time elapsed before work
commenced, or whether it was solely a fort or a royal habitation. And here
too questions are begged, questions that are neither petulant nor gratuitous.
Did the alleged survivors at Gordion include architects and engineers who
were capable of planning and supervising mm’s monumental construction,
the likes of which did not exist earlier anywhere in Anatolia, and only later
in Lydia? Where did the work force come from, the hundreds of men and
women required to build the tomb and to carry the countless tons of rubble
stone and clay filling, individuals who were free of other tasks, including pro-
tecting their (rebuilt?) homes from further Kimmerian raids? What secure
authority had the compulsion and the finances to both assume control of
the enterprise, with its problems of organization, payments, housing and
feeding, the gathering of timber, and to command an army to protect the
work force as well as the objects transported from other cities? And where
near east invited review: king midas’ tumulus at gordion 547
were the Kimmerians and why would they allow such booty to go unplun-
dered? Here, with Sherlock Holmes’ dog in mind, it is not an inconsequential
matter to note that it would have been a very curious incident indeed if
they did nothing, either being present or having recently passed through
Gordion, In short, the political, economic, and military forces and conceptu-
alizations implicit in the project’s planning and construction phases cannot
be perceived as entities existing in the chaotic post-destruction period of
Gordion’s history.
Mellink’s opinion concerning the late date of mm and the identity of its
occupant as Midas is not a new one, for it has been argued by E. Akurgal
since 1959 (although his views on this matter are not once mentioned in
the text). This position has the support of other scholars such as Snodgrass,
Karageorghis, and Hrouda, but not of Boehmer, H. ten Cate, H.-V. Herrmann,
Lloyd, Calmeyer, and Burney, as well as the reviewer, none of whom are cited
in the text. Rodney Young’s views on the chronological placement of mm are
summarized on page 102; he “did not believe that a tumulus on this scale
could have been made immediately after the Kimmerian catastrophe. He
thought the tumulus belonged to Midas’ predecessor … presumably named
Gordius.” This reviewer continues to agree with Young against the contrary
views of the contributors, A major Phrygian individual, probably a king, was
buried in mm by a rich and powerful state in its prime, during a period of
security and peace, not chaos. The time of the deposition will have been
some years before the destruction, perhaps even before the Kimmerians
were a potential threat, although on this particular issue we cannot speak
with certainty. In any event, the exact date could be close to the time
suggested by Young, or a decade or more later, sometime between the
years 720 and 700bc. Unfortunately the individual remains anonymous,
although his name could be Gordius. And it is probable that the impelling
force behind the enterprise was King Midas who, by erecting the largest
tomb and tumulus built up to that time, attempted and succeeded not
only to honor a revered relative but also to impress and overwhelm his
neighbors, friends and enemies alike, Sargon, Urpallu, and the North Syrian
and Urartian kings. The tomb and its overlying tumulus, I suggest, was
in addition to its sepulcharal function a self-conscious Midian-Phrygian
statement, an artifact cum message, a symbol of Midas’ political authority
and canniness. To this extent, and only to this extent, may Midas be brought
into a discussion of Tumulus mm, which letters signify not Midas Mound
but Midas’ Mound: a tumulus that will be identified in the minds of many
of us as an eternal monument not only, to Midas, its postulated builder, but
equally to Rodney S. Young its historical excavator.
548 chapter sixteen
References Cited
Abstract: Recent excavations at Gordion have revealed below the destroyed Phry-
gian city (ca. 700bc) an early Iron Age settlement with handmade coarse ware,
which is followed by a settlement that contains the earliest Phrygian pottery forms.
The handmade ware relates to that from Troy and the Balkans and is considered
firm evidence of the historically recorded migration of the Brygians into Anatolia.
A suggested chronology for the two early settlements is posited, based primarily
on information from Troy. This chronology is then examined together with the
information derived from preserved ancient traditions. A hypothesis is generated
regarding the chronology of the establishment of kingship in Phrygia. This event is
posited to have occurred in the late ninth century bc, and the historical King Midas
is considered to have been the fourth Phrygian king to reign.
From the modern scholarly perspective, the Anatolian Iron Age is said to
begin with the destruction and collapse of the Hittite and neighboring
states, in the years around 1200–1180bc, a chronology derived from texts
from Boğazköy, Ugarit, Meskene-Emar, and Egypt (for a summary of the evi-
dence see Hoffner 1992; Yon 1992; Caubet 1992; also Astour 1965; Freu 1988;
Otten 1976–1977; 1983). Destructions of the Mycenaean centers occurred
about the same time—and may have resulted from some of the same causes.
In both regions thereafter writing ceased to exist until the eighth century bc,
giving rise to the alternative term “Dark Age” for the period (e.g., Akurgal
1983; Muscarella 1988: 417). But whereas in the Aegean during the postde-
struction centuries, artifacts, burials, and architecture are archaeologically
documented, indicating the continuous presence of a population, material
remains are lacking in most areas of Anatolia. Such a condition signifies for
Anatolia a darker Dark Age and a less comprehensible cultural and histori-
cal situation than that which existed in the west.
Burned or abandoned settlements have been recorded at practically
every Late Bronze Age site excavated in central, southern, and eastern Ana-
tolia (Akurgal 1983: 75–78; Bittel 1976–1977: 39–43, 48–49; 1983: 26–28, 30–35,
* This article originally appeared as “The Iron Age Background to the Formation of the
Phrygian State,” Bulletin of the American Schools of Oriental Research 299–300 (1995): 91–101.
550 chapter seventeen
37–39; Muscarella 1988: 417–419; Drews 1993b: 8–11, fig. 1). The magnitude
of the political and social disruption is indicated by the fact that not a
single site is recognized to have been directly reoccupied; this process of
resettlement did not begin in most cases until centuries later.1
The identity of the forces that destroyed the Hittite polity still remains
unknown—although a number of candidates, e.g., the Sea Peoples (specif-
ically named as the destroyers by Rameses III), Muski, Kashka, or alliances
of these powers have been tendered and supported by various scholars. To
date neither archaeology nor textual remains in fact furnish definite answers
or even clues on this issue (Akurgal 1983: 72; Bittel 1983: 27; Hoffner 1992: 49;
Drews 1993b: 48–61). While indeed texts from destruction levels at Boğazköy
and Ugarit refer to military threats by land and sea against the Hittite and
allied states, it is uncertain if one may attribute them specifically to the
final destructions; and names of enemies for this time have not been recov-
ered (Astour 1965; Otten 1976–1977; 1983; Hoffner 1992). A text written in
the eighth regnal year of the Egyptian pharaoh Rameses III, ca. 1180/1176bc,
is both the last written reference to the Hittite state and a post quem non
chronological marker for the destructions (sometimes called the Catastro-
phe). It records that a massive force of diverse nationalities coming from
the ocean and islands in the sea invaded both Egypt and the Asiatic main-
land. The invaders—called by modern scholars Sea Peoples—are said to
have destroyed Cyprus, Carchemish, Arzawa, and Hatti. Rameses’ text is the
only specific reference to the Late Bronze Age catastrophe known, and it
must reflect some historical reality. This is maintained here even if, as has
been suggested, several years’ activities may have been conflated by Rame-
ses, or that the Sea Peoples never penetrated into the Hittite heartland but
only weakened the center for others to conquer (Bittel 1983: 46–47; Hoffner
1992: 46, 49).
The first historical documents mentioning a polity in Anatolia after the
destructions are from Assyria. In his first year, Tiglath-Pileser I (ca. 1115/1112–
1074bc) defeated a people called Mushku with their five “kings” (chiefs?),
on the western shore of the Tigris River in southeastern Anatolia, a territory
the Mushku had controlled for 50 years. This text documents the presence
1 Some scholars have argued against the existence of a gap in continuity between the Late
Bronze Age and the Iron Age (e.g., DeVries 1987: 6; Gunter 1991: 103, 106; Mellink 1991: 626).
There is, of course, the problem of continuous references in Assyrian texts to the Mushki
living somewhere in central or eastern Anatolia (Muscarella 1988: 417; Laminger-Pascher
1989: 16, 19–20). See below for discussion of the gap at Gordion.
the iron age and the formation of the phrygian state 551
of Mushku in eastern Anatolia since at least the 1160s bc, that is, within two
decades (or less) of the Late Bronze Age destructions. Since they are never
mentioned in Hittite texts, it is plausible that the Mushku arrived in Anatolia
during or not long after the time of the destructions (and it is plausible to
ponder, but not to assert, that they either may have played a role in the
destruction of Hatti and neighboring polities or entered the area after the
destructions when there existed no power to resist them).
More than 200 years later the Mushku appear briefly in Assyrian texts,
still mentioned as enemies. In the reign of Sargon II (722–704bc), over 150
years later, they are mentioned again, now called Muski, and ruled by a king
named Mita.
“Mita of Mushki” occurs as such only in the Sargon texts,2 but it is most
probable that he is the king Midas of the Phrygians mentioned in Phrygian
inscriptions and by Greek historians, first by Herodotus (VII.73; Muscarella
1988: 417; 1989: 333–334; cf. Laminger-Pascher 1989: 16–17, 23–24, 38, n. 55
who vigorously rejects the equation). The ancient historians record that
Midas reigned in Gordion, on the Sangarios River, that is in western—not
eastern—central Anatolia. If, as seems manifest, Mita is the Semitic variant
of the Phrygian name, then the appellations Phrygia and Mushki denote
the same polity. This dual nomenclature might signify that the king ruled
over two originally separate groups/peoples (as, e.g., Mellink 1965: 318–320),
the western one better known to the Greeks, the eastern to the Assyrians.
Whether the alleged two groups, if such were the case, were originally
related is presently unknown. Furthermore, the appropriate Phrygian form
Midas, not Mita, occurs in Phrygian inscriptions not only in the west but
also east and south of the Halys River (but not at Babylon [as Mellink 1965:
320]), which may or may not have been the (original) Phrygian-Mushki
boundary (Mellink 1965: 320–322; 1991: 623, 625). It is difficult to accept the
view that the Assyrians meant anything other than the polity that the Greeks
(and we) call Phrygians when they wrote of their encounters with Mita of
Mushki.
2 Postgate (1973: 23) mistranslates the Assyrian appellation Mita the Mushkian as “Mita
the Phrygian,” but on p. 24, n. 4, he paradoxically denies the equation of Mushki and Phrygia;
Hawkins (1982: 421) also mistranslates Mushkian as “Phrygian.” Roller (1984: 260, n. 16), slips
when she claims that the “name Midas” appears in Assyrian texts. Bartl (1994: 514, n. 39)
claims that a Mita of Mushki is mentioned in texts of Shalmaneser III. While I support the
Midas-Mita equation, I realize that on the one hand it is an interpretation, and on the other
it unnecessarily confuses the important issue that the Assyrians never use the term Phrygian
(see also n. 3).
552 chapter seventeen
Herodotus (VII.73) and Strabo (VII.3.2) preserved an old Iron Age tradi-
tion that either before or just after the Trojan war the Phrygians migrated
into Anatolia from Thrace or Macedonia, where originally they were named
Brygians. The fact that the Phrygian language is related to the Greek/Balkan
dialects supports the historical tradition (Haas 1970: 33, 58–59, 68; Neumann
1988; Drews 1993a: 9). But for the Mushku/i and their name there is no pre-
served tradition, only the Assyrian texts, which do not discuss earlier history
or traditions.
One modern hypothesis holds that the Mushki were related to the Bry-
gians, each constituting one of the many migrating tribes noted in ancient
sources, but that the former had migrated further east (Körte 1904: 9–10;
Akurgal 1983: 72–74; Diakonoff and Neroznak 1985: xi–xiii). Other scholars
see the Mushki as a distinct people, living in and deriving from the east (e.g.,
Mellink 1965; Laminger-Pascher 1989: 16–24). Notwithstanding these con-
flicting views, no scholar has been able to know the actual historical reality.
Perhaps there was a western B/ Phrygian conquest, or a peaceful assimila-
tion, of the eastern Mushki, whether the latter were related or not. Whatever
the historical background, however, if the late eighth century bc equating of
the names makes sense and is accepted, then at least by that time there was
but one polity, recognized by themselves and others as sharing ethnicity, a
common king, an ideology, even a common official language—manifested
by the wide distribution of texts in the Phrygian language. Whether or
not the Mushki originally spoke a different language is a question directly
related to that concerning their relationship to the Brygians.3
Up to the 1960s no site in central Anatolia, west, east, or south of the
Sangarios and Halys Rivers, was recognized to have yielded material of
any form that could be situated in the time between the Late Bronze Age
destructions and the eighth century bc (Muscarella 1988: 417–420). This
picture has been elegantly modified by archaeological research at Gordion.
Gordion is the one site in Anatolia, to date, where unambiguous information
bearing on cultural development within the Dark Age has been revealed.
encountered in any discussion of Phrygian history. Perhaps relevant is the fact that the Assyr-
ians, and modern historians, use the term Urartu for the polity identified by its inhabitants
as Biainili (Zimansky 1985: 51, 78). We do not yet know what the Phrygians called themselves,
but it certainly could have been something like B/Phrygians. Diakonoff and Neroznak (1985:
xii) deny this; but Lubotsky (1988: 13–14, 16) and Woudhuizen (1993: 6) interpret the word
vrekun in the Areyastis inscription at Midas City to read either (Lubotsky) “of the Phrygians,”
or (Woudhuizen) “the Phrygian,” in either case an ethnicon; the Greek Phrugioi would then
be a reflex of this word. This is ingenious and warrants more comment from other linguists.
the iron age and the formation of the phrygian state 553
From a deep test trench cut in 1950 hand-made (HM), coarse dark-ware
sherds were recovered below the eighth century, Early Phrygian, level that
was destroyed ca. 700 bc. It was not, however, until 1960 (and even then ten-
tatively: Mellink 1960: 251; Bittel 1983: 38–39; Sams 1992: 58–59)4 that the
significance of this ware was first recognized, namely that it was related
to pottery well known from Levels VIIb1 and 2 at Troy and also from the
Balkans, and that it documented the presence of a people living in central
Anatolia after the Late Bronze Age destructions and before the eighth cen-
tury bc.
In 1988, after a 14-year hiatus, field excavations recommenced at Gordion
under the direction of M. Voigt (see Voigt 1994 for the excavation results—
utilized here—and bibliography of the 1988–1990 campaigns). Because of
their significance for the purposes of this article, a summary of the finds
published to date is necessary.
Excavations below the destroyed Early Phrygian settlement revealed sev-
eral semisubterranean structures above the Late Bronze Age surface. They
were furnished with hearths, ovens, bins, and storage pits; this level was
labeled Early Iron Age, Phase 7B. HM pottery was retrieved in situ in and
around the structures, which seem to have been abandoned at different
times. This assemblage of architecture and pottery indicates, to date, the
only known HM culture settlement in Anatolia aside from Troy (in recent
campaigns at Kaman Kalehöyük and Boğazköy HM sherds have been
recovered—personal examination—but their contexts have yet to be ana-
lyzed; Yakar 1993: 8).
Directly above Phase 7B the investigators recovered a single structure
that utilized different building techniques and internal appointments from
those below; this level is called Early Iron Age, Phase 7A. The structure was
destroyed by a fire and the collapse sealed in a deposit of pottery, which
again represents a new ware, a buff fabric partly wheel-, partly hand-made.
It does not resemble and has no connection with the pottery of the previous
7B phase and, in fact, marks the first appearance of the typical shapes in
use throughout the Early Phrygian settlement down to ca. 700 bc. Above this
structure and cutting into it was another (partly excavated) structure and a
pit, each of which contained both HM and the buff wares, the former surely
indicating (as the burned house did not) that the Phase 7B HM continued
to be manufactured into the next phase.
4 Thus Young (1966: 276–277) in his report of a sounding to the Late Bronze Age level
mentions HM wares but does not discuss their significance. See Sams (1988: 9–10).
554 chapter seventeen
situation that forces a major modification of the 400 years’ gap in occupa-
tion previously postulated by Akurgal and supported by myself and others
(e.g., Akurgal 1983; Muscarella 1988: 417–420 with bibliography; Mellink 1991:
621).
So remarkable is the coincidence of the archaeological evidence (here
primarily pottery) with the preserved historical tradition that one is moved
to recognize in the Early Iron Age Phase 7B assemblage a precious docu-
mentation of the arrival—I think migration is the relevant word—of the
B/Phrygians from lands in the west (a view approached but not grasped by
DeVries 1987: 7, and Sams 1988: 9, 13; 1992: 59; 1994: 21). That the B/Phrygians
migrated from Europe has been accepted by most scholars (e.g., Körte 1904:
2; Mellink 1965: 317; 1991: 621; Haas 1970: 31, 36; Diakonoff and Neroznak 1985:
xi; Sams 1988: 10; 1992: 59; Laminger-Pascher 1989: 12, 15–16, 25, n. 27, 32, 34–
36—but to her not before the ninth and eighth centuries bc; Yakar 1993:
8–10, n. 9); a vigorous—but more asserted than demonstrated—rejection
of the European migration is given by Drews (1993a; 1993b: 65–66, 69);5 Car-
rington (1977: 117) considers the issue “open to doubt.” The material recov-
ered by archaeology considerably reinforces the historical reality of the
migration tradition: that Phase 7B documents the arrival of the B/phrygians
is probable, surely it is feasible.
5 Drews, of course, had no access to the latest archaeological research of M. Voigt but
in a recent publication (1993b: 65) he ignores the implications of the HM ware already
known from Gordion, which he cites only to dismiss; he ignored this information in 1993a.
Drews seems to conflate two separate matters, the alleged Phrygian migration itself and the
Phrygians’ alleged role in the destruction of the Hittites. To him, to reject the latter is to
reject the former. A major thrust of Drews 1993b is to deny the role of migrating people in
the catastrophe of ca. 1200bc, and he challenges most ancient references to such events.
Yet, casually and without discussion he automatically accepts Iron Age Greek migrations
(but never using the M-word) to Cyprus and Ionia, Etruscans to Lemnos, and Cretans to
Palestine, calling them “colonists” and “refugees”—labels denied to the Brygians (Drews
1993b: 59, 63, 68–69). The idea of a Phrygian migration has certainly not “been generally
abandoned by Anatolian archaeologists” (Drews 1993b: 65), as his own note 58 should have
indicated.
Relatively recent research is turning up evidence (pottery, statuettes, metal forms) that
indicate a period of strong contacts existed (whatever the causes) between the Balkans and
north and central Anatolia in the Chalcolithic period, from the late fifth through the fourth
millennium bc, probably in this case via the Black Sea. For bibliography see Thissen (1993:
207–237); several articles in the special issue on “Anatolia and the Balkans” in Anatolica
19 (Roodenberg 1993); and S. Steadman in this issue of BASOR (pp. 13–32). Because of the
millennia separating these contacts from the Brygian migration, it is not productive to
postulate at this time a long traditional mutual knowledge of the two areas. Yet, one is
intrigued by the events; future research may turn up more Balkan-Anatolian contacts.
556 chapter seventeen
An internal clue for the dating of the next period, Phase 7A, also lies in
the pottery. The shapes closely resemble those from late eighth century bc
contexts at Gordion, an indication, perhaps, that not too much time had
elapsed between the floruit of Phase 7A and the late eighth century bc.
If this suggested late dating survives future reviews, one would have to
consider the possibility that the mound had experienced a second gap (an
abandonment) in occupation after the 7B phase. I would estimate—there
is no other method at present—that not many decades elapsed between
the beginning of Phase 7A and that of Phase 6B; an initial date of ca. 850 ±
bc (at the earliest) for Phase 7A, and a time around 825/800 ± bc for the
initial Phase 6B construction cannot be too far from the historical reality.
This allows for a century or more (100–125 years: not so short a time as some
scholars would perceive) for the full development of Phase 6B, the major,
so-called Early Phrygian, period.6
This is a brief outline of the archaeological evidence known so far from
Anatolia. But other evidence for this period also derives from ancient histo-
rians.
Arrian (II.3) and Justin (XI.7, 3–5), among others (Roller 1984: 256–260,
and n. 4, for a summary of the sources), have preserved a tradition concern-
ing the establishment at Gordion of Phrygian kingship, i.e., the formation of
a unified Phrygian state, sometime in the Dark Age, within the period under
review. The tradition, as preserved, records that strife broke out among the
Phrygians and was resolved by a poor farmer, either Gordios, or his son
Midas—the sources disagree—who by arriving at a propitious moment at
Gordion in a cart or wagon drawn by an ox fulfilled a prophecy that conse-
quently led to his election as the king of Phrygia. In some sense this tradition
must surely reflect remembrance of an historic event.7
6 It would seem that the term Early Phrygian used to designate Phase 6B could be
misleading; Voigt uses it to refer to the historical period. The chronology presented here is
not accepted by the excavator of Gordion. Voigt (1994: 270) allows considerably more time for
Phases 7A and early 6B: she begins Phase 7A in the early tenth century bc, with no gap, and
she begins Phase 6B ca. 900/850bc. Yaker (1993: 11) believes that the 7A phase may represent
“the arrival of the Phrygians” in the late second, early first millennia bc; he paradoxically says
little of Phase 7B except that it represents Thracian 12th century bc elements and is “pre-
Phrygian” (Yakar 1993: 10, n. 10), which (in the context, confusingly) omits discussion of the
implications of HM and its possible Brygian connection.
7 Roller (1984: 263–270) presents a sununary of possible religious and folklore elements
preserved behind the scenes in the story, the relationship of Gordion to Gordios, the historic-
ity of Gordios (a fiction to her), the significance of Kybele, and so forth. None of this need be
discussed here except to note that the basic core element of the tradition—state formation
the iron age and the formation of the phrygian state 557
wagon in the same manner as did Rusa I, king of Urartu.8 In Sargon II’s
report to Assur on his eighth campaign in 714bc he records his capture of the
Urartian holy center of Musasir. Taken as booty, among many other items,
was a statue group of “Rusa, with two of his horses, (and) his charioteer,”
which was inscribed “With my two horses and one charioteer, my hand
attained the kingdom of Urartu” (Luckenbill 1926–1927: II, §173). What the
attainment entailed is not recorded; but just the year before, in 715bc, Rusa
survived a brutal revolt (Lanfranchi 1983) by escaping into the mountains,
and it was probably that event that was commemorated with the creation
of the statue group.9
What stands out is that the Phrygian king could have claimed that “with
my one ox and one charioteer, my hand attained the Kingdom of Phrygia,”
for that meaning is precisely what is preserved in the tradition. The histor-
ical parallel is certainly real, for a wagon and its connection to kingship
are associated with turmoil and a favorable resolution in both the Phry-
gian and Urartian events, although the specifics are different. In one case a
commoner who peacefully and innocently enters a city is made king, in the
other a king saves his reign by escaping, exiting, to the mountains. We are
dealing here with two different and unconnected incidents, one an histori-
cal report of a contemporary occurrence, the other preserved through local
tradition and perhaps not revealing the full circumstances that the wagon
episode entailed. Nevertheless, the possibility of condensation in the tradi-
tion does not compel a denial that an historical reality involving a wagon and
its occupant occurred. It would not be fruitful, though, to speculate about
the incidents that might lie behind the Phrygian episode based on the his-
torical Urartian experience.
8 Calmeyer (1974: 63, n. 57) also cites as a parallel Herodotus (III: 85–86) who reports
that Darius attained kingship with a ruse involving a horse and a groom. Here there is only
a horse, not a wagon, but there is an essential formal similarity to the Gordion and Rusa
events. Calmeyer (1974: 63–64) also links the Gordion wagon and its eagle (from Arrian) with
a deity and notes the wagon symbolizing the victory-bringing power of the highest deity.
Surprisingly, he does not pick up Körte’s (1904: 16) firm claim that the Phrygian wagon is
“Ein genaues Gegen stück” to the “leere Wagen des persischen Himmelsgottes …” Körte also
believed that the wagon was originally associated with Zeus and later associated with the
farmer king.
9 We are informed that to save his life Rusa escaped alone, but surely he had his chariot
and charioteer. Rusa suppressed the revolt and lived to commission the statue group and
later fight Sargon II. Lanfranchi does not mention the Musasir statue but Barnett (1982: 351)
mentions it in connection with the revolt. The statue group would then have been less than
one year old when taken to Assyria by Sargon.
the iron age and the formation of the phrygian state 559
What can one say about the nature of the prekingship society—for exam-
ple, what authority existed? were there tribal or clan chiefs at Gordion? was
Gordion the residence of a paramount chief? While it may satisfy neoevo-
lutionary theory to add still another chief or chiefdom to the many per-
ceived in anthropological literature, it may not be historically accurate to
do so.10 A neoevolutionary model would posit that the Phrygians should
have experienced first tribal and then chiefdom stages of government before
becoming a state, i.e., developing into a kingdom. Some anthropologists
reject this deterministic, integrative position and recognize the difficul-
ties in ascertaining not only what a chiefdom actually is, and the man-
ner in which it allegedly becomes, or is different from, a state, but also
that more than one state formation model exists. Ethnology can provide
analogies but not equations; the Phrygians were Thracians, not Polyne-
sians.
The act of election as a process of societal transformation need not be
an impediment to accepting the legend’s historical core. Aristotle (Politics:
1285: 3,7) discusses the existence both of elective, nonhereditary dictators
and elected, hereditary kings in Greek city-state formation (see also Runci-
man 1982: 357). In these instances the elected king had previously performed
some crucial services for his society. The tradition does not define the nature
of the political strife at Gordion, but the society was fundamentally dis-
rupted. Anthropologists suggest that a number of dramatic causes can result
in state formation, among them population growth, class stress, technolog-
ical or subsistence pressures, and external pressures or threats of warfare.
Which of those events (if any) may have caused the strife? All we can dis-
cern is that an individual—a peasant, not a nobleman—came to the aid of
his fellow Phrygians. We can also infer that he performed some significant
act—even fulfilling an oracle.
Nor does the tradition inform us about the role of the other Phrygian cen-
ters. If we are justified in using the HM ware as the criterion for determining
the presence of Brygians at other centers, we again ask questions that can-
not be answered: did the Phrygians at Boğazköy and Kaman Kalehöyük—
neither site geographically close to Gordion—and other centers still unrec-
ognized play a role in the formation of the Phrygian state? Was the strife
10 A good picture of the nature of pre- and poststate societies may be found in a number
limited to Gordion or were the other centers involved? How and when did
those centers accept the kingship organized at Gordion?
What information do the archaeological data yield regarding these prob-
lems? Can a pre- and postkingship pattern be observed in the excavated
remains at Gordion?
It is probable that the population settled in the initial Early Iron Age
phase, 7B, did not have a king. The material remains show that Phase 7B
was surely a village, as Henrickson (1994: 108) has posited that the pottery
was made by individual households. Hierarchical architecture has not been
exposed and no burials exist to reveal the mortuary furniture of an elite. No
archaeologist would argue from the remains that kingship existed in Phase
7B (the peasantlike character of the prekingship Phrygians in the preserved
tradition was noted by [Körte 1904: 15–16]).
The one structure fully excavated in Phase 7A was burned, and the charac-
teristic buff ware is interpreted (again by Henrickson [1994: 109–110]) to have
been more complex in production (mass produced?). Here, too, there are no
clues in the architecture or in burial customs to suggest conclusions about
the nature of the society. Whether the burning of the Phase 7A structure
reflects the strife that is noted in the tradition must remain an intriguing,
but yet another unanswered, question. Further excavation will surely pro-
vide information about the extent of the burning of that phase. But what is
significant about the existence of the HM and buff ware recovered together
in the later part of Phase 7A (above) is that some element of cultural conti-
nuity from the earliest settlement is revealed.
The first unambiguous, archaeologically documented indication of orga-
nized political and social activity in Gordion’s history was the initiation of
the massive building program around 825–800bc. The Phase 6B master plan
began tentatively, but within a short time represents a dynamic conception
absent in the earlier periods and it reflects the existence of an authority
commanding power and concomitant wealth, and an urban concept. In its
developed stage one encounters monumental architecture; massive fortifi-
cation walls (completely rebuilt at least once); royal burials (see below); the
presence of a writing system; specialists such as architects and engineers,
many artisans working in metal, ceramics, ivory, stone, textiles, and wood
carving; and also the availability of objects from foreign centers, probably
gift exchanges. It tentatively follows, then, that kingship could have arisen
perhaps late in Phase 7A or at the beginning of Phase 6B. If the beginning of
the 6B phase is indeed the actual first stage of the master plan, the evidence
could be interpreted to indicate late Phase 7A as the time when kingship
arose. In absolute dates this would be approximately 825 ±bc.
the iron age and the formation of the phrygian state 561
One hypothesis about Phrygian state formation that comes easily to mind
will not cause much opposition: the political transformation experienced
by Gordion-Phrygia would not have been that of primary state formation,
an indigenous, self-generated, and independent social evolution, but of
secondary state formation, in which state origins are attributed to pressures
or models derived from previously existing states. Such a development could
occur quickly and need not be gradual (Runciman 1982: 356). In the mid-
and late ninth century bc, kingships existed in neighboring areas, at least
in Assyria, North Syria, and Urartu, and perhaps also in Tabal. Aside from
Assyria, with which Phrygia (Mushki) had hostile relations (and perhaps
also normal contacts), there is no information concerning the intensity of
interaction between Phrygia and the other polities in the ninth century bc.
But surely their existence could not have remained unknown; one or all of
these states could have been a model for those who chose a citizen to be
their first king.
A potentially important historical implication is suggested by the chro-
nology advanced here for the advent of the Phrygian state (ca. 825 ±bc). The
dates correspond either to the last years of Shalmaneser III (858–824bc) or
to the reign of Shamshi-Adad V (823–811bc). For some time it has been rec-
ognized that during the reign of Shalmaneser III the Urartians consolidated
themselves and seem for the first time to have achieved a unified state, ruled
by one king (Zimansky 1985: 48–50). This event occurred at the very time
proposed here for the Phrygians’ consolidation under one king. Two sec-
ondary state formations would then have occurred in Anatolia in the second
half of the ninth century bc. More refinement on chronology is necessary;
but even if a higher chronology is indicated at Gordion, the proximity in time
of the two events will not be essentially altered. The populations of central
and northeastern Anatolia were in continuous conflict with Assyria, a ten-
sion that may have provoked (in part) parallel social transformation in each
area in the second half of the ninth century bc (for Assyrian aggression and
Urartu see Zimansky 1985: 48–50).
The first Phrygian king or his son (perhaps either the great-grandfather
or grandfather of the Midas who perished in the destruction of ca. 700bc)
or both should be considered as the mind(s) that conceived and initiated
the construction of the Phase 6B settlement. This major project could not
have been conceived and carried out in a noncentralized society. The early
modern historians of the Phrygians, Körte (1904: 13–14, 21), Swoboda (1912:
1590–1591), and Eitrem (1932: 1536–1538) interpreted the ancient sources to
indicate that the historical Midas (ruled ca. 738–696 bc) was in fact the
second of that name, the first having been mentioned in the traditional
562 chapter seventeen
story. They also believed, correctly, I think, that it makes sense to accept
Justinís version that a Gordios, not a Midas, was that first king.11
The earliest royal burial (so determined by its size, second only to MM,
and its rich contents) so far recognized at Gordion is that in Tumulus W,
plausibly dated to the years around 750±bc12 By this time some of the Phase
6B settlement’s structures were functioning units. Positing a few decades
for a reign, the beginning date for the buried king’s rule would have been
ca. 780± bc. Since many tumuli remain to be investigated, it cannot be
argued on archaeological grounds that the tomb contained Phrygia’s first
king—we do not yet know how and when this form of burial first came to be
used at Gordion. However, based on the chronology posited here, the man
buried in Tumulus W could have been either the second or third king to
reign at Gordion: Midas I or Gordios II.13
The historical importance of Tumulus W’s chronology (whatever one’s
opinion of that date is) is significant for two reasons: its proposed date
(750±bc) is a post quem non for the existence of kingship in Phrygia, and the
estimated date (780±bc) for the buried king’s first regnal year is a viable, if
hypothesized, earlier post quem non.
11 I had reached this view before going back and rereading these German scholars. It fol-
lows that I reject Roller’s opinions (1984: 260, 263) that Gordios was a fictitious name, e.g.,
Roller also posits that the first Phrygian king would have had a Balkan, not an Anatolian name
(Midas), a view that ignores evidence of the rather long sojourn of the Phrygians in Anato-
lia: the new excavation evidence interpreted here further argues against the need to seek a
Balkan name (see n. 3, above). However, I am informed that some linguists believe that Gor-
dion derives from gorod or grad—which if correct, signifies that the name Gordion/Gordios
is indeed Balkan.
12 Independently dated this late by both my research (Muscarella 1982: 8): 750/740 bc, and
that of R.M. Boehmer (1984: 259–260): ca. 740bc. Of the many excavated tumuli at Gordion,
about 11 date to the predestruction periods: G, P, Q, S, MM, W, Y, KIII, KIV; X and KY (not yet
fully published) have also been dated to this period by DeVries (1977: 9, n. 11), and Sams (1994:
192–196), although each has a different list (see also Muscarella 1967: 1–11).
13 This raises again the question of who was buried in Tumulus MM. I have argued
elsewhere (Muscarella 1982: 9–10), and still hold the view, that it could not be Midas because
he was alive when the burial was arranged, sometime between 720 and 700 bc. It follows
chronologically that it was he who must have ordered the tomb and tumulus to be built. Other
scholars, however. date MM to a time just after the destruction, and consequently consider it
to be Midas’ burial (e.g., DeVries 1977: 9; Mellink 1991: 634; Hawkins 1994: 273; see the judicious
comments of Sams 1994: 196). The only way for archaeologists to agree that MM contained the
burial of Gordios II (who presumably died before his son began to reign) is for its deposition
date to be raised to ca. 740bc at the latest—but which on the basis of present knowledge will
not satisfy any archaeologist. If neither Midas nor seemingly his father, Gordios II, is buried
in MM, the question of who is interred there remains intriguing. The burial was manifestly
arranged for a revered person who died in his 60s, a man Midas wished to honor by erecting
the largest tumulus hitherto known in Anatolia.
the iron age and the formation of the phrygian state 563
References
Akurgal, E.
1983 Das Dunkle Zeitalter Kleinasiens. Pp. 67–78 in Griechenland, die Ägäis und
die Levante während der “Dark Ages” vom 12. bis zum 9. Jh. v. Ch., ed. S. Deger-
Jalkotzy. Vienna: Österreichische Akademie der Wissenschaften.
14 I thank Kenneth Sams, Elizabeth Simpson, and Mary Voigt for their cooperation in
sharing their ideas—not always in agreement with mine—and for helping to make this a
better article than it would have been without their concern.
564 chapter seventeen
Astour, M.
1965 New Evidence on the Last Days of Ugarit. American Journal of Archaeology
69: 253–258.
Barnett, R.D.
1982 Urartu. Pp. 314–371 in The Cambridge Ancient History, vol. 3, part 1, eds.
I. Boardman, I.E.S. Edwards, N.G.L. Hammond, and E. Sollberger. 2nd. ed.
New York: Cambridge University.
Bartl, K.
1994 Die Frühe Eisenzeit in Ostanatolien und ihre Verbindungen zu den Benach-
barten Regionen. Baghdader Mitteilungen 25: 473–518.
Bittel, K.
1976–1977 Das Ende des Hethiterreiches aufgrund archäo-Zeugnisse. Jahresbe-
richt des Instituts für Vorgeschichte der Univ. Frankfürt a.M.: 36–56.
1983 Das archäologische Situation in Kleinasien um 1200 v. Chr. und während der
nachfolgenden vier Jahrhunderte. Pp. 25–65 in Griechenland, die Ägäis und
die Levante während der “Dark Ages” vom 12. bis zum 9. Jh. v. Ch., ed. S. Deger-
Jalkotzy. Vienna: Österreichische Akademie der Wissenschaften.
Bloedow, E.F.
1988 The Trojan War and Late Helladic III C. Prähistorische Zeitschrift 63: 23–52.
Boehmer, R.M.
1984 Uruk-Warka XXXVI: Survey des Stadtgebietes von Uruk. Baghdader Mit-
teilungen 15: 113–140.
Calmeyer, P.
1974 Zur Genese altiranischer Motive II. Der leere Wagen. Archaeologische Mit-
teilungen aus Iran 7: 49–77.
Carrington, P.
1977 The Heroic Age of Phrygia in Ancient Literature and Art. Anatolian Studies
27: 117–126.
Caubet, A.
1992 Reoccupation of the Syrian Coast after the Destruction of the “Crisis Years.”
Pp. 123–131 in The Crisis Years: The 12th Century B.C., eds. W.A. Ward and
M. Joukowsky. Dubuque, IA: Kendall/Hunt.
DeVries, K.
1987 Phrygian Gordion Before Midas. Pp. 6–12 in Anatolian Iron Ages, ed. A.
Çilingiroğlu. Izmir: Ege Universitesi Edebiyat Fakultesi.
the iron age and the formation of the phrygian state 565
Drews, R.
1993a Myths of Midas and the Phrygian Migration from Europe. Klio 75: 9–26.
1993b The End of the Bronze Age: Changes in Warfare and the Catastrophe ca. 1200
B.C.. Princeton: Princeton University.
Eitrem, S.
1932 Midas. Paulys Real-Encyclopaedia 15: 1526–1540.
Freu, J.
1988 La Tablette RS 86.2230 et la phase finale du Toyaume d’Ugarit. Syria 65: 395–
398.
Gunter, A.
1991 The Bronze Age: The Gordion Excavations Final Reports III. University Muse-
um Monograph 73. Philadelphia: University Museum, University of Penn-
sylvania.
Haas, O.
1970 Das Problem der Herkunft der Phryger und ihrer Beziehungen zu den
Balkanvölkern. Acta Antiqua Academiae Hungaricae 18: 31–69.
Hawkins, D.
1994 Mita. Reallexikon der Assyriologie 8, 3/4: 271–273.
Hawkins, J.D.
1982 The Neo-Hittite States in Syria and Anatolia. Pp. 372–441 in The Cam-
bridge Ancient History, vol. 3, part 1, eds. J. Boardman, I.E.S. Edwards,
N.G.L. Hammond, and E. Sollberger. 2nd ed. New York: Cambridge Univer-
sity.
Henrickson, R.
1994 Continuity and Discontinuity in the Ceramic Tradition of Gordion during
the Iron Age. Pp. 95–129 in Anatolian Iron Ages 3, eds. A. Çilingiroğlu and
D.H. French. Ankara: British Institute of Archaeology.
Hertel, D.
1991 Schliemanns These vom Fortleben Troias in den ‘Dark Ages’ im Lichte
neuer Forschungsergebnisse. Studia Troica I: 131–144.
Hoffner, H. A, Jr.
1992 The Last Days of Khattusha. Pp. 46–52 in The Crisis Years: The 12th Century
B.C., eds. W.A. Ward and M. Joukowsky. Dubuque, IA: Kendall/Hunt.
566 chapter seventeen
Korte, A.
1904 Geschichte Phrygiens. Pp. 1–27 in Gordion: Ergebnisse der Ausgrabung im
Jahre 1900, eds. G. and A. Körte. Jahrbuch des Kaiserlich Deutschen Archäol-
ogischen Institüts, Ergänzungsheft 5. Berlin: Reimer.
Laminger-Pascher, G.
1989 Lykaonien und die Phryger. Vienna: Österreichische Akademie der Wis-
senschaften.
Lanfranchi, G.B.
1983 Some New Texts about a Revolt Against the Urartian King Rusa I. Oriens
Antiquus 22: 123–136.
Lubotsky, A.
1988 The Old Phrygian Areyastis-Inscription. Kadmos 27: 9–26.
Luckenbill, D.D.
1926–1927 Ancient Records of Assyria and Babylonia I–II. 2 vols. Chicago: Univer-
sity of Chicago.
Mellink, M.
1960 Review of C. Blegen et al., Troy IV. Bibliotheca Orientalis 17: 249–253.
1965 Mita, Mushki and Phrygians. Anadolu Arastirmalari 2: 317–325.
1991 The Native Kingdoms of Anatolia. Pp. 619–643 in The Cambridge Ancient
History, vol. 3, part 2, eds. J. Boardman, I.E.S. Edwards, N.G.L. Hammond,
and E. Sollberger. 2nd ed. New York: Cambridge University.
Muscarella, O.W.
1967 Phrygian Fibulae from Gordion. Monograph Series (Colt Archaeological
Institute) 4. London: Quaritch.
1982 King Midas’ Tumulus at Gordion. Quarterly Review of Archaeology 3, no. 4:
7–10.
1988 Bronze and Iron. New York: Metropolitan Museum of Art.
1989 King Midas of Phrygia and the Greeks. Pp. 333–344 in Anatolia and the
Ancient Near East, eds. K. Emre, M. Mellink, B. Hrouda, and N. Özgüç,
Ankara: Türk Tarih Kurumu.
Neumann, G.
1988 Phrygisch und Griechisch. Vienna: Österreichische Akademie der Wissen-
schaften.
Otten, H.
1976–1977 Zum Ende des Hethiterreiches aufgrund der Boğazköy-Texte. Jahres-
bericht des Instituts für Vorgeschichte der Univ. Frankfurt a.M.: 22–35.
1983 Die letzte Phase des Hethitischen Grossreiches nach den Texten. Pp. 13–24
in Griechenland. die Ägäis und die Levante während der “Dark Ages” vom 12.
bis zum 9. Jh. v. Ch., ed. S. Deger-Jalkotzy, Vienna: Österreichische Akademie
der Wissenschaften.
the iron age and the formation of the phrygian state 567
Postgate, N.
1973 Assyrian Texts and Fragments. Iraq 35: 13–36.
Roller, L.
1984 Midas and the Gordion Knot. Classical Antiquity 3: 256–271.
Runciman, W.G.
1982 Origins of States: The Case of Archaic Greece. Comparative Studies in Society
and History 24: 351–377.
Sams. G.K.
1988 The Early Phrygian Period at Gordion: Toward a Cultural Identity. Source 7,
no. 3/4: 9–15.
1992 Observations on Western Anatolia. Pp. 56–60 in The Crisis Years: The
12th Century B.C., eds. W.A. Ward and M. Joukowsky. Dubuque, IA: Kendall/
Hunt.
1994 The Early Phrygian Pottery. The Gordion Excavations, 1950–1973: Final Re-
ports, 4. 2 vols. University Museum Monograph 79. Philadelphia: University
Museum, University of Pennsylvania.
Swoboda, H.
1912 Gordios. Paulys Real-Encyclopaedie 7: 1590–1594.
Thissen, L.
1993 New Insights in Balkan-Anatolian Connections in the Late Chalcolithic: Old
Evidence from the Turkish Black Sea Littoral. Anatolian Studies 53: 207–
237.
Voigt, M.
1994 Excavations at Gordion 1988–1989: The Yassihöyük Stratigraphic Sequence.
Pp. 265–293 in Anatolian Iron Ages 3, eds. A. Çilingiroğlu and D. French.
Ankara: British Institute of Archaeology.
Woudhuizen, F.C.
1993 Old Phrygian: Some Texts and Relations. Journal of Indo-European Studies
21: 1–25.
Yakar, J.
1993 Anatolian Civilization Following the Disintegration of the Hittite Empire:
An Archaeological Appraisal. Tel Aviv 20: 3–28.
568 chapter seventeen
Yoffee, N.
1993 Too Many Chiefs? (or, Safe Texts for the ’90s). Pp. 60–78 in Archaeological
theory: who sets the agenda?, eds. N. Yoffee and A. Sherratt. New York:
Cambridge University.
Yon, M.
1992 The End of the Kingdom of Ugarit. Pp. 111–122 in The Crisis Years: The
12th Century B.C., eds. W.A. Ward and M. Joukowsky. Dubuque, IA: Kendall/
Hunt.
Young, R.S.
1966 The Gordion Campaign of 1965. American Journal of Archaeology 70: 267–
278.
Zimansky, P.
1985 Ecology and Empire: The Structure of the Urartian State. Studies in Ancient
Oriental Civilization 41. Chicago: Oriental Institute of the University of
Chicago.
chapter eighteen
Abstract: For decades the date of the destruction of the Phrygian site of Gordion
in central Anatolia has been accepted to have occurred ca. 700 bc or later. In 2001
the Heidelberg Akademie der Wissenshaften laboratory released the results of
C 14 analyses of material from the destruction level indicating a date of 830 to
800bc. Members of the Gordion excavation team accepted the new chronology.
This high chronology significantly alters the dating of all the artefacts recovered
in the destruction level, affects knowledge of the specific Gordion phase in which
King Midas reigned and the proper assignment of royal tumuli burials to a specific
culture level, and mandates a major evaluation of Anatolian Iron Age history and
archaeology. The present study examines artefacts recovered from the destruction
level and the tumuli and concludes that the destruction occurred at the time
previously recognised. It is argued that the C 14 determination does not reflect the
historical reality.
The Background
* This article originally appeared as “The Date of the Destruction of the Early Phrygian
2 See Sams 1979, 46; 1993, 549, 551; DeVries 1990, 388–390; Muscarella 1995, 97; Voigt 1994,
dendrochronologically determined dates and subsequent retractions, see James 2002, 18.
572 chapter eighteen
over a massive clay deposit laid down over the DL and in some areas is
more than 5m deep.7 Voigt had earlier demonstrated that this rebuilding
occurred very soon after the destruction, and not, as had been maintained
by Rodney S. Young (and generally accepted over the years), after a long
period of elapsed time: the MP rebuilding began ‘almost as soon as the ruins
[DL] had cooled …’8 and there is ‘little or no gap between the [DL] fire and
the reconstruction’ of the MP citadel.9 Consequently the MP’s construction
is now to be considered as having commenced in the late 9th or early in
the 8th century bc, as pointed out recently by Strobel.10 In this scenario it
was the MP, not the EP, citadel that was the locus from which King Midas
reigned, although he would not have been responsible for the rebuilding,
which would have preceded his reign by six to nine decades. The earliest his-
torical reference to Midas is 717 bc, another, the last known to us, occurred
in 709bc.11 Based on ancient historians, Midas probably began his reign
ca. 740 or a little later, surely before the time of Sargon II. Consequently, the
rebuilder would have been Midas’ (very young!) father or his grandfather,
possibly even great-grandfather.
While the new C 14 chronology does not alter the 8th century artefact-
based dating recognised for the excavated tumuli—based both on den-
drochronology and internal artefact analysis, including imported material
(see below), it does require that these burials are now to be associated with
the middle-late phase of the MP floruit, and not as hitherto accepted with
that of the EP.
Given the unexpected and complex issues introduced by the ad hoc
acceptance of the Heidelberg C 14 analysis, it seems appropriate to (re)ex-
amine the artefacts recovered in the DL specifically with reference to their
archaeologically determined chronologies, that is, independently of the
accepted C 14 date. Acceptance of the new chronology concomitantly de-
mands that all the DL artefacts are to be considered 9th century productions.
This is all the more interesting inasmuch as Gordion excavation team schol-
ars had supported in various publications an 8th century artefact-derived
chronology for these same DL artefacts.12 (Some of the DL artefacts will not
paper delivered at the AOS meeting in Nashville, Tennessee: see AOS Abstracts, April 4th–7th
2003, 31.
destruction of the early phrygian period at gordion 573
be discussed here, viz. animal figurines.13 Represented here are at least four
different styles, at least one of which reflects an Aegean background, as Maya
Vassileva has demonstrated in her study of them; forthcoming).
The Fibulae
13 Young 1962, 166, pl. 48, figs. 21, 22, Prayon 1987, Taf. 21–25; Sams and Temizsoy 2000,
215, 239; also Sams 1994, 17, 19; Caner 1983, 6, 54.
17 Muscarella 1982, 9; 1986, 195; 1988a, 427, note 5; 1988b, 182–183; 1995, 97 and note 12;
24 Muscarella 1967, 1, 3; 1982, 9; 1988b, 183; 1995, 99; 1998, 151–153. For the imported objects
see Young 1981, 104–110, 121–123, pls. 51–57, 62–63; also Prayon 1987, 130–132, 185, Taf, 41, 42.
None of these derive from Urartu, pace Mellink 1981, 268; 1991, 637.
25 Mellink 1981, 270; DeVries 1990, 390; Caner 1983, 9, 67.
26 James 2002, 18; Manning et al. 2001, 2533–2544.
27 Kuniholm 1988, 8.
28 This dendrochronological date would neatly fit Eusebius’ beginning year for the reign
of Midas at 738bc—if the tomb were to have been constructed shortly after the cutting of
the logs. For years I have argued that Midas was not buried in Tumulus MM, that he built
the tumulus and tomb for a revered relative: see Muscarella 1967, 2; 1982, 9; 1988, 183; 1995,
99, note 13. This interpretation has been glossed over or ignored in publications and lectures
by the Gordion team who have often asserted that MM is the tomb of Midas. As of 2001 they
claim in memos and lectures that they now know for the first time ever that Midas was not
buried in MM.
29 Muscarella 1967, 6, 82; Caner 1983, 11; Sams 1994, 71, n. 64; Kohler 1995, 115–121, and fig. 54.
30 Temizsoy 1992, pls. I–X; 1993, pls. XI–XVIII—esp. XII, 2, XII, 11/13A, XII, 14A, and fibulae
D.
576 chapter eighteen
32 Muscarella 1967, 43, pls. III–VI; see Appendix A for a chart of the tumuli loci of the
‘discrepancy’ concerning the find spot within Megaron 4 of the XII, 9 fibula, B 1545: the
field notebook (written by the excavator) states it was recovered in the ‘burned fill above
the Phrygian floor’ in cut 4C—a specific reference to a DL burnt debris locus; the inventory
record, however, states that it came from another area ‘on the floor’ of Meg 4, cut 4S1. DeVries
believes that the clay overlay in the inventory recorded find spot (S1) ‘seems undisturbed’
but ‘was probably [sic] churned up’, while the clay overlying the notebook listed fibula (4C)
was ‘badly disturbed’ by later scavenging. The implication seems to be to cast doubt on
the conclusion that the fibula derived from the DL, that it could have been dropped by
578 chapter eighteen
recovered from stone packing below the floor upon which Megaron 4 was
built; it preserves only the catch and section of the arc, round in form, sug-
gesting it may be a XII, 13 or 14 form; the arc’s thinness suggest it is possibly
XII, 14.36 Megaron 4 was the latest structure to have been built before the
destruction, hence the fragment was deposited shortly before that event.
What is essential to recognise is that all the DL fibulae are the very same
types as those recovered in the 8th century bc tumuli W, III, IV, S, G and
MM, and include both early and late types.
In addition to the 14 imported examples from the DL, two others occur
in Tumulus G, and one in Tumulus Y, both 8th century tumuli. Eight of
these from the DL and those from Tumulus G and Y have swollen, leech
shaped arcs tapering to the ends and with flat catch plates (missing on the
Tumulus G example); slight rings occur at the ends of several examples.37
They are Aegean forms, Caner’s Type IVd, Sapouna-Sakellarakis Type IVc,
and, independently reinforcing the Gordion tumuli evidence, are dated to
a later intruder—but the absence of the studs, the damaged catch, and the missing pin
assuredly contradict this forced interpretation. Note that Sams (1994, 3) had no problems
with accepting a DL locus for the pottery he publishes from Megaron 4. The excavator
recorded disturbance resulting from the placement of the foundation of a MP building, and
did not doubt the DL origin of the fibula recorded in the field notebook. Fibula B 1454 derived
from the DL, along with EP pottery recorded here.
36 Caner 1983, no. 1179.
37 Caner 1983, nos. 84, 85, 88, 89, 92A–D, 93; Muscarella 1967, pls. XVI, XVIII, figs. 88–91—
one example is not published; 1989, 336–337, note 17; Kohler 1995, 87, pl. 21 C, D; 108–109, pl. 57
A. Three imported fibulae found in the fill of Tumulus B are not counted here but should be
noted, see Muscarella 1967, 6, 82, Pl. XVIII, 92; Kohler 1995, 21f., Fig. 9B–D, and dating the tomb
(p. 15) to the 7th century; Caner 1983, 181, no. 1191, dates it to the late 8th–early 7th century.
Caner 1983, no. 1184, a Near Eastern form, is not from the DL.
destruction of the early phrygian period at gordion 579
would then have been 15 imported fibulae in the DL.40 Caner catalogued
these fibulae as a central Anatolian variety of Near Eastern forms, Type V,
citing eleven close examples from the Phrygian period at Alishar; Pedde
categorised them in his broad Near Eastern C1.2 group.41 Fibulae of exactly
the same form also occur in central Anatolia at Kaman-Kalehöyük—but
unfortunately (as at Alishar) no information whatsoever exists in the site
publications concerning their loci and contexts that would help to situate
them within the local Phrygian sequence.42
Mouldings asymmetrically situated on the arc (B 1988c, d—not men-
tioned in Caner) characterise two of the imported fibulae in the DL; one
has a more pronounced asymmetrical moulding placement. I can find no
exact parallels, but Sapouna-Sakallerakis classifies them with her Type IV b
(all of which do not have asymmetrical arcs), allowing a range from Proto-
geometric (Vrokastro, her nos. 611 and 612—which have a longer catch plate
than either of the Gordion examples) to Geometric.43 Giesen cites them as
Type III, Cypro-Archaic II; she sees them as possibly Cypriote but leaves
open a general Aegean distribution, and dates them to the late 8th cen-
tury bc.44
A fibula from Terrace Building 7 (B 1587; not in Caner 1983) is missing its
catch plate but has what appears to be an obliquely rilled or grooved arc. I
can find no exact parallels, but it seems similar to an example from the fill
over Tumulus Y, to fibulae from Tumulus S1, to the rilled XII, 13 arcs from the
Mama Creek tumulus and to a Phrygian fibula from 7th century Ephesus.45
It may belong to Pedde’s Group B2.3, with parallels from North Syria—see
especially his nos. 143, 146–151, all dated to the 8th–7th centuries bc.46
A review of the arc forms and mouldings of the relevant Phrygian fibulae
in the tumuli and DL reveals no clear indication of recognisable differ-
ences (other than in refinements) in the chronologies of their constructions.
Both Young and Caner (who classified all examples from the tumuli and DL
together within two subtypes) made the same judgment with regard to XII,
Iron Horsebits
Three iron horsebits (Fig. 8) were found in the DL Terrace Building 2; the
cheek pieces are crescentic with a vertical central tang; rein holes exist on
the cheek piece and vertical tang. Young cited two examples, subsequently
noting that ‘there should have been four of them’ because he thought they
were associated with the ivory horse trappings for four horses, see below;52
in fact, three were recorded (Keith DeVries and Gareth Darbyshire, personal
communications). I have been able to find formal parallels in two distinct
geographical areas. They occur at several sites on Crete, Knossos, Prinias,
and Arkades, all bronze.53 These examples consist of crescentic cheek pieces
with rein holes, but a much smaller vertical tang and the rein hole is below
it. The examples in the Knossos North Cemetery are dated from the pottery
to the Late Geometric or a little later. On a stone relief in a tomb at the
Transcaucasian site of Astchadzor was carved a related form, here with a
less cresentic formed cheekpiece, and a vertical tang, all apparently pierced
for reins; A. Ivantchik dates it to the 9th century bc, but his evidence is
not certain.54 According to I. Medvedskaya (personal communication) an
Assyrian 7th century bc-form sword was recovered in the tomb and she cites
52 Young 1962, 166–167 pl. 48, fig. 26; Sams and Temizsoy 2000, no. 51.
53 Donder 1980, Taf. 8, 64, 65; Coldstream and Catling 1996, 213, Tomb 219, 210–213, fig. 168,
nos. 92, 102, 103a—here too, as at Gordion, there is an odd piece.
54 Ivantchik 2001a, 144, 149, Abb. 67, 5.
destruction of the early phrygian period at gordion 583
two Russian scholars who respectively date the tomb to the 9th–8th or the
7th centuries bc.
The bits found at Knossos occur together with a bronze bowl with lotus
handles, of the same form as one from Tumulus III and another of Egyptian
Blue from Tumulus P; other examples occur on Crete and in Cyprus, at Nim-
rud, Til Barsip, and Tell Halaf.55 G. Hoffman dates the Knossos (Tomb 219)
example to 745–700bc, and another from Amnisos to 745–710bc.56 Tumulus
P also contained Cypriote pottery, Black on Red II (IV), Cypro-Geometric III-
Archaic I vessels.57 Cypro-Geometric III is conventionally dated ca. 850–
750bc, Cypro-Archaic I to 750–600 bc, but some scholars place the chronol-
ogy higher, ca. 800–750 bc (Joanna Smith, personal communication). In
either dating, high/low, we do not know how long it took for the vessels to
reach and then be deposited in a Gordion tomb.
It may be significant that the crescent form of the DL cheek pieces is
exactly matched by an apparent nose guard from a horse excavated in a
manifestly non-Phrygian (nomadic!) burial at Gordion, Tumulus KY, dated
by Kohler close to (after?) the DL.58 Here the very same crescent shape is
formed as two units joined together as one piece. What this clear parallel
signifies, in both cases involving horse equipment, I leave open: except to
note that the certain late (late 8th–early 7th century) chronology of Tumulus
KY is to be kept in mind when dating the iron bits.
The iron horsebits from the DL most probably were manufactured some-
time in the 8th century bc. Sams believed (as did Young, above) that they and
the ivory horse trappings (see below) also found in TB 2, were used together
and were ‘immediately recognised as North Syrian products’ (but he gave
no parallels for the horsebits) and came to Gordion as a unit through a
diplomatic exchange (along with cauldrons); earlier he identified the North
Syrian centre as Carchemish.59 Indeed, the ivories very probably came from
North Syria, but I suggest that the iron horsebits probably derived from
another area. One may viably speculate that the horsebits arrived at Gor-
dion along with their horse(s), making an overland trip within Anatolia in
the 8th century bc—but from which direction?
Four ivory frontlets and five (Young thought there should have been eight,
for four horses) cheekpieces or blinkers (Figs. 9–10) as noted above were also
found in Terrace Building 2 near the iron horsebits.60 The frontlets, some very
fragmented, are all decorated with a winged solar symbol set above a frontal-
nude female who holds two winged sphinxes by their rear legs. The blinkers
all show a chimera, a winged lion with an en face human head projecting
from the shoulder, and a tail ending in a bird’s head, typical North Syrian
motifs of the 8th century bc.61 Add to this that the ivory lion’s head muzzle,
ear form and lolling tongue are also neatly paralleled on a Phrygian painted
lion that Sams assigned to the EP citadel; and a sphinx depicted on a vessel
from Tumulus P has a bird-ending tail.62
With the high chronology these DL ivories (and also their pottery motif
counterparts?) are now assumed to be imports from 9th century North
Syria—an assumption contradicting M. Voigt and R. Henrickson who, invok-
ing North Syrian parallels, dated the ivories to the 8th century bc.63 How-
ever, now brought forth as 9th century parallels for the ivory frontlet is a
bronze frontlet excavated on Samos, and, for the blinkers, two bronze exam-
ples excavated in Eretria, Greece (Voigt and DeVries, personal communica-
tions); the Samos frontlet and one of the Eretria blinkers bear inscriptions of
Haza"el, presumably the 9th century King of Damascus.64 But if one but looks
at the Gordion ivories and their alleged chronological parallels one instantly
recognises that the Eretria blinkers show a male, not a female, holding lions
by their hind legs; there is no chimera, and they have different end tang
shapes. One also sees that the Samos frontlet exhibits a different iconog-
raphy than the Gordion examples, with three nude females holding their
breasts, above a fourth. Further, all faces, mouths, eyes, hands, and hair, and
bracelets, anklets and necklaces, differ notably in style and body decoration
from those in the Gordion examples.
60 Young 1962, 166–167, pls. 46, 47, figs. 24, 25; Sheftel 1974, 408–425, pls. 60–62; Prayon 1987,
The winged sun forms on the Samos bronze and the Gordion ivory front-
lets are entirely different, except that their discs have an interior crossed
motif, a rosette in the former, crossed lines in the latter. This interior design
does not appear on any of the Nimrud ivories but it does occur on the reliefs
of two 8th century sites in North Syria, Sakcegözü and Zincirli. And it is on
Nimrud ivories of manifest 8th century date, not the Samos frontlet, that
exhibit the very same pendent palm below the sun disc.65
As for the date of the loci in which the alleged parallels occur, it is
noteworthy that the inscribed Samos frontlet was recovered in a 6th century
deposit, the Eretrian blinkers in a late 8th century stratum. What we have
here are objects apparently dated to the 9th century by inscriptions that
were recovered in levels dated hundreds of years later—all of which is not
necessarily relevant to the Gordion ivory pieces anyway, for the alleged
parallels have nothing stylistically in common with them. At present only
one site that may be brought into a discussion for formal and iconographical
parallels on the Gordion frontlets, and that is Nimrud, a site destroyed in the
late 7th century bc. Here as at Gordion are ivory frontlets depicting a frontal
nude female holding lions by their tails, and in at least one example framed
by a guilloche.66
65 Orthmann 1971, pl. 49, A/1, 50, A, 10, pl. 66, K/2; Sheftel 1974, 196–197.
66 Sheftel 1974, 417; Orchard 1967, pls. XXVIII–XXXI; Prayon 1987, 185–186.
destruction of the early phrygian period at gordion 587
Imported Pottery
A glazed juglet (Fig. 11) of Syro-Palestinian form was found inside a vessel
in Terrace Building 4.67 The juglet is a classic Iron II shape and can be
dated to the late 8th century bc, continuing into the 7th–6th centuries
(Larry Herr and Bruce Routlege, personal communications)—which neatly
informs us that it could not even have been a gleam in a potter’s eye in the
9th century bc.
There is also a small corpus of Proto-Corinthian sherds, not from the
DL, but some from the citadel. Although ‘not from a stratigraphically infor-
mative context’, DeVries assigned them to the EP period, assuming they
were picked up in pit digging; Sams, noting that they were recovered ‘scat-
tered and dislocated’, also concluded that they probably derived from the EP
level.68 Voigt and Henrickson disagreed, stating that the sherds (incorrectly
described by them as ‘semi-complete’) were recovered in the MP period.
They dated the sherds to ‘late 8th or early 7th century’; Sams dated them to
about 725–720; Mellink to 720–690 bc.69 The issue here is that some Gordion
67 Sams 1993, 552, pl. 94, 2; 1994, 64, 271, pl. 86, no. 786.
68 DeVries 1990, 390, fig. 25; Sams 1979, 47, fig. 3; 1994, 135; also Mellink 1991, 631, note 27.
69 Voigt and Henrickson 2000, 52; Mellink 1981, 269.
588 chapter eighteen
scholars interpret them as originally imported during the EP period (see the
stone reliefs, below), others that they were imported in the MP period.
The original loci of the sherds will never be determined, nor perhaps
may the lack of a more precise chronology for the terminal dates assigned
the sherds be resolved. But if the terminal date of ca. 690bc or even a
decade earlier holds up (i.e. 700 +/- bc), one is aware that each interpretation
regarding their original locus equally suggests/indicates either a date for the
time for the D1 or the time when the MP was coming to be or was already in
existence—and that is the late 8th or early 7th century bc.
Three bronze bull’s head attachments were recovered from the DL, two in
Terrace Building 1, one in Megaron 4; the two from TB 1 are a pair certainly
from a cauldron, which leaves one missing from Megaron 4 (see Note 35).70
They relate stylistically and structurally to the bull’s head attachments on
cauldrons from Tumulus W and MM of the 8th century bc,71 a date recog-
nised by DeVries and Muscarella for these and other bull’s head attachments
in the Near East.
Orthostate Reliefs
copied at a time that approaches ‘the date of the parallel material and thus
be as early as the ninth century’, no later than 800 bc.72
I argue with ease that awareness about the period of time available for
visual knowledge of and accessibility to the North Syrian reliefs is the objec-
tive and correct archaeological position, and this precludes a 9th century
conclusion.73 Furthermore, as Sams admits, it was in the late 8th century,
and not earlier, that we have information about mutual contacts. Sams’s
9th century date is not a substantive conclusion; it is but a guess (compare
his comments above about 8th century bc iron horsebits and other mate-
rial coming from North Syria in the late 8th century). Voigt and Henrickson
handled Sams’s 9th century date as an historical fact, which it patently is
not; their acceptance was unexamined.74 And this acceptance cannot be
invoked as evidence either to support a 9th century bc date for the local
carving of the reliefs, or as evidence for the date of the DL. One should recall
here Sams’s discussion of North Syrian influences on figures represented on
Gordion pottery, dated by him to the 8th century bc. He suggested that the
motifs could have been adapted from North Syrian reliefs ‘that were visible
at least until the end of the eighth century’ or even the early 7th century bc.75
Phrygian Inscriptions
nos. 60–63, do not derive from the pre-DL period as stated there.
592 chapter eighteen
88 Muscarella 1967, 8.
89 Young 1969, 255, 265, 296.
90 Lejeune 1970, 64; DeVries 1990, 64; Sams 1994, 176.
91 Muscarella 1989, passim, 337; see also above and note 50.
destruction of the early phrygian period at gordion 593
jump to the late 10th–early 9th centuries bc. I argue that the archaeology of
Gordion does not allow acceptance of these important conclusions.
Cimmerians
1991–1992. 565.
594 chapter eighteen
Voigt and Henrickson tell us that the ancient historians who refer to
Cimmerians and Phrygia ‘lived hundreds of years after the putative events’
and therefore cannot inform us about who destroyed Phrygia (Gordion).96
Only modern scholars, who are more competent and knowledgeable, can
do this correctly. But it is a matter of fact that the judgments of these
ancient historians, based on ancient texts available to them, were not far
from historical chronologies and data known to modern historians from
ancient Assyrian texts. Ivantchik persuasively notes that we cannot claim
to have in our possession all the original Assyrian documents, especially
letters to and from kings, that could have informed us whether or not
the Cimmerians were or were not in Anatolia between 714—when they
are first mentioned, in north-western Iran—and the 670s; our records are
incomplete.97 Decidedly pertinent in this context is a stone inscription from
Samos dated to 283–282bc that is fittingly cited by Ivantchik, for it sheds
valuable light on the issue of ancient knowledge of the Cimmerians.98 The
inscription records a boundary dispute between Samos and Priene, a city in
western Anatolia, and refers specifically to earlier, preserved and archived
documents reporting the Cimmerian invasion in the mid 7th century bc,
i.e. 400 years earlier. This inscription neatly documents for modern scholars
the fact of ancient record keeping by people who ‘lived hundreds of years
after the [not putative] events’.
Voigt and Henrickson cite DeVries as one who has ‘convincingly elimi-
nated the Cimmerians as the culprit’.99 This strong, but actually unsubstan-
tive claim is based solely on DeVries’ dating of the DL to 710bc, and naming
Sargon II as the culprit. But, albeit accepted by three Gordion scholars (but
only for a short time, until Heidelberg changed their minds) the 710 date is
a modern fabrication: no such event is recorded in any ancient, including
Assyrian, texts. We are left after all with the Cimmerians as the only histori-
cally documented invaders of Phrygia.
Voigt and Henrickson not only reject the Cimmerians as the culprits, they
also assert ‘there was little evidence for a battle’, and deny that it was a battle
that caused the DL in the first place (T. Cuyler Young agrees with this assess-
ment, personal communications). To Voigt, only the graphic evidence from
Hasanlu IV in north-western Iran, with its over 200 dead in the burnt and
bia’, IrAn XXXII, 1997, 91–108) rightly rejects the high C 14 determined chronology accepted
by the excavators of Rumeilah in Arabia (14th–9th centuries bc) because of much later-dated
Iranian influenced pottery recovered there (fig. 2, B, C, E, G). But Magee in turn dates this
pottery—and thus the site—still too high by probably a hundred years; it probably existed
not earlier than the 8th century bc (painted vessels from Sialk are not manifestly 9th cen-
tury bc). Moreover, Magee himself at his site Muweilah in the Emirate of Sharjah enthusi-
astically accepts here a C 14-based initial date of 10th/9th century bc (see for one report:
596 chapter eighteen
Addendum
In June 2003, after the present paper had been accepted for publication,
I read on the internet an article by K. DeVries, P. Kuniholm, K. Sams and
M. Voigt, ‘New Dates for Iron Age Gordion’, published in that month’s issue
of Antiquity. The article is short but mentions the C 14 dates ‘extending from
824–807bc …’, and it singles out as ‘independently support[ing] the late
9th century date’ the ivory horse trappings and the bronze griffin handle
(called a ‘protome’), citing Haza"el horse trappings and the Hasanlu bird
handle (‘protome’) as the relevant chronological parallels. Fibulae and other
artefacts are not mentioned. The authors assert that their conclusions indi-
cate ‘a severe gap between Phrygian material culture and the more modest
culture of Geometric Greece …’. The present paper, I suggest, demonstrates
that their conclusions both on the date of the Gordion artefacts and a now
revealed cultural unbalance of an alleged 9th-century Phrygia with that of
Greece cannot be accepted.
Acknowledgments
‘Excavations at Muweilah 1997–2000’, Proc. of the Seminar for Arabian Studies 31, 2001, 115–
130). And he claims as chronological support bridge-spouted vessels recovered there, asserted
by him to be Iron II forms—9th century bc. In fact, the Muweilah vessels have vertical, not
horizontal spouts (ibid., fig. 12), a characteristic not of Iron II, but of later vessels, dating after
800 bc (when Hasanlu IV was destroyed). Further, to seek a model for the columned hall exca-
vated at Muweilah one need not with the excavator look to far away 9th century bc Hasanlu
for a direct source: exact and geographically closer models existed in Iran in the 7th and 6th
centuries bc at Nush-I Jan, Godin II, Pasargadae, and Persepolis (Muscarella 1988a, 208–209,
note 3). Magee should not automatically have accepted the C 14 date assigned to Muweilah—
for the very same reasons he argued against accepting the C 14 date assigned to Rumeilah!
destruction of the early phrygian period at gordion 597
Bibliography
Abbreviations
ACSS Ancient Civilizations from Scythia to Siberia
AJA American Journal of Archaeology
AnatSt Anatolian Studies
BASOR Bulletin of the American Schools of Oriental Research
IEJ Israel Exploration Journal
IrAn Iranica Antiqua
IstM Istanbuller Mitteilungen
JFA Journal of Field Archaeology
QRArch Quarterly Review of Archaeology
Introduction
This paper continues the discussion presented in Muscarella 2003 (see also
Muscarella 2005/2006: 395, and note 4) concerning the date of the destruc-
tion level (DL) at Gordion that terminated the Early Phrygian (EP) period
there. I argued that the destruction occurred some time close to 700bc (+/-),
not in the late 9th century bc as maintained in publications and public lec-
tures by the Gordion Team excavators since 2001. The aims of the present
paper are to augment some of the issues I raised previously and to present
additional and relevant information, thus to expand the data available in the
published record. One of the stimuli that generated this review is the grow-
ing number of scholars who have uncritically (to me, without reflection)
accepted the 9th century bc destruction date, and thereby simultaneously
embraced the consequent profound historical and archaeological implica-
tions for first millennium B.C. Aegean and Anatolian archaeology and his-
tory. For example, Prayon (2004: 611) states that the New Chronology has a
“weitreichenden Konsequenz für die historisch-politischen wie auch künst-
lerischen Entwicklungenun Zusammenhänge,” which he then proceeds to
document; Prayon and Wittke (2004: 122–123) note that “das bisherige Bild
der phrygischen Kultur und des Phrygischen Reiches grundlegend verän-
dern ….” See also Kelp (2004: 286, 293); Strobel (2004: 259, 265–268); Genz
(2004: 221, 224); Dusinberre (2005: 4, 10, 220–222); Crielaard (2007: 223); and
Summers (2006: 2).
* This chapter originally appeared as “Again Gordion’s Early Phrygian Destruction Date:
ca. 715 +/- B.C.,” in Studies in Honor of Aykut Çinaroğlu, eds. E. Genç and D. Çelik (Ankara:
Ekici Form Ofset, 2008), 175–187.
602 chapter nineteen
1 The Gordion Team consists primarily of a quartet, Mary Voigt, Kenneth Sams, Keith
Dendrochronology
The starting point for this dramatic back dating of the destruction at Gor-
dion was generated solely from the C-14 laboratory report. But quite soon
thereafter dendrochronological evidence from structural timbers recovered
in the DL Terrace Building (TB) 2A was soon introduced as both further
and independent evidence supporting the 9th century date (Manning et
al. 2001: 2534). Here four scientists (a Cornell Team) report (in note 28 cit-
ing as the source M. Voigt “personal communication”), that “The last pre-
served ring from construction timbers in Terrace Building 2A … is now dated
ca. 883+4/-7bc,” and which terminal ring-date “approximates construction
of this building, which was destroyed … between 830 and 800bc” (italics are
mine; see also Voigt 2005: 30–31; DeVries 2005: 37 declared with ease that the
destruction occurred “not later than 805bc”!). The crucial message casually
introduced here (by another university laboratory) is that an unmeasurable
number of missing rings from burnt timbers is irrelevant in chronological
and archaeological investigations; they are not to be questioned by archae-
ologists seeking to determine the absolute date a structure was constructed.
This interpretative leitmotif continued over the years.
Earlier, Voigt (1994: 273) had provided an accurate archaeological analysis
of the absence of chronological value of the very same TB 2A roof beams.
She noted that “Kuniholm states” (personal communication?) that “the
latest ring on these timbers is 908 +/- 37bc” (on this specific date see CC3
below). But she correctly saw fit to record that no bark was preserved on the
samples, the number of missing rings is unknown, the beams were burnt
and also suffered excavation-process damage, and, further, that the “beams
represent re-used or stored timbers, cut long before this structure was built.”
This is a precise and neatly stated archaeological interpretation: but fully
ignored by the Cornell laboratory (in Manning et al. 2001). It should not go
unnoticed that Voigt’s 1994 disinterested interpretation of the TB 2A timbers
was made before the revelation of the New Chronology—and the later 2001
chronological claim was made after that revelation (see also Kuniholm on
CC 3, below; for others who raise the issue of wood re-use as a fundamental
component in dendro interpretation see James 1991: 323, Keenan 2006: 11,
Mielke 2006: 83, and (yes) Manning et al. 2007: 4). Thus, over the years the
date of the preserved rings went from 908 to 883 +/- bc—with no discussions
regarding the reasons for the changes.
DeVries et al. (2003: 1), and Voigt (2005: 30–31) suggested the possibility
of wood re-use for DL structural timbers. Contrary to the two other dates
previously reported for TB 2A—908 and 883bc, Voigt now writes that “the
604 chapter nineteen
latest ring [is] dated ca. 861bc,” again with no references to previously
published dates or reasons for the modification.2 The use of Manning et al. in
2001 of TB 2A timbers in support of a specific chronological determination
is therefore puzzling. For aside from the wood-condition problem along
with the correct archaeological conclusions articulated earlier by Voigt, the
new ring determination ignores, indeed contradicts, previously published
dendrochronological evidence and interpretation acquired from another
EP/DL structure, CC3. Wood recovered here manifestly demonstrated that
burnt beams cannot be used to furnish construction dates of structures at
Gordion. I quote Kuniholm (1988: 8), “It is clear [sic] that wood was re-
used at Gordion, sometimes many centuries after it was cut”—a conclusion
he determined from three pieces of wood here that “were cut about four
centuries earlier” in the 12th century bc, 400 years before eight other wood
pieces from the same structure.3 Concerning the later-cut CC3 timbers,
Kuniholm (1988: 8) informs us that “… the CC3 Master, composed of eight
samples,” [8 separate timbers?] ends in MMTRD 1826—sixty two years after
MMT [tumulus MM: Midas Mound].” That is, they were cut 62 years after
the cutting date of the outer logs surrounding MM: MMTRD 1764 represents
the terminal ring that grew just before the tomb’s logs were felled (ibid: 5).
All the final rings of the tumulus MM logs have not been preserved (even
though the bark is extant) but Kuniholm “thinks” he has the accurate date.
The cutting date of these logs is presented as 740 + 4/-7bc by Manning et al.
(2001: 2534); previously it had been dated to 718 + 1/-1 bc (see DeVries 2005:
43). To date the ca. 740 bc determination has not been revised (although
given past prouncements it may very well be.4 The 740 MM date is more
or less close to that suggested by some scholars based on artifact analysis
and ancient records Muscarella 1982: 9, idem 2003: 230–231, and note 28;
mented. For example, a recent email message to me (October 3, 2007) from an interested
party claims that Kuniholm (in a personal communication to him) believes “the last ring of
TB2a is 850bc” I think it best not to cite it in my text.
3 See James 1991: 323–324. See also Muscarella 2003: 227: here I mistakenly thought that
the CC3 wood derived from the post-DL, Middle Phrygian period. In 2003 I was thinking then
only about re-use per se, and therefore missed the chronological implications of the later-
dated timbers.
4 Indeed, one cannot object to modifications of chronology based on more accurate
readings of tree rings, but the changes should be explained, and, each time it should be
indicated that the date is indeed tentative. In one recent example, Kuniholm dated the cutting
year of logs at Ayanis to 655–651 bc, causing important and unexpected revisions to royal date
years. Years later he announced that the date was somewhere in the 670s: Manning et al. 2001:
2534, and A. Çilinğiroglu, in press.
again gordion’s early phrygian destruction date 605
Caner 1983: 9, 201). Missed by me and others, including the Gordion Team, is
that Kuniholm 1988: 6–7 noted in a chart (but not in the text) that six other
structures in various strata at Gordion contained wood re-used centuries
after the log’s cutting date.5
For the record, here are some chronological implications of Kuniholm’s
1988 report on CC3, working from his designated cutting date for tumulus
MM, and subtracting 62 years. One arrives at these possible dates for the
last preserved ring (not cutting) of wood beams in the DL, 682, 678, or 671bc.
But (not employing Gordion Team dendro analyses), these dates surely only
indicate that EP Gordion was destroyed sometime after these years in the
early 7th century bc (recognized by James 1991: 323–324—but not by me in
2003).
However, the information and conclusions published about CC3 in
1988 have now been thrown into the bulging file of altered and rejected
data labeled “Old Chronology.” In an email communication to me (June 29,
2007),6 Kuniholm rejected and modified his previously published conclu-
sions, and changed both the published data and his analysis of the 400-year
tree ring spread along with its chronological implications. He wrote that
the samples were “charcoal,” (in another email publication, July 12, 2007,
they were “horribly burned”), and further, that in 1988 he was then—nota
bene—“operating under the old assumptions about tumulus MM and the
DL, and in retrospect was trying to force a fit where there really wasn’t
any” (italics mine; this sentence speaks volumes about dendrochronological
analysis and conclusion-formation within the Gordion Team). In addition,
Kuniholm writes that he has since “found missing rings” that make “the
old ‘fit’ look nothing more than a random resemblance.” The result is that
“The fit with everything else at Gordion is now good. The end date for CC3 is
MTRD 1595 or 909+4/-7 bc” (italics are mine).7 Thus, the previously reported
400-year gap for the CC3 wood has now been officially narrowed; the latest
beams were actually cut (“end date”) only a century before the New Chronol-
ogy DL date, (and two hundred years before 700 bc). And responding to
my earlier email question about the revised dating of the earlier samples
from CC3, Kuniholm replied (in the same July 12, 2007 communication)
5 And although he mentions timber re-use here, none of the other issues concerning the
wood samples from TB 2A—that they were roof beams, the lack of bark, and missing rings—
are discussed.
6 I follow here the Gordion Team’s common use in publications of invoking a “personal
that they “must be re-used Bronze Age logs.” Translation: the older samples’
date has not changed—they were centuries-old, re-used timbers from the
Bronze Age, unlike the later beams, now interpreted to be fresh-cut later—
all neatly provided by a scientific laboratory. The concepts of missing rings
and re-use are considered meaningful only when they fit into a preconceived
scenario.
I repeat here what I wrote in 2003 (page 227), that re-use of wood at Gor-
dion over long periods of time absolutely precludes any chronological use
of terminal rings as reflecting “cutting dates” for wood recovered in Gor-
dion structures. The Gordion Team ignores this fact. One of the significant
problems in this matter is that they have unreflectively and unequivocally
accepted, without independent analysis and verification, every conclusion
issued from the Cornell University laboratory. Given Kuniholm’s (2007) can-
did (and latest) explanation of his ad hoc methodology used in arriving at
his 1988 interpretation, it is appropriate to ask: are scholars now (again) obli-
gated to believe that this revised “scientific” review does not equally operate
from more recent assumptions about the DL, and that it too is not another
ad hoc attempt to “force a fit,” this time with the dogma of the New Chronol-
ogy?
In the final analysis archaeologists cannot invoke the currently pro-
claimed 909bc date (last preserved treering) to establish the time of the
destruction of CC3 and the DL—because of the possibility of wood re-use,
and because the latest preserved ring says nothing about the date of con-
structing a building.8 Indeed, although preached and widely accepted oth-
erwise, it is a false dichotomy to claim that dendrochronology is an objective
science to be privileged over “subjective” archaeology. It is therefore signif-
icant (and surprising) to find this correct position precisely stated recently
by the Cornell Laboratory staff—for the first time: “Dendrochronology is not
an exact science” (Manning 2007: 3).
Until the published (and unpublished) reports cited above are reinvesti-
gated and published in detail, giving all information related to the condi-
tion of the wood involved, the presence or absence of bark, the evidence
and reasons for both the original and subsequent date changes, and the
9 I came upon Manning 2007 after I had essentially finished writing this paper, courtesy
of D. Keenan. Aside from informing us (subtly—I had to read it several times to understand
it) of the review to be undertaken, the report provides a clear discussion about dendro
methodology and the many problems involved in analyses: see also Mielke 2006: 77–84.
10 It is relevant here to remind scholars of the (if correct!) spectacular discovery by
Kuniholm of a central joist within a modern house on the Black Sea that was cut 6200 years
ago (Aegean Dendrochronology Project December 2001 Progress Report: 8). Had this house been
burned down 50 years ago and recently investigated, no doubt some dendrochronologists
would have dated the house to the early Neolithic period.
608 chapter nineteen
dendro examination at Gordion has proved (yet again!) that timbers can-
not be used as chronological markers. In the most recent Cornell University
Weiner Laboratory report (December 2007: 3) is the statement by Sturt Man-
ning that: “The date for the last preserved ring” of juniper logs from Building
A, from the MP (i.e. post DL!) citadel, is 991 + 4/-7bc. But unfortunately his
report refrained from confronting the fact that the MP period is dated by
the Gordion Team post DL, to the late 9th early 8th century bc (but to me
late 8th at the earliest, or the early 7th century bc), and that either way we
have here evidence of reused beams; and that a beam with its last ring dat-
ing almost a century earlier than those from the EP period buildings TB 2A
and CC3 (above), is not employed here using the same methodology to date
the MP period to the 10th century bc!
Dendrochronological dating of tumuli has suffered the same blinkered
mis-treatment: in several instances we are commanded to accept as an abso-
lute chronological determinant the last preserved ring of a tomb’s timber
construction. Thus Strobel (2004: 271, 276 and note 10) reveals to archaeolo-
gists that the plundered and destroyed Kayran Mevkii tumulus (a Gordion
tumulus) was constructed in the 9th century bc (862 + 7/-3)—because of a
(not fully published) dendrochronological dating of burnt logs—with no
bark; and no tomb artifacts were preserved. In the same sentence (p. 276)
we are also commanded to know that there exists other Phrygian tumuli
constructed in the 9th century B.C, tumulus W (see below), and also the, to
me, 8th century bc Mamaderesi tumulus (another Gordion tumulus), which
date is known to him “auf die Funde;” in a footnote (10) he mentions the 80
fibulae recovered. The chronology of the tumuli as proclaimed by Strobel,
is of major importance to his general Early Chronology project, and he sim-
ply announces that the 80 fibulae were made in the 9th century bc (pace
Muscarella 2003: 231). This chronology can be based only on an a priori deter-
mination about the DL (and consequent tumulus construction) chronology:
i.e., a circular argument. The Mamaderesi fibulae include Types XII, 2, 9,
13, 13A, 14, and 14A, only one of which type occurs in tumulus W (9th cen-
tury to Strobel), and there are no 7A examples here, although they are quite
common in W. A number of the Mamaderesi fibula types were recovered in
manifestly 8th century bc tumuli, such as (even to Strobel) MM, which con-
tained XII, 7, 9, 11, 13, and 14 examples. The late 8th century or slightly later
tumulus S-1 (viz. DeVries 2005: 39–40, notes 4–6, 43; Kohler 1995: 115–140)
also has two XII, 14A buckles (ibid. 127–128, pl. 66), a late form (compare
Strobel 2004: 276 and note 10). An appropriate question raises itself based
on artifact analysis: why should archaeologists not assert that tumulus MM
must also be dated to the 9th century bc, inasmuch as it shares with the
again gordion’s early phrygian destruction date 609
DL and Mamaderesi tumulus XII, 9, 13, and 14 fibulae forms? (See also the
discussion of the clay foundation deposit below.)
For the record: that tumulus MM did not contain the burial of king
Midas—because of its construction date determined by dendro (more or
less) and artifact analysis—and that it belonged to an earlier ruler, probably
Midas’ father, has been argued for a long time: by R.S. Young after its exca-
vation, and subsequently by a good number of other scholars, including me,
for decades (see Muscarella 1982: 9, idem 1995: 99, note 13, idem 2003: 231,
note 28). This has been ignored by the Gordion Team in lectures and writing:
viz. Kohler 1995: 192, 228; Manning et al. 2002: 2534; DeVries 2005: 42; DeVries
et al. 2003: 2; DeVries et al. 2005: 45. Voigt (in Yildirim and Gates 2007: 311)
writes as if we now know for the first time ever that Midas is not buried in
MMT. And Strobel (2004: 276) indicts me for continuing to maintain there
is a relationship between MM and King Midas and his family. This specious
charge requires no response, but I will state simply what has often been rec-
ognized as obvious: I believe a long with other archaeologists, that MM is
manifestly a royal burial based on its labor intensive size and construction,
its locus, and its artifact deposition and chronology coincide with/are quite
close to that of the historical King Midas/Mita. Its chronology strongly sug-
gests however, that Midas himself is not buried there.
The Gordion Team (DeVries et al. 2005: 46) asserts its position loud and
clear, “The radiocarbon and dendrochronological evidence provides a firm
and consistent absolute chronology for ninth [Early Phrygian] and eighth
century [Middle Phrygian] Gordion” (italics are mine). (What is consistent
is their circular-citation of each other’s publications—but very few of other,
relevant publications.) For analytic criticisms and rejections of the “scien-
tific” consistent claims blithely made by DeVries et al. 2005, see Keenan
2004, idem 2006, James 1991: 322, idem 2002. Artifacts are not mentioned
as evidence in the above quoted claim, but are mentioned in the article
itself (p. 45), with the claim that they “independently indicate a late 9th cen-
tury” New Chronology date; see also DeVries et al. 2003: 2, Voigt 2005: 31, and
idem, in Yildirim and Gates 2007: 311. These artifacts include fibulae, horse
trappings, vessel attachments, ivory horse pieces, attachments, pottery, and
orthostates, all of which are chronologically evaluated in Muscarella 2003:
237–245, although not confronted by the Gordion Team. On an (attempted)
objective basis, i.e. not a priori using the old date of ca. 700bc for the DL
as a guide, I argued that the DL artifacts are 8th, not 9th century produc-
tions. I have not changed my mind, and those who disagree should confront
each of my evaluations. Strobel (2004: 278), consistent in asserting, but not
demonstrating a 9th century date of each DL artifact, does not.
610 chapter nineteen
Fibulae
11 I use standard terminology here and not Caner’s 1983 unnecessary new terminology,
intended to replace Blinkenberg, Muscarella, Boehmer, etc., and pace Strobel 2004: 276,
note 5.
again gordion’s early phrygian destruction date 611
The Phrygian fibulae from several of the tumuli parallel the DL fibulae.
The most common type from the burials, Type XII, 7A, occurs in tumulus W,
KIII, KIV, S and G; XII, 5 in KIII; XII, 9 in KIII, KIV, MM, S-1, and Mamaderesi;
and Type XII, 14 in KIV, MM, S-1, and Mamaderesi. In the Ankara tumulus
excavated by Sevim Buluç, Type XII, 7A, 9 and 14 forms also occur together
(Muscarella 2003: 233). All these tumuli are dated to the second half of the
8th century. A review of earlier scholars’ determinations on the relative
construction dates of the tumuli yields the conclusions that MM, P, KIII,
and KIV are relatively chronologically close to each other, W is earlier, and
all date to the 8th century bc. DeVries (2005: 43) dates tumulus MM before
Mamaderesi. Strobel (2004: 267, 276–277) dates tumulus KIII and P close to
the time of the DL, and for tumulus W makes another throwaway assertion
of a date: “bestimmte Fundstücke belegen ein Datum im späten 9. Jh. V. Chr.”,
but which Fundstücke are not shared with us. Wittke (2004: 257, note 361;
258, 268) dates tumulus W to 850bc and KIII to the first half of the 9th
century bc. Crielaard (2007: 223) dates W to ca. 850 bc “on the basis of recent
radiocarbon dating,” which in fact does not exist for this tomb (although he
may have meant the recent DL carbon dating). All these chronologies follow
manifestly from the authors’ acceptance of the New Chronology—for not
one provides detailed artifact analyses. Tumulus W is indeed the earliest,
in my opinion dated ca. 750 +/- bc (Muscarella 1982: 8, idem 1995: 97, and
idem 2003: 229; also Kohler 1995: 191–192).
Post 2001 the Gordion Team also supports a 9th century bc date for
tumulus W (verbal communications) primarily because of the presence
therein of XII, 7A fibulae, which to them are chronologically paralleled by
their presence in the DL—and therefore 9th century bc. If this is the case,
then KIII, KIV, and G are also to be dated to the 9th century. But what of the
other later DL fibula forms, Types XII, 7, 9, and 14, all of which occur in other
tumuli, especially MM, which tumulus all the involved scholars date more
or less to ca. 740 bc?
For a discussion of the chronological relationship of the Gordion tumuli
based on artifact analyses, the comparisons among them, and conclusions
that they were not constructed far apart in time one another within the
2nd half of the 8th century bc, see Muscarella 1982: 8, idem 1989: 337–342,
and idem 2003: 229–230. Whether these conclusions are correct or not, the
analyses were based on an analysis of the finds and not on a preconceived
date for the DL. Strobel (2004: 276–277) disagrees with this chronology, but
without meaningful discussion of the parallels brought forth. In the final
analysis, scholars will have to decide which chronological interpretation in
fact “drehen sich immer wieder im Kreise.” The vital point that I repeat here
612 chapter nineteen
is that all the Phrygian fibulae forms recovered in the DL are also present in
the tumuli: that are accepted by all parties to have been constructed in the
8th century bc. Moreover, no Phrygian fibulae recovered in the West (none
are Type XII, 7A) have been dated by pre-8th century bc (Muscarella 1989:
338–339, and note 21).
Numerous other artifacts recovered in the DL have yet to be published,
and thus remain to be analyzed for contributions to the chronological issues
(Maya Vassileva is presently engaged in publishing the metal remains).
Another locus at Gordion for the fibulae remains to be evaluated, one that
reinforces the Old Chronology. Only recently did I become aware that there
are more fibulae available for chronological discussions than I had hitherto
realized, that for some inexplicable reason I had missed. I argue that these
fibulae must be brought into the discussion about the date of the DL, and
that they reinforce my original artifact analyses and dating.
Mary Voigt’s excavations (Voigt and Henrickson 2000: 52) produced a sig-
nificant fundamental adjustment in our knowledge and understanding of
the nature of the cultural and chronological sequence at Gordion follow-
ing the destruction. She determined that the citadel was resettled above the
DL, not 150 years later (as posited by Rodney Young), but quite soon after,
“with little or no gap” (Muscarella 2003: 227–228). In DeVries et al. 2003: 2 (for
which Voigt was a co-author) the rebuilding time was modified to “a gener-
ation after the destruction.” Later, however, Voigt (2005: 32) reinforced her
original conclusion (ignoring her 2003 position), claiming that the recon-
struction “began immediately” after the EP destruction.
If this stratigraphical/chronological adjustment holds up it seems the
labor-intensive and well-planned rebuilding process began soon after the
destruction of the EP citadel, although whether within a year, or a few years,
or a decade (more or less) remains unknown. This well-organized rebuild-
ing was accomplished first by leveling the DL debris and some walls, then
covering most of the DL with a massive clay deposit two meters thick, within
which were set the stone rubble foundations for the structures forming the
settlement that was built directly on top of the clay (see Voigt 1994: 273, pls.
25.5, 6.2; 2005: 33, fig. 3–7A, B). Closely copying the plan of the EP citadel—
which is another indication that rebuilding occurred soon after the destruc-
tion (Voigt 2005: figs. 3-3, 3-4)—the post-clay settlement is designated Mid-
dle Phrygian.
again gordion’s early phrygian destruction date 613
Figure 1.
12 DeVries 2005 includes a modern reconstruction of the South Cellar’s stratigraphy (the
section, Figure 4–2, was made in Philadelphia, not Gordion), and it presents confusing dating
and loci of the Greek pottery excavated there. It is disturbing that DeVries made no mention
in this article of his and K. Sams’ claims made years earlier about the loci and associated
again gordion’s early phrygian destruction date 615
Figure 2.
Arrowheads
problems relating to the Greek sherds under discussion (as if for the first time). For in fact, in
earlier publications, DeVries claimed that the pottery derived from pre-destruction contexts
“although none is from a stratigraphically instructive context,” and Sams claimed that the
fragments had been recovered “scattered and dislocated” within the city and elsewhere! I
discussed these issues in Muscarella 2003:241–243, with references. It is not cited by DeVries.
For chronological problems at Boğazköy regarding imported Greek pottery, see Muscarella
2005/2006: 394. I suggest that we have not heard the last about the South Cellar stratigraphy
or of the loci of the Greek sherds.
616 chapter nineteen
with fragments of gold foil[;] 5903 B1196: bronze arrow point” (R.S. Young
Field Book, Vol. 78: 14). B1196 is a socketed trilobate arrowhead. Another
(7876 B1503; Figure 2, right) was found below a white floor covering the
clay: “… work continued digging clay. Beneath the white floor … was found
a well-preserved bronze arrowhead, c 4cm long. 3 flanged” i.e. a trilobate.
(R.S. Young Field Book, Vol. 114: 50–51).
There is a third example to be considered (89 SF# 16 Operation 11: a bilo-
bate). This arrowhead was discovered in 1989 by M. Voigt, who generously
gave me the following information (verbal 1992, and in recent email com-
munication): “The arrowhead from OP 11 came from a drain that is part of
the construction of YHSS 6B or (early) Phrygian … The stub [on which the
arrow was recovered] is under a drain that was built as part of YHSS 6A, the
construction phase that was mostly standing when the fire [DL] took place
… The problem is that RSY dug down to the level of the PAP/6B and then
below it to the wall stub in 1963 and both lay exposed till we cleared it in
1989 … a lot of people went right across that area for many years of exca-
vation so you can’t exclude the possibility that the arrowhead was dropped
there [post 1963]. We found an Attic black figured sherd on the exposed sur-
face [which?] as well. Since the arrowhead was tucked between stones, it
could be in situ but if it is a critical part of any argument, the possibility
of it being dropped cannot be excluded.” Indeed, I accept this conclusion
as objectively viable. I also add that the arrow could have been dumped—
along with the fibulae and the two arrowheads mentioned above—with the
clay deposit. This arrowhead is best left in abeyance—neither included nor
excluded in the discussion. I record it as evidence of a possible third example
of a nomadic arrowhead from the DL.
The interesting and most important fact about the arrowheads found in
the clay is this: not a single excavated bronze socketed arrowhead is known
from excavations anywhere in the Near East that pre-dates the 7th century bc.
Some surely may have been present since the very late 8th century bc, but
up to the present none this early has been identified. This chronological
situation has been fully documented by Derin and Muscarella 2001: 197, pas-
sim, and Muscarella 2006: 157–158. Thus, the arrowheads in the clay deposit
manifestly document a post-late 8th century bc date for the DL. Indeed, sock-
eted arrowheads traditionally have been culturally associated with intrusive
nomadic peoples (first recorded in the late 8th century bc—the earliest date
is 715 bc, for the Cimmerians) but they were soon thereafter employed by
other cultures (Derin and Muscarella 2001: 197–203; Muscarella 2006: 157–
158). However, their presence at Gordion in a post-late 8th century bc clay
deposit context surely suggests a Cimmerian presence, as I and others have
again gordion’s early phrygian destruction date 617
argued for some time, pace the Gordion Team’s rejection (Muscarella 2003:
247–249; Muscarella 2005/2006: 395. For possible Cimmerian presence at
Norşuntepe and Boğazköy, both because of artifacts recovered including
socketed arrowheads. And in northern geographical locations, see Derin and
Muscarella 2001: 199–200; for a probable nomadic burial at Gordion in tumu-
lus KY, where two horses were buried, but no arrows (or Phrygian fibulae!)
were recovered, see page 195).
Fibulae do not lie (fib); to this (growing) list of truth speakers we now add
bronze socketed arrowheads. Both occur within the clay deposit/platform
overlying the DL, constructed soon after the destruction and almost simul-
taneously with the erection of the MP settlement. Both the fibulae and the
arrowheads date the DL to ca. 700 +/- bc, a time just preceding the beginning
of the MP period. The New Chronology decreed for the DL must therefore be
rejected because: the C-14 data have been accepted ad hoc without confir-
mation from a second laboratory’s analysis and discussion of statistics (see
also above); the offered dendrochronological conclusions and their inter-
pretations are flawed; and archaeological analysis of the artifacts preserved
in the DL and the clay do not support the New Chronology, they contravene
it. Fibulae may indeed be cited as an axiomatisches Argument, they most
certainly are one of archaeology’s best chronological indicators, a classic
Leitfossil, as are socketed arrowheads. Such realia cannot casually be ignored
or manipulated.
Bibliography
J. Bjorkman 1999: “How to Bury a Temple …,” in Studies in the Civilizations of Nuzi
and the Hurrians, vol. 10, ed. D.I. Owen and G. Wilhelm: 103–122.
E. Caner 1983: Fibeln in Anatolia, München.
J.P. Crielaard 2007: review of Wittke 2004, in Bibliotheca Orientalis LXIV, 1/2: 220–226.
Z. Derin and O. White Muscarella 2001: “Iron and Bronze Arrows,” in A. Çilingiroğlu
and M. Salvini, Ayanis 1, Istituto per gli Studi Micenei ed Egeo-Anataolici, Rome
K. DeVries 2005: “Greek Pottery and Gordion Chronology,” in Lisa Kealhofer ed., The
Archaeology of Midas and the Phrygians Recent Work at Gordion, Philadelphia:
37–55.
K. DeVries, P. Kuniholm, G.K. Sams, M.M. Voigt June 2003: “New dates for Iron Age
Gordion,” Antiquity 77 (Internet: http://antiquity.acuk/ProjGall/devries.html).
K. DeVries, G.K. Sams. M.M. Voigt 2005: “Gordion redating,” in A. Çilingiroğlu and
G. Darbeyshire, eds., Anatolian Iron Ages 5: 45–46.
E. Dusinberre 2005: Gordion Seals and Sealin, University of Pennsylvania Museum.
H. Genz 2004: “Erste Ansätze zu einer Chronologie der frühen Eisenzeit in Zen-
tralanatolien,” in Die Aussenwirkung des späthethitischen Kulturraumes, M. No-
vák et al., eds. Ugarit-Verlag: 219–228.
618 chapter nineteen
Abstract: Metal artifacts constitute the main component of the cultural material
known from the ancient state of Urartu. The great majority of these artifacts avail-
able to archaeologists for cultural analysis derives from plundered sites, which
proveniences, whether from sites in modern Turkey, Armenia, or northwestern Iran,
are unknown. In fact, only a minority was obtained from excavations at known sites
and tombs. At the same time there is a large quantity of publications on Urartian
metal artifacts, most of which are in museums and private collections in Europe,
the United States, and Japan. Most of the publications, many by museum person-
nel, ignore the reality of the lack of provenience and present subjective conclusions
regarding provenience and interpretation. This paper discusses the consequences
of the non-archaeological methodology of these publications, the nature of which
many archaeologists remain unaware, and how this behavior has interfered with
achieving an accurate knowledge of Urartian culture.
Scholars attempting a cultural study of Urartian metal artifacts experience
immediately several problems. First is acquiring the large number of publi-
cations spread over many venues. Then when investigated, one soon recog-
nizes that the majority of the material published has no archaeological prove-
nience, rather they have a large number of museum, dealer and collector
provenances in many nations. These non-archaeological provenances, a cul-
tural artifact of modern consumerism and greed, signify that only a limited
corpus of archaeologically documented data are available for meaningful
study, juxtaposed to a larger corpus that does not possess this quality. And
further, one soon becomes aware that, notwithstanding the fundamental
significance of the provenance/provenience, locus/non-locus relative pro-
portion reality, this issue remains basically unperceived, certainly under-
appreciated, and at times rejected in scholarly discussions. Only a very few
scholars call attention to it. Others note the disproportionate nature of the
excavated data merely as an aside, and then proceed to argue an archaeo-
logical interpretation that ignores the ontological backgrounds of the dis-
tinct corpuses. The consequence is that an entrenched subjective, non-
anchored, anything-goes publication pattern has come to prevail, and its
2 Kellner 1979, 152, Abb. A; Kellner 1980, 206, Taf. I, and Kellner 1993, pl. 61.2; also Born
Concerning the votive plaques (figs. 7, 8, pp. 287, 294–295), Kellner easily
informs us that its provenience “is known at least approximately to be near
Malazgirt” (italics added), pace the captions for the two objects that give it as
“Giyimli?” a site not near Malazgirt. In an earlier publication, Kellner5 gen-
erously confided that his Malazgirt source was in Germany, the brother of
the man who “ausgeackert” the finds (whether at Gyimli or Malazgirt doesn’t
matter), which included fibulae and bronze belts.6 The pedagogical message
here is that archaeology cannot function without the help of curators who
keep in touch with innocent peasants plowing their fields.
Kellner furthermore informs us (p. 287) that his Munich cache of objects
did not derive from a tomb, but “probably from an Urartian settlement in
or near a fortification” (found under a plowed field). These archaeological
provenience data are provided by a museum director while sitting in his
Munich office, and whose readiness and ease to share what he knows neatly
illustrates much of “said to be” Urartian metal artifact scholarship.
The Fundkomplex is a beloved term defined by those who use it as a
corpus of material painstakingly kept together along with the site’s name,
by the innocent but archaeologically astute peasants who casually found
the objects, and then passed them to the honest dealers in Germany; but
smuggling and smugglers within and outside of Turkey are not mentioned.
In Merhav’s catalogue there is a group of 10 Urartian silver and bronze
objects (pp. 206–207, 217–225), most inscribed with the names Ishpuini or
Menua (pp. 221–225). They were earlier chronicled as deriving “aus einem
Fundkomplex” by Kellner who saw them in a Kunsthandelsgalerie in 1975,
and later described to be “Ein sensationeller Fund …”7 Writing in his office
in Munich, Kellner posited that they all probably derived from Inushpa’s
royal tomb, the locus of which he blithely identifies as probably Liç, near
Patnos. Merhav, sitting in her Jerusalem office, disagrees;8 to her they derived
from a “temple treasure” not a tomb locus. Indeed, Liç was plundered, then
excavated as a salvaged site, where a bronze belt was in fact recovered (see
below): but what has this to do with the material Kellner saw in Germany,
Merhav in Jerusalem? We are expected to be grateful to get such valuable
specific archaeological data generously provided by museum directors and
curators (not to mention dealers).
5 Kellner 1982, 84; Caner 1998, vii accepts the Malazgirt label.
6 See also Kellner 1991, No. 70.
7 Kellner 1975–1976, 57, 66–67; Kellner 1991, 8; also Calmeyer 1986, 81, note 10; on this see
9 For criticism of the common occurrence of Schatzfunde and the geschlossenen Fund-
komplex in Urartian scholarship, see Haerinck and Overlaet 1984, 54, 56; Muscarella 1988,
395–396, 398, note 6, 423, notes 13, 14; Muscarella 2000, 146–147, 213, note 54; Garrison 1994,
150; and Simpson 2005—here with a neat discussion of museum anti-archaeological prac-
tices. Seidl (Seidl 2004, 17) correctly warns us about Komplexe, but then accepts others (Seidl
2004, 17, 65, note 37, 132).
10 Biscione 1994; Yildirim 1991.
11 Rehm 1997, 225.
12 Vanden Berghe and de Meyer 1983, 207–208, Nos. 177, 178.
13 Dezsö and Curtis 1991, 107–119, figs. 1, 19, 20: excavated at Nimrud; see also below.
urartian metal artifacts: an archaeological review 627
XVIII—for the latter see below) also publish two other probably Assyrian
examples curated in German museums and in the catalogue labeled Urar-
tian by Kellner.
I further suggest that decoration on some objects exhibited are modern
forgeries, or at least to be considered suspicious (see also below): pp. 87,
Nos. 40, 41, 44; 92, Nos. 51, 52; 122, No. 8; 126–127, fig. 8, No. 9; 129, No. 11 (a
helmet excavated at Ayanis, is of the very same form and bears an ancient
[sic] incised scene):14 166, No. 2; 167, No. 4. D. Collon15 suggests that the gilt
relief scene on the silver bucket, pp. 220–221, No. 22, may be modern—I am
not sure and reserve judgement; however, she does not indict the scene on
p. 129, No. 11. Problem pieces include pp. 106, No. 67, 108–109, Nos. 71–73.16
And there is naked dissimulation. In an essay on bronze candelabra, Mer-
hav mentions six excavated complete examples, four from Karmir Blur, one
each from Toprakkale and Altintepe (pp. 262–267, figs. 10–12, Nos. 10a, b). She
then introduces a bronze example curated in her institution, the Jerusalem
Museum (pp. 264, 270, No. 11), that has “come to light” (another museum
euphemism; here designating an artifact saved by simple peasants from
the dark earth! Like some of the examples mentioned, it is typically Urar-
tian), consisting of an upright shaft placed over a three-legged base, each
of which has a recumbent lion on the curved feet; here only two lions are
preserved, and the missing one has been restored (three candelabra from
Karmir Blur lack the lions). The candelabrum in Jerusalem bears an inscrip-
tion of Menua, son of Ishpuini (p. 357).17 Although the isolated lion excavated
at Kayalidere is correctly identified as deriving from a candelabrum (pp. 265,
274–275, fig. 1), the equally excavated lion from Aznavur-Patnos, also dis-
associated from a candelabrum18 is not accorded an excavated attribution
in her text. In fact, it is only mentioned once (with no bibliographical ref-
erence), buried in a caption for one of the Jerusalem candelabrum lions
(p. 280, No. 5), where it is casually noted, “Similar [sic] in shape and details is
a bronze lion found [not excavated!) at Patnos.” The Jerusalem lion is shown
in front and back view, but not the side view, as accorded the Kayalidere lion.
There is more to consider. The candelabrum was donated to the Museum
in 1972, and in the previous nineteen years the museum had possessed it, it
14 Derin and Çilingiroğlu in Çilingiroğlu and Salvini 2001, 181, fig. 13.
15 Collon 1993, 126.
16 For earlier discussion and references of these objects, see Muscarella 2000, 148–152; for
other forgeries of Urartian art see Muscarella 2000, 147–154; Haerinck and Overlaet 1984, 53,
55 also raise the issue; see also below.
17 Also Salvini 1991.
18 Boysal 1961, 199, figs, 1–3.
628 chapter twenty
of museum (and also collector’s) catalogues, far more than their editors
desired.
The data excavated from the catalogue at large concisely reveal the state
of affairs of Urartian archaeology, and specifically define the complex issues
that utterly compromise a study of the culture’s metal artifacts. Not enough
scholars recognized the catalogue’s underlying message and archaeolog-
ical fictions. One example is the otherwise important historical work of
Salvini,22 one of the very best scholars of Urartian history and language, who
here demonstrates how a scholar can be so readily seduced by a catalogue’s
charm. In his discussion of Urartian metal artifacts (“Die Bronzekunst” chap-
ter), paradoxically only a few excavated artifacts are cited and discussed, and
plunder matters not at all.23 Privileged however, is the material published in
the Jerusalem exhibition catalogue, which to him is “eine Summa unserer
Kenntnisse hinsichtlich der Bronzen und der urartäischen Metallarbeieten
im besonderen.”24 This claim is less a hyperbole than uncannily correct, but
only ironically, and not in its intended sense. One scholar who confronted
some of the issues raised here was M. Garrison in his review of the Jerusalem
catalogue. He recognized its usefulness and importance, but stressed the
non-excavated, plundered nature of the material exhibited, how this fact is
basically ignored, and that some of the contributing scholars “run[s] the risk
of legitimating the very forces that have brought such havoc to the study of
Urartian art.”25
The Jerusalem exhibition of Urartian artifacts together with its catalogue
was not the first such event; several other museum-generated exhibitions
anticipated it in form, content and spirit, albeit less extravagantly. German
museums in particular specialized in exhibiting Urartian artifacts accom-
panied with catalogues in the late 1970s and early 1980s, the first exhibi-
tion setting an example for the others. The first exhibition occurred in
Munich in 1976,26 where nearly 200 objects exhibited derived from muse-
ums, private collections, and dealers (the names of the latter two cate-
gories are identified only by initials), which set the pattern for all suc-
ceeding exhibitions. The accompanying catalogue also provides us with
many examples of “angeblich” or “velleicht” from X, or just plain absolute
“Transkaukasien,” museum-generated proveniences. Noteworthy was the
22 Salvini 1995.
23 Salvini 1995, 170–182; on this issue see Kroll’s apt criticism, Kroll 1997, 204.
24 Salvini 1995, 171; see my review of Salvini, Muscarella 1997b, 737.
25 Garrison 1994, 151.
26 Kellner 1976.
630 chapter twenty
29 Kohlmeyer and Saherwala 1984, 34, Abb. 18, 19; Muscarella 1992, 27; Seidl 2004, 132,
note 743 asserts that I did not read the Berlin text; she has. But ignored in her reading is
that both cauldron and stand are unexcavated, and that an outline for a stand per se would
be expected, but by itself (for no other information is given) does not demonstrate that the
Berlin stand and cauldron belonged together before it was plundered. Seidl did not challenge
the alleged “Grabkomplex.”
30 Vanden Berghe, de Meyer 1983.
31 Özgen 1983, 117, fig. 18; Özgen 1984, figs. 8, 24, 27–33, 36, 45; see also Özgen’s Ph.D. 1979
published by The Ancient Orient Museum, Tokyo.33 Illustrated are 127 Urar-
tian metal objects purchased by Japanese private collectors, who are ac-
corded “deep gratitude” for allowing their possessions to be revealed (but
whose names are not vouchsafed). Indeed, the catalogue publishes a valu-
able bibliography of Urartian studies in many languages (985 entries), but a
number are of works that merely mention Urartu; for an extensive bibliog-
raphy see P. Zimansky.34
The distribution of Urartian artifacts excavated outside their homeland is
important for the study of cultural influence, political engagement or trade
beyond Urartu’s borders, particularly concerning the west, and metal arti-
facts have been the main source of research in most investigations of long-
distant exchanges. Since the 1960s I and H.-V. Herrmann have argued that
examples of large bronze cauldrons bearing bull or siren protome attach-
ments excavated outside the Urartian homeland, at sites in Turkey, Italy
and Greece, which many scholars had concluded over many years were
imports from Urartu (thereby demonstrating widespread knowledge of that
culture in Phrygia and the West) were in fact not Urartian productions. We
argued from a stylistic analysis that they were manufactured and exported
from elsewhere, probably from North Syria, specifically that the cited bull
attachments quite clearly are not the same in style and manufacturing
technique as the Urartian examples, and the sirens are not characteristic
of Urartian style.35 Merhav disagrees, and reprises the old interpretation
that these cauldrons recovered abroad indeed are of Urartian manufac-
ture, including some from Toprakkale, from whose workshops they were
exported West, to Gordion, Italy, and Greece.36 I have not changed my mind
on this important issue—no cauldrons recovered at any site beyond Urartu
arrived from Urartu: of those published in the Jerusalem catalogue, some
are Urartian, one bull attachment (p. 238, No. 31) was excavated, two oth-
ers, unexcavated, most probably have an Urartian background (pp. 240–241,
Nos. 35, 36). Of the other plundered examples, Nos. 30 and 32 (pp. 236, 238),
are Urartian (Nos. 37 and 38, mentioned above, are ancient, but not Urar-
tian). And to me, the cauldron with achments in Munich, is not Urartian
Muscarella 1988, 28, note 4 and Muscarella 1992, 22. I made it clear none derived from that
site: pace Seidl 2004, 3, note 7, where she claims the contrary; she only cited a much earlier
(1962) article of mine.
urartian metal artifacts: an archaeological review 633
(pp. 242–243, No. 39); it was plundered and has no known archaeologi-
cal provenience. The two terracotta cauldrons with bull head attachments
(pp. 239–240, Nos. 33, 34) also have an Urartian background.
In the context of recognizing and defending contacts between Urartu and
the West as active historical events, the museum curator R.D. Barnett leaps
(very high) when he suggested that the throne of King Midas of Phrygia
sent to Delphi could have been made, not by Phrygians, but by Urartians;
that King Urpalla at Ivriz is decorated with Urartian design encrustations,
which designs are also to be seen on plaques excavated in the Ephesian
Artemisium: all of which is serious, indeed embarrassing, distortions of the
archaeological/historical record.37
Up to the present only one, or possibly two, examples of Urartian artifacts
have been recovered outside of the homeland—if, of course, my attribution
arguments about the cauldron attributions are not contradicted by future
excavations. One, manifestly, is a bronze bell from a sanctuary on the Aegean
island of Samos (B474), which has a number of inscribed Urartian examples
as parallels.38 The other is a statue also excavated at Samos (B1217), which
a number of scholars argue is Urartian; I am not certain because its style
is not easy to determine, but cannot categorically refute the attribution.
The Samos bell, and apparently the statue derived, ultimately at least, from
Urartu; how they got from there to a sanctuary on Samos is as intriguing as
it is unknown.
A number of publications report on the occurrence of unexcavated Urar-
tian metal artifacts from loci outside of the states’ known boundaries, or
identify their derivation from sites within the Urartian sphere. I refer here
to scholars who know with ease the precise site or geographical location
of plunder-derived, smuggled artifacts shown to them by dealers or collec-
tors, and share this information. Such determinations most often are gener-
ated in museum offices. It is not necessary to cite every occurrence of this
Leitgedanke, common, practice; a few examples will suffice to document the
wide parameters available to bazaar archaeologists.
In two publications, R. Ghirshman39 shared with colleagues what to him
was an important archaeological discovery. He reports that five Urartian
objects, blinkers and horse bits, all but one inscribed, housed in an Iranian
other objects curated in his museum. He knew immediately when he saw and purchased
urartian metal artifacts: an archaeological review 635
(Nagel 1963, Nagel 1970) in German dealers’ shops that two statuettes derived from one
Fundkomplex, Hacilar. After they were exposed by me as obvious modern forgeries, his
archaeologist wife defended them vigorously as indeed deriving from ancient Hacilar: both
knew—and therefore did not think it necessary to provide us with comparanda: perhaps
because none exist. Nor did they deign to discuss in print the manner in which the said-to-
be plundered neo-Hacilar figurines left Hacilar, Turkey, and reached Berlin: for details and
bibliography see Muscarella 2000, 136–137, 438–439, Nos. 13, 14. The present Berlin Museum
curators and director continue to obey the museum code, for at the time of this writing
(2005) both figurines are still on exhibit—to protect an honored curator, the museum, and
to educate visitors about neo-Hacilar, which in Berlin is merely a semantic difference from
ancient Hacilar.
45 Kellner 1980, 206; Kohlmeyer and Saherwala 1983, 38, Abb. 26; Vanden Berghe and de
catalogued.
47 Taşyürek 1975a, 151, note 3; Taşyürek 1975b, Nos. 1, 19; Taşyürek 1977, 157.
48 Kellner 1991, Nos. 63, 102, 103; the Dedeli attribution for these were rightly challenged
by Curtis (Curtis 1996, 121). Note that a bronze belt was excavated at Dedeli, see Kellner 1991,
No. 373 (below).
636 chapter twenty
49 Taşyürek 1973, 204, Kellner 1991, No. 37 (another belt is listed as “umgebung von Malaz-
recently incised to a helmet is Born and Seidl 1995, Abb. 1–22, 96, 97: “ein Unikum,” (p. 43);
see also Muscarella 2000, 184, No. 42.
58 Kellner 1980, 211–212; Kellner 1993, 326, pl. 60–63; Dezsö and Curtis 1991, 115–116, pl.
XVIII.
59 Kellner 1980, 210–211, Taf. XI–XII; accepted as ancient by Born and Seidl (Born and Seidl
1995, 96, Abb. 85); see Muscarella 2000, 152–153, No. 31.
60 Rehm 1997, 105–108, note 36, fig. 1, Abb. 196–199.
61 Rehm 1997, 105, note 36.
638 chapter twenty
62 Maass 1987, 84, No. 13, Taf. 9,2; 88, No. 18, Taf. 8; Muscarella 2000, 148–149, Nos. 15, 16,
64 Kellner 1980, 207–208, Taf. II–VI, 210; it is the same as Calmeyer in Merhav 1991, 126–127,
No. 9; published as genuine also by Born and Seidl 1995, 26–27. See Muscarella 2000, 149, 460,
No. 17; Seidl 2004, 70–72, Abb. 32.
65 Calmeyer 1986, 81, Abb. 4, Taf. 15–19; see reference to Muscarella in note 56.
66 Kellner 1980, 211–212, Taf. XIII–XIV; Muscarella 2000, 150, No. 19.
67 Kellner 1993, 325.
68 Kellner 1993, 330.
640 chapter twenty
striding animals and addorsed calves’ heads at the handle juncture. Of the
animals, he knows they are Urartian although he casually notes, “there is
little at present to compare them.”69 He shares with us, however, that the
addorsed heads find “a close parallel” to a stone example in the Louvre—
“said to have been acquired in 1898 from the region of Van” (he doesn’t tell
us who said it). A bronze ladle, excavated at Tell Farah, Israel, was published
as an import from Urartu by the archaeologist R. Amiran solely because of
an apparent stylistic accord of its addorsed heads with the British Museum’s
alleged “Urartian” heads.70 Amiran then utilized the “Urartian” ladle exca-
vated in Israel as evidence to suggest that it might prove that the siren
cauldron attachments excavated at Gordion, Phrygia, indeed derived from
Urartu. Here an archaeologist accepts instructions to see around corners, to
recognize how unexcavated, but perceptively recognized, “Urartian” metal
artifacts are capable of revealing ancient activities, here otherwise elusive
long distance trade contacts.
The ladle and mirror were accepted as Urartian by Zimansky, who fur-
ther calls to our attention that Barnett’s mirror is “the first Urartian mirror
known.”71 He also cites T. Kendall’s publication of another unexcavated, pur-
chased “Urartian” bronze mirror, in the collection of the Museum of Fine
Arts, Boston;72 presumably, we should accept this as the second Urartian
mirror known. Kendall “immediately” recognized it to be Urartian, but did
not consider it relevant to note that there are no Urartian parallels. Despite
these museum attributions, both objects are not Urartian, neither in style
nor manufacture, and no mirror or similar ladle has been excavated to date
at an Urartian site. Also, a bronze belt published as Urartian by Merhav,
which Urartian attribution is accepted by Zimansky, was surely not made
in Urartu, but probably somewhere in the Caucasus.73 Taşyürek74 also eas-
ily recognized several objects in his Museum, all acquired from nomadic
dealers, as being Urartian artifacts: a bronze lion, stag figurine, and a female
statuette—found, according to a dealer, in the province of Muş, and naming
the exact locus. Taşyürek also identified a terracotta bull figurine as Urartian;
he presented no Urartian parallels—because there are none.
The Urartian metal industry produced a prodigious quantity of metal
artifacts (minus, of course, those discussed just above), as much as any
ancient culture known to us. Most forms are conveniently presented and
illustrated with bibliographies in the essays dealing with specific categories
in the Jerusalem catalogue, and also in the other catalogues discussed above.
For a chronological study, one should begin with M. van Loon, who not
only for the most part uses excavated material, but also intelligently covers
the range of metal artifacts:75 also G. Azarpay, who concentrates on some
excavated material, but doesn’t discuss the full corpus.76 Then one goes to
Seidl 2004. Also, of course, the examples of excavated metal artifacts from
specific sites, viz., Ayanis and Karmir Blur.77 I record here in list form the
metal artifact corpus available to us: utensils made of bronze, iron, silver and
gold: vessels of many forms, bowls, jugs, askoi, beakers, situlae, cauldrons
with bull attachments; small bronze statuary in the round (rare), sculpture
associated with thrones, furniture and candelabra; jewelry of bronze, iron,
silver and gold, fibulae, straight pins with figured heads, body ornaments
such as earrings, brooches, bracelets and armlets, medallions, pectorals;
weapons of many kinds made of bronze and iron: numerous swords, daggers,
spears and arrows, helmets, shields, quivers, breastplates; many bronze belts
most decorated with iconographic or geometric scenes; votive plaques of
different shapes, but best known from the many hundreds associated with
Giyimli; horse and chariot gear of great forms and variety: yoke and wheel
equipment, studs, horse bells, breastplates, blinkers, nose plates, bits, etc. A
good number of these have excavated backgrounds.78
To succinctly exemplify the dilemma facing anyone studying these arti-
facts even within the excavated corpus, I cite a few prominent examples that
illustrate the confusion and difficulties involved when one seeks informa-
tion about proveniences and cultural contexts.
Many hundreds (Caner 1998 catalogues 813 examples) of unexcavated
fragmented bronze plaques, some decorated in Urartian style and iconog-
raphy, but others not, have been associated by many scholars with the
plundered site of Giyimli, south of Lake Van.79 Giyimli was subsequently
raphy of Urartian metal artifact publications and sites where artifacts have been excavated
is in Seidl (1988, 1, partial, and 2004). Specific essays concerned with metal artifacts at large
include Merhav 1991, Zimansky 1998 (199–227), and Derin and Çilingiroğlu in Çilingiroğlu &
Salvini 2001. A dazzling bibliography of Urartian metal artifacts exists in Rehm 1997: 165–323,
but is very difficult to access as it is embedded in countless (repetitious) notes.
79 Kellner 1982, 80. Caner 1998 is a lengthy, painstaking description of hundreds of plaques
642 chapter twenty
especially his catalogue (Kellner 1991); Curtis 1996, Rehm 1997, 203–209, and Seidl 2004, 133–
168.
85 Curtis 1996.
86 Collon 1993, 127.
urartian metal artifacts: an archaeological review 643
92 Merhav 1995.
93 Sevin and Kavakli 1996, figs. 26–27.
94 van Loon 1966, 43–44, figs. 5a, b; Salvini in Merhav 1991, 11.
urartian metal artifacts: an archaeological review 645
(see below). Sargon’s catalogue of artifacts, his booty, lists enormous quan-
tities of metal artifacts,95 among them hundreds of bronze shields, vessels,
chariots, weapons, and so forth; also a “silver bowl of Rusa” is mentioned
(see also the silver bowls in the Jerusalem catalogue inscribed with royal
names: pp. 224–225, 356, Nos. 27–34). Kendall and Barnett believe that all the
objects inventoried by Sargon were Urartian productions.96 Salvini claims,
probably correctly, that Sargon’s catalogue is a precise listing of the booty
taken; and further, that while objects from different cultures are specifically
reported, most of the temple’s contents were Urartian productions.97 Belli
more cautiously posited that one cannot “draw far-reaching conclusions”
from the catalogue “[regarding] the material culture of Urartu,” as many
of the objects listed derived from other cultures—Tabal, Assyria, Habhu,
attributions mentioned in Sargon’s text.98 Belli’s conclusion is sensible and
viable, it cannot be dismissed, for we know that members of several polities
dedicated votives to the Urartian deity—or that booty from these lands was
dedicated by Urartians there. Rusa’s silver bowl and statue with chariot, and
the lion protome shield are obviously Urartian metal artifacts, manifestly
documenting Urartian productions here. But it cannot be determined how
many of the other recorded artifacts may be accorded the same attribution.
Some important recently excavated metal objects merit consideration.
From Ayanis (yes; not said-to be-from), a site excavated for years by A.
Çilingiroğlu, derives precious archaeological information and data on Urar-
tian architecture and metal artifacts, carefully excavated and their specific
loci recorded. A good number of bronze and iron artifacts have been recov-
ered, increasing the quantity of excavated material considerably. More arti-
facts have since been excavated: as of 2004 the total numbers are 18 shields,
several decorated with typical incised circular lion processions, one dis-
playing a lion projecting from its center, exactly like those depicted on
the Musasir temple relief (noted above).99 Here is a list: 52 quivers, some
with their arrows inside; 16 helmets; two cauldrons (none with attach-
ments); 2 jars; 4 sikkatu, all inscribed; 174 spear heads; 931 arrows; and 28
gold rosettes, palmettes, and buttons. And only 15 % of the site has been
excavated. Another excavated object is a magnificent inscribed shield with
95 van Loon 1966, 84–88, Salvini in Merhav 1991, 12–13, and Belli in Merhav 1991, 20–21,
example of an inscribed artifact assigned to “Luristan” has been excavated there, and they
are thus not “heterogener Herkunft”; see Muscarella 1988, 120, note 6.
104 Merhav 1991, 201, 217, 224, figs. 1.2, Nos. 17, 25, 27; see also Rehm 1997, 167–169.
urartian metal artifacts: an archaeological review 647
inscriptions on metal compared to about 500 on stone (now see Seidl 2004
for the most comprehensive listing).
There is one significant caveat that remains to be investigated. I have not
encountered anyone who has considered whether or not all the unexcavated
inscribed metal artifacts were in fact inscribed in ancient Urartu. No one can
deny for a second the absolute reality that a forger is able to skillfully carve
any Urartian inscription he chooses,105 all the more so a small one, a king’s
name on an ancient metal artifact—to enhance in one act its monetary
and cultural value to scholars and museum staffs (one envisages a smiling
forger imagining: think of all the royal Fundkomplexen they will recognize
and write about!). One could legitimately argue that until all unexcavated
inscribed inscriptions are tested—but not by a museum laboratory—it may
be impossible to argue that an ancient hand wrote the inscription. A warn-
ing is a good place to end a review filled with warnings.
Addendum
105 Or any inscription—see e.g. Muscarella 1988, 283–284, note 4, and Muscarella 2000, 183,
No. 39.
648 chapter twenty
Bibliography
Amiran, R. 1966: Two Luristan Fibulae and an Urartian Ladle from Old Excavations
in Palestine, Iranica Antiqua 6, 88–91.
Azarpay, G. 1968: Urartian Art and Artifacts: a Chronological Study (Berkeley).
Barnett, R.D. 1965: An Urartian Mirror. Anadolu Araştırmaları 2, 1–2, 51–54.
Barnett, R.D. 1982: Urartu. In Cambridge Ancient History 3.2, 314–371.
Belli, O. 1997: Der Bronzegurtel des Königs Rusa in Malatya Museum. Anadolu 23
(Akurgal Festschrift volume), 293–297.
Belli, O. 1999: The Anzaf Fortress and the Gods of Urartu (Istanbul).
Biscione, R. 1994: Recent Urartian Discoveries in Armenia: The Columbarium of
Erevan. Studi Miceni ed Egeo-anatolici 34, 115–135.
Born, H. and Seidl, U. 1995: Schutzwaffen aus Assyrien und Urartu (Philipp von
Zabern).
Boysal, Y. 1961: Anzavur’da Definecilerin Meydana Çikardıǧı: Eserleri. Belleten 25,
199–212.
Calmeyer, P. 1986: Zu einem Ziselierten Bronzehelm des 8. Jahrhunderts. Archäolo-
gische Mitteilungen aus Iran 19, 79–86.
Caner, E. 1998: Bronzene Votivbleche von Giyimli (Rahden/Westf.).
Çilingiroğlu, A. and Salvini, M. (eds.) 2001: Ayanis 1 (Rome).
Collon, D. 1993: Review: Kellner 1991. Bulletin of the School of Oriental & African
Studies 56, 1, 125–127.
Curtis, J. 1996: Urartian Bronze Belts. Zeitschrift für Assyriologie 86, 118–136.
Dezsö, T. and Curtis, J. 1991: Assyrian Iron Helmets from Nimrud now in the British
Museum. Iraq 53, 105–126.
Erzen, A. 1974: Giyimli Bronz Definesi ve Giyimli Kazisi. Belleten 38, 191–213.
Garrison, M. 1994: Review of Merhav 1991. Ars Orientalis 24, 149–151.
Ghirshman, R. 1964: Deux oreillers en bronze des rois d’Urartu. Artibus Asiae 27,
49–60.
Ghirshman, R. 1966: Deux bronzes des rois d’Urartu. Artibus Asiae 28, 219–222.
Gropp, G. 1981: Ein Pferdegeschirr und Streitwagen aus Urartu. Iranica Antiqua 16,
95–123.
Haerinck, E. and Overlaet, B.J. 1984: Zur Funktion Einiger Uraratäischer Bronzege-
genstande. Iranica Antiqua 19, 53–70.
Hoffman, H. 1971: An Urartian decorated Bronze Strip from Diyarbakir. In D.G. Mit-
ten et alii (eds.), Studies Presented to George M.A. Hanfmann (Cambridge, MA),
69–71.
Kellner, H.-J. 1975–1976 (= 1980): Ein datierter Silberfund aus Urartu, Anadolu 19, 56–
67.
Kellner, H.-J. (ed.) 1976: Urartu: Ein Wiederentdeckter Rivale Assyriens (München).
Kellner, H.-J. 1977: Pectorale aus Urartu. Belleten 41, 481–493.
Kellner, H.-J. 1979: Eisen in Urartu. In Akten des VII lnternationalen Kongresses für
iranische Kunst und Archäologie (Berlin), 151–156.
Kellner, H.-J. 1980: Bemerkungen zu den Helmen in Urartu. Anadolu Araştırmaları
8, 207–213.
Kellner, H.-J. 1982: Gedanken zu den Bronzen Blechvotiven in Urartu. Archäologis-
che Mitteilungen aus Iran 15, 79–95.
urartian metal artifacts: an archaeological review 649
Kellner, H.-J. 1991: Gürtelbleche aus Urartu. (Prähistorische Bronzefunde XII, 2) (Stutt-
gart).
Kellner, H.-J. 1993: Ein Assyrischer Beutehelm? In M. Mellink et alii (eds.), Aspects of
Art and Iconography: Anatolia and its Neighbors. Studies in Honor of Nimet Özgüç
(Ankara), 325–328.
Kendall, T. 1977: Urartian Art in Boston: Two Bronze belts and a Mirror. Boston
Museum Bulletin 75, 25–55.
Kohlmeyer, K. and Saherwala, G. 1984: Frühe Bergvölker in Armenien und im Kauka-
sus (Berlin).
Kroll, S. 1979: Urartu. Das Reich am Ararat (Hamburg).
Kroll, S. 1980: Urartu. Ein wiederentdeckter Rivale Assyriens (Muenster).
Kroll, S. 1997: Review: Salvini 1995. Die Welt des Orients 28, 203–208.
Maass, M. 1987: Helme, Zugehör von Wagen und Pferdgeschirr aus Urartu. Archäol-
ogische Mitteilungen aus Iran 20, 65–92.
Merhav, R. 1981: A Glimpse into the Past. The Joseph Ternbach Collection (Jerusalem).
Merhav, R. (ed.) 1991: Urartu A Metalworking Center in the First Millennium BCE
(Jerusalem).
Merhav, R. 1995: Gold and Silver Pins from Urartu: A Typology and Methods of
Manufacture. Tel Aviv 21, 1, 129–143.
Muscarella, O.W. 1978: Urartian Bells and Samos. Journal of the Ancient Near East
Society of Columbia University 10, 61–72.
Muscarella, O.W. 1981: Giyimli Plaques. In O.W. Muscarella (ed.), Ladders to Heaven
(Toronto), 175–178.
Muscarella, O.W. 1988: Bronze and Iron (Metropolitan Museum of Art).
Muscarella, O.W. 1992: Greek and Oriental Cauldron Attachments: A Review. In
G. Köpcke, I. Tokumaru (eds.), Greece between East and West: 10th–8th Centuries
bc (Mainz), 16–44.
Muscarella, O.W. 1997a: Review of Zimansky 1998. Bulletin of the Asia Institute 11,
205–210.
Muscarella, O.W. 1997b: Review of Salvini 1995. Bibliotheca Orientalis 54, 5/6, 734–
737.
Muscarella, O.W. 2000: The Lie Became Great (Styx).
Nagel, W. 1959–1960: Ein urartäischer Helm aus dem Argisti-Magazin. Archiv für
Orientforschung 19, 144–147.
Özgen, E. 1983: The Urartian Chariot Reconsidered: I. Anatolica 10, 111–131.
Özgen, E. 1984: The Urartian Chariot Reconsidered: II. Anatolica 11, 91–154.
Piliposyan, A. 1996: An Urartian Bronze belt. Studi Micenei ed Egeo-Anatolici 37, 123–
129.
Piotrovsky, B. 1970: Karmir Blur (Leningrad).
Rehm, E. 1997: Kykladen und Alter Orient (Karlsruhe).
Salvini, M. 1980: Iscrizioni cuneiformi Urartee su Oggetti di Metallo. Studi Micenei
ed Egeo-Anatolici 22, 181–189.
Salvini, M. 1991: Una nuova iscrizione urartea. Orientalia 60, 4, 334–346.
Salvini, M. 1995: Geschichte und Kultur der Urartäer (Darmstadt).
Seidl, U. 1988: Urartu as a bronzeworking center. In J. Curtis (ed.), Bronzeworking
Centres of Western Asia c. 1000–539BC (London), 169–175.
Seidl, U. 2004: Bronzekunst Urartus (Mainz am Rhein).
650 chapter twenty
Sevin, V. and Kavakli, Ersin 1996: Van/Karagündüz. Arkeoloji ve Sanat Yayınları (An-
kara).
Simpson, E. 2005: Tall Tales: Connoisseurs, and The Fabrication of Archaeological
Context. Source 24, 2, 28–41.
Tanabe, K. et alii 1982: Studies of the Urartian Bronze Objects from Japanese Collec-
tions (1). Bulletin of the Ancient Orient Museum 4 (Tokyo).
Taşyürek, O. 1973: The Karahasan Bronze Urartian Belt in the Adana Regional Mu-
seum. Anadolu 17, 197–205.
Taşyürek, O. 1975a: Some Inscribed Urartian Bronze Armour. Iraq 27, 151–155.
Taşyürek, O. 1975b: The Urartian Belts in the Adana Regional Museum (Adana).
Taşyürek, O. 1976: Urartian Figurines and Metal Vessels in the Adana Regional and
Gaziantepe Museums. Türk Arkeoloji Dergisi 23, 2, 106–115.
Taşyürek, O. 1977: The ‘Dedeli’ Bronze Urartian Belt in the Adana Regional Museum.
Türk Arkeoloji Dergisi 24, 1, 157–160.
van Loon, M. 1966: Urartian Art (Istanbul).
Vanden Berghe, L. and de Meyer, L. 1983: Urartu een vergeten cultuur uit het bergland,
Armenië (Gent).
Yildirim, R. 1991: Urartian Belt fragments from Burmageçit. In A. Çilingiroğlu and
D. French (eds.), Anatolian Iron Ages (Oxbow Monograph 13), 131–148.
Zimansky, P. 1998: Ancient Ararat (New York).
PART TWO
* This article originally appeared as “The Archaeological Evidence for Relations between
Greece and Iran in the First Millennium B.C.,” Journal of the Ancient Near East Society of
Columbia University 9 (1977): 31–48.
1 Hans-Volkmar Herrmann, “Frühgriechischer Pferdeschmuck Vom Luristantypus,” JdI
83 (1968), 1–78. (I wish to thank Gunter Kopcke and Joan Mertens for making important
suggestions in the writing of this paper.)
656 chapter twenty-one
2 For a discussion of Achaemenid remains and influences in Greece see D.B. Thompson,
“The Persian Spoils in Athens,” in The Aegean and tbe Near East (New York, 1956), ed. Saul
Weinberg, 281–291; Herbert Hoffmann, “The Persian Origin of Attic Rhyta,” AntK 4, 1 (1961), 21–
26; The article by Anna Roes, “Motifs iraniens dans L’ Art grec archaique er classique,” Revue
Arch. IV (1934), 135–154, juxtaposes too many motifs and cultures and presents conclusions
too broad to be of value. In BCH 85, 2 (1961), “Chronique des Fouilles,” 722 and pl. XXV, is
a report from Olympia of a bronze pointed helmet with an inscription identifying it as an
object captured from the Medes—i.e. from the Persian army in the early 5th century bc (see
also my comment in AJA 73, 4 (1969), 479). And at least two Achaemenid gold lion bracteates
have been found on Greek soil, at Dodona and on Samothrace: BCH 80 (1956), 300, fig. 2;
Arcbaeological Reports 1965–1966, 19, fig. 33. J. Börker-Klähn in ZfA 61, 1 (1971), 138–139, notes 31
and 32, fig. 15b, publishes a drawing of an apparently Achaemenid seal that was claimed
many years ago to have been found at Marathon: unfortunately there is no documentation
to support this claim. For a discussion of the lack of information regarding the find-spot
of the gold Achaemenid bracelet in the Karlsruhe Museum (“said to be” from Corinth),
see my “Unexcavated Objects and Ancient Near Eastern Art,” in Mountains and Lowlands
(Undena Press, California, 1977), ed. Louis D. Levine and T. Cuyler Young, Jr., 195. This latter
work includes a list of excavated Achaemenian Kleinkunst and their proveniences, 192–196.
Andrew Oliver, Jr. in “Persian Export Glass,” JGS XII (1970), 9, believes that glass fragments
excavated at Olympia were “imported … from Persia …,” a conclusion yet unproven. For a
discussion of a 4th century quasi Achaemenid relief excavated in Athens see A.D.H. Bivar,
“A Persian Monument at Athens, and its Connections With the Achaemenid State Seals,” in
the W.B. Henning Memorial Volume, ed, M. Boyce, I. Gershevitch, (London, 1970), 43–61. Bivar
concludes that the relief was sculpted by a Greek—an opinion I share (at least we can state
the relief was not sculpted by an Achaemenid Persian.) That non-Persians manufactured
Achaernenid-style objects is proven by the Tomb of Petosiris reliefs from Egypt; see my
comments in BASOR 223, (October, 1976), 72; and Unexcavated Objects, 193f., No. 100.
relations between greece & iran in the first millennium bc 657
3 For a discussion of North Syrian exports to the West, see my “Near Eastern Bronzes in
the West: A Question of Origin,” in S. Doehringer and David Mitten, eds., Art and Technology,
(Cambridge; Mass., 1970), 116 f. See also note 67 for Phrygian and Assyrian imports.
4 Berta Segal, “Greece and Luristan,” BMFA 41 (1943), 22 f., fig. 3; J.W. Brock, Fortetsa (Cam-
bridge, 1957), 199, pl. 114, “unmistakable Luristan type”; John Boardman, The Cretan Collec-
tion in Oxford (Oxford, 1961), 150, “related to Persian (‘Luristan’) bronzes” that came “via
North Syria”; idemn., The Greeks Overseas (Penguin, 1964), 89; Roman Ghirshman, The Arts of
Ancient Iran (New York, 1964), 331, fig. 406; Herrmann, Pferdeschmuck, 26, fig. 20, “luristanis-
chen Fassung”; Pierre Amandry in Le Rayonnement des Civilisations Grecque et Romaine …,
Huitième Congrès International d’ Archéologie Classique (Paris, 1963), 487, considers the Cre-
tan pendant to be one of the few objects “sûrement iraniens” found in the West. P.R.S. Moorey,
Catalogue of the Ancient Bronzes in the Ashmolean Museum (Oxford, 1971), 26, alone to my
knowledge rejects an Iranian attribution for this pendant: “… if anything [it is] Caucasian.”
Note that Moorey claimed that Jacobsthal had changed his mind that this object was “sub-
Luristan,” but it is my impression that Jacobsthal was referring only to the Perachora buckle
(infra); see Paul Jacobsthal, Greek Prins (Oxford, 1956), 77, n. 1. Compare the Fortetsa pendant
to the same type of objects attributed often to Luristan or Iran, Andre Godard, Les Bronzes
du Luristan. (Paris, 1931), pl. 32, nos. 116, 117, pl. 35, nos. 148, 151; H. Potratz, “Das ‘Kampmotiv’
in der Luristan-Kunst,” Orientalia 21/1 (1952), pls. VI–X, nos. 23–35.
5 Cited many times, e.g., Segal, “Greece and Luristan,” fig. 2; Ghirshman, The Arts of
Ancient Iran, fig. 406; Herrmann, Pferdescbmuck, 26, fig. 20; Moorey, Catalogue, 276ff., 280;
idem, “Ancient Persian Bronzes from the Island of Samos,” Iran 7 (1974), 191, fig. 1 (hereafter
Moorey 1974); Peter Calmeyer, Datierbare Bronzen aus Luristan und Kirmanshah (Berlin,
1969), 101; for Iranian proveniences, Amandry, Le Rayonnement des Civilisations, 487; R.M.
Boehmer, “Zur Datierung der Nekropole B von Tepe Sialk,” AA 1965, 811f.: Judy Birmingham,
“The Overland Route Across Anatolia,” An. St. 11 (1961), 192; Boardman, Greeks Overseas, 89.
658 chapter twenty-one
6 Ulf Jantzen, Ägyptiscbe und orientalische Bronzen aus dem Heraion von Samos: Samos
VIII (Bonn, 1972), pl. 74, B896, pl. 72, B1282; for the latter see also Günter Kopcke, “Heraion
von Samos: Die Kampagnen 1961/1965 im Südtemenos,” Ath. Mitt. 83 (1968), 291f., fig. 33,
pl. 123. Birmingham, Overland Route, 192, mentions “the Ibex” at Samos as Iranian but gives
no clarification with regard to what object she refers; nor is there any further information
concerning the “horse handle fitting and the heavily bossed harness and armour pieces” from
Samos likewise attributed to Iran.
7 See for example the notes above and my review of Jantzen, Samos VIII in AJA 77 (1973),
237; Moorey, Catalogue, 280; idem 1974, 191, 193. Jantzen catalogued the Samos mountain goat
as Assyrian on pp. 70, 73; and see note 4 for Moorey’s rejection of the Cretan pendant as
Iranian. Herrmann and J. Börker-Klähn in their reviews of Jantzen, Samos VIII, in Gnomon 47
(1975), 398, 399, and OLZ 70/6 (1975), 544, respectively, also accept the Cretan pendant and
the Samos mountain-goat as Iranian. Börker-Klähn, ibid., 540f., claims Samos B589 is Iranian,
an opinion I cannot accept.
8 Jantzen, Samos VIII, pl. 74, B1161, pl. 75, B1278, and pl. 58. Jantzen attributed the bell and
Herrmann, Gnomon 47 (1975), 398, and Börker-Klähn, OLZ 1975, 545, accept the Samos hell as
Iranian. For excavated examples, see E. Negahban, A Preliminary Report on Marlik Excavation
(Tehran, 1964), fig. 132; R. Ghirshman, Fouilles de Sialk (Paris, 1939), pl. LVI, S833; Hasanlu,
unpublished, 72.151, two examples. For bells said to be from Iran (all unexcavated except no. 12
from Sialk) see Jan Bouzek, “Openwork ‘bird-cage’ Bronzes,” in J. Boardman & M.A. Brown,
eds., The European Community in Later Prehistory, Studies in Honor of C.F.C. Hawkes (London,
1971), 80, fig. 8. Bouzek does not cite the Marlik example and does not distinguish the Samos
pomegranate bell from the others; see pp. 88 and 94.
relations between greece & iran in the first millennium bc 659
therefore, that we may with some confidence accept an Iranian source for
the Samos bell unless future excavations document a more widespread
distribution for the type.
The bronze pendant, unlike the bell, has to date no published parallels
from excavations in Iran or, to my knowledge, elsewhere in the Near East.
However, similar types of objects have been reported from clandestine dig-
ging in Iran, and on the basis of style an Iranian attribution for the pendant
is not precluded. Therefore, it is suggested that the pendant be accepted, at
least on a tentative basis, as an Iranian product, with the understanding that
we cannot accept proveniences offered by dealers as an historical reality.10
As for the goats, I have noted elsewhere that no excavated parallels are
presently known.11 Nevertheless, as with the pendant, the goats fit into
an Iranian background stylistically, and the few exact parallels known to
me are claimed to be from Iran (which without the stylistic underpinning
would be of doubtful value). On this basis, therefore, it is suggested that the
goats derived from Iran. It is of some interest to note that the goats were
excavated in five different areas at Samos and that each is slightly different
in height, width and body structure, which surely indicates that they should
be considered as seven separate objects, rather than as parts of a single unit.
These two groups, a total of thirteen objects, are to my mind the examples
that may legitimately be brought forward in discussions concerned with Ira-
nian material excavated in the West. If they are separated into two groups
it is because I am deliberately taking a conservative position in the conclu-
sion that the first category is more firmly grounded—because derived from
excavations—in an Iranian background than the second; in the final analy-
sis the division may be more theoretical than factual.
Turning now to a direclty related subject, the recognition of specifically
Iranian motifs manifesting themselves in Greek art, we find ourselves in
a more difficult type of research. It is more difficult because the motifs
alleged to be Iranian have been translated into Greek forms and style and
are thereby at least one step removed from the original model. The problem
for those concerned with these orientalizing (Iranianizing) motifs is to
10 Muscarella, AJA 77 (1973), 237, Moorey 1974, 192, and Herrmann, Gnomon 47 (1975),
398, accept these objects, including the Samos example, as Iranian. For problems related to
proveniences offered by dealers, see my “ ‘Ziwiye’ and Ziwiye: The Forgery of a Provenience,”
JFA 4/2 (1977), 197–219.
11 Muscarella, AJA 77 (1973), 236, where the two paraliel pieces are mentioned; Moorey
1974, 192 f., and n. 56, and Herrmann, Gnomon 47 (1975), 396f., also accept the Samos goats as
Iranian.
660 chapter twenty-one
12 Pferdeschmuck, 3–18, figs. 1, 4, 7–10, with parallels given in footnotes; P.R.S. Moorey,
“Towards a Chronology for the ‘Luristan Bronzes’,” Iran 9 (1971), 123, seems to support Cal-
meyer’s suggestion. See also J.A.H. Potratz, Die Pferdetrensen des Alten Orient (Rome, 1966),
figs. 32: f.g, 46: c, 60, pl. LVI, 133–135.
13 Pferdeschmuck, 31 f., n. 117, fig. 26.
14 E. Negahban, “The Seals of Marlik Tepe,” JNES 36/2 (1977), 99f., figs. 19–26; for unprove-
nienced examples attributed to Iran, see P.R.S. Moorey, Ancient Persian Bronzes in the Adam
Collection (London, 1974), 177 f., nos. 175–179. The Caucasian example cited by Herrmann in
note 117 is not close enough; the Marlik examples were not available to Herrmann but give
excellent support for his conclusions. In part they also satisfy Roes’ suggestion that the Near
Eastern derivatives for the Greek examples would someday be recovered; Anna Roes, “Pro-
tomes Doubles et Tètes d’ Animaux Géminées,” Revue Arch. 35 (1932), 206.
relations between greece & iran in the first millennium bc 661
The third category of objects consists of pottery vessels from East Greece
decorated with horizontal animal friezes and filler ornaments (Figure 12);
as I have argued elsewhere, they depend, ultimately at least, on an Iranian
source.15 Evidence for the Iranian origin of the animal frieze exists both on
excavated vessels from Marlik and on other vessels, which although clan-
destine finds, are stylistically surely Iranian (Figure 13). I am able to present
here another metal vessel, hitherto unpublished and in a private collection,
which in basic shape, construction, and general decorative scheme, is simi-
lar to the Metropolitan Museum example of Figure 13, and which probably
came from the same area in Iran (northwest?).16 (The neck part of this vessel
should be examined to see if any repairs have been made in modern times.)
The suggestion that the Greek animal frieze in horizontal zones may depend
on Iranian prototypes has recently received welcome support from Moorey.17
With regard to filler ornaments one must be on guard against isolating
minor details on Greek pottery and casually interpreting them as examples
of influence from one or another Near Eastern culture.18 However, it may not
be a mere coincidence that rosettes were commonly used as filler ornaments
on Greek pottery of the 8th and 7th centuries bc and that they were equally
popular on earlier and contemporary Iranian metal and terracotta vessels
15 O.W. Muscarella, “A Bronze Vase from Iran and its Greek Connections,” MMA Journal 5
(1972), 43 f.
16 Ibid. 49 f., n. 80, Heeramaneck collection [now in the Los Angeles County Museum of
Art]. I did not know the owner of the vessel when I wrote note 80. This vessel is also cited by
P.R.S. Moorey, “Some Elaborately Decorated Bronze Quiver Plaques Made in Luristan,” Iran
13 (1975), 25, n. 42. (N. B. In my article “A Bronze Vase from Iran …” I inadvertantly neglected
to point out an important parallel for the metope type of decoration represented as a band
on the upper part of the bronze vessel and on other Iranian vessels [41f., note 23; fig. 9].
This very same type of decoration occurs on the fringe of the kilt worn by the boxer and
the archer on the Hasanlu gold bowl. Also note that the Objects Conservation Department
of the Metropolitan Museum of Art believes that the vessel represented in figures 13–14 of the
bronze vase article is apparently modern: one of the motifs that convinced me it was ancient
was the metope type banded decoration! For this piece see also Oscar White Muscarella,
Unexcavated Objects, 172, no. 10).
17 Iran 12 (1974), 195. Note that Moorey in Iran 13 (1975), 25, considers both the Metropoli-
tan Museum and the Heeramaneck vessels to be laterl in date (Iron Age IIIB) than sug-
gested by me (ca. 1000–800bc, i.e., Iron Il)—which would make them closer in time to the
Greek examples I claim to be derivative; Peter Calmeyer, Reliefbronzen in babylonischem Stil
(Munich, 1973), 205, n. 442, c, on the other hand, dates the Metropolitan Museum vase earlier:
“wohl zu spät datiert ….”
18 See also the comments of Herrmann, Pferdeschmuck, 33, n. 121; 34, n. 130; J. Weisner,
“Zur orientalisierenden Periode der Mittelmeerkulturen,” AA 1942, 392, n. 4, 437; Jack Benson,
Horse Bird and Man (Amherst, 1970), 67; Boardman, Greeks Overseas, 80.
662 chapter twenty-one
(cf. Figures 13 and 14).19 Since the quarter-rosette has been attributed to Iran,
may we not assume the same source for the full rosette?20 Moreover, during
the same period bees were also represented on both Greek and iranian
vessels.21 Ought we to assume an independent, coincidental use of this motif
in both cultures, or can we hypothesize that the earlier Iranian examples
were the source for the Greek (mainly Cretan) representations? I do not find
it difficult to conclude that Greek artisans may have been made aware of
both rosettes and bees by the same Iranian vessels that were decorated with
animal friezes.
The fourth category that should be included in our discussions is a frag-
mentary bronze spout on which stands a horned animal in the round; it
was excavated at Aetos on Ithaca (Figure 15), and Herrmann argues that
the piece “ist ohne Luristanvorbilder schwer denkbar.”22 If a spout of this
type had been excavated in Iran no doubts as to its indigenous origin would
have been raised; in fact, a terracotta vessel with a very similar spout sup-
porting an animal was excavated in a late 9th century context at Hasanlu
(Figure 16).23 Given this striking parallel, I believe we may easily accept
Herrmann’s conclusion that the Ithaca spout was an adaption or an Iranian
vessel shape. And in this context it should be remembered that it is a histor-
ical fact that Iranian vessels were capable of reaching the West (Figure 2).
One wonders whether the animals on both the Aetos and Hasanlu spouts
19 Ghirshman, Sialk, pl. 83: A D, 87: S1548, 90:9; Negahban, Marlik, figs. 109, 136, pl. XVI;
Muscarella, A Bronze Vase, figs. 1 f., L. vanden Berghe, Archéologie de l’Iran Ancien (Leiden,
1969), pls. 107: c, 150: b, 171: a, 143: d. Compare also rosettes on pottery from Kultepe, T. Özgüç,
Kültepe and its Vicinity (Ankara, 1971), pls. 17:3a, b; 21:1b; 22:6; on pp. 86f. they are dated to the
6th century bc, too late to relate them to the earliest Greek examples.
20 R.D. Barnett, “Oriental Influences on Archaic Greece,” in S. Weinberg, ed., The Aegean
and the Near East, 230; Ekrem Akurgal, The Art of Ancient Greece (New York, 1966), 194, won-
ders if the Greek rosettes came from Urartu because of their occurrence on siren attachments,
which to Akurgal are of Urartian origin, to me North Syrian; see Muscarella, Near Eastern
Bronzes, 110 f., for a summary of various opinions regarding their origin.
21 For a discussion of bees and “birds,” sometimes represented together with rosettes,
on Iranian vessels, see my A Bronze Vase, 42, notes 24 and 25 and Fig. 13 here. See also
T.J. Dunbabin, The Greeks and their Eastern Neighbors (London, 1957), 45, and Muscarella,
Phrygian Fibulae from Gordion (London, 1967), 60 f., where it is suggested that there might
be a direct connection between the geometric styles of Greece, Anatolia, and Iran: but this
is a complex topic that needs much research and thought before it is accepted as a definite
historical event, rather than as an hypothesis worthy of serious consideration.
22 Herrmann, Pferdschmuck, 30 f. By “Luristan” Herrmann means western Iran; 6, n. 26.
23 Ghirshman, The Arts of Ancient Iran, fig. 25; cf. also his fig. 102. Herrmann’s parallels
cited in his n. 110 are not so obvious. The idea of an animal or bird perched on a spout occurs
also in Phrygia, E. Akurgal, Phrygiscbe Kunst (Ankara, 1955), pl. 24b, but the Hasanlu example
is obviously closer to the Aetos example than are the Phrygian examples.
relations between greece & iran in the first millennium bc 663
served apotropaic roles and whether the role as well as the vessel form was
borrowed by the bronze workers of Aetos.
Two other types of objects excavated in Greece warrant special consider-
ation at this point of our discussion because the possibility that they devel-
oped from an Iranian source is still unresolved. The ambiguity exists because
these objects are not stylistically defined enough to allow one to come down
definitely on the side of those who argue that they are examples of Iranian
influence, or, equally, that they are not. Herrmann discerns Iranian influ-
ence in a fine bronze horsebit and on other objects that depict the “master
of animals” motif, in particular on a bronze horsebit from Messinia and ivory
and lead figurines from Sparta.24 The Iranian parallels he cites are obviously
pertinent, and one’s first reaction is to accept them. But since the motif
occurs throughout the Near East—whence the Greek examples no doubt
derive—the “master of animals” motif per se must be excluded from con-
sideration as a specifically Iranian influence.25 At the same time, inasmuch
as the Messinia horsebit is V-shaped, it is by no means impossible that in
this instance the motif may also reflect an Iranian source. Thus, although
we should not include the Messinia horsebit within the corpus of definitely
Iranianizing material, we should not categorically exclude it.
One is equally tempted to agree with Herrmann in deriving from Iran the
many fenestered spheres surmounted by an animal or bird that have been
excavated in Greece.26 Examples of similar spheres are reported to come
not only from Iran but also from the Caucasus.27 Though it seems that the
24 Pferdeschmuck, 18 f., figs. 14, 15; cf. his figs. 11 and 12.
25 Moorey, Catalogue, 27. It is not clear if this motif was known in the Greek Bronze Age;
Benson, Horsee Bird and Man, 46 f., 147, n. 48. I leave out of discussion here any mention of the
Italian horsebits that have been related to Iranian prototypes (e.g. Herrmann, Pferdeschmuck,
13 f., notes 56, 57 [add T.J. Arne, “Luristan and the West,” ESA 9 (1934), 283]) both because
the issues of possible Italian-Greek-European relationships is outside my competence, and
because of Moorey’s reservations, Catalogue, 27. Note also that I have consciously accepted
above the Cretan open-work pendant with a “Master of Animals” motif as Iranian (contra
Moorey) because of specific formal parallels of the whole object with examples considered
to be Iranian.
26 Pferdeschmuck, 32, with Greek, Iranian and Caucasian references in note 118. See also
spheres attributed to Iran, some of which functioned as bells (cf. Figure 17),
appear to be closer to the Greek examples than those from the Caucasus, we
cannot know for sure which were available to the Greeks. Nevertheless, in
this instance it seems certain that either Iran or the Caucasus—or both—
was the source of inspiration; even if we cannot pinpoint a specific region,
at least we have evidence of influence reaching Greece from a general area.28
This brief discussion on the search for the oriental provenience of the
“master of animals” motif and the fenestered spheres leads to a related sub-
ject for study. Scholars are now becoming aware that it is sometimes diffi-
cult, if not impossible, to recognize in which specific culture of the Near East
a particular orientalizing motif or imported object may have originated, and
whence it came to the West. The evidence from controlled excavations—as
opposed to the unverifiable proveniences offered by dealers—has demon-
strated that several types of objects were manufactured, or at least used,
at more than one near eastern region. Unless there is unambiguous evi-
dence, such as obvious stylistic indications, for one of these regions to be
indicated as the probable oriental source, it is methodologically safer to
conclude that the object had a general near eastern background, rather
than to guess at a specific source. We thereby avoid distorting the evidence
and assigning credit for cultural contributions to the West to one culture
when in fact it may actually belong to another.29 In the following passages
I shall comment upon some of those objects and motifs that to my mind
have been erroneously attributed solely to Iran, to the exclusion of other
regions.
We begin with a number of bronze objects excavated on Samos and
Rhodes, and called at different times mace heads, scepters, furniture fittings,
and cosmetic containers. (It is probable that these similarly shaped objects
28 Equally, if not more, difficult to place solely in Iran are the many bronze animal figurines
with suspension loops (Anhänger) excavated in the West. They have a wide distribution in
the Near East and the Caucasus; Herrmann, Pferdeschmuck, 31, nn. 111–115; see also an example
from Hasanlu; M. Rad, A. Hakimi, The Description and Results of the Scientific Excavations of
Hasanlu. ‘Solduz’, (Teheran, 1960), fig. 1 opposite page 72 [in Persian]. It is also possible that
the addorsed heads of Samos VIII, pl. 73, B1130 may be accepted as Iranian-Caucasian in the
broad sense; Moorey 1974, 192, sees it as Caucasian, Muscarella 1973, 237, as possibly Iranian;
see also on the same page my comments regarding the open-work bird, BB762. For references
to material common both to Iran and the Caucasus, see F. Hančar, “Kaukasus-Luristan,” ESA
9 (1934), 47–112.
29 For further discussion on near eastern proveniences and problems raised in attribution
see Muscarella, Near Eastern Bronzes, 109 f.; Herrmann, Pferdeschmuch, 33, and his “Urartu
und Griechenland,” Jdl 81 (1966), 79–141.
relations between greece & iran in the first millennium bc 665
30 Birmingham, Overland Routes, 192, 187, figs. 7–10; Calmeyer, Datierbare Bronzen, 91f.,
figs. 94, 95; Kopcke, Heraion von Samos, 294, pl. 126:4; R.H. Dyson, Jr., “In The City of the
Golden Bowl,” ILN (Sept. 12, 1964), 375, fig. 10; E. Gjerstad, Swedish Cyprus Expedition 4/2
(Stockholm, 1943), fig. 24, no. 11; R.D. Barnett, “Layard’s Nimrud Bronzes and their Inscrip-
tions,” Eretz Israel 8 (1967), 4 f., pl. VIII (called Syrian); Jantzen, Samos VIII, 56f., pl. 50 (called
North Syrian). Note that at Hasanlu at least one “footed” example contained kohl, the “footed”
end was sealed with a wooden plug, and a kohl stick was excavated juxtaposed, which means
that it could only be a cosmetic container; others found there are either not so clearly distin-
guished or are apparent mace heads.
31 For a discussion of proveniences and possible sources of manufacture, see Calmeyer,
Datierbare Bronzen, 91f. and the map of fig. 94, which should have the Cypriote and Hasanlu
examples added. See also Herrmann, Gnomon 47 (1975), 396.
32 Jantzen, Samos VIII, 81 f., pls. 79, 80; Hans Möbius, “Kaukasische Glocken in Samos,”
in E. Sprockhoff, ed., Marburger Studien, (Darmstadt, 1938), 150–166, pls. 66–69; Calmeyer,
Datierbare Bronzen, 111 f.; idem, “Glocke,” In RLA (Berlin, 1969), 427–431. Möbius and Calmeyer
suggested that the Samos bells came from the Caucasus; Moorey, Catalogue, 138, suggested
they came either from the Caucasus or Urartu; Bouzek, Openwork, 88, sees the Samos bells
“related more directly to Caucasian bells … but there need not be any direct connection”;
Herrmann, Pferdeschmuck, 31, recognized the bells’ wide distribution. For other near eastern
bells see Guitty Azarpay, Urartian Art and Artifacts (California, 1968), 25, fig. 6; L. vanden
Berghe, La Nécropole de Khurvin (Istanbul, 1964), pl. 37, also pl. 29, no. 216; R.H. Dyson, Jr.
“Hasanlu and the Solduz and Ushnu Valleys …,” Archaeologia Viva 1 (1968), 90, upper left;
Potratz, Die Pferdetrensen, 161, n. 1; Gordion, unpublished, 1910 B300, from the destroyed
Phrygian level.
666 chapter twenty-one
of a Luristan bit,33 even though he was aware that the closest parallels are
an electrum example from Nimrud and several representations depicted
on Assyrian reliefs. To Herrmann, both the Rhodes and Nimrud examples
were derived from Luristan. Based on stylistic grounds, expecially the lack
of a ground line and the galloping position of the horse, Calmeyer and
Moorey have rejected an Iranian attribution for the Lindos bit, and consider,
correctly I believe, that it is Assyrian and came from there.34 This conclusion
obtains equally for the similar bits that were excavated on Samos.35 That
Iranian horsebits were probably the inspiration for the Assyrian examples
is not to be denied, but this is not at issue here: the horsebits from Samos
and Rhodes came from an Assyrian workshop and they cannot be cited as
examples of Iranian imports in the West.
In my review of Samos VIII I stated that a fragment of a bronze vessel
excavated on Samos was to be considered an Iranian import.36 I now real-
ize that although very similar vessels have been excavated at War Kabud
and at Tepe Guran in western Iran, other examples have been excavated at
Uruk in Mesopotamia.37 Most of the known examples of this bronze ves-
sel shape, those excavated and those from dealers’ shops (Figure 20), are
reported to come from Iran (which is why I originally attributed the Samos
vessel there), but the excavated examples from Uruk must also be intro-
duced into the discussion. At this time we do not know whether the Uruk
vessels were Iranian imports into Mesopotamia, although I suggest that
this conclusion is probable. In any event, we do not know at present if
the vessel came from an Iranian or from a Mesopotamian shop, and it is
33 Pferdschmuck, 22 f., fig. 17; 13, n. 53. In Gnomon 47 (1975), 397, Herrmann changed his
mind and sees the Samos bit as a local copy of an Assyrian bit.
34 Calmeyer, Datierbare Bronze, 114; Moorey, Iran 9 (1971), 123f.; idem., Iran 2 (1964), 194;
194. I believe that Moorey’s fig. 4 is not correctly drawn as there should be a sharper carination
at mid-point; cf. Kopcke, Heraion von Samos, pl. 127:1.
37 L. vanden Berghe, “La Nekropole de War Kabud,” Archaeologia 18 (1967), 60f.; J. Meld-
gaard et al., “Excavations at Tepe Guran, Luristan,” Acta Archaeologica 1964, fig. 30. Compare
the War Kabud vessels to Eva Strommenger, Gefässe aus Uruk … (Berlin, 1967), pl. 32, espe-
cially no. 78. Of some interest is the juxtaposition of the spouted vessel and and carinated
example from Tepe Guran, both types of which have been excavated at Samos.
I would also now claim that the knobbed mace from Samos (Jantzen, Samos VIII, pl. 51,
B574: upside down?), which I suggested in AJA 77 (1973), 236, “has good parallels in Iran,”
also has good parallels elsewhere and therefore might better be labelled near eastern; see
Calmeyer, Datierbare Bronzen, 108.
relations between greece & iran in the first millennium bc 667
a discussion of these objects in Greece and the Near East; see also vanden Berghe, Khurvin,
63, pl. 27; Özgüç, Kültepe and its Vicinity, 101, pl. XXXIV.
44 Ghirshman’s attempt, Arts of Ancient Iran, 341, to link bucchero pottery in Italy with grey
wares in Iran, and winged creatures in Greece to those from Iran, are equally unacceptable
and only further complicate the issue under study here. However, his discussion on p. 343
regarding the use of pins as votive offerings in Greece and Iran is probably correct, except
that this idea was prevalent all over the Near East, at least with regard to fibulae, see Oscar
White Muscarella, “Fibulae Represented on Sculpture,” JNES 26 (1967) 84f.
668 chapter twenty-one
A few other objects need only be mentioned here for they have all been
discussed by Moorey who rejects the Iranian background alleged by others.
The best known of these is the bronze “buckle” excavated at Perachora
in Greece by Humfry Payne. Moorey believes that the piece actually has
affinities with the Caucasus; it seems certainly to be near eastern and, if
not necessarily Iranian, then possibly Caucasian.45 Moorey also rejected as
Iranian two fine bronzes excavated on Samos. One (B1211) consists of two
confronted equids, and is considered by him to be near eastern, with no
specific source suggested: the other (B1130) consists of addorsed wolf heads
each devouring a bull head, and is assigned to the Caucasus. While neither
object would cause surprise if it were excavated in Iran, one could agree
with Moorey that an Iranian attribution, to the exclusion of other regions, is
not compelling, and, therefore, I tend to support his position.46 In addition,
Moorey has correctly denied an Iranian attribution to a bronze object from
Olympia, which Herrmann (apparently unknown at the time to Moorey)
had implicitly claimed to be related to a mirror handle said to be from
Luristan.47
Whereas after critical analysis an Iranian attribution might be rejected
for the objects just discussed, most, if not all, have at least certain features
that allow such attribution to be understandable. This is not the case with a
particular motif that has been presented as an example of Iranian influence
in the West, and for which the evidence offered is nonexistent. Ghirshman
has cited the mounted archers employing the “Parthian Shot” on one of
the well-known Cretan bronze shields and maintained that the “motif can
only have come from Iran.”48 However, to my knowledge this motif is not
45 Moorey, Catalogue, 26, nn. 4, 5; Herrmann, Pferdeschmuck, 27, also rejected an Iranian
attribute claiming that its origin is not clear (but cf. Gnomon 47 [1975], 399); Boardman,
Greeks Overseas, 89, thought that it “may have come” from Iran; Amandry, Le Ravonnement
des Civilisations, 487, believed it to be Iranian.
46 Jantzen, Samos VIII, 74 f., “Luristan,” pl. 75; Moorey, Iran 12 (1974), 191f.; cf. Muscarella AJA
77 (1973), 237, “… they could tentatively be considered Iranian,” which is still true although
modified by the Caucasian factor. Herrmann, Gnomon 47 (1975), 398f., also rejected an
Iranian attribution.
47 Moorey, Catalogue, 26; Herrmann, Pferdeschmuck, 30.
48 Arts of Ancient Iran, 341, fig. 432. A bronze bowl in the Ashmolean Museum that
depicts the Parthian Shot is to my mind incorrectly listed as having derived from Olympia:
F. Studniczka, Jdl 22 (1907), 165; H.T. Bossert, Altsyrien (Tübingen, 1951), 803; Dunbabin,
Greeks, pl. VII, 1. In fact, the bowl was purchased by the Ashmolean in 1903 and attributed
by the dealer to Olympia, which means that it cannot be legitimately cited as an example of
an import to Olympia or to any other site (information about the acquisition supplied to me
by P.R.S. Moorey).
relations between greece & iran in the first millennium bc 669
depicted in Iranian art before the Sasanian period, but it is depicted in the
earlier art of Assyria, Urartu, and on Cypro-Phoenician bowls.49 Further, it is
mentioned as an Urartian technique of fighting by Sargon II of Assyria in his
report on his eighth campaign in 714 bc.50 To relate the representations on
the Cretan shields specifically to Iran, where to date there is no evidence for
its depiction in the first millennium, and to ignore the multiple near eastern
examples, is a distortion of the facts.
We come now to the final category of objects to be discussed in the
context of possible Greek-Iranian contacts. Instances occur where a motif
or object excavated in a first millennium bc site in Greece has parallels in
the contemporary Near East as well as in the earlier second millennium bc
Greek Bronze Age. In this connection we may cite, besides the terracotta
boots mentioned above, double-headed animals, two animals sharing one
head, animal-headed pins, and a bird mounted on the back of a horse, and
possibly still others. These motifs are represented in Greece both in the
round and painted on pottery;51 all have been cited by some scholars as
examples of Iranian influences on Greek art—improperly, it will be argued
here.
In these cases it is difficult to determine whether we are witnessing
the reappearance of motifs locally preserved, recognized and deliberately
developed from the earlier period, or whether in fact we are witnessing a
fig. 1. Sulimirski and others, e.g. R.H. Dyson, Jr., “Problems of Protohistoric Iran as Seen from
Hasanlu,” JNES 24 (1965), 208, and T.C. Young, Jr., “The Iranian Migration into the Zagros,”
Iran 5 (1967), 20, have assumed that the warriors on this relief are Scythians, a conclusion
I consider to be dangerous. There are no identifying inscriptions on the relief, which was
sculpted before the time we first learn of Scythians in the Near East; see also K. Jettmar, Art of
the Steppes (New York, 1964), fig. 44. R.D. Barnett, “Assyria and Iran …,” Survey of Persian Art
14 (1967), 2997, sees the figures on this relief as Iranian, inferring that they are Zamuans, to
my mind not a proper—because it is a guess—conclusion.
For other examples of the Parthian Shot depicted in art, see Eva Strommenger, 5000 Years
of the Art of Mesopotamia (New York, 1964), figs. 242, 243, practiced by Arabs; Potratz, Pfer-
detrensen, Pls. IX:15, XXXIV:7; D. Randall Maciver, Villanovans and Early Etruscans (Oxford,
1924), pl. 38: 1, 2; M. Rostovzeff, “The Parthian Shot,” AJA 47 (1943), 180f., n. 15. In Urartu it
is represented on a belt in the Ashmolean Museum and on another in the Norbert Schim-
mel Collection; Oscar White Muscarella, ed., Ancient Art, The Norbert Schimmel Collection
(Mainz, 1974), no. 133, panel 20. Note that the archer on the Cretan shield is represented rid-
ing backward in the saddle, which also occurs in Sasanian art; R, Ghirshman, Parthians and
Sassanians (London, 1962), figs. 248, 250, 251.
50 D.D. Luckenbill, Ancient Records of Assyria and Babylonia (Chicago, 1927), 2, para-
graph 158, describes Urartian horsemen turning around in the saddle when fighting.
51 Segal, Greece and Luristan, 72 f.; Herrmann, Pferdeschmuck, 33f.
670 chapter twenty-one
new appearance, solely derived from recent stimuli from the East. Were the
motifs copied by Greek artists of the first millennium bc who discovered
them on their own soil as a consequence of an interest in their Bronze
Age heritage, or were the motifs observed and adapted because they were
first noticed on oriental objects? Or, to mention yet another possibility, did
the Greek artists discover a motif as a result of knowledge from their own
past but used it in art only after exposure to oriental exotica? This last
suggestion seems too elusive to document without textual evidence and
may actually beg the question. For we would actually be arguing that the
oriental stimulus was the primary source and the historically significant
one; cultural continuity would have played no creative role, and the term
orientalizing would be the correct term to describe the action.
Thus, when Herrmann suggests that the double-headed animal, which he
recognized had a Bronze Age history both in the Near East, and in Greece,
re-emerged in late Geometric Greek art because of contemporary influence
from Iran (“der Zweiten Rezeption”), he is actually arguing that the motif
is an example of oriental influence, not of cultural continuity.52 The Greek
examples of this motif (Figure 21) appear close in form to contemporary
examples attributed to Iran (Figure 22), but it is argued here, pace Her-
rmann, that the issue of cultural continuity cannot be set aside; it must be
an integral part of the discussion. To my mind, once it has been established
that a given motif used in first millennium bc Greek art had both an earlier
local history and a contemporary life in the Near East, we cannot presume
that the latter occurrence was the stimulating force. To accept as fact either
that the oriental source was the cause for the Greek artist’s use of a motif,
which was already known to him, or that it was cultural continuity that was
the cause, is in the final analysis sophisticated guesswork. Therefore, this
charged category of material should be kept separate in our minds subject
52 Pferdeschmuck, 34 f., nn. 130–131; cf. also Benson, Horse Bird and Man, 70, who believes
that the animal frieze was known to 8th and 7th century bc Greeks from their Mycenaean
heritage but that they decided to use them in art only after they were exposed to oriental
models; i.e., that the motif is an orientalizing one according to the position taken here. To
the three possible explanations suggested for the use of a motif in the first millennium bc
one might add another: thc motif was independently borrowed from the Near East twice,
once in the Bronze Age, the second time centuries later (an idea suggested by Roes, Revue
Arch. 4 [1932], 198). This theory, however, also rejects continuity and actually argues for the
new arrival in the first millennium of the motif, i.e. that it is an example of orientalising. For
evidence of first miliennium bc physical contacts with Mycenaean remains, see A.M. Snod-
grass, The Dark Ages of Greece (Edinburgh, 1971), 394 f., and Benson, Horse Bird and Man, 109f.,
115 f., 121 f.
relations between greece & iran in the first millennium bc 671
53 Pierre Amiet, La Glyptique Mesopotomienne Archaique (Paris, 1961), pl. 26, 425, pl. 66:
890; Ernst Herzfeld, Iran in the Ancient East (London, 1941), 163, fig. 278, below; Arthur Evans,
The Palace of Minos IV 2 (London, 1935), figs. 575–577; Humfry Payne, Necrocorinthia (Oxford,
1931), 52, fig. 12, pl. 16:4; Muscarella, Norbert Schimmel Collection, no. 15; Ghirshman, Arts of
Ancient Iran, figs. 581, 582.
54 Oriental Influences, 232, pl. XXI: 1, 2, 3; Arts of Ancient Iran, 316, figs. 383, 384, and page 335
genuine; I know of another example of this motif on a disc pin of first millennium date in
a private collection that seems to be genuine; see also Muscarella, Unexcavated Objects, 184,
no. 156. For an Achaemenian example of this motif, see John Boardman, “Pyramid Stamp Seals
in the Persian Empire,” Iran 8 (1970), 35, fig. 12. The gold appliques in 7000 Years of Iranian
Art, no. 456, and Kunstschätze aus Iran (Zurich, 1967), no. 868, illustration no. 18a, may be
genuine—but I have not examined them personally.
56 For excavated examples in Iran, see Ghirshman, Sialk, pl. XXIX, 1, top; Negahban, Marlik,
fig. 131 (cf. fig. 85); at ieast one example, with a ram’s head, comes from Hasanlu, unpublished:
70–50. For others attributed to Iran, see André Godard, Les Bronzes du Luristan (Paris, 1931),
pl. XXXII; Moorey, Catalogue, 191 f., 197 f., pls. 50, 53; Herzfeld, Iran in the Ancient East, 154f.,
fig. 275, pl. XXXI; for the Caucasus, see F. Hanfčar, “Die Nadelformen des Kaukasusgebietes,”
ESA 7 (1932), figs. 10, 17, 18.
672 chapter twenty-one
the second millennium bc.57 Again given these facts, Herrmann’s suggestion
that an Iranian source sufficiently explains the motif in Greece must be
rejected.
Years ago Berta Segal suggested that the representation of a bird on a
horse’s back was dependent on Iranian prototypes; the motif occurs in
first millennium Greek art in the round and on pottery.58 Both Benson and
Herrmann have rejected Segal’s conclusions, the former because he believes
that the motif developed within Greece itself and is an example of cultural
continuity from the Mycenaean period, the latter apparently because it is
too causal a motif.59 There is yet another reason for rejecting the suggested
Greek-Iran relationship—namely, that in the Near East Iran was not alone
in depicting the motif in art.60
How did the objects and motifs arrive in Greece? It could indeed be argued
that suggestions concerning the routes by which the material travelled from
the various areas within Iran can only be speculative.61 Yet the possible
routes available to us for serious consideration are not so many and may be
discussed with some profit; it is even possible to point to a specific, known
route as probably the one on which goods moved from East to West.
for a recently discovered example of this motif, see Marvin H. Pope, “The Scene on the
Drinking Mug from Ugarit,” in H. Goedicke, ed., Near Eastern Studies in Honor of W.F. Albright
(Baltimore, 1971), 395, fig. 2. Compare also Pierre Amandry’s study of the widespread motif
of the reclining goat, “Un Motif ‘Scythe’ en Iran et en Grèce,” JNES 24/3 (1965), 149–160.
Amandry rejects a specific near eastern background for the Greek use of this motif and raises
the questions (without attempting to answer them) whether it was an independent use in
the first millennium or whether it resulted from continuity from the Mycenaean period.
Benson, Horse Bird and Man, 58, 150, n. 92, cites the motif as an example of a true “Mycenaean
renascence.”
60 Muscarella, Near Eastern Bronzes, 120, fig. 11, North Syrian style; see also Segal, Greece
and Luristan, fig. 121; Herzfeld, Iran in the Ancient East, fig. 286; Pope, The Scene on the Drinking
Mug, fig. 2.
61 Here we may note the statement of M.I. Finley, Early Greece: The Bronze and Archaic
Ages (New York, 1970), 39, “Archaeology alone, it cannot be repeated often enough, rarely can
reveal the mechanism of foreign relations even when it unearths great quantities of foreign or
foreign-inspired goods.” And Dunbabin, Greeks, 15, speaking about historical interpretation of
archaeological material, says: “The historian who is to attempt to use archaeological material
must have a clear idea how much to expect and how to go about it … a superficial treatment
will lead to many pitfalls, mistakes of method, misunderstandings and misconceptions,
errors of detail and errors of scope, and particularly, ignorance of what is possible and of
the limitations on what archaeological evidence may be used to establish.”
relations between greece & iran in the first millennium bc 673
While there can be little doubt that central Anatolian goods travelled
overland across Anatolia to the West in the 8th century bc,62 there is no
evidence to suggest that goods originating in western and northwestern Iran
also travelled by this route. And there are two archaeological indications
that suggest that Iranian goods did not travel by this route. The first is
negative, but significant: It seems clear on the basis of recent research that
Urartian goods did not travel West across Anatolia, nor, it would seem, did
they travel West by any route.63 Thus, given our present knowledge, there
is no reason to assume that the overland Anatolian route extended as far
east as Urartu. Following upon this, how then may we hypothesize that
goods originating in Iran moved north through Urartu in order to cross
Anatolia by a route which seems not to have existed? We cannot; and if
the northern route is rejected, we must equally reject an alleged Black Sea
route to the West.64 We must look elsewhere, and this brings us to the second
archaeological indication mentioned above.
In the 9th century bc a vigorous trade existed between northwestern Iran
and Assyria and North Syria. The evidence consists of a large quantity of
imported Assyrian and North Syrian objects excavated at Hasanlu, many
of which are still unpublished.65 The existence of these imported objects
conclusively documents an overland route connecting North Syria, and ulti-
mately the Mediterranean Sea, on one end, and Assyria and Northwestern
Iran across the Zagros passes on the other. Whether this route was used in
the 8th and 7th centuries is, of course, not proven. But inasmuch as Assyria
continued to be interested in Iran during this period and was still a power
from the late 8th through most of the 7th century, it is probable that the
route was not discontinued. To summarize, we have archaeological evidence
62 For a summary of routes suggested for the movement of material over long distances
within the Near East to Greece, see Birmingham, Overland Routes, 185f.; Bouzek, Openwork,
94, dismisses an overland route from the Caucasus across Anatolia.
63 For evidence suggesting that no Urartian material reached the West and a rejection
of conclusions to the contrary, see Hans-Volkmar Herrmann, “Urartu und Griechenland,” Jdl
81 (1966), 79–141, and Muscarella, Near Eastern Bronzes, 111f., 122. Recently S. Salvadori, “An
Urartian Bronze Strip in a Private Collection,” East and West 26, 1/2 (1976), 106, ignoring the
recent literature and relying on Pallottino, still maintains that Urartian goods reached the
West in quantity.
64 This route was championed by Barnett, Oriental Influences, 228.
65 For published information on Assyrian and North Syrian contacts with Hasanlu, see
Dyson, Problems of Protohistoric Iran, 199, and Muscarella, “Hasanlu 1964,” BMMA 25/3 (1966),
121–135. The subject is more fully explored in my forthcoming A Catalogue of tbe Hasanlu
Ivories.
674 chapter twenty-one
for only one general route by which Iranian goods apparently reached the
West: across the Zagros Mountains via Assyria and North Syria.
Related to our concern regarding trade routes is our interest in discov-
ering which groups or categories of people transported the material to the
West, and why. Was some of the material carried long distance from Iran
by an Iranian trading group, or were they carried in stages by groups of for-
eign merchants? Was some of the material transmitted from the emporia on
the North Syrian coast to Greek cities by Greek merchants, or foreign mid-
dlemen, or both, and did the objects arrive as items of trade? Were some
of the objects brought to the West by Greek or near eastern merchants, or
travellers, or pilgrims as votive objects?66
Leaving the question of motifs aside, we may hypothesize that some of
the objects discussed above had a votive value while others seem to have
been secular in function—although, it must be stressed, these distinction
are highly subjective. Of the supposed votive objects we might include the
open-work Luristan type pendant from Crete (Figure 1), the bronze Luristan
finial (Figure 3), and perhaps the mountain goat (Figure 4) and the seven
goats (Figure 7) from Samos. Of the apparent secular objects there are
the spouted vessel (Figure 2), the bronze pendant (Figure 6), and the bell
(Figure 5), all from Samos, But even if our tentative distinctions between
votive and secular objects are correct—which is by no means certain, as
we cannot claim to know how each object was used or whether a given
object had different values under different conditions—we are ultimately
incapable of reaching conclusions concerning who brought each object. For
it could be argued on the one hand that merchants bring secular goods and
individuals (not necessarily merchants) bring votive goods; and on the other
hand that merchants alone may bring secular objects to sell and votive ones
to dedicate in order to secure personal safety and good trade.
66 Note the views of John Boardman, R.D. Barnett, T.J. Dunbabin, and W.L. Brown that
some of the oriental objects were made in the West by near eastern craftsmen; see Muscarella,
Near Eastern Bronzes, note 9; for Boardman see also “Orientalien auf Kreta,” in Dadalische
Kunst auf Kreta im 7. Jahrhundert (Hamburg, 1970), 14–25, and “The Khaniale Tekke Tombs, II,”
BSA 62 (1967), 63 f. My aim in note 9 was not to deny the recognized value and importance for
scholars to know if near eastern craftsmen actually migrated to the West and worked there.
Rather it attempts to state that, given the fact that we do not know if migration of craftsmen
occurred, in the final analysis we are forced to concern ourselves with what we have, which
are the objects themselves. Are they oriental—imports whether made locally or abroad—
and are they orientalizing, resulting from an imported idea or motif adapted to local style
and needs? At present, I see no need to speculate about travelling Iranian craftsmen.
relations between greece & iran in the first millennium bc 675
67 In addition to the throne sent to Delphi by King Midas (Herodotus I:14), other objects
of Phrygian origin that reached the West are as follows: from Samos there are three fibu-
lae, five belt buckles, possibly a belt fragment, apparently two bowl handles with spools,
and some sherds. Phrygian fibulae have also been excavated at Olympia, Argos, Perachora,
Tegea, Sparta, Lindos on Rhodes, and from Lesbos; bronze omphaloi have been excavated
at Olympia, Argos, Perachora, Delphi, Lindos, and on Cyprus. For further discussion, see
Muscarella, Phrygian Fibulae from Gordion, 59 f., Appendix C, 80f.; idem AJA 77 (1973), 236;
Jantzen, Samos VIII, 48 f. (Through a misadventure I reversed the numbers five and eight
when discussing the Phrygian and East Greek belt buckles published by Jantzen and men-
tioned in my review in AJA 77 (1973), 236; there are in fact five buckles with cast mouldings,
eight with added ones, not vice-versa as published. Thus there are, as listed above, apparently
eleven Phrygian objects on Samos beside the sherds. Fibula B473 (misprinted in the review
as 453), at least three belts (pl. 47), and two plaques (pl. 49), are probably East Greek. These
six objects plus the eight buckles with added mouldings are local, East Greek; which leaves
six objects listed by Jantzen as Phrygian as stili of unknown origin.)
Assyrian imports to the West include, from Sames, six statuettes (B1218, B165, B1594, B773,
B1214, B779), very probably another (B1217), four horse bits (pl. 61), for a total of ten or eleven
objects (AJA 77 [1973], 237). There is also a horse bit from Lindos (Herrmann, Pferdeschmuck,
22f., fig. 17), and a cylinder seal from Olympia, E. Unger in Max Ebert, Reallexikon der
Vorgeschichte 4/2 (Berlin, 1926), Taf. 162, c. There is also the stele of Sargon from Cyprus;
P. Demargne, The Birth of Greek Art (New York, 1964), 277; F. Dunbabin, Greeks, 41, refers to a
bronze plaque from Olympia as an object that “may be Assyrian.” To my mind this plaque is
North Syrian; Muscarella, Near Eastern Bronzes, 116 f., no. 2. On page 39 and note 5 Dunbabin
gives the impression that he considers the distinctly Assyrian seal from Olympia (above) to be
north Syrian, but this may be a misunderstanding on my part. Whether the Assyrian objects
recovered in the West were sent by Assyrian kings is a question worth considering; certainly
the Sargon stele on Cyprus would seem to have resulted from a royal command.
676 chapter twenty-one
That Iranian objects and motifs passed over land and sea and that they
eventually touched Greek soil has been established. And we may claim with
some certainty that our historical knowledge of the West is broadened, for
we now know that there was a distinguishable Iranian spice in the unique
stew of Greek culture.
Addenda
After this paper went to press I came across J. Bouzek, “Macedonian Bronzes:
Their Origins, Distribution and Relation to Other Cultural Groups of the
Early Iron Age,” Památky archeologiké 65 (1974), 278–341. Bouzek states, as
I do in the text above, that the pomegranate shaped bell and the crea-
tures on a shank attached to a disc have an Iranian background (305, 323,
330); also that the Assyrian horsebits from Samos and Lindos are not Ira-
nian. However, contrary to the position taken by me, he claims that Samos
bronzes B1211 and B1130, and the “maceheads/furniture attachments” are Ira-
nian (323).
To Note 2:
An Achaemenid horsebit was excavated at Athens; H. Lechat, BCH 14 (1890),
385f., fig. 1.
relations between greece & iran in the first millennium bc 677
Archaeological remains indicate that from the very beginning of the Greek
and Cretan Bronze Age, Aegean cities maintained contact with neighboring
lands in Egypt and the Near East. These contacts are indicated by imported
objects and materials excavated at sites on the Greek mainland and on var-
ious islands in the Aegean Sea, and, especially during the Late Bronze Age,
by Greek material found in Egypt and the East. After the collapse of the
Mycenaean civilization and the onslaught of the Dorian invasions, some-
time in the twelfth century bc, foreign imports are no longer encountered
on Greek sites, and there is no evidence that Greek goods continued to be
exported. These facts, coupled with the lack of monumental architecture
and the sparse remains from Greek sites, clearly indicate that Greece suf-
fered a cultural decline and that ties to the East had been severed. A change
occurs at least by the tenth century bc, if not slightly earlier, for a number
of sites in the Near East have yielded Greek Protogeometric pottery, a sure
indication that earlier contacts were being renewed. But during this period
the flow of goods apparently moved in only one direction, West to East, and
it was not until the late ninth century, at the earliest, that a reciprocal rela-
tionship began. It is at this time that the earliest near eastern material is
recovered in post-Mycenaean Greece, in particular at Athens. Aside from
these early finds—a bronze decorated bowl, faience beads and plaques,
and gold earrings either imported or made under near eastern influence—
the archaeological record is silent until approximately the mid-eighth cen-
tury bc. Beginning at this time, and continuing for some decades of the
seventh century, there is a quickening of imports to Greece, and it is dur-
ing this period that the greatest number of near eastern objects arrived.1
* This article originally appeared as “Urartian Bells and Samos,” Journal of the Ancient
Muscarella, “Near Eastern Bronzes in the West: The Question of Origin,” in S. Doehringer,
D.G. Mitten, and A. Steinberg, eds., Art and Technology (Cambridge, Mass., 1970). 109f.;
H.V. Herrmann, “Hellas (orientalischer Import im frühen Griechenland),” Reallexikon der
Assyriologie (1975), 4:303–311.
690 chapter twenty-two
2 There are approximately 79 examples known to date. From the Greek mainland there
are 38 oriental and 9 Greek examples, all excavated at sanctuaries, plus 3 Greek and 1 oriental
examples in museums that are attributed to Greece (a Greek example in Boston is the mate
of an example from Olympia); Delos and Rhodes each have one oriental type, and 6 are
from Italy, from tombs. Two aberrant examples were excavated at Olympia and 4 aberrant
examples come from Salamis, Cyprus. From the Near East there are 8 sirens from Gordion
and one example reported from Alishar, on the Araxes River. A total of 10 examples have
been attributed to the Van area but all derive from dealers or private owners; one example
has been attributed to Nimrod. Recently, the Munich art market had for sale a cauldron with
siren attachments in situ of a type hitherto not encountered. For relevant bibliography see
note 3, below.
3 Bibliography and discussions in detail concerning the origins of sirens and griffins are
to be found in Muscarella, Near Eastern Bronzes, 109 f., 116; idem, “The Oriental Origin of
Siren Cauldron Attachments,” Hesperia 31/4 (1962), 317–329; H.-V. Herrmann, Die Kessel der
orientalisierenden Zeit (Berlin, 1966), 25 f., 50 f., 131 f.; idem, “Urartu und Griechenland,” JdI 81
(1966), 79–141; idem, Hellas, 303 f., 306 f.; P. Amandry, Gnomon 41 (1969), 796f.; V. Karageorghis,
Excavations in the Nekropolis of Salamis III (Nicosia, 1973), 97f.; P. Amandry, “L’Art ourartien
et ses relations avec le monde grec,” Proc. of the Xth Inter. Congress of Classical Arch. (Ankara,
1978), 5 f.
urartian bells and samos 691
Urartian influences.4 But far more important than the issue of the origin
of these groups of objects per se is the fact that most scholars who have
championed their Urartian background have taken the position that Urartu
played a significant role—the major role some would argue—in the near
eastern encounter with Greek culture in its formative stage.
A vigorous reaction to this position, the “Urartu-These,” began to develop
in the 1960s, and a growing number of scholars maintain that the sirens
and griffins, and still many more objects—bronze vessels, horse blinkers,
plaques, ivories, and assorted motifs—derived from North Syria, not from
Urartu. Therefore, it follows that it was not Urartu but the North Syrian
cities that played the crucial orientalizing role in Greece.5 In addition, the
contributions of Phrygia, Assyria, and Iran have been better defined.6
This writer has long maintained that there are no recognizable Urartian
objects or motifs on Greek soil, nor in the West in general, and that “Urartu
seems, to my mind, to have played no recognizable role at all” in the trans-
mission of oriental art and motifs to the West.7 H.-V. Herrmann has referred
to the alleged presence of Urartian material in the West as an “archäologis-
che Legende,” and he notes “dass Urartu beim Export nach Griechenland
keine nennenswerte Rolle gespielt zu haben scheint.”8
4 See note 3 for references. U. Jantzen, Griechische Greifenkessel (Berlin, 1955), reported
about 250 examples; G.M.A. Hanfmann, Gnomon 29 (1957), 241–246, added over 60 more;
others continue to be excavated (e.g., AA 1978, 3, 396, fig. 15). Over 220 have been excavated
on Greek soil, 24 from Italy, 5 from France, and one from Spain. One example was excavated at
Susa, two are in the Ankara Museum, one in Izmir, and 2 gold examples have been attributed
to the site of Ziwiye, in Iran.
5 Especially Muscarella, The Oriental Origin, and Near Eastern Bronzes; Herrmann, Die
Kessel, and Urartu und Griechenland. See also R.S. Young, “A Bronze Bowl in Philadelphia,”
JNES 26 (1967), 150; J. Muhly, “Homer and the Phoenicians,” Berytus 19 (1970), 49; I.J. Winter,
“Phoenician and North Syrian Ivory Carving …,” Iraq 38 (1976), 14, 16f., 21; idem; “Carved
Ivory Furniture Panels from Nimrud …,” MMAJour. 11 (1976), 42; idem, AJA 80 (1976), 202;
Amandry, L’ Art ourartien, 5 f. A few scholars still support an Urartian origin for sirens and
griffins; S. Korti-Konti, “Orientalizing Bronze Work in Greece,” Athens Annals of Archaeology
4/2 (1971), 281 f.; S. Salvadori, “An Urartian Bronze Strip in a Private Collection,” East and West
26, 1/2 (1976), 106; C. Burney, The Ancient Near East (Ithaca, 1977), 183; A.M. Bisi, “Elements
anatoliens dans les bronzes nouragiques de sardaigne,” Proc. of the Xth Inter. Congress of
Classical Arch., (Ankara, 1978), 356 f.
6 For Iran, with bibliography, see O.W. Muscarella, “The Archaeological Evidence for
Relations Between Iran and Greece in the First Millennium B.C.,” JANES 9 (1977), 31–48; for
Phrygia and Assyria, see 32 and 47, n. 67; also Herrmann, Hellas, 306, 310f.
7 Near Eastern Bronzes, 122; see also The Oriental Origin, 325f.
8 Urartu und Griechenland, 80; Hellas, 396. Note also Amandry, L’Art ourartien, 7: “… on
ne trouve dans le monde méditerranéen aucun objet qui soit indiscutablement d’origine
ourartienne.”
692 chapter twenty-two
9 U. Jantzen, Samos VIII, Ägyptische und orientalische Bronzen aus dem Heraion von
Klähn, OLZ 70/2 (1975), 545; Amandry, L’ Art ourartien, 4, 6; G.M.A. Hanfmann, in Bi. Or. 30,
3/4 (1973), 199, alone accepted Jantzen’s attributions.
11 See note 10 and H. Kyrieleis, “Orientalische Bronzen aus Samos,” AA 1969, 2, 166f.,
in Urartu und Griechenland, 126 f. he said of it “das berechtigte Anwartschaft darauf hat, als
urartäisches Werk zu gelten …”; and in Gnomon 47 (1975), 400, he states that it “bleibt als
gesichertes urartäisches Importstück …”; but in Hellas, 306, he seems more hesitant, B1217
“mag eine vereinzelte Ausnahme sein,” i.e., may be the only Urartian object known in the
West. Amandry in L’ Art ourartien, 6, also expressed reservations about the Urartian origin of
this piece; however, he suggests that a bronze horse from Samos, B492, could be Urartian, a
position I do not share.
13 Samos VIII, 81 f., pls. 79, 80. A total of 20 bells were excavated on Samos, 12 of which are
Muscarella, AJA 77 (1973), 237, and JANES 9 (1977), 39 f., n. 32, for further references; H. Möbius,
“Kaukasisches Glocken in Samos,” Marburger Studien (Darmstadt, 1938), 156f. Even J. Bouzek
accepted the Samos bells as Caucasian, “Macedonian Bronzes …,” Památky archeologiké 65
urartian bells and samos 693
of the bells from Samos (B474) stands out from the rest (Figure 1): it has
eight facetted sides with a double row of rectangular apertures separated
by a raised horizontal ridge and with another ridge at the base; the top
area consists of a heavy rosette moulding on which is a solid loop set on
a low platform. At the time the bell was published two similar examples
were available for study. One (Figure 2) is the octagonal bell with a single
row of rectangular apertures and raised horizontal ridges, two at the base,
one at the top, reported to have been found at Alishar on the Araxes River
together with two other bells and a bull and siren cauldron attachment.15
The uncontrolled find, made by Kurds in 1859 in a rock chamber (a tomb?),
precludes any objective information concerning the nature of the deposit,
or whether all the objects were indeed found together. Nevertheless, the
octagonal bell is inscribed with the name of the Urartian king Argishti, Ar-
gi-iš-ti-i ú-ri-iš-hi ‘From the arsenal of Argishti’ (whether the first or second is
not certain and˘ need not concern us here), and the bull attachment claimed
as part of the find is certainly of Urartian manufacture. Further, one of the
other bells, uninscribed (Figure 3), is also octagonal, although with only two
rectangular apertures, and it is clearly related in form to the Argishti bell.
But apparently because the third bell seems to be a typical Caucasian type
(Figure 4), and because the other bells are fenestered, a feature which is
considered to be Caucasian,16 and because Alishar is situated on the border
of Urartu, in Transcaucasia, scholars failed to identify the facetted bells as
Urartian. In any event, and for whatever reasons, the Alishar Argishti bell
was not cited by those who studied the Samos bells.
The second bell of concern to us was excavated at the Urartian site of
Karmir Blur and was published in 1955 (Figure 5).17 It is octagonal and has
a single row of rectangular apertures and raised horizontal ridges, two at
the base and one below the top; in fact, it seems to be a mate to the Alishar
example, except that it is uninscribed. This bell was also ignored by those
seeking parallels for the Samos bells, and B474, along with the other seven
examples published by Jantzen, was assigned to the Caucasus.18
(1974), 305, 309, 323, 333, 335; idem, Graeco-Macedonian Bronzes (Prague, 1973), 84, 87, 89,
fig. 25: no. 4 is B474 (infra), but it is an inaccurate drawing.
15 See P. Calmeyer, “Glocke,” in Reallexikon der Assyriologie (1969), 3: 429f., fig. 5; B.B.
Piotrovskii, Urartu: The Kingdom of Van and Its Art (London, 1967), 82f., figs. 58, 59.
16 Of course, the Caucasian apertures are typically triangular in shape, not rectangular:
sources for the Samos bells, but neither the Alishar nor the Karmir Blur bells were cited.
P.R.S. Moorey, A Catalogue of the Ancient Persian Bronzes in tbe Ashmolean Museum (Oxford,
1971), 138, mentioned both Urartu and the Caucasus as possible sources for the Samos bells—
but he could not have had B474 in mind as it had not yet been published.
19 Oktay Belli, “Van Bolge Müzesindeki Çivi Yazilı Urartu Tunç Eserleri,” Anadolu 4–5
(1976–1977), 198 f., fig. 8, said to come from Patnos, fig. 9 has the Menua inscription; fig. 10
has the Argishti inscription.
20 P.R.S. Moorey, Ancient Iran (Ashmolean Museum, 1975), 25, pl. XIV; its height is 8.5cm.
Moorey claims the bell came from “Western Iran,” but this attribution comes from a dealer
and therefore has no archaeological value.
21 N. Spear, Jr., A Treasury of Archaeological Bells (New York, 1978), 111f., figs. 121, 122; height
8.7 cm.
urartian bells and samos 695
This bell is also octagonal with a double row of apertures and ridges in
the appropriate places, and it too has a heavy rosette moulding at the top
surmounted by a solid loop set on a low platform; the iron clapper is now
missing but the iron suspension bar is extant (Figure 11). And it is inscribed
with the name of Argishti, in the very same manner as on the Alishar and
Van bells. The Metropolitan Museum bell is clearly of the same type as
the Ashmolean bell, but more significantly, in all features, except for the
inscription and its size—it is 7 millimeters shorter and ca. 2 to 3 millimeters
wider—it is a duplicate of B474 from Samos.
Given the readily identifiable characteristics of the Urartian bronze bells
under review here, it should be relatively easy to recognize other exam-
ples, or adaptations, in the West—if they exist—and, indeed, at least one
adaptation of an Urartian bell excavated on Greek soil exists. Spear, who
published the Metropolitan Museum bell, as well as those from Alishar,22
perceptively called attention to their resemblance in all essential details to
a small bronze bell from Pherae in Thessaly, Greece (Figure 12). This bell
has seven facets, a row of rectangular apertures, a raised ridge at the base, a
heavy undecorated moulding at the top surmounted by a solid loop set on a
low platform, and the ends of the suspension bar project from its holes.23 To
my mind the Pherae bell is a copy of an Urartian bell, rather than an import.
If so, it is at present the first certain example of a local Greek copy of an
Urartian work. Perhaps it might be rash to conclude that the Pherae bell is
directly dependent on Samos B474, but it is relevant in this context to note
that there were extensive contacts between Macedonia and Thessaly and
Samos in the eighth and seventh centuries bc, as we know from the impor-
tant research of J. Bouzek.24 Thus it is archaeologically acceptable to posit
that the Pherae bell was made as a result of contact with Samos, reinforcing
Bouzek’s conclusions. At the same time, Bouzek believes that there is good
evidence for suggesting that there was contact between Macedonia and the
Caucasus,25 which means that we must equally allow for a more direct trans-
mission.
At this point in our discussion two questions arise: How does the presence
of an Urartian bell on Samos affect the thesis that there are no examples of
Urartian art in the West? How does the recognition that the Pherae bell is
adapted from an Urartian prototype affect the conclusion that Urartu played
no recognizable role in the formation of the Greek orientalizing taste and
style?
Concerning the first question, it is of course obvious that the position
must be modified to include B474; at least one Urartian object reached the
West—or to those who accept the statuette B1217 from Samos as Urartian,
there are now two objects. But one bell (and one putative statuette) does not
make “influence,” and I believe it would be irresponsible to refer to Urartian
cultural contacts with the West in any manner other than as the evidence
permits. At present the evidence is minimal. Reacting to Jantzen’s strong
statement with regard to the alleged quantity of Urartian art on Samos, Herr-
mann said that “-die von J. aufgestellte Behauptung das sich in Samos ‘das
urartäische Kunsthandwerk besonders deutlich dokumentiere’ (80), ist auf
jeden Falle unhaltbar.”26 I see no reason to modify this conclusion. Further,
the conclusion that no Urartian work of art or motif has yet been recog-
nized on mainland Greeze at any important sanctuary—Athens, Olympia,
Delphia, Perachora, Argos, Sparta, Ptoion, Tegea, Ithaca, etc.—still obtains
and is still significant.
With regard to the second question, I believe that here too a conservative,
cautionary posture should be maintained. Because there exists on Greek soil
one bell that was manufactured with an Urartian model in mind does not in
any meaningful way indicate that Urartu can now be included among those
near eastern cultures, the products of which stimulated Greek artists and
artisans.27 Yet the Pherae bell with its “Urartianizing” features exists, albeit
in an area north of the major Greek centers, and as such it deserves to be
cited as an example—the sole example I believe—of a Greek object copied
from an Urartian model.
26 Gnomon 47 (1975), 400. Further, given the evidence presently available, I see no need
to alter previously expressed conclusions that neither Urartian nor Iranian goods travelled
across Anatolia to the West; Muscarella, The Archaeological Evidence, 45f., nn. 62, 63.
27 Amandry, L’ Art ourartien, 7, stated it neatly: “Même si quelques produits des ateliers de
la région de Van sont parvenus en terre grecque, on ne décèle, ni dans l’art grec du 8e siècle,
ni dans le premier art grec orientalisant—celui des styles protocorinthien et protoattique—
aucun trait qui soit ourartien.”
I wish to express my sincere thanks to Oktay Belli, V. Lukonin, B.B. Piotrovskii, B. Philip-
paki, U. Jantzen, P.R.S. Moorey, and Nathaniel Spear, Jr., for their courteous and prompt coop-
eration and their permission to publish the bells from the Van Museum, Alishar, Karmir Blur,
Pherae, Samos, and the Ashmolean Museum.
urartian bells and samos 697
Figure 3. Bell said to come from Figure 4. Bell said to come from
Alishar; Hermitage Museum. Alishar; Hermitage Museum.
698 chapter twenty-two
Figure 6b.
urartian bells and samos 699
There are no internal historical sources for the history of the people the
Greeks called Phrygians. The earliest reference to the Phrygians in western
historical sources occurs in the text of Herodotus who lived in the 5th
century bc (I.14, VII.73). Herodotus (1.14) records that King Midas of Phrygia
dedicated his royal throne to the Delphic sanctuary and that he himself
saw it still on view.1 The date of Midas’ dedication is specifically situated by
Herodotus to a time before the dedications of King Gyges of Lydia (reigned
ca. 685–650 bc).2 The implication of Herodotus’ statement is that Midas
reigned at a time before Gyges, and this chronological positioning almost
certainly establishes a pre-early 7th century date for Herodotus’ Midas. That
Herodotus’ King Midas of Phrygia lived in the 8th century bc and is the very
same person cited by the late 8th century Assyrian King Sargon II (between
the years 717 and 709bc) as King Mita of Mushki has long been accepted by
a majority of scholars, including the earliest to write about the Phrygians3
(G. Rawlinson: according to Bittel 1950: 76, but without references; Winckler
1901: 136, 283f., 287; Koerte 1904: 9f., 17, 18ff., 21; Kroll 1932: 1538 f.; Friedrich
1941: 863; Bittel 1942: 67f.; 1950: 76; 1970: 135; Akurgal 1955: 113, 120 ff.; Goetze
1957: 202; Dunbabin 1957; 64; Mellink 1965: 317 f.; Muscarella 1967: 59 f., 72,
n. 1; Herrmann 1975: 310; Roller 1983: 300).4
Herodotus’ firm chronological statement allows us to accept as an his-
torical fact—a hard fact—that it was the 8th century bc, King Mita of the
* This chapter originally appeared as “King Midas of Phrygia and the Greeks,” in Anatolia
and the Ancient Near East: Studies in Honor of Tahsin Özgüç, eds. K. Emre, M. Mellink,
B. Hrouda, and N. Özgüç (Ankara: Türk Tarih Kurumu, 1989), 33–44.
1 Koerte 1904: 21, suggests that the throne was not Midas’ but was one sent for the god’s
Muscarella 1988: 179 f. Postgate 1973 is ambiguous. He translates Mushki as Phrygian from
Sargon’s letter but on p. 24, n. 4, he claims that he does not “imply that Mushku and Phrygia
were identical”.
704 chapter twenty-three
5 The oracle sought could have been related to Midas’ problems with the Assyrians, or
Both in the quality of manufacture and in the skill and artistic nature
of their designs the Phrygian wooden furniture represents some of the
most aesthetic—worth-seeing—artifacts known from antiquity, and they
objectively inform us of the woodworking capabilities of the 8th century
Phrygians. I suggest that it is highly probable that the throne Midas sent to
Delphi was made of woods of different color, and was probably elaborately
inlaid in geometric patterns (an idea indepentetly reached by Jeffery 1976:
148). The throne may have been made in the very same woodworkers’ shops
that made the Tumulus P and MM furniture.
Two late Greek sources record that King Midas of Phrygia married a Greek
princess. Aristotle (fr. 611,37, ed. V. Rose) calls her Hermodike and says that
she “cut/struck the earliest coinage for Kyme”.6 Pollux (Onamastikon IX.83)
names her Demodike, the daughter of King Agamemnon of Kyme, and he
notes that she was but one among several others who were alleged to have
been the first to strike coins.7 Both sources cite Kyme in Aeolis, on the west
coast of Asia Minor, as the princess’ home and Pollux specifically identifies
her father as being king there.
Given the late date (albeit derived from earlier sources) of the accounts,
the fact that coinage is mentioned, and that there were presumably 7th cen-
tury, as well as 6th century Phrygian kings named Midas (Herodotus I.35;
Koerte 1904: 25f.), it remains uncertain that the Midas cited is the Midas-
Mita of the 8th century bc, and not a later one (although it is clear that
the early Midas-Mita was in the minds of the authors). Scholars are divided
between those who assume that this Midas is the 8th century king (a major-
ity: Koerte 1904: 21, 23; Kroll 1932: 1539; Friedrich 1941: 388; Akurgal 1955: 125;
Huxley 1959: 94.f.; Barnett 1967: 14; Boardman 1973: 86; Mellink 1965: 317;
Sams 1979: 47; Roller 1983: 300; Diakonoff and Neroznak 1985: xiii; Prayon
1987: 118, 193), and those who cautiously leave the question open (Dunbabin
1957: 64; Lejeune 1969: 188; Young 1981: 253), Jeffery (1976: 238) assumes he
was a later, ca. 600bc, Midas.
6 “They say that Hermodike, wife of the Phrygian king Midas, was very beautiful and
also that she was wise and a skilled craftswoman and struck the earliest coinage for Kyme.”
(Translated by Donald F. McCabe.)
7 “Another example of local pride is the dispute about coinage, whether the first one to
strike it was Pheidon of Argos, or Demodike of Kyme (who was wife of Midas the Phrygian
and daughter of King Agammemnon of Kyme), or Erichthonios and Lykos at Arthens, or
the Lydians (as Xenophanes says), or the Naxians (as Aglosthenes thought)”. (Translated by
Donald F. McCabe.)
706 chapter twenty-three
8 Kagan on p. 359 cites Weidauer as dating the deposition to 626bc at the latest, but she
coinage (Balmuth 1971: 5ff., pls. 2, 3). The thrust of all this is to demonstrate
that scholars are not in a strong position when they categorically deny the
existence of some form of coinage in Asia Minor in the 7th century bc, earlier
than 625 bc (as Akurgal 1983).
It follows from the above comments that it is not a solecism if one accepts
as viable the conclusion that Midas-Mira married a Greek princess in the
late 8th century bc (or a few years into the 7th century)—a princess who
may have played a role in the introduction of coinage to Kyme at some
time (still to be determined) after the death of her husband (ca. 696bc).
The Kymean princess could have been a young woman when she married
Midas-Mita and a middle-aged or older widow when she involved herself
in the economic affairs of her home city. If we can trust some elements
in the pseudo-Herodotus Life of Homer11—which offers further information
about Midas’ marriage—Midas’ inlaws, presumably King Agamemnon and
his queen, survived his death (that is, if we can trust this part of the story,
if being sceptical of the rest), for they are supposed to have commissioned
an epitaph for his tomb.12 If they survived Midas’ death, so could their
daughter—who presumably would have left Gordion at the threat of or as
a result of the Kimmerian invasion. Another possibility to be considered,
but one that cannot be forced, is that the coinage connection is an error or
a later misunderstanding (Huxley 1959: 94), joined somehow to an original
account of the marriage.
A second issue of concern for the marriage is the chronology of Agamem-
non. Koerte (1903: 19) and Wade-Gery (1952: 7) date him to ca. 700 bc, but
their chronology seems based on the acceptance of his being the father-in-
law of the 8th century King Midas (see also Roller 1983: 300; cf. Jeffery 1976:
238). If it can be firmly established that kingship at Kyme did not continue
into the 7th century bc (v. Büchner, PW, supp. 1, 1903, “Kyme”, 2475), we might
have in the reference to a king in Kyme a good indication that he was a con-
temporary of Midas, and could have been his father-in-law.
The problem may be approached from another direction, one that I
suggest supports the claim for a late 8th century marriage. For this evidence
we return to the throne dedication. An archaeological interpretation of the
11 Vita Homeri Herodotea in Homeri Opera, ed. T.W. Allen, vol. V, 198, lines 130–140, for the
epitaph commission (Donald McCabe located this reference for me); Jeffery 1976: 198. Plato,
Phaedrus, 264D, gives a shorter version; see Koerte 1904: 22, and Wade-Gery 1952: 65, n. 21.
12 Koerte 1904: 22, considered the epitaph to be a fiction. There is no discussion in the
dedication and the dynamics informing it not only reinforces the tradition
of the marriage, it also yields information about the nature and the degree
of Phrygian-Greek relations.
Although it is not an impossible view to uphold that the Phrygians
learned of the Delphic oracle from the Cilician or north Syrian coastal cities
(e.g. the ubiquitously cited al Mina), there are no compelling indications
for such an assumption, and there is no historical or archaeological evi-
dence for direct Phrygian contacts there.13 Phrygia indeed had political and
cultural contacts with inland cities of north Syria, especially Carchemish,
with whose king, Pisiris, Midas made a short-lived treaty in 717 bc (Lucken-
bill 1927: para. 8). Undoubted north Syrian artifacts and motifs have been
recovered at Gordion, and north Syrian motifs influenced Phrygian potters,
collectively attesting to this relationship (Muscarella 1967: 67 f.; Sams 1974:
181ff., 191; 1978: 233ff.; 1979: 46; Prayon 1987: 183 ff.). Except for a sherd at Tell
Halaf, no certain Phrygian material has been recovered in north Syria or the
Levant.14 And the Phrygians never penetrated to the Cilician coast, because
of Assyrian interference (Luckenbill 1927: para. 18, 42; Postgate 1973: 28, 30,
32). There is no evidence that in the 8th century bc Phrygia maintained trade
routes to the Cilician or Syrian coasts.
From an archaeological and geographical perspective it makes more
sense to recognize that the Phrygians learned of the oracle—and other
Greek matters—from the west coast Greek cities, a region that was closer
than the sourthern ports, and where there were no geographical or mani-
fest political barriers to contact (see Koerte 1904: 21 ff.; Friedrich 1941: 888).
Given these realities, joined with the archaeological record (infra), it seems
certain that Midas’ throne must have travelled (by cart!) from Gordion via
an overland route west to the Greek coast—and thence by ship to Greece.15
Further, a considerable number of Phrygian artifacts were recovered in the
Heraion sanctuary on Samos (infra), and Boehmer (1973: 166) has cogently
suggested that some of this material represents 8th century dedications of
13 A XII, 14 sub-type fibula was excavated at al Mina, but it is clearly post-8th century bc
in date: Boehmer 1972: 63. Birmingham 1961: 189, incorrectly cites a XII, 13 fibula from Zincirli.
14 Akurgal 1966: 118, figs. 90–92, claims that bronze vessels from Tell Halaf are Phrygian, but
none of the examples cited has a Phrygian parallel, and handles at right angles to the spout
occur outside of Phrygia. In 1967: 67, I cited a painted jug from Carchemish as Phrygian: Sams
1974: 192; 1978: 235, seems to reject the attribution. I also here reject my 1967: 67, comments
regarding right angled handles to spouts that occur in north Syria as definitely Phrygian.
15 Pace Winckler 1901: 284 f. (see also Kroll 1932: 1534) there is no evidence that Phrygia
was a sea power. The ship that transported the throne was probably Greek.
king midas of phrygia and the greeks 709
16 We know nothing of the dynamics of the relationship: who initiated it; how many Greek
cities were involved; how much physical contact occurred. Nor do we know if the marriage
preceded the Delphic dedication.
17 These fibulae are not mentioned by Sams 1979: 47; Young 1963, Roller 1983: 301 or Prayon
A small number of Greek sherds, but representing five or six vessels, of 8th
century bc date have been recovered at Gordion. Although they all derive
from later levels (Sams 1979: 47) it is highly probable, as Sams states, that
they were displaced from the destruction level and in fact reached Gordion
during the reign of Midas. Sams prefers to see the vessels as coming from
the south, from Cilicia or the Levant, not the west—which he admits is
the “most obvious channel;” but his preferred direction warrants review:
more probably, most likely, the vessels travelled overland from the west, the
“obvious channel”—along the same route that carried the fibulae and the
throne, and more material still to be discussed.19
Over this same route from the Greek east was most probably transmitted
the alphabet, from Greek minds to Phrygians (because the Greeks called
their alphabet Phoenician, not Phrygian), although the opposite direction
has been considered (Barnett 1967: 20; Young 1963: 363, no. 54; 1969: 264,
294: but cf. p. 296). Whence it reached Phrygia is of course in great part a
philological matter, and the archaeologist takes risks in offering opinions.
Young (1969: 253ff., 264 f. 294) vigorously argued for Cilicia or the north
Syrian coast and rejected the west coast of Asia Minor for the learning area
(see also Snodgrass 1971: 352; Sams 1974: 192). Diakonoff and Neroznak (1985:
3) on the other hand argued for the opposite view; they rejected the Levant.
From an archaeological perspective, Diakonoff and Neroznak’s thesis best
fits into the background, and it does not contradict the philological data.
The transmission of the alphabet from Greece to Phrygia, the fact that
Phrygians chose this alphabet in which to communicate (perhaps because
of a common ancestral background?), rather than the Assyrian cuneiform
(adopted by the Urartians), or the Luwian hieroglyphs (employed in Tabal
and north Syria), remains the most significant cultural event resulting from
the relationship between the two people: and the Phrygian choice manifests
the closeness of that relationship.
overland route across Anatolia has been concerned with explaining how alleged Urartian
material allegedly reached the west, e.g. Birmingham 1961, and others; see Muscarella 1977:
45 f.
19 I must cite—but wish to reserve judgment on—the conclusions of A. Snodgrass and
Prayon (1987: 163, nn. 677–680, pl. 33d) that a horseman depicted on an ivory plaque from
pre-destructions Gordion wears a Greek helmet, specifically the closed Corinthian form (to
Snodgrass he is a Greek, to Prayon he is a Phrygian wearing a Greek (style) helmet). If indeed
the helmet can be explained only in Greek terms (that the artist’s rendition and intention
is not misunderstood by us), then the plaque would be an important piece of evidence
for a Greek cultural presence at Gordion. At present, I do not see the certainty of a Greek
connection, hence its parenthetic inclusion here.
king midas of phrygia and the greeks 711
20 While the imitations are culturally significant in terms of Greek adaptations of Phrygian
material, they are not manifestly dated to the 8th century bc; the Greek artisans may of course
have used 8th century Phrygian models.
21 Strangely, not a single example of XII, 7A fibulae, a common sub-type, occurs outside of
Phrygia. Note, however, that a broken fibula arc from Argos (Waldstein 1905, pl. 87, no. 903)
could be a XII, 7A form as the arc seems to be a crescent and the end tapers. Neither Boehmer,
Caner, nor the writer have cited the fragment.
712 chapter twenty-three
mouldings) come from Lindos on Rhodes, and Samos (with double pins),
while XII, 5 fibulae occur at a number of island and mainland sites.
XII, 9 fibulae, a sub-type of XII, 7, embellished with studs on the arc,
occur in three 8th century tumuli at Gordion, MM, KIII, and KIV, but also
in Tumulus S1, which may be post-destruction in date; they are also found
in the destruction level and in almost all post-destruction levels on the
city mound. At Boğazköy they occur in levels dated to the 8th and the 7th
centuries bc (Boehmer 1972: 56). In Ankara some were recovered from an
8th century tumulus (Buluç 1979: pl. 9). XII, 9 fibulae were also dedicated on
Greek island and mainland sites. An example from Marino in Latium may
have reached Italy by indirect trade, from a Greek source; it was deposited
in a 7th century grave (Muscarella 1967: 19, no. 26; Boehmer 1972: 57, n. 394).
While the majority are dated in Phrygia to the 8th century bc, enough
examples are attested there from later levels to indicate that the sub-type
continued to be manufactured in the 7th century, and apparently later.
R.M. Boehmer (1972: 50f., n. 323) has thoroughly reviewed the chronology
of the find spots at the Greek sites where Phrygian fibulae have been recov-
ered. His conclusion is that none of these sites has a clear-cut context that
allows one to argue with certainty that a Phrygian fibula reached there in the
8th century bc rather than in the 7th: viz. Sparta, Lindos, Argive Heraeum,
Perachora, Olympia, Samos; and that at other sites, viz. Chios, Aetos (Ithaca),
Lesbos, they are 7th century arrivals.22 At the same time, however, on the
basis of formal similarities to those from Phrygian 8th century contexts, he
accepts an 8th century date for a double-pin XII, 7 fibula from Samos, two
XII, 7 examples from Lindos, and the same date for XII, 9 fibulae from Lin-
dos and the Argive Heraeum (Boehmer 1972: 53, 54, 57, nn. 388, 392, 396; 1973:
166). I agree, and I would further suggest that a XII, 9 example from Olympia
and one from Paros, and perhaps some (if not all) of the XII, 5 fibulae from
the Argive Heraeum, Sparta, and perhaps Lindos could be 8th century bc
arrivals there. All the examples cited here could very well be curated 8th
century dedications placed with later material in the sanctuary deposits, a
view accepted, in part at least, by Boehmer.
Sub-types XII, 13 and XII, 14 fibulae are among the most common Phry-
gian examples excavated both within Phrygia and in the Greek sphere.
They occur in quantitiy at Gordion in 8th century tumuli and in all the
22 Boehmer (1972: 57, n. 397) considers some examples of his XII, 9 Β, found at some of
these sites, to be local copies, and therefore post-8th century bc. Of course, we do not know
whether this is true or not.
king midas of phrygia and the greeks 713
23 Pn copies of Phrygian fibulae see Muscarella 1967: 37ff.; Boehmer 1972: 50, 58, 60, 63,
1973: 149 f.
24 Boehmer (1973: 149 f.) publishes a terracotta fragment from Argos with a XII, 14 fibula
11, 12), at Boğazköy (Boehmer 1972: 72f., no. 180, dated to the 7th century,
cf. Neve 1972: 183, fig. 11, dated to the 8th century), and at Ankara (Özgüç
and Akok 1947: figs. 23–26; Buluç 1979: pl. 14: 8th century). The Tumulus P
examples have handles with a plain arc, except for the terminal mouldings;
the example from Boğazköy has a central moulding; Ankara lacks a handle.
Other, later, examples from Gordion have multiple mouldings on the arc
(Muscarella 1967: pl. XV, 77, 78).26 A number of belts, or buckles with missing
belts, have been excavated in the west at Chios, Ephesus, Bayrakli, and
Samos (Boardman 1961: 179ff.; Jantzen 1972: pls. 45–47). Except for some
handles and a belt fragment from Samos (Jantzen 1972: B593, 606, 613, 616,
1691, 1289),27 all the others seem to be East Greek copies of Phrygian belts.28
And it must be admitted that all the buckles from Samos could be post-
8th century bc in date. They all have multiple mouldings on the arc, a
late feature at Gordion (Muscarella 1967: 26), although not apparently at
Boğazköy (see note 26 here).
Typical Phrygian omphalos bowls and bowls with bolster attachments
and ring handles joined to horizontal rim bands with vertical spools, known
from Gordion and Ankara, have been excavated in the west, in some cases
at the same sites that preserved Phrygian fibulae. Although in the past there
has been confusion regarding which bowls are actually 8th century Phrygian
imports and which local adaptions, the full publication of three 8th century
tumuli at Gordion (Young 1981) and the work of A. Knudsen (1964) has made
it easier to resolve these problems.
At least four Greek sanctuaries preserve examples of Phrygian omphaloi
and bowls with ring handles and bolsters: Samos, the Argive Heraeum,
and Perachora. From the Heraion on Samos derive the remains of two
ring-handled bowls (Jantzen 1972: 54, pl. 50, B494, B413), accepted as 8th
century bc by Jantzen, Knudsen (1964: 63, n. 10), Muscarella (1973: 236;
1977: 47, n. 67), and Herrmann (1975: 310). The remains of a horizontal rim
band and spool that derive from Olympia (Fürtwangler 1890: 136, no. 852)
have been accepted equally as 8th century Phrygian by Knudsen (1964:
64, n. 10) and Herrmann (1975: 310); so have horizontal rims with spools
26 But compare a XII, 14A fibula, with multiple mouldings on the arc, from a grave dated
to the 8th century from Boğazköy: Boehmer 1972: 64f., fig. 28.
27 Compare Samos B606 and B1691 to Boehmer 1972: no. 180, and fig. 28; and B613 and B616
to Muscarella 1967: pl. XV, 77, 78. Note the moulding formation and the ridged terminals. For
the Samos belt compare Özgüç and Akok 1974, fig. 25, and Neve 1972: fig. 11.
28 Boardman 1961, 185, has noted the different form of the catch area: holes on the East
representing three bowls from the Argive Heraeum (Young 1963: 362; Herr-
mann 1975: 310).29 Finally, two plain omphaloi from Perachora (Payne 1940:
pl. 55, nos. 1 and 4) are most probably Phrygian imports (v. Dunbabin 1957:
152; Birmingham 1961: 190, n. 4; Muscarella 1967: 60, 72, n. 4: cf. these bowls
to Young 1981: pl. 72, MM 133, MM 137, MM 139, MM 141, MM 145, and pl. 90,
W 11, W 13, W 14).
Note also that bowls with banded rims and bolsters have been excavated
on Cyprus and at Magnesia in Asia Minor (Akurgal 1955: 81, 82, pl. 60a;
Knudsen 1964: 63).
In addition to the above, petalled decorated omphalos bowls have been
excavated at Olympia, Perachora, and the Argive Heraeum (Fürtwangler
1890: pl. 52, no. 880; Payne 1940: pl. 52; Waldstein 1905: pl. 114, nos. 1975, 1976),
and rim bands or just spools derive from Samos (Jantzen 1972: pl. 50, B1397,
B1633), Lindos (Blinkenberg 1931: pl. 30, nos. 718, 719), and Delphi (Perdrizet
1908: 78, fig. 270). All of these bowls seem to be Greek imitations of Phrygian
examples, as noted by Knudsen (1964: 63f., no. 10), Muscarella (1973: 236),
Mellink (1981: 236, n. 82): but compare Akurgal (1955: 82), Birmingham (1961:
190), Muscarella (1967: 60; 1977: 47, n. 67), who cited some of these examples
as actual Phrygian imports.30
The bona fide Phrygian bowls cited above all match 8th century exam-
ples from Gordion, and it seems difficult to reject a conclusion that they
reached Greece during that time. At three of the sites that preserved Pry-
gian bowls, Samos, Olympia, and the Argive Heraeum, Phrygian fibulae were
also recovered. However (as Kenneth Sams reminds me), banded rim and
omphalos bowls occur at Gordion in post-destruction 7th century bc tumuli,
and to be archaeologically correct we cannot exclude the possibility that
some of the bowls reached Greece after Midas’ reign. So little is known of
post-destruction 7th century Phrygian history that it is not certain whether
Phrygia continued to maintain strong and direct relations with Greece. But
inasmuch as we are aware of 8th century contacts, at least it may be argued
that some of the bowls could have reached Greece at that time. No stronger
statement is warranted.
We now turn to the bronze cauldrons with bull head attachments, two
to a vessel. At Gordion two bull cauldrons derive from the 8th century
Tumulus W (W 1, W 2), three from Tumulus MM (MM 1, MM 12, MM 13),
29 For Gordion parallels to these spools and rims see Young 1981: pls. 60, 66: MM 60, MM 64.
30 Note that no bona fide examples of Phrygian bowls with petalled decoration have
been recognized in the west: but if local copies exist perhaps Phrygian imports were used
as models.
716 chapter twenty-three
a single corroded example from Megaron 4, and a pair from a cauldron was
excavated in the Terrace Building in the destruction level of the city mound,
i.e. from the early years of the 7th century bc (Young 1981: 102, 112, 199 f.,
figs. 67, 68, 117, pls. 50, 59, 87, 88, 95B).31
The Gordion cauldrons have been attributed either to Phrygia or to north
Syria, the latter as imports. The two cauldrons from Tumulus W were con-
sidered to be local product by Young (1960: 231; 1981: 219), but north Syrian
by Herrmann (1966: 128; 1980: 576; 1984: 23), Kyrieleis (1977: 76), DeVries
(1981: 221), and Prayon (1987: 122ff.). The three Tumulus MM cauldrons were
accepted as Phrygian by Young (1958: 151), Akurgal (1961: 202), van Loon
(1966: 105, n. 119), DeVries (1981: 222), Prayon (1988: 124 ff.), and Muscarella
(1987: 185). Herrmann (1966: 84, n. 27, 122, 128, 129; 1980: 576 f.; 1984: 23)32
accepted MM 12 and MM 13 (he actually knew only one example at that
time) as Phrygian, but he considered MM 1 to be north Syrian; Kyrieleis
(1977: 76) apparently only considered MM 11 and MM 13 to be Phrygian; and
the Terrace Building pair were accepted as Phrygian by Kyrieleis, Herrmann,
DeVries, Prayon, and Muscarella. The Megaron 4 example may be North Syr-
ian (DeVries 1981: 221; Prayon 1987: 124).
At least MM 12 and MM 13, the Terrace Building pair, and possibly MM 1
may be accepted as probable Phrygian products, the characteristics of
which, especially the T-shaped clamp attachment, have been given by
Kyrieleis, DeVries, and Prayon.33
Herrmann (1966: 129; 1975: 310) considered it possible that bull attach-
ments from Argos (Waldstein 1905: pl. 75, no. 25), from Athens (de Ridder
1896: 198, fig. 164), one from Samos (Jantzen 1972: pl. 77, B1266), and three
examples from Olympia (Herrmann 1966: pl. 43, A25, A26, pl. 51, A33) were
also Phrygian exports. But except for the Samos example, which is accepted
as possibly Phrygian by Kyrieleis (1977: 87), DeVries (1981: 222), Prayon (1987:
125f., 211), and Muscarella (1988: 185), all the others are surely not: the Argos
example could be Near Eastern or Greek, and the Athens and Olympia exam-
ples are surely Greek (Muscarella 1968: 14, n. 20; Kyrieleis 1977: 87).34
31 Birmingham 1961: 191, refers to bull cauldrons from Ankara—but in fact she was exam-
ining MM 12 from Gordion. In Muscarella 1988: 185, I mention two examples from the METU
excavations at Ankara, but there is actually one example and of a type other than those dis-
cussed here: Buluç 1979: pl. 13, fig. 6.
32 Herrmann makes his determination here with caution; see also Muscarella 1968: 13f.
33 For other cauldrons with the T-shaped clamp see Young 1981: pl. 58; Buluç 1979: pl. 13,
figs. 4, 5. Kyrieleis and DeVries have made it easier to distinguish Phrygian from North Syrian
bull attachments, a problem I raised in my 1968 paper.
34 Note that in 1980: 577, Herrmann called the Argive example north Syrian; and he did not
king midas of phrygia and the greeks 717
Another cauldron with bull head attachments has been introduced into
the literature as being Phrygian, the well-known example in Copenhagen,
which although without provenience is usually assumed to have come from
Cumae in Italy (see Akurgal 1961: figs. 33, 34). While most scholars convinc-
ingly attribute the cauldron to a north Syrian workshop (Herrmann 1966:
123, 128, 129, n. 45; 1980: 576; Kyrieleis 1977: 74, 76; DeVries 1981: 221), both
Young (1960: 231) and Birmingham (1961: 191) have called it Phrygian. It is
surely not Phrygian and could very well be north Syrian (cf. Muscarella 1968:
14f.). To summarize, there is but one bull attachment—which equals one
cauldron—from the west that seems to be a Phrygian export, the example
from Samos.
A review of Phrygian-Greek cultural relationship and contact must also
bring into the discussion a miscellaneous group of artifacts recovered in the
west that have been assigned, primarily by Herrmann, to Phrygian work-
shops. We have seen that Herrmann in his 1966 and 1975 contributions
attributed, incorrectly it was claimed, a number of bull attachments to Phry-
gian workshops. In his 1984 paper he has argued that it is possible to recog-
nize even more bronze Phrygian exports to the west, thereby considerably
increasing the inventory of exports, and at the same time expanding the
range of categories dedicated by Phrygians. Two of the objects purported
to be Phrygian derive from Samos, two from Olympia, and one each from
Delphi and Lindos.
The Samos objects consist of a winged siren-like attachment for a caul-
dron, and an ibex. The attachment was first published by Walter (1963: 294,
pl. 342a), and was assigned by him to Phrygia, an attribution accepted by
Herrmann (1984: 19f.). It is an odd piece indeed, not a typical siren attach-
ment of the oriental or Greek form, and it was recognized by Kopcke (1963:
284, no. 57) as a provincial work, probably of local manufacture.35
mention here the Olympia and Athens examples: probably as a reaction to Kyrieleis 1977, a
critique of Hermann’s attributions.
35 Birmingham 1961: 190, mentioned “siren figures” excavated at Samos. Probably she was
discussing this (single) example. See also A.C. Brookes, “Greek Siren Cauldron Attachments,”
AthMitt 99, 1982: 608, who also mistakingly claims that siren attachments derive from Altin-
tepe. Birmingham (191) also calls the classic siren attachments “Urartian-Phrygian” which is a
meaningless term. In a recent article F. Isik (AnatStud 37, 1987: 163–178) attempts to document
8th century bc cultural influences of Urutu on Phrygia, and to some extent, vice-versa: here,
for example (169), by claiming—without documentation—Phrygian influence on Urartian
painted pottery, which in fact is rare and usually dated post-8th century. Işik (169) considers
the siren cauldrons from Tumulus MM at Gordion to be Urartian exports (“gegenständlich
belegt”), Unfortunately, Işik does not cite the extensive literature attributing these cauldrons
718 chapter twenty-three
The second Samos object was also first published by Walter (1963: 294 f.,
pl. 342c), and also assigned by him to Phrygia; it is not mentioned by Herr-
mann. Jantzen (1972: 70, pl. 72, B1282) considered it to be Assyrian, but
Kopcke (1968: 291f., no. 119) correctly recognized its apparent Iranian (“Berg-
lands”) background (Muscarella 1973: 237).
In his 1984 paper Herrmann also considered an odd, crudely executed
attachment of a male figure with outstretched arms (1966: 32, 49, 81 f., pl. 20,
A15) and a Greek-like (to my eyes) siren attachment discovered in 1965 at
Olympia (1984: 19f., pl. 5, A15a) to be Phrygian productions. In 1966 (82 f.,
nos. 43, 54, pl. 27) he discussed a siren attachment from Delphi and one
from Lindos. For both he suggested that they might be East Greek, but he
also suggested that they could be Phrygian; in 1984 (18, 20) he rejected the
East Greek attribution and claimed them for Phrygia.
For all the above attachments Herrmann correctly noted that they “erhe-
bliche stilistische Unterschiede aufweisen,” but he urged that this fact is not
a problem, inasmuch as Phrygian art known in its homeland, aside from pot-
tery, is not homogeneous. Indeed, Herrmann is correct in this observation
about Phrygian art. But in a meaningful archaeological discourse it cannot
occur that any odd, unidentifiable, piece found in the west, artifacts that
have absolutely no parallels either in Phrygian art or among its excavated
artifact forms which description obtains for all the attachments discussed
here may casually be assigned to Phrygia. Such attributions do not lead to
historical reality.36
Summarizing in brief the historical and archaeological evidence for
Phrygian-Greek contact limited to the period of Midas-Mita, we have at
Gordion Greek pottery and fibulae, and probably a Greek princess; and in
Phrygia at large we have a Greek alphabet. In the Greek sphere there is
Midas’ throne at Delphi, and bronze fibulae, belts, probably some bowls, and
a cauldron,37 all dedications to various Greek deities at several sanctuaries.
From an analysis of the archaeological finds and of the historical sources we
have been able to recognize conscious knowledge of and friendly interac-
tion with Greek society, both on the west coast and the mainland. Among
to north Syria: he does cite Herrmann 1966 for the list of the Gordion cauldrons but he offers
no evidence that he has read Herrmann-who argued for a north Syrian origin.
36 See my discussion on the problem of Median art and the assignment of stray artifacts
the Greek sanctuaries the Heraion on Samos has in quantity and variety the
largest number of Phrygian dedications preserved: three or four fibulae, at
least one belt, and perhaps more, two bowls, and a cauldron; Paros has one
fibula; Lindos four or five fibulae; Olympia has fibulae and a bowl; Argos has
several fibulae and three bowls; Perachora has bowls. Surprisingly—because
we want to have material here to inform us about other kinds of objects
besides a throne that may have been dedicated—we have, to my knowledge,
no Phrygian material preserved at Delphi.
I do not think we can speculate at the present why Samos has more
material than that from the other sanctuaries. The reasons could be simple
ones, that Samos is closer to Phrygia and was easier for a Phrygian to visit;
or perhaps the extant material at other sanctuaries is merely a result of
accident, much having disappeared over the centuries. It is also possible that
here Hera was identified with Kybele, but we do not know.38
In the late 8th century bc Phrygia under the rule of its dynamic King
Midas was actively involved in international affairs. Midas intrigued and
sought alliances to the east and south with Urartu and Tabal and with
the north Syrian states, also with Assyria, first as an enemy and then after
709bc apparently as an ally. Equally, Midas involved Phrygia in political and
cultural affairs with the west, with the Greek states on the coast and with
the Greek mainland itself.
Phrygians travelled long distances overland and by sea to honor Greek
deities, and perhaps Greeks visited Gordion and other Phrygian centers.
King Midas, an international personality, and the Greeks were well ac-
quainted with each other.39 And the historical reality of this acquaintance
38 The Heraeum yielded dedications from Egypt, North Syria, Assyria, Iran, possibly the
Caucasus, and one object, a bell, from Urartu. The sanctuary was manifestly well known
to Near Easterners who must have had special reasons for dedicating votive material there.
Whether each people had the same reason is of course unknown to us.
39 Here we may mention the conclusions of K. DeVries in his provocative paper “Greeks
and Phrygians in the Early Iron Age”, in From Athens to Gordian, ed. K. DeVries, Philadelphia,
1980: 33–50. DeVries argues that there was a correspondance, not fortuitous, between the
“ways of life and certain attitudes” of the Phrygian court society and that revealed in the
Homeric epics and he implies that the epics may actually reflect a “living counterpart”
in Phrygia (especially pages 40 ff.). This conclusion is exciting and stimulating: but much
of it is based on subjective archaeological and personal interpretation—for example, that
hundreds of “not free” women worked in the Terrace Building at Gordion, and that gift
giving may account for the presence in Tumulus MM of vessels bearing personal names.
Indeed, the Homeric epics could in part have been based on the megaron society at Gordion,
but there is no hard evidence that this was so, or that other such societies. Still unknown
720 chapter twenty-three
Bibliography
E. Akurgal
1955: Phrygische Kunst. Ankara.
1961: Die Kunst Anatoliens, Berlin.
1966: The Art of Greece, New York.
1983: “Zur Datierung der ältesten ionischen Münzen mit Löwenkopfdarstellun-
gen,” in Beiträge zur Altertumskunde Kleinasiens, R.M. Boehmer, H. Haupt-
man editors, Mainz: 1–11.
M. Balmuth
1971: “Remarks on the Appearance of the Earliest Coins,” in Studies Presented to
George M.A. Hanfmann. D.G. Mitten, et al., eds., Cambridge.
R.D. Barnett
1967: “Phrygia and the Peoples of Anatolia in the Iron Age,” CAH fascicle.
J. Birmingham
1961: “The Overland Route across Anatolia in the Eighth and Seventh Centuries
B.C.,” AnatStud 11: 186–195.
K. Bittel
1942: Kleinasiatische Studien, Istanbul.
1950: Grundzüge der Vor- und Frühgeschichte, Tübingen.
1970: Hattusha, New York.
Ch. Bunkenberg
1926: Fibules grecques et orientales, Copenhagen.
1931: Lindos: les petits objets, Berlin.
J. Boardman
1961: “Ionian Bronze Belts,” Anatolia 6: 179–189.
1973: The Greeks Overseas, London.
R.M. Boehmer
1972: Die Kleinfunde von Boğazköy, Berlin.
1973: “Phrygische Prunkgewänder des 8. Jahrhunderts v. Chr. Herkunft und Ex-
port,” AA: 149–172.
to us may not have contributed their essence. At the same time, DeVries’ suggestions are
not to be eliminated from our thoughts, especially given the evidence presented in this
paper.
king midas of phrygia and the greeks 721
1983: “Eine tabalische XII, 9 Fibel aus Assur,” in Beiträge zur Altertumskunde
Kleinasiens, R.M. Boehmer, H. Hauptmann eds., Berlin: 75–82.
S. Buluç
1979: Ankara Frig Nekropolünden üç Tümülüs Buluntulari, Habilitation Thesis,
Ankara.
E. Caner
1983: Fibeln in Anatolien PBFXIV/8, Munich.
A. de Ridder
1896: Catalogue des bronzes sur l’Acropole d’Athenes. Paris.
K. DeVries
1981: contributions in Young 1981.
T.J. Dunbabin
1957: The Greeks and their Eastern Neighbors, London.
J. Friedrich
1941: “Phrygia,” in PW, Stuttgart: 782–891.
A. Fürtwangler
1890: Die Bronzen und die übrigen kleineren Funde von Olympia, IV, Berlin.
A. Goetze
1957: Kleinasien, Munich.
H.V. Herrmann
1966: Die Kessel derorientalisierenden Zeit, OL VI, Berlin.
1975: “Hellas,” in RLA Band IV: 303–311.
1980: “Kessel,” in RLA Band V: 573:578.
1984: “Kesselschmuck, nachträge,” AthMitt: 17–33.
G. Huxley
1959: “Titles of Midas,” GrRomByzSt 2,2: 85–99.
U. Jantzen
1972: Ägyptische und orientalische Bronzen aus dem Heraion von Samos, Samos
VIII, Bonn.
L.H. Jeffery
1976: Archaic Greece, London.
722 chapter twenty-three
D. Kagan
1982: “The Dates of the Earliest Coins,” AJA 86, 3: 343–360.
A. Knudsen
1964: “From a Sardis Tomb …,” Berytus XV: 59–69.
A. Koerte
1903: “Agamemnon,” in PW, Supp. 1, Stuttgart: 19.
1904: Gordion Ergebnisse der Ausgrabung im Jahre 1900, Berlin.
G. Kopcke
1968: “Heraion von Samos: Die Kampagnen 1961/1965 in Südtemenos,” AthMitt 83:
250–314.
W. Kroll
1932: “Midas,” in PW, Stuttgart: 1526–1540.
H. Kyrieleis
1977: “Stierprotomen orientalisch oder griechisch?,” AthMitt 92: 71–89.
M. Le Jeune
1969: “Discussions sur l’alphabet phrygien,” SMEA 10: 19–47.
D.D. Luckenbill
1927: Ancient Records of Assyria and Babylon I, Chicago.
M. Mellink
1965: “Mita, Mushki and Phrygians,” Anadolu Arastirmalan 2: 317–325.
1981: contributions in Young 1981.
O.W. Muscarella
1967: Phrygian Fibulae from Gordion, London.
1968: “Winged Bull Cauldron Attachments from Iran,” MMA Jour 1: 7–18.
1973: review of Jantzen 1972 in AJA 77, 2: 236–237.
1977: “The Archaeological Evidence for relations Between Greece and Iran in the
First Millennium B.C.,” JANES 9: 33–57.
1982: review of Young 1981 in Quarterly Review of Archaelogy 3, 4: 7–10.
1988: “The Background to the Phrygian Bronze Industry,” in Bronzeworking Cen-
tres of Western Asia 1000–539 B.C., J. Curtis ed., London: 177–192.
P. Neve
1972: “Bericht über die Deutsche Boğazköy-Expedition im Jahre 1970”, TurkAD 19,
1: 175–186.
H. Payne
1940: Perachora I, Oxford.
king midas of phrygia and the greeks 723
P. Perdrizet
1908: Fouilles de Delphes, V, Paris.
J.N. Postgate
1973: “Assyrian Texts and Fragments,” Iraq 35, 1: 13–36.
F. Prayon
1987: Phrygische Plastik, Tübingen.
L. Roller
1983: “The Legend of Midas,” Classical Antiquity 2,2: 299–313.
K. Sams
1974: “Phrygian Painted Animals: Anatolian Orientalizing Art,” AnatStud 24: 169–
196.
1978: “Schools of Geometric Painting in Early Iron Age Anatolia,” in Proc. of the
Xth International Congress of Classsical Archaeology, Ankara: 227–236.
1979: “Patterns of Trade in First Millennium Gordion,” Arch. News 8, 2/3: 45–53.
E. Simpson
1983: “Reconstructing an Ancient Table …,” Expedition 25, 4: 11–26.
1986: “Royal Wooden Furniture from Gordion,” Archaeology 39.6: 40–47.
A. Snodgrass
1971: The Dark Age of Greece, Edinburgh.
M. van Loon
1966: Urartian Art, Istanbul.
H.T. Wade-Gery
1952: The Poet of the Iliad, Cambridge.
G. Waldstein
1905: The Argive Heraeum, New York.
H. Walter
1963: “Die Ausgrabung irn Heraion von Samos (1952–1962),” ADelt 18, 2: 286–296.
H. Winckler
1901: Altorietalische Forschungen, 2nd series, Leipzig.
R.S. Young
1958: “The Gordion Campaign of 1957,” AJA 62, 2: 139–154.
1960: “The Gordion Campaign of 1959,” AJA 64, 3: 227–243.
1963: “Gordion on the Royal Road,” ProcAmerPhilSoc 107, 4: 348–364.
1969: “Old Phrygian Inscriptions from Gordion,” Hesperia 38, 2: 252–296.
1981: Three Great Early Tumuli, Philadelphia.
chapter twenty-four
Plates I–Vb
The 8th century bc was truly one of the major periods in the history of
the western world. During this century, primarily in its second half, the
Greeks, fully emerged from a Dark Age, commenced active trade and multi-
level cultural exchange with the Near East, the Orient. The evidence for
these activities is primarily the creation of the Greek alphabet, and also the
presence of manifold oriental objects and motifs recovered in Greek soil.
One of the most prevalent classes of oriental material that reached Greece
in the 8th century bc was the large bronze cauldron fitted with winged
siren1 and winged bull-head attachments. More significant in a cultural
sense is the fact that the imported cauldrons and attachments were copied
by local craftsmen. While not alone among the imported material copied
and adapted to satisfy local needs, the Greek-made cauldrons, especially
those with sirens, constitute some of the best evidence available of the phe-
nomenon known as orientalizing.
Cauldrons with attachments have been of special interest to me since
1957, when I saw my first siren and bull cauldrons in the great Tumu-
lus MM at Gordion (pl. Ia), which encounter led to my 1962 paper on sirens.
Although much has been written about cauldrons since 1962 (especially
Herrmann’s potent and laudable monograph, 1966a), some of the ques-
tions asked remain unanswered, some questions were never asked, and
conclusions about their origins are still contested. It seems appropriate,
Review,” in Greece between East and West: 10th–8th Centuries B.C., eds. G. Kopcke and I. Toku-
maru (Mainz: von Zabern, 1992), 16–45.
1 I will call them sirens as a convenient, well-known term: see Amandry 1956, 244, n. 14;
Herrmann 1966a, 28; Wartke 1985, 90. In this study I shall not cite or discuss individual
differences among the sirens or their identification and background, as these matters are
discussed elsewhere. I wish to thank Susan Pattullo for reading and commenting upon this
paper.
726 chapter twenty-four
2 On May 11, 1959, Kunze told me he wished to remove no. 21 from his list; no. 39 is
eight from the royal tomb at Gordion, Tumulus MM (pl. Ia; infra), arriving
at a total figure of 78 examples in the corpus.3 This figure did not include
two aberrant attachments from Samos, which Herrmann believed did not
correctly belong to the class of sirens (29, 147, 187).4
Jantzen (1967, 91ff.) argued for the addition of the Samian pieces, recog-
nizing them as local, orientalizing productions of an oriental form. In his
review of Herrmann 1966a, Amandry (1969, 797) also accepted the Samian
pair as legitimate members of the class (see also Rolley 1984, 278); he further
added another example from Delphi (Rolley 1967, no. 140; pl. Ib) and the
four aberrant hammered sirens from Salamis on Cyprus (Karageorghis 1973,
97ff., figs. 18–23, pl. 244), accepting them as local concepts of the class. At
the same time, Amandry removed from the corpus Kunze’s and Herrmann’s
nos. 7, 8, and 9 (of the alleged Van group), maintaining that these pieces have
never been seen and, in fact, may never have existed. These additions and
subtractions yielded a total number of 82.5
Herrmann in 1984 (21) added a pair on an unexcavated cauldron in
Munich (infra; pl. IIa), but he provided here no total figure. Barnett (1986,
112) noted Amandry’s 1969 deletions, the additions of the Samian pair and
the one from Delphi, but not the four from Cyprus; he also cited the recently
published siren in the Glencairn collection in Bryn Athyn, Pennsylvania
(Romano and Pigott 1983)6 but not the Munich cauldron. Although his fig-
ures add up to 80, Barnett claimed 81 sirens for the corpus.7 Finally, Rolley
(1984, 280) added three more sirens (said to be Greek) from Delphi (unpub-
lished).8
My review of the figures yields the following results. Appropriately start-
ing with Herrmann’s 1966a figure of 78, I would remove his no. 51, a bull figure
799, thinks that no. 41 is close to the Gordion examples: I accept the former but not the latter
conclusion.
7 Note that Barnett misstated that Herrmann catalogued 78 oriental sirens: this figure, of
course, included both Greek and oriental examples. Barnett also misstated the actual number
of sirens excavated in Greece (see infra).
8 In Filippakis et al. 1983, 131, sample nos. 76 and 77 are listed as oriental sirens from
Delphi, and on 132 sample no. 78 as Greek: if the first two are the same as those called Greek
in 1984 by Rolley, the Greek and oriental lists presented in this study would be skewed. Future
publication may resolve the apparent problem.
728 chapter twenty-four
from Argos (pl. Vc–d, Kunze’s 1950 addition; see Herrmann 1966a, 79 f., 153,
156), because I find no clear evidence that a siren was necessarily originally
present. Kunze (1931, 12), Amandry (1956, 245, n. 17), and Rolley (1984, 283)
envisioned a siren with a bull placed on each wing, the latter ingeniously
citing as an exact formal parallel the cauldron from Salamis, which has
a griffin on each wing of a siren (eight griffins with four sirens). To me
the case is not proven.9 Following Amandry’s suggestion, nos. 7, 8 and 9
should be removed (but to be re-added if ever they surface). This leaves
74 examples. To this figure we should add the two examples from Samos,
the four from Delphi cited by Rolley (1967, 1984), the four from Cyprus
(recognizing their uniqueness), the one in the Glencairn collection, and
parenthetically the clay head in the Ashmolean Museum.10 We are also
obliged to add, tentatively accepting them to be genuine, the two sirens on
the cauldron that surfaced on the Munich art market (pl. IIa).
These figures add up to 87 bronze sirens in the corpus,11 88 sirens, if we
accept the Ashmolean Museum clay head. A number are clearly aberrant,
non-canonical, but clearly (or seemingly) expressing the same form and
concept: those from Samos and Cyprus, two from Olympia, A14 and A15
(Herrmann 1966a, 32; idem 1984, 17ff.), and the two on the Munich art market
cauldron.
9 Blegen 1939, 430, stated that there are “traces apparently of a hand” on the plinth’s
base; Kunze 1950, 96 says that on the object “ruht eine Hand flach auf.” I see in the photo
a raised area to the bull’s right and a projection: a thumb? Rolley 1984, 283, cites a similar
example from Delphi but did not supply an illustration. In any event, no siren with bulls
on its wings exists. Gabelmann 1964, 6, 8, accepts Kunze’s conclusion and goes further: he
believes that a bronze lion figure and a siren in the Petit Palais (Herrmann 1966a, 155, no. 32,
59, no. 62) belong to the same cauldron and thus represent another example of animal-siren
juxtaposition: which to him is Urartian. Each of these assertions is unacceptable: the Petit
Palais collection is a modern, not an ancient, assemblage. Ingrid Strøm in her paper also
accepts the presence of a siren (infra, 52). And since the sirens come from North Syria (infra),
she further concludes that the juxtaposed animal handle, and other handles (“Kesseltiere”),
are also from North Syria (54, n. 29): but there is no archaeological or stylistic evidence to
support this conclusion (Muscarella 1970, 115).
10 Kunze 1950, 101, removed the head from his 1931 list, but later (apud Herrmann 1966a,
28, n. 8) reversed himself again. In CVA Oxford fasc. 2 (1931), 55, pl. 2, 2–4, it is called a siren
head. Levi 1945, 290, said it “belongs to a vase;” Boardman 1961, 60, no. 258, also calls it a siren
head; on 52 he cited Kunze no. 11 and the 1950 retraction. Amandry and Barnett implicitly
accept Herrmann’s rejection of the head; in any event, it is not a bronze siren attachment.
11 After the completion of this paper I was informed that a siren tail had been excavated at
Isthmia; it will be published by I. and A. Raubitschek. This brings the corpus up to 88 bronze
sirens, 89 for the full corpus.
greek and oriental cauldron attachments: a review 729
How many of these sirens were excavated, and where? Concerning ourselves
first with the oriental sirens recovered in the Greek sphere, we recall that
Herrmann listed 21 examples from Olympia, two of which were recovered on
a cauldron with protomes (1966a, 11ff., 30f., 57, 187, pl. 1; idem 1984 mentions
20). Agreeing with Brandes (apud Herrmann 1966a, 55, n. 32, 83, n. 24),
Brookes (1982, 609f.), and Rolley (1984, 278), I would remove Herrmann’s
nos. A14, A15, and A15a (Herrmann 1966a, 187; Herrmann 1984; Taf. 5: 1–3)
from the oriental list and transfer them to the Greek category. The hair form
and herring-bone pattern of A14 are not the same as those of any oriental
siren, and the fact that it is aberrant does not make it oriental (cf. its round
base to Herrmann’s pl. 46, 1). A15 and A15a—two distinct pieces despite the
similar numbers assigned—fit into the same category: the arms and hair
on both are not obvious oriental forms.12 This leaves 18 oriental sirens from
Olympia, and an estimated maximum number of 15 cauldrons (accepting
A3, A4, A5, A5a, and A10, A11 as pairs).
From Herrmann’s Delphi list (57f.) of 13 examples, two of which were
recovered on a cauldron I would omit no. 37, the well-known bearded siren
in Copenhagen (pl. IIb); purchased from a dealer, it has no provenance,
Delphi or elsewhere (infra).13 Agreeing with Brookes (op. cit.) that it is not
oriental, I omit no. 44, in the Louvre. For the same reason I would omit
nos. 42, 43, 45, and 47 (as Brandes, op. cit.). Only at first view may no. 43
deceive, but the hair and face form and, especially, the attenuated and non-
articulated hands and arms against a short wing do not point very far east
(see also Rolley 1984, 278). No. 45 was described by Perdrizet (1908, 81) as
having no arms (but is considered to be oriental by Filippakis et al. 1983,
131, sample no. 21). No. 47 is a thin fragement with no head preserved.
After adding Rolley’s 1967 (no. 140) example, we arrive at the figure 8 for
the number of oriental sirens from Delphi, with an estimated maximum
number of seven cauldrons.14
There are two oriental sirens on a cauldron from the Ptoion. The two
single, detached examples from Argos and Delos (nos. 52, 53, 50 and 55)
equal two cauldrons. A single example from Lindos on Rhodes (no. 54), one
cauldron, is not easy to attribute (especially lacking autopsy). Originally,
12 Kyrieleis 1967, 7, claimed A14 was Urartian; Herrmann 1984, 20, claimed A15 and A15a
to be Phrygian; to Kilian-Dirlmeier 1985, 247, A15 is Syrian, A15a Phrygian; see also Herrmann
1966a, 83, n. 24.
13 It was once offered to the Metropolitan Museum of Art for sale by a Dr. Hirsch.
14 Note that I accept no. 39 as oriental, contra Brandes; see also n. 8 above.
730 chapter twenty-four
15 Herrmann 1966a, 82, believed that nos. 43 and 54 were made in the same workshop,
mainland. Of the five listed from Olympia (91f., 102), only three were actually
excavated there, nos. 2 and 3, a pair, and 10. But no. 11, in Boston, is clearly a
mate to no. 10 (Herrmann 1966a, pls. 32, 33), and is thereby rightly included
with the Olympia finds. No. 6 in the British Museum, however, is a purchased
piece with no provenance (pace Herrmann 1966a, 92, 95, n. 8).17 To the four
excavated examples I add the three removed from the oriental list (nos.
A14, A15, A15a), for a total of seven Greek-made sirens from Olympia, which
equals five cauldrons.
To Herrmann’s list of two Greek sirens from Delphi, nos. 1 and 4, I add
the five removed from the oriental list (nos. 42, 43, 44, 45 and 47)18 and the
three reported by Rolley 1984, for a total of 10 Greek sirens from Delphi,19 and
about 8–10 cauldrons. Athens has four sirens on 4 cauldrons, nos. 5, 7, 8 and
9?, not five as Herrmann claims (1966a, 29, n. 9, 102, 113), for his no. 12, in
Munich, is not an excavated piece (v. Kunze 1931, no. 43). Herrmann’s no. 9
is a problem piece, for only a head exists and in a poor state of preservation.
Nevertheless, it may be a siren as claimed and is included here tentatively
pending autopsy. And Samos has two examples on one cauldron.
We arrive thus at the figure 23 as a base number of excavated Greek-made
sirens from the Greek sphere: 21 from the mainland and 2 from Samos. If one
considers the Lindos siren to be Greek/East Greek, the figure is 24, and the
oriental figure would be reduced to 30. But it would be too harsh a position
to defend if we were to exclude from the above list Herrmann’s 1966a, nos. 6
and 12, from two cauldrons, in London and Munich. Both are Greek-made
and surely derived—if not scientifically excavated—from somewhere in
the Greek mainland. And although the clay Ashmolean head from Crete is
not unambiguously a clear-cut member of the siren class, I include it here
tentatively, as “possibly a fragment of a siren.”20
The total count of Greek-made sirens is thus 26, from about 21–23 caul-
drons, a figure more than double the number hitherto recognized. These
figures do not include the four locally made sirens from Salamis on Cyprus,
for this region warrants a distinct cultural designation.
17 It was purchased in 1914 from a Greek who lived in Paris; it was he who said it “came
attribution. Neither Amandry nor Herrmann cite Kunze’s no. 31, hence I do not include it
(what and where is it?).
19 See Filippakis et al. 1983, 132, sample no. 78 for, apparently, one of these sirens.
20 On Crete griffin head protomes were locally made in clay, Hanfmann 1957, 243; Board-
There are still only nine excavated sirens, from two sites, in the Near East:
four each on two cauldrons from Tumulus MM at Gordion (pl. Ia), and an
isolated example from Alishar (Verachram), a site on the Iranian side of the
Araxes River (Herrmann’s nos. 10–18).21 No examples have been excavated in
the heartland of Urartu (infra).
To summarize: 76 of the 8822 sirens in the corpus were excavated (in the
Near East, Cyprus, Greece, Italy), about 86 %. Of the corpus, 57 (30 oriental,
26 Greek, and Lindos), ca. 64%, derived from the Greek sphere (the Italian
examples are not included here; see supra), which is the same percentage
recorded by Ipsen and Herrmann; and of these 57 examples, representing
an estimated 47 to 49 cauldrons, about 45% are of Greek manufacture.
Olympia and Delphi have both oriental and locally-made cauldrons; Ptoion,
Argos, and Delos preserve only imports; and Athens and Samos preserve
only locally made examples. Significant is the 45 % figure: it indicates both a
more intense degree of local copying, that is, of orientalizing, and, perhaps
a larger number of Greek workshops in existence in the late 8th and early
7th centuries bc than previous studies had suggested.
21 Amandry 1969, 798, joins Herrmann’s nos. 2 and 3 in Istanbul with no. 62 in Paris,
suggesting that they with a missing siren were originally on one cauldron as at Gordion.
Herrmann (1966a, 6 f., 77, 87, n. 13, 143, n. 2) previously had joined no. 62 to the Barberini
cauldron.
22 I include in these summaries the Ashmolean clay head and the London and Munich
the region where the primary workshops flourished24 (see also Strøm’s paper,
infra, 54). That a number of workshops functioned there is evidenced by the
variety in details of the sirens, although we are incapable of matching indi-
vidual sirens to specific North Syrian centers (Herrmann 1966a, 73).25
Unfortunately, some scholars have been misled to believe that a number
of sirens derived from Urartu and, while accepting a primary North Syrian
involvement, have argued that Urartu must have had at least a partial role in
their manufacture (e.g., Brandes apud Herrmann 1966a, 55; van Loon 1966,
110; Wäfler 1975, 225, n. 1298; Amandry 1969, 799; idem 1978, 5 f.; Boardman
1980, 65). But in fact, only the example from Alishar in northwestern Iran
derives from Urartu-controlled land, and its presence there represents a
rare import from the west.26 The Berlin “Toprakkale” siren, long assumed to
have been excavated, is now known to have been purchased in Europe (not
in Van, as Piotrovskii 1967, 37). Consequently, “fehlt damit die wesentliche
Stütze für die urartäische Provenienz der gesamten Gruppe” (Wartke 1985,
87, nn. 2 and 3, 92; see also Muscarella 1988b, 28 f., n. 4; also Herrmann 1966a,
59, n. 36).
Notwithstanding the archaeological and art historical evidence pre-
sented since the early 1960s against an Urartian center, it is still asserted that
siren cauldrons were made in Urartu and exported west,27 through Anatolia
and into the Greek sphere (e.g., Rolley 1967, 15; idem 1984, 286; Karageorghis
24 One must think also of Riemschneider 1959, 66, who in an odd article was perhaps the
first scholar to argue against a major Urartian center; see also Riemschneider 1965, 105f.; Ipsen
(apud Lehmann-Haupt 1931, 495 f.) also cited North Syrian elements; so did Lehmann-Haupt
(497), but he claimed an Urartian origin for the sirens. Note that Kilian-Dirlmeier’s article
(1985) on foreign material excavated in Greek sanctuaries should be used with caution. Her
list of the individual objects recovered (244 ff., and in charts and maps) is of value, but her
assignment of them to specific oriental cultures is not always accurate; the number of objects
claimed to derive from each culture as well as their percentages within the totality of finds
in the sanctuaries will have to be revised.
25 Herrmann 1966a, 66 f., does of course cite parallels between the sirens and specific
North Syrian reliefs, but he holds back from claiming that siren workshops existed in the
centers with the parallels. See also Winter 1988, 198, but she gives more emphasis to this
matter than Herrmann himself did.
26 Note that Riemschneider 1965, 105 f. and Amandry 1969, 798 and 1978, 5 (both of whom
believed that some sirens came from Urartu) thought it possible that they arrived there as
imports.
27 Part of the problem facing classical and other scholars is that they had been led
to believe, incorrectly, that Urartu conquered North Syria. But Urartu only reached the
northeastern area of Malatya: none of the major centers was ever conquered by Urartu. Note,
e.g., Barnett 1969, 146, who claims that Tell Rifa"at is in Urartu! See Dyson and Muscarella
1989, 19 and n. 93.
734 chapter twenty-four
1973, 113, n. 9; Barnett 1982, 367; idem 1986, 112 ff., although here implying
some role for North Syria; Thimme 1982, 134; Zimansky 1985, 97; Işik 1987,
169).28
In the 1970s a cauldron with two siren attachments and a tripod stand
surfaced on the ever expanding Munich art market (pl. IIa; Muscarella 1978,
62, n. 2; Kellner 1976, 74, pl. 4, no. 99; Kroll 1979, 1980, 46 f., fig. 38). The
original publications of the cauldron were presented in exhibition cata-
logues of alleged Urartian art and artifacts consisting of recently acquired
objects in the possession of museums, private collections, and apparently
dealers.29 Along with other objects of unknown background the cauldron
was baptized Urartian by those who organized the exhibition.30 The fact
that the cauldron has no archaeological provenance (pace the label “Prov.
Transkaukasien-Urartäisch”), and that the sirens cannot by any manipula-
tion of manifestly Urartian art styles be attributed to an Urartian workshop
were totally ignored—an old story:31 but one we shall ignore.32
The cauldron’s siren heads differ in execution from all the others known:
the tilt of the head, the hunched neck and shoulders, the unlidded eyes and
short brows, the cursory execution of the hairform and lines, the slit mouth,
and the circular pattern on the chest and shoulder. The wing and tail and the
position of the arms and hands look oriental,33 the face less so.34 Although
puzzling, the Munich sirens appear (without autopsy) to be genuinen; one
would welcome a disinterested laboratory analysis.
But aside from the question of the cauldron’s origin and the uniqueness
of the sirens, there is another problem to be addressed. Only two sirens are
28 See also the confused article, Korti-Konti 1971, 281ff. Also, Frankenstein 1979, 269, 271.
29 Kellner’s no. 92, a vessel, is listed as belonging to one G.N. In the Ghent catalogue it
turns up as no. 194, there listed as belonging to a Swiss private collector. Does this mean it was
sold after its exhibition in Munich? Kellner 1976, 10, laments the recent plunder of Urartu—
“eine zweite Zerstörung Urartus, seiner Kultur und Geschichte;” and he calls his museum’s
exhibition “ein Abschied von einer vorgeschichtlichen Kultur …”: but he ingeniously refrains
from discussing the Munich museum’s role in the “Zerstörung” and “Abschied.”
30 St. Kroll has since rejected the attribution of the Munich cauldron to Urartu (personal
communication).
31 See Herrmann 1966b, 124 f., n. 148. It seems over the past years that if an object arrives
was made in a workshop other than those that made the other sirens.
greek and oriental cauldron attachments: a review 735
35 Which is a position I tentatively hold: see Muscarella 1981a, 50f. and supra.
36 I give but a few examples: Muscarella 1980, 56 f., nos. 119, 172, 177; Muscarella 1970, 120,
fig. 11; van den Berghe and De Meyer 1982–1983, nos. 183, 184 (see n. 39 below); Kyrieleis 1969,
76ff.; Kohlmeyer and Saherwala 1984, fig. 19; Herrmann 1966a, 4, fig. 6; Layard 1853, figs. on
178 f.; see also Liepmann 1968, 55 f.
37 This argument cannot be applied to the production of sirens, at the very least because of
style. One would appreciate disinterested laboratory analysis to demonstrate that the tripod
belongs to the cauldron.
38 Barnett 1986, 113, mistakenly placed Kyrieleis among the North Syrian advocates; he also
seems to imply that Herrmann believed that Urartian cauldrons with sirens reached Greece.
736 chapter twenty-four
95, 96, n. 53, 98).39 A few more views are worth mentioning for the record.
Wäfler (1975, 254, n. 1298) completely excludes North Syria as the Heimat
and posits that they were made in Tabal and/or Phrygia, Urartu (the Berlin
“Toprakkale” example), and Assyria; and Calmeyer (1973, 130) excludes both
North Syria and Urartu, positing an origin in east Anatolia.
The sirens believed to be of Assyrian manufacture are an example in the
British Museum and the one in Copenhagen (Herrmann’s nos. 1, 37, pls. 22,
23; supra).40 Beginning with Kunze (1931, 271, no. 1), the London siren, with
its round, plump face and chest, has rightly been singled out as standing
outside the canonical forms of the corpus and is assumed to have derived
from Assyria (van Loon 1966, 110; Kyrieleis 1967, 12 f., n. 68; Amandry 1969,
798; Muscarella 1970, 111; Rolley 1984, 278).41 Herrmann (1966a, 63 f., 148;
1966b, 119; 1980, 575) believed it to be the prototype from which the North
Syrian group developed, an assertion rightly rejected by Kyrieleis (ibid, 12 f.)
and Amandry (ibid).42
The Copenhagen head (pl. IIb) also stands out from the canon in the style
of the head and beard, the wing and tail patterning, and in the position of
the hands, which hold onto the rim of the cauldron. Kunze also thought
this attachment to be Assyrian, and most scholars—Herrmann, Kyrieleis,
Rolley, even Akurgal (1968, 38)—have followed him; only Riemschneider
(1959, 66) and Amandry (but not decisively—1969, 798) believed this siren
to be Urartian. I do not see in this head an Assyrian background but rather
a product of a North Syrian workshop (1962, 327, pl. 104b; 1970, 126, n. 17).
The lack of a mustache (at least it is not apparent or prominent), the hair
configuration and the spiral volute on the tail point to North Syria, from a
workshop that apparently was aware of Assyrian art.43
39 See also Seidl 1988, 172 f., n. 1. But Kyrieleis 1967, 13, n. 68 is correct in rejecting an
Urartian attribution for the Erlangen tripod statuette; see also Moorey 1973, 83ff., 90. Moorey
(84 f.) also cogently relates the face and hair of the Erlangen statuette to that of the British
Museum siren, infra; cf. Herrmann 1966a, 66.
40 Both exhibit ears, which are not found on other Near Eastern examples; some Greek
uity from North Syria, but that it and other sirens were of Urartian manufacture. It follows
that the Assyrians copied at least one North Syrian cauldron.
43 I call attention here also to an unexcavated bucket with six bearded, divine siren-figures
(Muscarella 1981b, 269 f.), which on stylistic grounds I attributed to a North Syrian workshop,
and which I compared to the canonical siren attachments; see n. 70.
greek and oriental cauldron attachments: a review 737
The class of winged bull head attachments excavated in the Near East can be
divided into two main groups based on stylistic details and manufacturing
techniques. One can recognize a homogeneous Urartian group and a general
Near Eastern group that is considerably diversified, but in which we may
discern two or more cultural backgrounds (Muscarella 1968, 10 ff.; 1970, 111 ff.;
1989, 262f.). The Urartian and Near Eastern groups are quite distinct, yet
paradoxically many scholars have failed to observe the distinctions.
Thirteen examples of the Urartian group have been excavated: four each
on cauldrons from Altintepe, Toprakkale,44 and Guschi, the last two convinc-
ingly reconstructed from the incomplete records; and an isolated example
from Alishar, where, as reported above, a siren was also recovered (Amandry
1956; Hanfmann 1957; Herrmann 1966a, 118, n. 10; Piotrovskii 1967, 36 ff.; Mus-
carella 1968, 7ff., n. 2).45 Not a single example of an Urartian bull attachment
has been excavated outside of Urartian territory. This geographic unity par-
allels the group’s distinct stylistic characteristics, which include a separate
casting of the head and the wing-tail unit, a prominent forelock with hair
curls, distinct muzzle incisions, and the placement of four outward-facing
bulls to a cauldron.
In recent years three bronze cauldrons, each preserving Urartian bull
head attachments, and a number of detached Urartian bull head attach-
ments have surfaced, all courtesy of the antiquities market. One of the
cauldrons, together with a tripod stand, is now in the Karlsruhe Museum
(Thimme 1982, 129ff.; van den Berghe and De Meyer 1982–1983, no 174). It
has three attachments symmetrically placed on the rim. A second, smaller
cauldron is in Japan (Tanabe et al. 1982, pl. 1) and has only two attachments
placed opposite each other. The third, also small, was originally advertized
for sale in London (Apollo, April 1981, 13). Long distance trade eventually
44 Note that the Toprakkale head published by Amandry 1956, 239ff., pls. 24, 26, d, as in
the N.K. collection is now in the Burrell collection, Glasgow, AfO 19 (1959–1960), 189, fig. 1;
Scottish Arts Review 6, 4, (1958), 27 f.; Amandry 1978, 4: I wish to thank James K. Thomson
of the Burrell collection for answering questions on this matter. Also note that via modern
long distance trade routes the head travelled through at least one other dealer after leaving
N.K.
45 Potentially very important information is not given in van Loon 1989 regarding a
collection of material on loan to the Oriental Institute of the University of Chicago between
1970–1979, which may have derived from Guschi: why did it take twenty years to publish this
collection, and why weren’t the owners questioned about their acquisition?
738 chapter twenty-four
brought it to a dealer in Japan (Tanabe et al., 1982, pl. 2; see also Maass
1978, 11ff., fig. 7a, b; Thimme 1982, 130, 132, fig. 5). This cauldron has four
attachments.
The detached attachments include an example in the Mildenberg collec-
tion (Verdier 1981, no. 18),46 and an example in the Louvre (van den Berghe
and De Meyer 1982–1983, no. 176). A small example in the British Museum
(Amandry 1956, 260, pl. 32, 3; Herrmann 1966a, 118, n. 10, 128, n. 44) purchased
in 1893 from a vendor, who said it came from Urmia, may also belong to this
group. Solid cast, it is quite small, 3cm. by 2.6 cm; the base is flat and has
eight small, circular depressions arranged in two rows of four. In style—the
forelock and the hook incision on the muzzle—it is Urartian, but if it is an
attachment, now separated from its wing and tail, it is the smallest known.47
The general Near Eastern group is less easily characterized because of the
variety of the head forms and styles and the heterogeneous shapes of the
attachment units, which are either naturalistically rendered, more or less,
as a wing and tail, or are abstracted into a T-shape. Nevertheless, certain
shared features exist: the head and attachment unit are cast together, the
heads are usually embellished with an oval or triangular forelock, and they
were fitted two to a cauldron. There is also a ring cast with the unit to hold
a loop handle, a feature lacking on the Urartian attachments.
Bull head attachments with these characteristics, often recovered on
their cauldrons, have been excavated in Anatolia and North Syria. In Ana-
tolia they occur at Gordion in Phrygia (13 examples), in Tumulus D in
Bayindir-Elmali in Lycia (4 examples),48 in a tumulus burial at Niǧde, south
of the Halys River (4 examples), and at the Urartian site of Karmir Blur in
Soviet Armenia (1 pair). In North Syria a pair derives from Tell Rifa"at (found
pressed together with no indication of a cauldron), one from Zincirli, and
one example was purchased years ago in Aleppo.
Of the thirteen Gordion examples at least eight—four pairs on four caul-
drons: three from a tomb, Tumulus MM, and one from the destroyed city
mound (MM 1, 12, 13, TB 1)—have bull heads that are very similar to each
other, with a triangular forelock, and a T-shaped abstracted wing and tail
attachment. Two other pairs, on two cauldrons from a rich, probably royal
46 It is not a mate to the Karlsruhe and Japanese examples with three and two attach-
ments.
47 John Curtis kundly furnished the information given here; he also supplied the weight:
66.5 grams.
48 Only one cauldron is published; I was informed that two cauldrons were excavated; see
49 Is Niǧde in Tabal, and thus is the Phrygian material evidence of Phrygian presence? Or is
the find evidence of gift-exchange or trade with a friendly ally? See Muscarella forthcoming.
50 The finds from these tumulus burials are both extraordinary and unexpected. Aside
from apparent 8th century Phrygian material, the tombs contained griffin protomes (infra)
and one silver and three ivory statuettes of East Greek/western Anatolian style that are
conventionally dated to the second half of the 7th century bc. Were Phrygians buried here?
See n. 49. A preliminary publication of these finds in Acar and Mazur 1989. Pace the quote of
von Bothmer, each piece is not “purely Phrygian.”
740 chapter twenty-four
51 In 1968, 7, 13, I incorrectly mentioned only one example, an error I picked up from
Saherwala 1984, 34, that Urartian workshops existed outside of Urartu proper and exported
their productions to Urartu, and the suggestion of Riemschneider 1965, 105f., and Amandry
1978, 4, that Urartians imported cauldrons: both views, of course, given to explain the alleged
presence of sirens in Urartu.
53 Galerie Arete, Antike Bronzen (Zurich, n.d.), no. 21; Isler 1982, 80, pl. 15, 4.
greek and oriental cauldron attachments: a review 741
54 Herrmann 1966a, 157, n. 7 sees a connection with a Greek attachment in the Stathatos
collection. Kyrieleis 1977, 82 compares the latter to 01 A33, which is more acceptable.
55 There is also a small silver bull head in Munich, Maass 1978, 13, fig. 8a, b. Is it a protome
or a cauldron attachment?
56 At first it would seem that only two examples exist, but no. 177 has a broken tail that is
not present on one of the sales catalogue examples. This suggests that there are two pairs—
subject to further information.
742 chapter twenty-four
Vessels with three bull heads exist. A terracotta vessel with three bull head
protomes excavated at Karmir Blur (Piotrovskii 1967, pl. 26) and probably to
be dated to the second half of the 7th century bc may be of local manufac-
ture. But here the heads are placed below the vessel’s rim, and they lack a
wing and tail.57 There also exists an unexcavated vessel with three bull head
protomes, lacking a wing and tail, placed around its rim (Kellner 1976, pl. 2,
no. 92; Kroll 1979, 1980, 34, fig. 25; van den Berghe and De Meyer 1982–1983,
no. 194). In all publications it is asserted that this vessel is Urartian; Kellner
even claimed that it came from Patnos! (But he did not say he would return
it there.) But why is it Urartian? And what is its date?58
May we invoke one possible Urartian (Karmir Blur) and one unexcavated
vessel of unknown provenance, both with three protomes, not attachments,
to demonstrate that a stray bronze cauldron in Karlsruhe is normative?59
And what of the cauldron with two attachments? May we objectively reject
Meyer’s conclusions, which are at least disinterested? Answers do not force
themselves upon us, and one may cite the Karlsruhe and Japanese evidence
only parenthetically as a source of evidence concerning Urartian practice.60
Turning to the unexcavated non-Urartian group, we encounter a problem
concerned with cultural attribution, a problem we shall experience yet again
when discussing the excavated finds from the Greek sphere. It is appropriate
to begin with the Copenhagen cauldron because it plays a prominent role in
the literature. Almost consistently, the cauldron has been assumed by schol-
ars to have been excavated at Cumae in Italy, an assumption that sometimes
produces skewed historical conclusions (Boardman 1980, 169; Coldstream
1977, 206).61 In fact, the cauldron was purchased in the international port of
Naples in 1900. Consequently, it has no known provenance.
The cauldron has been assigned to a number of cultural areas within the
Near East (an indication of the subjective state of art historical analysis)—
57 See also a fragmentary vessel that preserves a bull head protome in the Van Museum
Age.
59 One could assume that the 7th century Karmir Blur terracotta vessel is either an import
or a local copy of a foreign vessel: but I know of no such vessel with these protomes; see n. 51.
60 I am aware that this is not a world-shaking problem: but it is one that focuses on an
ever-present problem in our discipline, that of the methodology employed to gain cultural
information, especially with reference to using unexcavated material in that quest.
61 Coldstream suggests that the cauldron was sent to Greeks in Italy by a Phrygian royal
family; see infra and Muscarella 1989. The fact that the cauldron was purchased in Naples, a
port for many foreign ships, cannot signify it was found in Italy.
greek and oriental cauldron attachments: a review 743
but one may emphatically state that wherever it was manufactured, its
workshop will not have been in Urartu (as claimed by Thimme 1982, 133—
who cites as Urartian parallels unexcavated Near Eastern style bull attach-
ments; compare Amandry 1956, 243; Herrmann 1966a, 122; Akurgal 1968,
48—reversing an earlier attribution to Urartu; Liepmann 1968, 54 f., n. 37;
Calmeyer 1973, 130; and Maass 1978, 14f.: all of whom recognized its non-
Urartian background). Van Loon (1966, 106) inexplicably attributed the caul-
dron to Cyprus; others to Phrygia (Young in DeVries 1981, 219, n. 1; Coldstream
1983b, 206) or to North Syria (Herrmann 1966a, 122, 128; Kyrieleis 1977, 74).
Still others have recorded their hesitancy to place it within a specifically
known cultural center, North Syria for example (Muscarella 1968, 14 f.; Maass
1978, 14f.; Calmeyer 1973, 130).
Stylistic criteria recorded above indicate that Phrygia must be removed
from consideration, which leaves North Syria or another nearby center. I
suggest that we continue to employ the non-specific general term “Near
Eastern” (Muscarella 1968, 14f.; see also Winter 1988, 198) to identify those
attachments that are clearly neither Urartian nor Phrygian, nor manifestly
North Syrian, in which area there was regional diversity (for North Syrian
style and diversity see Winter 1988).
We may posit at least three North Syrian workshops, not solely because
of the find spots (two sites, plus the Aleppo purchase), but because of the
differences in the head form among the attachments. They give us some idea
of the variety of related styles/forms being produced in cauldron workshops
within North Syria. Variation in form and execution within a stylistic unity
is a characteristic of other metal workshops and also of ivory carving from
that area (Winter 1988, 194ff.).62
A concern with origin also must pay attention to the accident of archae-
ological recovery, in this case to the lack of excavated material that might
indicate the existence of other manufacturing centers outside of Phrygia
and North Syria, in adjacent south central Anatolia for example. Indeed,
the finds from "Ain Gev and Ai Radanah mentioned above are telling argu-
ments against casually offering attributions to the presently known centers.
But while compelled to admit that unrecognized workshop centers prob-
ably existed, we are constrained from hypothesizing where they were and
purport to know their products (Tabal is becoming a popular guess-area).
62 See also Winter 1981, 101 ff.; Winter 1983, 179 ff. Note a bronze wing and tail attachment
from Hama (Riis and Buhl 1990, 120 ff., no. 390). It has a ring at the rear, but if there was
originally a protome, it is missing.
744 chapter twenty-four
In short, the Copenhagen cauldron mayor may not have been made in a
North Syrian workshop; it is a Near Eastern cauldron. The same conclusion,
I believe, must obtain also for the Karmir Blur attachments and those from
Tumulus Wand Megaron 4 at Gordion. The Tumulus W examples have
triangular decoration above and below the forelocks, the same decoration
found on the sirens; but the head-wing-tail units are claimed not to be cast in
one piece, apparently a unique feature for a non-Urartian bull attachment.
If they were made in North Syria, it will not have been in one of the three
workshops mentioned above.
It follows, accordingly, that Herrmann’s North Syrian “Cumae Gruppe”
(1966a, 128f.; 1984, 23 f.) is misleading, not merely because of the improper
use of Cumae, but because the alleged group has no demonstrated stylis-
tic unity (on this v. also Maass 1978, 14ff.; Rolley 1984, 282). Included in
the group are sirens of disparate forms, and, I suggest, attachments made
in several distinct cultural centers reflected in the differences in styles—
Greek, North Syrian, Phrygian, and Near Eastern. The same mixture of styles
compromises Herrmann’s “weitere späthethische Werkstätten” (1966a, 128 f.;
1984, 24): Greek, Cypriot, Phrygian, Near Eastern. What is rather extraordi-
nary is that Herrmann in 1966 did not cite the excavated bull attachments
from North Syria in his discussions of the cauldrons and their attributions to
that area; they were first mentioned by him in 1980 (576; see also 1984, 23 f.,
where he assigns the Aleppo head to the “Curnae Gruppe,” the Zincirli and
Tell Rifa"at heads to a second and third North Syrian workshop).
The anonymous writer in the sales catalogue advertizing the Near Eastern
cauldron later loaned to the Zurich University collection (n. 53) identified
it as “Urartu, um 700 v. Chr.,” but then proceeded to associate it with North
Syrian attachments, the latter opinion accepted by Herrmann (1984, 26) and
tentatively by the author (1988b, 263, n. 1). In style, the bull heads could indi-
cate a North Syrian background,63 but the placement of four non-Urartian
heads, two embellished with raptors, is unique and not archaeologically
documented anywhere. It follows that we rather should avoid attributing
this orphaned cauldron, whose homeland may never be known.
The same conclusion is also in force for the equally unique Berlin caul-
dron with its two bull attachments (pace van den Berghe and De Meyer
1982–1983, 108, and Kohlmeyer and Saherwala 1984, 34 f., who infelicitously
63 For example, compare it to a stone bull head in Woolley and Barnett 1952, pl. B47:
muzzle swelling, eye form, round forelock, and head proportion. Herrmann 1966a, 122, n. 22,
cites this stone bull head as a parallel to Olympia A31, which it is not.
greek and oriental cauldron attachments: a review 745
label it Urartian; cf. Herrmann 1984, 26, n. 59a, who rightly rejects this
attribution). There are no close parallels for the stylized heads and they are
difficult to attribute.
Both the Brummer and Ashmolean detached winged bull heads are prob-
ably Near Eastern (not Urartian, pace Kyrieleis 1977, 73 on the latter head).64
Here too we may posit that they were made in North Syria or somewhere in
Anatolia. The Brummer head seems to be Near Eastern in form but is unlike
any of the others known to me.65
The Munich attachment is indeed very similar in almost all features to
those on the Copenhagen cauldron, as Herrmann, Kyrieleis and Maass have
recognized (the latter scholar alone not accepting a North Syrian back-
ground). The only obvious difference is the wing-tail attachment unit: natu-
ralistic wing-like outline and herringbone pattern on the Copenhagen fig-
ures, smooth and plain on the Munich one. Although the attachment is
unexcavated, the three above mentioned scholars (also Kilian-Dirlmeier
1985, 248) accept an Olympia provenance. This claim, like the one that the
Copenhagen cauldron was excavated in Cumae, complicates an already con-
fused situation; in both instances we do not know if they were plundered in
Greece, Italy, or the Near East.
The Barcelona attachment is probably, as Pallejá Vilaseca suggested, an
orientalizing Greek product (cf. Herrmann 1984, 24); it was purchased for a
private collection, and one cannot state that it was archaeologically recov-
ered in Spain. Perhaps also reflecting a Greek, perhaps East Greek or western
Anatolian, background are the two/four unique bull attachments exhibited
in Paris and Ghent. Wherever their origin, however, stylistic determination
precludes an Urartian attribution, as published (v. Kyrieleis 1977, 79, 82 for a
discussion of the Greek background of the Stirnband; for the wing structure
see Akurgal 1961, fig. 125, 153, 154, 210). The example in Brussels (pl. IVa) has
a plain wing and tail, small upright horns, and a flat elongated head ending
in a thick muzzle; it is not Near Eastern.66 Although it is by no means certain
that the Metropolitan Museum-Schimmel pair derived from Iran (even if it
was purchased there, Muscarella 1968, 7), stylistic analysis does not argue
against such a background (ibid, 17f.). The pair remains unique and may be
late manifestation, 6th century bc.
64 The style of the head and its casting technique are foreign to Urartian bull head
attachments.
65 The closest parallel I know is a goat (?) attachment in the British Museum, BMQ 37, 3/4
Bull head attachments excavated at sites in the Greek sphere are just about
equal in number to those excavated collectively in the Near East, Olympia,
12 (Herrmann 1966a, 114ff.; 1984, 22ff., A24a, A34a);71 Delphi, 8 plus (Per-
drizet 1908, nos. 327–332; Syria 35, 1958, pl. 6c, d; Rolley 1984, 280 mentions
several unpublished oriental examples); possibly one from Athens (Herr-
mann 1966a, 129);72 Argos, two (Waldstein 1905, pl. 75: 23, 25);73 Amyclae, one
67 See Özgen 1984, 91 ff., where some of these objects are published; in n. 4 on 119, Özgen
refers to his dissertation on some of the material in the collection. Note that neither in the
1984 article nor in the dissertation are bull heads mentioned. The bull heads are on exhibition
in the University Museum.
68 One of the heads, inv. 67-35-3 is claimed in the Museum records to fit onto three holes
winged animal head attachments (bulls? but no horns are visible). Objects like this one and
that mentioned in note 43 are known only from modern plunder, not from excavations.
71 Note that from Herrmann’s 1966a and 1984 total of 14 examples I believe we should
remove A28 and A34 because they are not clearly attachments, as Herrmann 1966a, 117f., 121,
126, himself realizes and which there he thinks might be Greek works.
72 Is this actually a cauldron attachment? See Amandry 1956, 245, n. 17.
73 There is a winged attachment on pl. 123, no. 2204, that deserves a better publication: is
(Muscarella 1968, 9, figs. 6, 7); Delos, one (Rolley 1973, 513 ff., figs. 20, 21);
Lindos, two (Blinkenberg and Kinch 1931, nos. 704, 706), and Samos, 11 or
more (Jantzen 1972, pls. 76, 77; Herrmann 1966a, pl. 52, 2; AM 74, 1959, Beil.
71: 1; Kyrieleis 1977, 83f., pls. 35–37; Isler 1978, 77, no. 16, pl. 38).74 These
figures add up to 38 bull attachments, plus the unpublished Delphi examples
and perhaps others from Samos.75 Only one example, Olympia A31, was
recovered in situ on a cauldron.
Cyprus has an isolated example from Idalion, two on a cauldron probably
from Kourion, and two units of three each on a cauldron from Salamis
(Catling 1964, 154f., pl. 21, c; Karageorghis 1973, 108 ff., figs. 25–28, pl. 246).
With the exception of Herrmann 1966a (infra),76 all scholars agree that
Greek workshops produced cauldrons with bull attachments, copied, of
course, from the oriental imports and paralleling the history of siren caul-
drons. But the literature reflects much disagreement and confusion once
attempts are made to isolate the imports from the locally made pieces. This
problem is well formulated by Kyrieleis (1977, 80) as “Die Frage ‘Orient oder
Griechenland’?” This question leads naturally to another, one that also fol-
lows from the previous discussion of Near Eastern workshops: whence in
the Near East did the imports depart?
In 1968 (15, n. 70), in reaction to Herrmann’s claims (1966a, 124, n. 23, 126;
1966b, 136) that not only were all the bull attachments from Olympia oriental
imports, but that in fact there is no evidence that the Greeks copied them,
I countered that indeed there were only about four attachments from the
Greek sphere that were imports. I cited the example from Amyclae, Samos
BB 740, Olympia A24, and no. 23 from Argos; and I cited a few about which
I was uncertain, Delphi no. 329, Argos no. 25, and the Cypriot example from
Idalion. It was also noted, and those scholars who have paid attention to
style and technique of manufacture agree, that not a single example of a bull
attachment recovered in the Greek sphere relates to the distinct Urartian
group. Since 1968 more discussion has ensued, and a review of my list of
imports and those of others follows.
First of all, it should be appreciated that using traditional stylistic anal-
yses does not produce firm conclusions. This is because, as recognized by
Herrmann (1966a, 121ff., 124) and Kyrieleis (1977, 71 f., 85 f.), it is far easier to
perceive and isolate distinctive traits in the style and execution of human
heads—eyes, mouths, hair, facial cast—than of bull heads. While to my
mind (and to Kyrieleis’) the majority of bull attachments in Greece are of
Greek manufacture, a certain number seem not to be. But which ones? For in
the course of studying the evidence and attempting objective attributions,
I realized that I changed my mind several times, which indecision precisely
indicates the problem.77
The oriental (Near Eastern) origin of the Amyclae attachment is accepted
by Herrmann, Kyrieleis and Maas. Kyrieleis (1977, 87) also agrees that Samos
BB740 is an import, but to him from Assyria, an attribution that has no stylis-
tic underpinning. I suggest it fits into the same Near Eastern background
that produced the Copenhagen cauldron and the Munich example (see also
Calmeyer 1973, 130). The presence and position of the volute on the wings,
however, allows us to make a more specific claim, namely that the Samos
attachment points to a North Syrian workshop. I am still inclined to think
that Olympia A24 and a recently discovered similar example, A24a (Herr-
mann 1984, 22, pl. 6: 1, 2)78 are Near Eastern; the form and position of the ears
and horns are very like those on the Tell Rifa"at heads, although the Olympia
heads are more attenuated. Kyrieleis (1977, 86, n. 93) is uncertain whether
A24 is Greek or oriental, and he calls it a “Zweifelsfall,” noting the lack of a
forelock.
I no longer think that Argos no. 23 is oriental, mainly because of the dis-
cussion of the Stirnband/Stirnrosette by Kyrieleis (1977, 79, 81 f., 84). Kyrieleis
(1977, 87) also shares my uncertainty whether Delphi no. 329 and Argos
no. 25 are Greek or oriental. Both pieces have the characteristics of the
“orientalische Typus,” including the forelock, but, echoing my own feel-
ings, Kyrieleis wonders whether “es sich um etwas provinzielle griechlis-
che Nachahmungen orientalischer Vorbilder handelt.” The execution of the
heads and wing shapes of both examples do readily fit into a Near Eastern
background.
After reexamining the Idalion head (not autopsy), I have no doubts but
that it is a local work (pace Herrmann 1966a, 123, n. 26),79 as are all the other
Cypriot examples cited (for oriental influence on the local bull attachment
production see Herrmann 1966a, 123f.; Liepmann 1968, 54 f.).
Syrian trade with Cyprus. Liepmann 1968, 55, sees it as being close to Herrmann’s Cumae
Group.
greek and oriental cauldron attachments: a review 749
Of the twelve examples recovered at Olympia, three, aside from A24 and
A24a, warrant further comment: A31, A34, A34a. Herrmann (1966a, 120 f.,
124; 1984, 23, n. 38), noting among other features the dewlap folds, strongly
argued that Olympia A31 was not a Greek work. Among those who discussed
the piece after 1966, only Maass (1978, 17) agreed it was oriental; Rolley
(1973,515), and later Kyrieleis (1977, 76f.), with extensive analysis, argued that
it was made in Greece.80 This is the position I took in 1968 (14, n. 20) and still
maintain, all the more so after Kyrieleis’ cogent analysis.
Likewise, I maintain that Olympia A34 is a Greek work. Originally per-
ceived as Greek by Herrmann (1966a, 121, 125f.),81 he reversed himself in
1984 (25). In the 1984 article Herrmann published another bull attachment
from Olympia, one similar to A34, that he labelled A34a, and which he also
attributed to the Orient. Maass (1978, 14) considered A34 to be securely
Greek, while Kyrieleis (1977, 79, 86) accepted A34a (he did not mention A34)
as oriental. To Kyrieleis, A34a and the unexcavated attachment in Munich
are the only certain (!) oriental examples in the Olympia repertory.82 My eyes
cannot find stylistic parallels that would securely place A34 and A34a in the
Near East (see also n. 71).
Delphi no. 373 is surely Greek, as noted by all the above cited schol-
ars, even now Herrmann (1984, 23, n. 38), who rejected his 1966 oriental
attribution.83 Delphi 327 and 328 (pls. IVb–c, Va–b) are considered to be
North Syrian by Herrmann in all his publications, and also by Kyrieleis (1977,
87),84 who sees a very close relationship to the Zincirli and Aleppo heads. I
doubted this conclusion in 1968 (14, n. 20), and I do not see the relation-
ship suggested by Kyrieleis: cf. pls. IIIa–b, IVb–c, Va–b. C. Rolley informs,
me that no. 327 preserves a small section of what must have been a short
80 No one has mentioned the claim by Herrmann 1966a, 7, 25, 142 that the cauldron to
which the attachment was joined (they were found separately) has placements for five bull
attachments—a feature not hitherto documented, neither in the Near East nor the Greek
sphere. Herrmann (142) denies that bull cauldrons ever appear with protomes—could this
cauldron be an exception?
81 Kyrie1eis 1977, 86, claimed that Herrmann recognized no Olympia attachments as
at Olympia. While Herrmann thinks A34 may not be an attachment (see n. 71), he makes no
such claim for A34a, which seems to be an attachment.
83 Kyrieleis 1977, 84 f., sees it as East Greek; he is uncertain whether it belonged to a
cauldron.
84 There is no indication that either Herrmann or Kyrieleis have actually examined the
Delphi attachments. Delphi nos. 330 and 331 remain unpublished and are not discussed here.
I wish to thank Claude Rolley for sending me photographs of Delphi Nos. 327 and 328.
750 chapter twenty-four
T-shaped attachment plate and evidence that there was once a ring at the
back of the head; no. 328 preserves only the head. No forelock is evident in
the photographs, although no. 327 is badly damaged. I cannot relate these
heads to the Near East, but as they are both poorly preserved, I will leave the
issue of origin open.
The Delos example, considered to be North Syrian by Herrmann (1975a,
307), is to my mind certainly Greek (also Rolley 1973, 513 ff., no. 18, figs. 20, 21).
An example from Lindos (Blinkenberg 1931, no. 706) remains unpublished.
It was described as having a wing and tail like the Lindos siren attachment
no. 705 (Herrmann’s no. 54, supra); not being published, one cannot discuss
it. Lindos no. 704, a double protome attachment, is clearly not Near Eastern
(cf. Herrmann 1966a, 125, 129, pls. 44, 54).
We turn now to Samos where, as may be expected, there is also contro-
versy concerning attribution: local or import? I consider three of the eleven
bull attachments from Samos to be Near Eastern imports, BB740 (supra),
B1266, and, highly probably, B438 (Jantzen 1972, pls. 76, 77). Herrmann (1980,
577f.), Kyrieleis (1977, 87) and Kilian-Dirlmeier (1985, 250) agree that B1266
is an import from Phrygia (v. also Muscarella 1989, 341). All also agree that
B348 is North Syrian; surely, if not a close local copy of an oriental model,
it must be Near Eastern, either from North Syria or another close-by work-
shop. Kyrieleis (1977, 80, 83f.) argues that B1482 (Jantzen 1972, pl. 77), B161,
B171, B172, and B303 are all Greek or East Greek, a conclusion I accept. B51
(Jantzen 1972, pl. 7) is a fragment lacking wing and tail (and face?), yet both
Herrmann (1975a, 307) and Kyrieleis (1977, 87, n. 101) see it as North Syrian,
Kilian-Dirlmeier (1985, 252) as oriental; I am not convinced it is oriental.
Calmeyer (1973, 130) accepts B348, B51, B1482, and B1266 as oriental, to him
from east Anatolia (but not Urartu). Samos then has, I suggest, at least one
example from Phrygia (B1266), one from North Syria (BB740) and one (or
more) ‘general Near Eastern’ (B348)—representing at least three imported
cauldrons.
Summarizing the Greek-oriental discussion, it is concluded in this review
that of the bull attachments recovered in the Greek sphere, no more than
six, representing six cauldrons, appear to be imports from the Near East:
Amyclae, Olympia A24, A24a, and Samos BB740, B1266, B348. Five others are
of uncertain origin and require more investigation and better publication:
Argos no. 25, Delphi nos. 327, 328, and 329, Samos B51.
Kyrieleis (1977) has nine definite imports in his list, representing nine
cauldrons, Amyclae, Olympia A34a, and the unexcavated bull attachment
in Munich; the same attachments from Samos as on my list plus B51; Delphi
nos. 327 and 328. He is uncertain about Olympia A24, Argos no. 25, and
greek and oriental cauldron attachments: a review 751
Delphi no. 329 (the last two are also on my uncertain list). However, one
reacts to the respective views, what seems almost certain is that Samos has
both imported and locally made bull cauldrons and Amyclae has an import;
Olympia, Delphi and Argos may have both.85 Furthermore, only two imports,
Samos BB740 and BB1266 may be attributed with some certainty to specific
Near Eastern polities, North Syria and Phrygia.
Griffin Protomes
Although the griffin protome is inextricably linked to the history of the siren
cauldrons (v. supra), it remains a major and vertiginous problem (along with
the lion protome).86 I am speaking about the origin, oriental or Greek again,
of the actual protome, not about the griffin motif or concept, whose origin is
manifestly Near Eastern (but not Urartian! Muscarella 1962, 319 ff.; 1981a, 50 f.;
Herrmann 1979, 137ff.). The question is whether oriental cauldrons arrived
in Greece with hammered griffin protomes and sirens as an ensemble, or
wether the protomes were added to the siren cauldrons in Greece to satisfy
(unknown) local needs.87 I will not discuss this problem here because I have
little to add (v. the discussion supra on the Munich siren cauldron), except
to note that Benson’s (1960, 65) and Herrmann’s (1966a, 137 f.) claim that no
cauldrons with siren attachments together with griffin protomes have been
recovered in the Near East—this ensemble occurring only in the West—still
obtains.88
Cast griffin protomes do of course occur in the Near East, and since
1979 a few more have surfaced; none are earlier than the earliest examples
recovered in Greece, and none are considered to be of local manufacture.
In addition to the already recorded cast examples from Susa, Miletus, and
Ephesus (Herrmann 1979, nos. 217, 131, 232, 242, 304, 349, 350, 143, 177 and 400
85 Olympia seems to have a majority of Greek bull head attachments, while Delphi may
possibly have a majority of oriental ones—just the reversal of the siren proportions; but this
will only be resolved by full publication.
86 To give but one example of the problem of attributing lions to a source, in 1981a, 52. I
stated that an oriental lion protome from Olympia, B4999, was North Syrian: based on a photo
published in ILN. The protome is now published in detail with good photos in Herrmann 1981,
72 ff., pls. 4–7. I now am less sure about its North Syrian origin as it is not typical for that area.
While I do not exclude that area, I cannot assert it.
87 Muhly 1988, 338, confuses the origin issue between hammered and cast griffin pro-
tomes; cast protomes are accepted by all as Greek, it is the hammered ones that are contested,
Greek-made or oriental import.
88 See n. 20.
752 chapter twenty-four
[?]), and examples in the Ankara and Izmir Museums89 (ibid., nos. 215, 245
and 302), we now have more excavated examples from Ephesus (Herrmann
1984, 32; Bammer 1984, 201, figs. 63, 64), from Samos (unpublished), two
from Ikiztepe near Ushak in western Turkey (unpublished), and two from
the extraordinary find from Bayindir-Elmali (Antalya Museum June 1989, 34,
187, nos. 29, 30).90 These last two are described as being cast; their date of
manufacture remains to be investigated.91
Herrmann (1984, 26ff., 32f.) has listed newly recovered hammered and
cast griffin protomes from the Greek sphere, from Olympia, Samos,
and Dodona, and some newly surfaced unexcavated examples. Hammered
and cast examples (unpublished) exist at Isthmia. There are now over 450
griffin protomes known to exist (cf. Muscarella 1981a, 47).
Chronology
Herrmann (1966a, 3f., 148) sees the beginning of siren cauldron production
in the Near East sometime in the third quarter of the 8th century bc; he
believes that export to Greece also began at that time (see Strøm’s paper,
infra, 47f.). Chronological evidence available for close review from a Near
Eastern center exists solely in Tumulus MM at Gordion. There is continuing
debate about the precise time of the tomb’s deposition. I have dated tumulus
MM to a time before the destruction of the site in ca. 696 bc, to ca. 720–700 bc
(1982, 9f.), while other scholars (ibid., 9) date the deposition either imme-
diately after the ca. 696bc destruction, or to a time before a later possible
destruction date, ca. 675 bc.92 Herrmann (1966a, 86 f., 129) also supports the
earlier dating of Tumulus MM to arrive at his conclusions.
The same chronology for the same reasons, namely their presence in
Tumulus MM, obtains for the Phrygian bull attachments (Herrmann 1966a,
129f.). But Tumulus W, ca. 750–740bc (supra), contained, as already noted,
89 Hanfmann 1957, 241, says the Izmir example comes from Labranda.
90 No cauldron is mentioned, or evidence of other fragments.
91 They derive from Tumulus C (the bull cauldrons came from Tumulus D). The griffin
heads look similar to the hammered heads in Herrmann 1979, G25 and G32, dated by him to
the early decades of the 7th century bc.
92 Without discussing the archaeological remains, Snodgrass 1971, 350 accepts a ca. 675bc
date for the destruction of Gordion, and he dates Tumuhls MM to ca. 685/680 bc. If this late
chronology were accurate, it would affect our perception and knowledge not only of Phrygian
history but also of the nature of Greek-Phrygian relations. Thus, Snodgrass sees Phrygians
playing a minor role in Greek orientalizing because of his low chronology; he does not discuss
the Midas gift.
greek and oriental cauldron attachments: a review 753
two bull cauldrons that seem not to be Phrygian, but North Syrian or ‘general
Near Eastern.’ Whatever their cultural origin, they are the earliest dated bull
cauldrons known to exist, and it seems from the Gordion sequence that bull
cauldrons were manufactured earlier than siren cauldrons.
How long production of bull cauldrons continued in Phrygia is unclear.
We may plausibly assume that the presence of the three examples in the
debris of the destroyed citadel indicates production down to the early 7th
century bc, if they were not curated heirlooms. Precise dating of the Bayin-
dir-Elmali tumulus, the deposition of which could date to a time in the 7th
century, perhaps even to the second half, will have some bearing on this
issue (see nn. 50, 91). Only a general late 8th–early 7th century chronology
appears to exist for the Urartian bull attachments, and one of the alleged
key chronological supports has been eliminated. Evidence for a possible
late 8th–early 7th century date is the poorly documented find from Alishar,
where a bell inscribed with the name of Argishti was apparently recovered;
but it is uncertain whether this is the first (ca. 780–750bc) or the second
(ca. 712–685 bc) king with that name (Muscarella 1978, 64, fig. 2). The Alt-
intepe tomb was for some time thought to date to ca. 740–732 bc because
of the presence of an inscription allegedly of Urikki of Cilicia. But it has
been known since Klein 1974 (87ff.) that the name Urikki in fact does not
appear, which eliminates Altintepe as a chronological datum.93 Toprakkale
was founded by Rusa II (ca. 685–670bc) not Rusa I (van Loon 1989, 268).94
The tomb is thus early 7th century in date—although the cauldron could
have been an heirloom. For subjective relative chronologies, or sequence of
manufacture, of the Urartian bull attachments based on style, see the differ-
ent views of Hanfmann 1956, 212; Azarpay 1968, 53; Verdier 1981, 31; Thimme
1982, 133; van Loon 1989, 268.
Of the North Syrian finds, only the examples from Tell Rifa"at may be said
to yield chronological information, although it is not very specific: 9th–7th
centuries bc (Winter 1988, 198, 208). But an 8th century date is supported by
the "Ain Gev clay example, which form presupposes the existence of metal
bull cauldrons at this time.
The Tumulus W evidence documents that bull cauldrons were being
manufactured in the Near East in the mid-8th century bc, and other finds
from Gordion, Urartu, Israel, and North Syria indicate a continuous produc-
tion and use to the end of the century and later.
93 Klein 1974 was ignored by DeVries 1981, 220; Brookes 1982, 608; and van Loon 1989, 268.
94 See also Salvini 1988, 130, 135 f.
754 chapter twenty-four
A review of North Syrian history in the late 8th century bc may yield
information concerning the chronology both of local production and the
period of export, assuming, of course, that North Syria was the source of the
siren cauldrons and of some of the bull cauldrons. Winter (1976, 17 ff.) has
noted that the production of North Syrian ivory and metal artifacts seems
to taper off, or may have ceased, in the late 8th century bc. She links this
situation to the Assyrian destruction of North Syrian cities by Sargon II,
beginning in 720bc with Hamath, followed by Carchemish in 717 bc, Malatya
in 712 bc, and so forth. These destructions must have interfered with local
workshop productions. The date of the destructions suggests the possibility
that the export of the cauldrons to the Greek sphere could have been in force
before the Assyrian onslaught and ceased, or at least declined, sometime
in the penultimate decade of the 8th century bc. This chronology is not
readily confirmed from stratigraphical excavations on the Greek side where
griffin cauldrons—but not those with siren and bull attachments—are first
depicted in art in the late 8th (presumably) and early 7th centuries bc
(Herrmann 1966a, 1f., figs. 1–4).
Methodology
95 Part of the problem here is that there may have been more manufacturing workshops
in different areas of the Greek sphere than hitherto recognized; see Kyrieleis 1977, 77f., 79f.,
83 ff.
greek and oriental cauldron attachments: a review 755
Function
Although we are not able to infer all the implications connected with the
presence of the oriental cauldrons in the West, their ultimate function
seems relatively easy to ascertain. Evidence from many excavations in the
Greek sphere indicates a pattern: the cauldrons occur only in sanctuaries
of various deities, male and female (see Strøm’s paper, infra, 52 ff., fig. 5),
and they served as votive objects deposited as dedications (for a different
interpretation, banqueting, see Strøm’s paper, infra, 55 f.). It is fundamental
to recall here that the dedication of cauldrons to deities at their sanctuaries
greek and oriental cauldron attachments: a review 757
was a Greek custom in existence before the time when the first oriental
examples arrived (Coldstream 1977, 334ff.), and continued in later periods
(Herodotus IV. 152).
In the Near East the cauldrons apparently had functions that were similar
to those known in the West. At Gordion, siren and bull cauldrons were
deposited in tombs, and bull cauldrons were also recovered on the citadel.
Recent cogent research on the contents of Tumulus MM (Simpson 1986, and
1990, 69–87; v. also Moorey 1980, 195) suggests that most likely the cauldrons,
and accompanying vessels, were used in a funeral banquet for the deceased
before they were placed in the tomb (see Strøm’s paper, infra, 55). At first
examination, it may seem that a tomb deposition precludes a dedicatory
function, which is Herrmann’s (1966a, 149) meaning when he speaks of a
“sepulkrale Funktion” in Phrygia as opposed to use as “Kultgeräte” in Greece.
This conclusion, however, may be too restrictive. In Tumulus MM, and in
the burials at Niğde and Bayindir-Elmali, after use in the banquet, which
itself may have been a religious ceremony, the cauldrons may have been
deposited either as utensils for personal use by the deceased, or as votives to
be presented as dedications (gift exchange) to gods encountered in the new
life.96 In other words, albeit not manifest, a dedicatory function for cauldrons
may have existed in Phrygia. This interpretation may also obtain for the
cauldrons deposited in tombs in Cyprus and Italy.97
As for Urartu, the cauldron with bull attachments from Toprakkale was
recovered in the ruins of the Haldi temple (Barnett 1950, 1 ff.), which locus
indicates a dedicatory function for cauldrons there. Secondly, at Kayalidere
a stone base facing the temple has three triangularly arranged hoof-shaped
depressions for a tripod (Burney 1966, 72, pl. 8b). Furthermore, we recall that
the inventory list of Sargon II’s booty from Musasir mentions cauldrons on
tripods among the dedications in the Haldi temple (van Loon 1966, 85).98
All the other excavated Urartian cauldrons derive from tombs, certainly
at Altintepe, and most probably also at the sites of Guschi and Alishar. In
Urartu, then, we have evidence of bull cauldrons placed both in sanctuaries
and in tombs.
96 Along with the bag of fibulae placed behind the deceased’s coffin, Muscarella 1967b, 1,
53, 51 f., n. 19. For fibulae as votives see Muscarella 1967a, 85 f., and Muscarella 1986, 199.
97 The fact that animal bones were found in one of the Vetulonia cauldrons does not, of
relief illustrating the sack of the Haldi temple (van Loon 1966, 44), but the fact remains that
cauldrons were placed in the temple complex.
758 chapter twenty-four
The examples from North Syria, Zincirli and Tell Rifa"at, and at "Ain Gev
in Israel, derive from occupational areas and furnish us with no specific
information with regard to secular or religious use. They are similar in this
respect to the bull cauldrons from the Gordion citadel.99
Distribution
99 Plain cauldrons are depicted being used in an apparently secular banquet scene in the
time of Sargon II (Herrmann 1966a, 3 f., n. 7, fig. 5), but it does not necessarily follow that
cauldrons with attachments—which may have given a charged value to the cauldrons—were
likewise used in the Near East or in Greece.
greek and oriental cauldron attachments: a review 759
made for and acquired both by people of rank and the gods.100 Because of the
lack of data it is not known whether a different social, religious, or political
status was accorded respectively to siren and bull attachments and if so,
whether this affected the cauldrons’ destinations (or Greek perceptions).
The social and economic organization of the North Syrian kingdoms
and their neighbors and royal relationships to workshops remain unknown,
although we may assume the workshops were not totally independent of
state control. But we do know that royal giftgiving had an ancient history
in the Near East, including North Syria (Winter 1988), and royal gift giving
signifies some state control of artisans and workshops (Winter 1988, 205;
Zaccagnini 1983).
North Syrian cauldrons—two with four sirens from Tumulus MM (query:
were four sirens of higher social and religious value than two sirens?) and
possibly the two with bull attachments from Tumulus W—arrived at Gor-
dion in the 8th century bc. While K. Sams (1979, 46) accepts the possibility
that a number of other North Syrian objects that reached Gordion at this
time were “the result of a kind of diplomatic protocol between Phrygia and
North Syrian heads of state,” he asserts that the siren cauldrons there “were
trade items.” I would posit, however, that given both their deposition in a
royal tomb and their high status value, it is tenable to conclude that the
movement of the cauldrons to Phrygia resulted from diplomacy, not trade.
They surely represented a gift, part of an exchange, from a North Syrian king,
and possibly are material reflections of an alliance, like that between Phry-
gia and Carchemish in 717bc, or of another (earlier?) unrecorded alliance (v.
Sams 1979, 46; Winter 1988, 209).
Relevant, albeit circumstantial, evidence supports this interpretation.
Tumulus MM also contained two extraordinary animal-headed situlae that
probably arrived there from either Assyria or North Syria (v. Muscarella,
forthcoming). Textual evidence exists in the Near East that animal-headed
vessels were a standard diplomatic gift in the 2nd millennium bc (Dunham
1989).101 While we lack first millennium textual data on this custom, it is not
improbable that it continued, and that the Gordion situlae were royal gifts to
the Phrygian king. On Assyrian reliefs of Sargon II situlae of exactly the same
form are depicted being dipped into cauldrons in order to obtain liquid for
100 See Coldstream 1983b, 201, who argues that Greek terracotta vessels found on Crete
and Cyprus were of “high prestige value” and were gifts from Greek noblemen (208). Kolaios’
dedication of a griffin cauldron to Hera is a good example of this feature.
101 Unfortunately, I did not know of these texts when I wrote Muscarella forthcoming—
102 Simpson 1986, 47. This view would support a North Syrian origin for the situlae (Mus-
carella forthcoming). It is also possible that the cauldrons and situlae arrived at Gordion
separately, as gifts from different kings, and were there used together.
103 The throne was most probably made of inlaid wood (Muscarella 1989, 334). Would
Midas have sent other wooden furniture to Delphi, like that from Tumulus MM, wood which
would no longer be preserved? To date no Phrygian material has been excavated at Delphi:
without the historical record we would not know of the Phrygian dedication. Could this gift
have played a role in the genesis of the rich-Midas legend?
104 For Phrygian material recovered in Greece see Muscarella 1967b, 59ff., and a revised
survey in 1989, 337 ff. In this latter work (339) I reported that two “unpublished” XII, 13 fibulae
with double pins were seen by me in the Olympia museum. They had in fact already been
published by Philipp 1981, 310f., pl. 69, nos. 1115, 1116. Philipp is correct that in Muscarella 1967b,
32, n. 35, I was referring to these two examples; when no. 1115 was first published in 1890 (as
no. 370) the double pin was not shown, and thus it was cited by me, ibid. 31, n. 31a, as a single-
pin example.
greek and oriental cauldron attachments: a review 761
be seen in the context of royal gift-giving and thus are formally interrelated.
Phrygia and North Syria maintained political and cultural contact (Sams
1974, 181ff.; idem 1979, 45f.). Now, while the Phrygians were knowledgeable
about Greek affairs (Muscarella 1989), we do not know what the North
Syrians knew of the Greek sphere. Here archaeologists lack information
on crucial questions: did the manufacturers or the commissioners of the
cauldrons know their destination (Greek sanctuaries), and did the Greeks
know their source? But if gift-giving was the mechanism in force, then we
would infer that the Greeks did know the source(s), and that some Syrians
were informed about Greece.
Thus to invoke the Phrygian model to explain the presence of oriental
cauldrons in Greece is to assume that the North Syrians knew something
of Greeks and their sanctuaries, either indirectly through contacts with
the Phrygian court, and/or through contacts, commercial or diplomatic,
with Greeks on the Mediterranean coast. Furthermore, it assumes that they
desired to communicate with Greek deities.
Although an assumption and not an inference, this conception is by no
means untenable, especially given the quantity of North Syrian material in
Greece (Muscarella 1970, 116ff.; Winter 1988, 196 ff., 207); and it further allows
us to perceive a connection between the two events mentioned above. If one
or more North Syrian kings gave cauldrons as gifts to the Phrygian court,
then he/they might also give gifts elsewhere; and if one oriental king gave a
gift to a Greek deity, then other kings may well have done the same, for the
same general reasons.
Inferences and assumptions do not necessarily add up to historical real-
ity, but I argue that the Phrygian model with its foundation in Herodotus
could legitimately be adduced to explain how perhaps some oriental caul-
drons reached Greece: as gift-dedications.105 It would be rash to go beyond
the “some,” because obviously other mechanisms may equally have been
involved.
105 Herodotus (I. 14) states that after Midas’ dedication at Delphi a gap of some 30–60 years
ensued before another oriental king, Gyges (died ca. 650 bc), dedicated an object. This claim
is not a problem—for Herodotus is of course speaking about Delphi, not other sanctuaries;
and even for that site he may have been reporting only about important kings remembered
hundreds of years after the fact by his local informers. Note that Herodotus’ account informs
us that at least in some cases the Greeks were aware of the specific oriental sources of
the imports in the 8th and 7th centuries: see also Herod. I. 45–55 for 6th century Lydian
“imports.”
762 chapter twenty-four
I. Winter (1988, 210f., 212) has argued that al Mina was an outlet for goods
from North Syria, and that the presence of Greeks there indicates that “North
Syrian goods were being distributed as the products of trade, as well as the
result of booty, tribute and giftgiving.” This view of the apparent mecha-
nisms allows for gift-giving (but with no specific examples indicated). With
her broad claims I have no quarrel in general, but they beg a number of
questions, questions that I doubt can be answered meaningfully at present:
which goods that reached the Greek sphere were gift, which trade items?
Were some cauldrons with attachments gifted, others traded, and if so, was
trade part of their manufacturing strategy? Who within North Syria (and
elsewhere) sold and profited from the sale, the court or private merchants, or
both? There are other questions: where were these cauldrons sold, in North
Syria, on the coast, or in Greece? To whom were they sold, to other Near East-
erners and/or to Greeks? We may also ask why are the cauldrons not found
in private Greek tombs? And what did the Greeks, if they purchased caul-
drons, give as payment?106 Not knowing the answers to all these questions
precludes definitive statements about the movement process.
Moreover, to assert that either Greek or Phoenician ships transported
the cauldrons west tells us nothing about how and why they got to these
ships, whether for trade or giftgiving, nor who owned them at the time of
embarkation. This is why it is insufficient to suggest that Greeks brought the
cauldrons west (Herrmann 1966a, 149; Cold stream 1977, 360; Winter 1988,
211) or that Near Easterners carried them (Dunbabin 1957, 35 ff., 49; Muhly
1970, 48f.) without confronting the questions raised here. Winter (1988, 212)
aptly cites a reference to a 7th century bc merchant from Carchemish, but
the reference does not inform us of his responsibilities or of his wares.
Trade, then, while it may be posited as a possible process of distribution,
is not readily demonstrated, even if we follow the discussions in scores of
books and articles on trade patterns and models. Thus, while I have found it
possible to ascertain a Phrygian-Midas model for the movement of certain
high-class objects, I have not been able to comprehend a trade model—
which failure, however, does not exclude trade from being considered as a
possible process.107
106 Coldstream 1982, 265 suggests that Greek silver was used as payment.
107 That a successful Greek merchant, Kolaios, could dedicate a cauldron with griffin
protomes at a Greek sanctuary (Hera on Samos) in the 7th century bc is indicated by
Herodotus (IV.152). But this act conforms to an old Greek custom: a Greek commissions and
then dedicates a Greek-made cauldron to a Greek deity. Thus I believe that this knowledge
cannot reasonably allow archaeologists to invoke a “Kolaios model”—which would obtain
greek and oriental cauldron attachments: a review 763
for Greek dedications alone—to justify the conclusion that oriental cauldrons were also
commissioned/purchased by Greeks in the 8th or 7th centuries bc for dedication at Greek
sanctuaries.
Nor ought we be quick to accept the “Denkmodell” of H. Kyrieleis in Kyrieleis and Röllig
1988, 37–61, with a philological discussion by W. Rollig (62–75). Here Kyrieleis publishes a
superbly decorated North Syrian bronze horse plaque found in a 6th century context on
Samos in 1984. It is inscribed in Aramaic with the name of Haza"el and is thus dated to
the late 9th century bc (49; Röllig, 62 ff., 71 ff.). Reference is also made to one of two bronze
North Syrian blinkers excavated around 1900 at Eretria on Euboea (Muscarella 1970, 116,
fig. 10) that after cleaning revealed the same inscription as on the Samos piece discussed
(49, 55 f.; Röllig, 69 ff.). A. Charbonnet earlier (1986, 117, 140ff., fig. 3) published a different
interpretation of the Eretrian inscription and accorded it an 8th century date (see below).
Kyrieleis presents two crucial assertions: the two objects arrived on Greek soil in the 9th
century bc as a result of trade, or better, as gift-exchange transactions between enterprising
Greek merchants and a local ruler; and the objects were later dedicated at the sanctuaries
by Greeks, not by Near Easterners (56 ff., 59, 61). Furthermore, as a consequence of these
assertions he postulates what are surely major historical conclusions concerning early Greek-
Near Eastern contacts and the date of the creation of the Greek alphabet (58ff.). But we are
dealing here with intuitive guesses; we are not able to accept it as archaeological explanation
for events that no archaeologist is capable of documenting. We do not know when the objects
reached Eretria and Samos (8th century bc?), or the political and cultural background of the
donors. And I strongly argue that the two objects under review may not, cannot, be presented
legitimately as evidence for early, 9th century bc, oriental writing on Greek soil. Note that
Charbonnet 1986, 117, 143, 145, 157, believes that the Eretrian plaque arrived at the site in
the 8th century bc and that it was dedicated by an Aramaean who inscribed the plaque on
the spot: also guesswork, in part based on a mis-translation of the text. It is also relevant to
recall here that in his discussion of the Hundehalter figurines that were excavated on Samos,
Kyrieleis 1979, 44 f., 46, 48, argued that surely Near Easterners, not Greeks, dedicated these
oriental objects to Hera. Note that both Strøm and Röllig, infra, 48, 97, accept an 8th or 7th
century date for the deposition of the two pieces.
108 But the alleged “foundation deposit” cited from a tomb at Khaniale Tekke (Board-
man 1970, 15; idem 1980, 57; Coldstream 1977, 100; idem 1982, 267) does not clearly reflect a
764 chapter twenty-four
recently J. Shaw (1989) has published strong evidence for the presence of
Phoenicians on Crete, an eastern sanctuary at Kommos.
None of these scholars posits immigrant craftsmen to explain the pres-
ence of oriental cauldrons in Greece, and, in fact, Coldstream and Boardman
accept the cauldrons as imports. One might be tempted to suggest that those
bull attachments that scholars find difficult to call either Greek or orien-
tal are hints of Orientals working in Greecereflecting their adjustment. But
this view ignores the reality that the problems here may be an artifact of
modern perceptions and analyses. There is ample evidence that in the Near
East throughout its history artisans moved either voluntarily or under orders
(Zaccagnini 1983), but it would be infelicitous to suggest that oriental arti-
sans came to Greece to manufacture cauldrons for dedication.
Near Eastern custom. Frankenstein 1979, 274 ff., argues that the mechanism of movement of
Phoenician goods west was trade, not immigrant craftsmen.
Section Two
Artifacts
chapter twenty-five
“Notes on Some Distinctive Types of Bronzes from Populonia, Etruria,” Proc. Prehist. Soc., 1956,
p. 131.
2 P.E. Botta, E. Flandin, Monument de Ninive, II (Paris, 1849), Pls. 103, 106 bis. Some
females represented on objects from Luristan wear what appears to be long straight pins
at the shoulders, E. Porada, The Art of Ancient Iran (New York, 1965), p. 89, and Fig. 60.
These representations are very similar to the straight pins worn by some Greek ladies on
the François vase, H.L. Lorimer, Homer and the Monuments (London, 1950), p. 379, Fig. 54.
Note that it is possible the goddess with mirror represented on the Hasanlu gold bowl wears
straight pins on her shoulders, Porada, op. cit., p. 101.
3 The horned catch is one of the most characteristic features of Phrygian fibulae, see
Oscar White Muscarella, Phrygian Fibulae from Gordion, 1965, an unpublished dissertation
in the University of Pennsylvania Library (hereafter Phrygian Fibulae).
4 Fibules Grecques et Orientales (Copenhagen, 1926), pp. 213f. (hereafter Fibules); Phrygian
would seem, that they all belong to one national group. Yet the garment
of only one of the bearers is shown open its whole length, in front view,
the two borders clasped by a fibula. In order to show the manner in which
the fibula is worn, the sculptor has depicted this man with arms raised
supporting a sack on his shoulders; the other gift bearers have their arms
before them. Thus it would appear that there has been a deliberate attempt
to call attention to the fibula, and I suggest that the sculptor wished to
indicate that Phrygians were passing by.6 We gain here an insight into the
practice of an ancient sculptor, or his supervisor, who wished to identify a
specific group of people, the Phrygians, possibly for a non-reading public. He
simply chose to illustrate one of their most characteristic products, a fibula.
If this interpretation is correct, I suggest that the procession represents
gift bearers from King Midas in the year 710/709bc, the year Phrygia was
attacked from Cilicia by Sargon’s agents.7 The frieze incidentally proves that
the Type XII,7 fibula was used in the late eighth century bc.
An exact parallel for the use of a fibula as an indicator of nationality is
provided at Persepolis by the Apadana reliefs and by the dais of Artaxerxes I
in the Throne Hall. On the Apadana walls many groups of allies and sub-
jects of the Persian king are represented bringing gifts, perhaps for a new
year’s festival. Only one of these groups is depicted wearing fibulae (Pl. III,
Figs. 2, 3), and only two of the fourteen men supporting Artaxerxes dais wear
fibulae.8 There can be little doubt that the sculptors were aware that certain
peoples characteristically wore fibulae and so represented them. Therefore,
we may assume that the ancient passerby, unlike the modern scholar, eas-
ily recognized the nationality of the fibula wearers. Barnett claimed that the
fibula wearers were Phrygians; and Bittel supporting E. Schmidt, maintained
that they were from Central Anatolia, in particular from Alisar, even though
he could find no clearly similar examples there. One type of fibula repre-
sented on the Apadana reliefs, has a ribbed arc, a spring and a hand catch
(Pl. IV, Fig. 3); it is apparently of the same type as, or similar to, those on
6 R.D. Barnett, “Early Greek and Oriental Ivories,” JHS, LXVIII (1948), 9.
7 D.D. Luckenbill, Ancient Records of Assyria and Babylonia, II (Chicago, 1927), §41, 43.
8 E. Schmidt, Persepolis, II (Chicago, 1953), Pls. 35 A, B, 108, 109, 112. See also R.D. Barnett,
“Persepolis,” Iraq, XIX, 1 (1957), 67; K. Bittel, “Fibeln in Perspolis,” Vorderasiatische Archäologie,
Ed. by K. Bittel (Berlin, 1964), pp. 39 ff., Pl. 9. It is interesting to note that the arcs of the
Persepolis fibulae are worn both upright and pendent. This demonstrates that there was a
choice. Some fibulae would look better with the arc upright, viz. Iraq, XXII (1960). Pl. XXX,
p. 3, a gold fibula from Ziwiyeh; and most of the fibulae from Boeotia, in Greece, Fibules,
pp. 187 ff.
fibulae represented on sculpture 769
earlier North Syrian reliefs (infra). Another type of fibula, represented both
on the dais and on the Apadana reliefs, has a plain arc and what seems to
be a hinge rather than a spring (Pl. III, Fig. 2). Fibulae of both types, those
with hand catch (Near Eastern), and those with hinges, were in common use
in Western Iran and in the South Caucasus.9 The hand catch and hinge are
crucial in this discussion, for neither was employed extensively in Anatolia
although some imported pieces have been found.10 Therefore I suggest that
the fibulae wearers should tentatively be identified as an Iranian or perhaps
a South Caucasian people. The dress of the group does not contradict this
opinion, for they are dressed like the other groups on reliefs at Persepolis,
one of which Schmidt has identified as Medes.
We have presented two examples of fibulae employed by sculptors to des-
ignate particular national groups, that at Khorsabad and those at Persepolis.
The well-known and often illustrated relief at Ivriz presents us with a differ-
ent situation in which the fibula is not used as an indication of nationality.
On this relief, King Urpallu of Tabal (fl. 738 bc) stands before a god.11 Urpallu
wears elaborately decorated clothing. The outer garment is secured by a
Phrygian fibula of a variety similar to examples found at Gordion and Ankara
in Phrygian contexts.12 This particular variety has different arc forms and
is characterized by a double pin masked by a rectangular transverse plate,
rather than the usual single exposed pin. The plate is detachable and serves
to shield the pins; two tubes sheath the points (Pl. IV, Fig. 4). Fibulae with
9 Pace Bittel, op. cit., pp. 40, 41, n. 16. See n. 10. Also in 1964 at Hasanlu a Pd. II (fifth-fourth
century bc) cist tomb yielded a hinge type and a spring fibula found on the same skeleton
(unpublished).
10 Oscar White Muscarella, “A Fibula from Hasanlu,” AJA, LXIX 2 (1965), 234, nn. 11, 12;
p. 235, n. 20. Add Alishar, OIP, XIX, 266, Fig. 353; XXIX, 439, Fig. 493; XXX, 111, Fig. 106, 2041;
180, Fig. 201, d25, e776. A few imported Near Eastern fibulae have been found in the later
levels above the destroyed Phrygian city at Gordion, Phrygian Fibulae, Appendix D, pp. 220f.
In 1964 some Near Eastern and hinged types of fibular were found at Hasanlu in late seventh
century and fifth and fourth century contexts (Periods III and II), Oscar White Muscarella,
“Hasanlu 1964,” BMMA, Nov. 1966, p. 135, Figs. 37, 38. It should be noted here that the fibula
on the Khorsabad relief discussed in the text above has nothing to do with the fibulae on the
Persepolis reliefs. As I hope the texts makes clear, one is Phrygian, the others “Near Eastern,”
see nn. 3 and 4 supra. Recently W. Culican in The Medes and Persians (New York, 1965), p. 99,
erroneously calls the Persepolis fibulae Phrygian, perhaps following Barnett Iraq, 1957, p. 67.
11 The most recent and best illustration of the relief is to be seen in E. Akurgal, The Art of
10. Only a few of the Gordion examples have been published, viz. AJA LXII, 2 (1958), Pl. 25,
Fig. 20; Expedition, VI, 2 (1964), lower left figure on p. 37. See also Phrygian Fibulae, p. 7, Fig. 1;
22; 33, Figs. 8, 9; 63, Fig. 16; 75; 127.
770 chapter twenty-five
a fixed transverse plate, which has undoubtedly been copied and modified
from the Phrygian examples, have also been found.13 The fibula represented
on the Ivriz relief has an arc of XII,9 type, according to Blinkenberg’s classi-
fication. Although no examples of this particular type have yet been found
with double pins and detachable transverse plates, many with single pins
do occur in Phrygian contexts of the late eighth and seventh centuries bc at
Gordion and elsewhere14 (Pl. IV, Fig. 5).
Since Urpallu was a king of Tabal, an eastern neighbor of Phrygia, we
cannot call Urpallu a Phrygian simply because he wears a Phrygian fibula.15
We may assume in this case that the fibula was an import to Tabal worn by
the king as a piece of jewelry and so depicted on the relief by the sculptor.
Another representation of King Urpallu exists on a relief found in frag-
ments at Bor, northeast of Ivriz, and now in the Istanbul Museum (No. 837).
As exhibited at present, the relief has been restored and a Phrygian fibula is
worn over the chest. The reason for adding a fibula to the restored area was
recently explained to me in a letter from Bay Necati Dolunay, Director of the
Archeological Museums of Istanbul: the authorities decided to use the Ivriz
relief as a model for restoration and, since Urpallu wears a fibula there, it
was decided to copy it on the Bor relief. Whether the latter originally had
a fibula is not known. Through the kindness of Bay Dolunay, I illustrate the
relief both before and after restoration (Pl. V, Fig. 6).
Several examples of fibulae on reliefs have also been found in North
Syrian cities, not far from Ivriz: one fibula from Zincirli on a relief of a queen,
one on a female who is part of a royal couple from Marash, and two others on
a stone fragment of clothing from Carchemish (Pls. VI–VIII, Figs. 7–9).16 All
these fibulae are of the same type, semicircular with a ribbed arc; they are
found in many parts of the Near East including Iran, from the eighth century
and later.17 They were commonly worn in daily life. When represented on the
13 Fibules, pp. 217 f., Type XII, 10; U. Jantzen, “Phrygische Fibeln,” Festschrift für Friedrich
Matz (Berlin, 1962), pp. 39 ff., Pl. 8, Nos. 1, 2, Pl. 11, Nos. 1–3; D.G. Hogarth, et al., Excavations
at Ephesus (London, 1908), pp. 151 f., called a handle in the publication; R.M. Dawkins, et al.,
The Sanctuary of Artemis Orthia at Sparta (London, 1929), Pl. 83 c; Phrygian Fibulae, pp. 48ff.;
Alisar, “OIP,” VII, 96, Fig. 76, 2855.
14 See n. 5. Also J.M. Birmingham, “The Overland Route Across Anatolia,” Anatolian Studies
XI (1961), 186 f.
15 Cf. Barnett, op. cit., pp. 8 f.; Phrygian Fibulae, pp. 145ff., nn. 8 and 9.
16 F. von Luschan, Ausgrabungen in Sendscherli, IV (Berlin, 1911), Fg. 236, Pl. 54; E. Akurgal,
Späthethitische Bildkunst (Ankara, 1949), Pl. XL; C.L. Woolley, and R.D. Barnett, Carchemish,
III (London, 1952), 240, n. 2, Pl. B64C.
17 As D. Stronach already pointed out, “The Development of the Fibula in the Near East,”
Iraq, XXI, 2 (1959), 189; we add Marash to his list. The relief fibulae under discussion have
fibulae represented on sculpture 771
reliefs under discussion, they have no significance beyond the fact that the
sculptor was representing accurately the daily dress of his subjects. Their
value for the archeologist is chronological for they are well dated artifacts of
the late eighth and seventh centuries bc.
Fibulae are often represented on other groups of sculptures from the
Near East, and in these cases with interesting implications. These sculp-
tures are exorcizing plaques or amulets used in rituals to ward off sickness.
The amulets are decorated with scenes in relief; one of these scenes has a
demon as a central figure, and various objects, including a fibula, are often
represented in the neighboring field (Pl. VIII, Fig. 10).18 These objects were
offerings, votive gifts, for the demon represented on the amulet.
Classical archeologists are familiar with the fact that fibulae played a
role as votive gifts, as offerings to a deity, in sanctuaries of Greece and Asia
Minor.19 Exactly what this role was still remains obscure. Little light is shed
on the subject by Herodotus’ etiological story in Book V, 87, 88. He describes
how the Athenian women murdered the sole survivor of the war with Aegina
by means of their peronai, and subsequently dedicated their peronai in
sanctuaries at Argos and on Aegina. It is not clear in these passages whether
Herodotus is writing about straight pins or fibulae. However, since he refers
to the objects being dedicated at temples, and since almost every Greek
sanctuary of the first millennium bc has yielded large numbers of fibulae,
I suggest that he is referring to fibulae rather than to straight pins. It is
also possible that Herodotus has joined two seperate stories, one concerned
with the murder, where straight pins may have been the weapon, the other
concerned with votive offerings, where fibulae were used.
incorrectly been called Asia Minor (i.e. Phrygian) types: Fibules, p. 29; Lorimer, op. cit., p. 354;
E. Akurgal, “Chronologie der Phrygischen Kunst,” Anatolia, IV (1959), 116.
18 J.B. Pritchard, The Ancient Near East in Pictures (Princeton, 1954), Figs. 658, 660; H.H. von
der Osten, “A Babylonian Amulet,” BMMA, XIX (1924), 145 ff.; “Zwei Neue Labartu-Amulette,”
AfO, IV (1927), 89 ff.; F. Thureau-Dangin, “Ritual et Amulettes contre Labartu,” RA, XVIII
(1921), 172 ff., Pl. 1. For a more extensive bibliography on these amulets see H. Klengel, “Neue
Lamaštu- Amulette aus dem Vorderasiatischen Museum zu Berlin u. dem Br. Museum,” MIO,
VII (1959), 334 ff.; for fibulae, pp. 338 f., Fig. 1a, b; pp. 344 ff., Fig. 6a, b, and n. 29. See also by
the same author “Weitere Amulette gegen Lamaštu,” MIO, VIII (1960), 24ff., for fibulae p. 27,
Fig. a, b. Klengel alone seems to be aware of the fibulae representations but comes to no
conclusions about them.
19 Fibules, pp. 19 ff., 35, 106, 182, 193 f.; R. Hampe, Frühe Griechische Sagenbilder (Athens,
1935), pp. 3. 90 ff.; Lorimer, op. cit., pp. 249 f.; A.M. Tallgren, “Caucasian Monuments,” ESA, V
(1930), 178 f. A fine archeological example of fibulae used as votive gifts may be seen in the
gold and silver sheetmetal types found in the basis at Ephesus, Hogarth, op. cit., p. 98. Pl. V, 6,
117, Pl. XI, 12. These objects could not actually be used to fasten clothing and must have been
dedicated as fibula simulacra.
772 chapter twenty-five
20 Besides the examples discussed in this study see Fibules, pp. 28ff. Some more recent
seventh centuries bc, C. and A. Koerte, “Gordion, Ergebnisse der Ausgrabungen in Jahre 1900,”
JdI, Supplement No. 5 (Berlin, 1904), pp. 45 ff., 76 ff., 101 ff.; AJA, LXI, 4 (1957), 327; AJA, LXVII, 2
(1958), 149; AJA, LXIV, 3 (1960), 227 ff.; Hampe, op. cit., p. 4; Notz. degli Scavi, 1880, p. 77; 1895,
p. 167; E. Gjerstad, Early Rome, II (Lund, 1956), 89. There is no strong evidence forthcoming
from Near Eastern graves to demonstrate that fibulae were placed there in large numbers.
22 “Death and Netherworld According to the Sumerian Literary Texts,” Iraq, XXII (1960),
59 ff. These texts deserve special study by anthropologists and archeologists who are inter-
osted in why objects were placed in graves.
23 Thucy. iii, 584; Herodotus v, 92; Euripides Orestes, 115ff., 1453ff.; Sophocles Elektra, 431ff.;
Plutarch Aristides. xxi; Tacitus Annals iii. 2; Vergil Aeneid xi. 195f.
fibulae represented on sculpture 773
The exorcising amulets, which were the starting point of this discussion
on fibulae as votive objects, document the fact that fibulae did have religious
value connected with at least one deity in the Near East during the seventh
century bc. And quite relevant in this context are some interesting fibulae
found in Palestine and Iran, probably of the late eighth and seventh cen-
turies bc.24 These fibulae have a demon’s head (a Pazuzu?) at one end of the
arc and a female (a goddess?) or bird at the other end. Note that the relation-
ship of the demon with a fibula also occurs on the amulets. Is it not possible
that these fibulae have a votive or at least an apotropaio, function? I suggest
therefore that we may conclude on the basis of the amulets and these fibulae
that in the Near East as well as in the Greek world fibulae had some special
religious function. Whether the very same value was applied in both areas
cannot be known at present (especially with respect to fibulae in graves),
and more research will have to be done before one may state in which area
the idea originated. Nevertheless, it is suggested that the evidence presented
here has brought forth information about a hitherto unrecognized relation-
ship between the Greek world and the Near East in the early first millennium
bc.
24 R. Ghirshman, “La Fibula en Iran,” Iranica Antiqua, IV, 2 (1964), Pl. XXV, 13–15, also
Pl. XXIV, 11; R.S. Lamon, and G.M. Shipton, Megiddo, I (Chicago, 1939), Pl. 71, No. 72. I owe
the Megiddo reference to Professor Edith Porada. These Iranian fibulae were not known to
me when I wrote “A Fibula from Iran”; see n. 10.
774 chapter twenty-five
Fig. 6. Relief from Bor, Before and After Restoration. Istanbul No. 837.
fibulae represented on sculpture 779
PHRYGIAN OR LYDIAN?*
* This article originally appeared as “Phrygian or Lydian?” Journal of Near Eastern Studies
l. of lock-plate 5.2 cm., w. 1.4 cm., diam. of hemispheres on lock-plate ca. .2cm., l. of chain 4cm.
Karat tests show a variation: the obverse of the lock-plate: 18K, one of the curved pieces in
the catch: 18K, the tube/tongue of the double-pin mechanism: 16K, one of the hollow balls
at the end of the chain: 20+ K. I wish to thank the owner of the fibula for lending it to the
Metropolitan Museum of Art and Miss Kate Lefferts and Mr. Ed Rowe of the Conservation
Department of the Metropolitan Museum who, among others, generously gave much of their
time in discussing the fibula.
Minute fragments of black silk were seen by me under a microscope: on the point of one
of the scrolls on the lock plate and on the underside of one of the two prongs of the double-
pin mechanism. Mr. William Young of the Boston Museum of Fine Arts also found such a
fragment on the catch; he identified it as raw silk. I am most grateful to him for informing me
of his discovery.
Because it is fair to discuss openly and objectively all information one has about an
unexcavated object I must state here that some people think the fibula is a forgery, at
least in part. To be sure, when I first examined it I was puzzled by certain aspects of its
construction, its differences from the bronze examples known to me, the presence of a tassel,
and the decoration on the lock-plate. I discuss these features in the text. Obviously, I do not
think the object is a forgery but, if it develops that it is modern, the problems concerned
with the purchase of works of ancient art are more complex and frightening than even
now recognized. I would hope, in any event, that the comments made in this paper about
artistic continuity and about Phrygian and Lydian relationships are valid, independent of
the publication of the gold fibula.
2 C. Blinkenberg, Fibules Grecques et Orientales (Copenhagen, 1926—hereafter cited as
784 chapter twenty-six
Fibules), pp. 204 ff., 222 ff., Fig. 255; Oscar White Muscarella, Phrygian Fibulae from Gordian
(London, 1967—hereafter cited as Phrygian Fibulae), pp. 24f., Figs. 63–74.
phrygian or lydian? 785
three disks; these grommets are smoothly made apparently in a form. Each
of the five edges of the moldings is masked by a wire; in three cases the wire
is vertically grooved throughout, in two cases it is horizontally grooved. A
plain wire collar around the arc exists just above the moldings where solder
might have been visible. Therefore, each end molding consists of thirteen
parts. The center molding apparently consists of two grommets and two
disks, masked by three wires. In addition, a wire collar around the arc exists
at each side, making a total of nine parts. The quarter moldings are made
from two half grommets, one wire mask, and two collars at the ends, for a
total of five parts for each quarter. All the wires and collars show seams. A
red coloration is visible at all the areas where soldering occurred. The total
number of parts in the arc and moldings consists of at least fifty-five separate
parts.
The double-pin mechanism consists of two pins each soldered to a solid
vertical rectangular bar. This bar is itself soldered to a pierced horizontal
solid bar placed perpendicular to it which in turn is soldered to a spool
decorated in two zones with grooves. The spool seems to have been made
in one piece, but two thin semi-circular strips were placed over the top and
bottom of the spool (barely visible in Figs. 3, 5). The horizontal pierced bar
is soldered to the spool only at the ends, leaving the central area clear, and
leaving two shallow grooves for passage of the prongs on the lock-plate. In
addition, a very small piece of wire was placed as a reinforcement at both
corner junctures where the bar joins the spool (one of these is now missing).
The vertical bar holding the pins is also soldered at one end to the spring.
This spring is not functional and seems to have been made by twisting a strip
of wire decorated with incisions into the desired shape. Eighty-four (one is
missing) perfectly formed very small granulations were added to the edge
of the spring. Above the spring is a gadrooned disk, or spring-plate, perhaps
made in two pieces, with an additional grooved wire masking its edge. A thin
grooved strip of wire masks the join of the spring to the spring-plate (visible
in Figs. 1, 3), but we may assume that there was also solder at the join. On
the upper part of the spring-plate a short hollow tube that functioned as a
tongue was soldered; it has a lap-join that was soldered.
A braided wire chain consisting of many interconnected loops of wire was
loosely attached to the hole in the horizontal bar connected with the spool
by means of a wire loop (Figs. 2, 5). This hole has on one side, the outer
visible one, a thin wire collar. At the upper end of the chain is a free-moving
spool or grooved collar, and at a point where the chain divides into four
units is another more elaborate collar, also free-moving. This latter collar
is made up of two pieces and has wire covering two of its edges. Finally,
786 chapter twenty-six
Plate I
four hollow spheres cover the ends of the four braided wires; although
their collars appear to have been added, I could see no seams. Thus, we
have a total of at least thirty separate parts plus the many chain loops,
plus eighty-four granulations used in the construction of the double-pin
mechanism.
The last unit to be described, the lock-plate, has a total of sixty-seven parts
in its construction. The plate itself is formed from a strip bent up on three
phrygian or lydian? 787
Plate II
sides and bent further at the right partially to inclose a fourth side. One
end of the plate is forked to fit behind the spool soldered to the spring; the
thin tines each had a strip of wire soldered to the rear for reinforcement
(one is now missing) (just visible in Fig. 2). A U-shaped unit, higher at one
end, was soldered to this plate and it alone consists of six pieces (Figs. 3,
4): the U-shape itself, two wires forming an arched recess that decorates
the outer face of the U-shaped recess, two hollow hemispheres each framed
with a collar placed in this outer arched recess, and seven thin strips added
at several places as reinforcements. And finally, there are the two tubes,
rolled and soldered, within the recess to hold the double pins. Although the
tubes are eight mm. in length, the pin ends penetrate only two mm. The
obverse part of the lock-plate is decorated on the outer edge with twenty-
eight hollow hemispheres that slightly vary in size and wall thickness; on
several of them may be seen a minute hole visible through a microscope.
There is also a zone, formed by one piece of wire bent to make a rectangle,
788 chapter twenty-six
Parallels for the fibula are known from sites in Anatolia and elsewhere.
The main features that characterize the fibula, aside from the precious
material and decorative moldings, are the removable bar or lock-plate and
the nonfunctional spring and double-pin mechanism that is shielded by the
latter. Two types of fibulae that make use of a lock-plate are presently known:
those with a fixed, permanent bar that joins the two ends of the arc; and
those with a removable decorated bar, constructed in the very same manner
as the one found on our gold fibula (Figs. 6, 7). In both types double-pins
rather than the far more common single pin were customary, and the nature
or shape of the fibula’s arc and decoration was varied.3
Fibulae of the first type have been excavated at Olympia, the Argive
Heraeum, and Sparta in Greece; at Ine, Alishar, and Ephesus in Anatolia;
and on Samos. Some of these fibulae appear to be Phrygian, others are
apparently copies and not Phrygian, and they all date to the late eighth and
seventh centuries.4
Fibulae of the second type are known from the Phrygian sites of Gor-
dion (Figs. 6, 7) and Ankara in contexts dating them to the late eighth cen-
tury bc; the eighth century fibula carved on the well-known Ivriz rock relief
depicting King Urpallu may also be of this type.5 Other isolated examples
were excavated at Olympia and Samos, and at least one is in the Ankara
Museum.6
3 I summarize here information available in more detail in Phrygian Fibulae, pp. 15, 17f.,
pp. 217 f., No. XII, 10e from Delphi may be a buckle.
5 Phrygian Fibulae, pp. 15, 17, 19 f., 22, 24, 26, 30, n. 11, 33, n. 51, 39, Figs. 8, 9, 16–21, 47, 48,
63, 65.
6 Ibid., pp. 17, 22, 30, n. 14 and 21, 32, n. 35. The examples from Olympia and Ankara do
Plate III
It is the latter type with which we are concerned, for our fibula without
doubt belongs in this category. The detachable lock-plates and other details
of this type of fibula, particularly those from Gordion, Ankara, and Samos,
are constructed in the same manner as noted on the gold fibula: two tubes
to hold the double-pins, a recessed setback to inclose the catch, a catch that
790 chapter twenty-six
fits under the upper tube leaving the lower one free, a nonfunctional spring
and a decorated spool set below the spring that fits into a forked cut-out on
the plate.
The gold fibula has crisp, sharp moldings and catch, parallels for which
are easily found in eighth century Gordion and elsewhere.7 It was character-
istic of the eighth and early seventh century bronze fibula types with extra
moldings on the arc (Types XII, 13, 13a, 14, and 14a) to exhibit finer workman-
ship than was customary for the later examples. However, several fibulae
with the same neat sharp workmanship do occur on the City Mound at Gor-
dion in the sixth and later centuries.8
On the basis, therefore, of parallels in typology, and because of the fine
workmanship of the arc, moldings, and catch, it would be possible to con-
clude that the gold fibula was made in Phrygia in the eighth century or the
first decade of the seventh century bc. Nevertheless, certain questions that
arise from a style analysis seem to suggest that such a conclusion cannot be
the only viable one.
One of the questions concerns the gold twisted wire tassel. No fibulae
of a Phrygian type known to me have a tassel of any sort. It is indeed
possible that tassels of a perishable material were used and have not been
recovered or recognized in excavation, but, of course, this is not verifiable.9
We may conclude only that there is at present no evidence that tassels were
characteristic of eighth century fibulae.
Twisted gold wire chains like the example on our fibula seem to have
been made over a relatively long period of time. One finds good parallels
in Iran, Anatolia, and the West dating to the second and first millennia bc
and continuing down into the Hellenistic and Roman periods;10 therefore,
the chain by itself cannot help us chronologically.
7 Ibid., pp. 21 ff., Figs 47, 48, 63–65; see also Fibules, pp. 220ff., especially Figs. 249, 252,
255.
8 Phrygian Fibulae, pp. 23, 25. See also my discussion of the finds from Tumulus S at
1
Gordion on pp. 4 f., 10, n. 22 and 23, and Figs. 49, 50, 52, 53, 67–71.
9 In 1968, right after I first saw our gold fibula, I examined two bronze fibulae from
Tumulus MM at Gordion that were travelling with “The Art Treasures of Turkey” Exhibition.
One was a Type XII, 7, the other a Type XII, 14 example and both had double pins and lock-
plates. In each example I noticed a neat hole directly behind the spool below the spring:
these holes could have held perishable tassels that are now missing; see the catalogue for the
exhibition, Smithsonian Institution, Washington D.C., 1966, Fig. 102.
10 E. Negahban, A Preliminary Report on the Marlik Excavation (Tehran, 1964), p. 57, Fig. 127;
R. Ghirshman, The Art of Ancient Iran (New York, 1964), p. 114, Fig. 151; F.H. Marshall, Cat-
alogue of Jewellery in the British Museum (London, 1911), Pls. IV (550), LII (2442–2443), LIV
(2606, 2581), LVIII (2720), LXIX (3024); P. Jacobsthal, Greek Pins (Oxford, 1956), Nos. 643, 647;
phrygian or lydian? 791
Another question concerns the wire scroll motif. I do not know of any
such motif, or related one, that may be presented as occurring on the lock-
plate of any excavated fibulae.11 Those lock-plates known to me are deco-
rated with cut-outs or lattice patterns, or with studs that pierce the plate.12
The lock-plate on the fibula shown in Fig. 7 is decorated with a separate cir-
cular diaper pattern apparently soldered into place; and, indeed, if solder
was employed here, this feature would be a parallel in technique, if not in
motif, to the scroll decoration on our fibula. Scroll patterns on terracotta
architectural plaques are well known in seventh or sixth century contexts
in Anatolia. They occur at Gordion on the City Mound in post-destruction
levels, and at Sardis;13 a common source is implied. It is possible that these
post-eighth-century scroll patterns might supply us with a clue that our
fibula is post-eighth century in date.
Hollow-headed hemispheres are soldered around the scrolls on our
fibula, but there are no rivets that pierce the plate as was common on eighth
century fibulae. A good late seventh or sixth century example of the use of
hollow gold hemispheres, apparently soldered into position on a gold plate
background, is the jewelry found at Aydin in western Anatolia, just south of
Sardis.14 Again we have a clue that we are dealing with a post-eighth century
object.
Finally, there is the construction of the fibula itself, the fact that the arc
is a hollow tube, and that the moldings are constructed of several units
and added individually and soldered in position. Again, no parallel for this
technique is forthcoming. All the many hundreds of bronze fibulae from
Anatolia are cast solid together with the moldings. One might, of course,
H. Hoffman, P. Davidson, Greek Gold (Brooklyn Museum, 1965), pp. 36f., Fig. 1; O.M. Dalton,
The Treasure of the Oxus (London, 1964), Pl. XXI, Nos. 151, 152, p. 40; P. Amandry, “Orfèverie
achéménide,” Antike Kunst, I (1958), 14 f., n. 47, Pl. 12, No. 32; P. Demargne, The Birth of Greek
Art (New York, 1964), Fig. 493.
11 See Negahban, op. cit., p. 47, and Figs. 65, 67 for similar flat wire decoration. Related
wire decoration may also be seen on some earrings recently excavated in the Athenian Agora,
E.L. Smithson, “The Tomb of a Rich Athenian Lady,” Hesperia, 37, 1 (1968), Pl. 31, Fig. 77b, Pl. 32,
Figs. 77a, b. Both examples cited here are early first millennium bc in date.
12 Phrygian Fibulae, pp. 17, 19, Figs. 19–22, 47, 48, 63; Fibules, p. 216, Figs. 243, 244.
13 G. and A. Koerte, “Gordion, Ergebnisse der Ausgrabungen im Jahre 1900,” JdI, Supple-
ment No. 5 (Berlin, 1905) (hereafter cited as Gordion), pp. 155f., Figs. 138, 139, p. 164, Fig. 148;
R.S. Young, “Gordion–1950,” Univ. Mus. Bull., 16, 1 (1951), p. 6, Pl. III; “Progress at Gordion,” Univ.
Mus. Bull., 17, 4 (1953), p. 15, Fig. 8; T.L. Shear, Sardis, X (Cambridge, 1926), Pl. XI; H.T. Bossert,
Altanatolien (Berlin, 1942), No. 183. The pattern may be Greek in origin.
14 A. Dumont, “Notes sur des bijoux d’ or trouvés en Lydie,” BCH, 3 (1879), pp. 129f., Pls. IV,
suppose that it was economy that suggested using a hollow arc because gold
was expensive. It can only be stated, however, that the few gold fibulae found
to date at Gordion (Fig. 8), all from the eighth century, are types with plain
arcs and all are cast solid.15 One site in Asia Minor that has yielded gold or
electrum fibulae is Ephesus, where about seven examples were found in the
Basis and are ca. 600 bc in date.16 One of these fibulae (Hogarth’s Pl. V, No. 3),
a XII, 14 type, invites our attention as it seems (from the photograph) to
have a hollow arc; the other examples illustrated all seem to have solid arcs.
Whether this fibula actually has a hollow arc is not known to me, as the
text gives no information on the matter; I also cannot discover the present
location of the fibula. Certainly, if the arc is hollow, it would be an important
parallel for our Type XII, 14 fibula.
Collecting all the above observations together and attempting to inter-
pret them culturally and chronologically suggests to me the following con-
clusion: the fibula is either an eighth or early seventh century bc Phrygian
object of a hitherto unknown technique for that period, or it is a later copy
of a Phrygian product. If the latter possibility proves correct, I would sug-
gest that it is Lydian and that it was made sometime in the sixth century bc.
In other words, the fibula could have been made any time between the late
eighth and the sixth century bc. My inclination, however, is to think that
the latter conclusion, that the fibula is sixth century, and probably Lydian,
is actually the correct one.
The gold fibula is not an isolated example of an object that illustrates
the problem of sorting out eighth from sixth century bc or Phrygian from
Lydian products. For what emerges from a study of the artifacts excavated
in post-Kimmerian Gordion is that several eighth-century forms and motifs
continue apparently unchanged into archaic times, the sixth century. And
when one examines the small but growing number of objects becoming
available from Lydia, one notices in some examples a dependence on Phry-
gian culture. At one time I was under the impression that the late seventh
and sixth century Lydian parallels to Phrygian objects reflected a knowl-
edge of eighth century culture there, but I now believe that the evidence
suggests Lydia borrowed objects and ideas from contemporary Phrygian cul-
ture.
15 Gold jewelry was not common at Gordion until the sixth century, Phrygian Fibulae, p. 9,
17 Phrygian Fibulae, passim, pp. 78 f., Appendix A and B. Problems of dating Types XII, 13
of Anatolian tombs, cf. Phrygian Fibulae, pp. 17, 26, 30, n. 11 with comments by E. Akurgal,
Phrygische Kunst (Ankara, 1955), pp. 83, 130; Akurgal dates all the Ankara tumuli to the
seventh century bc. Makridy’s Tumulus II, apparently the third tumulus which he dug, is
seventh or early sixth century in date, see n. 25.
20 G.M.A. Hanfmann, “The Ninth Campaign at Sardis (1966),” BASOR, 186 (1967), p. 29,
Fig. 14, shows a fibula from Sardis dated to the Protogeometric period. On p. 36, n. 7 a parallel
from Skyros is cited; better parallels may be seen in Fibules, Type IV, 11, examples of which
794 chapter twenty-six
found in a Lydian level at Sardis dated to the late seventh–early sixth cen-
tury bc; whether these were products of local Lydian workshops or were
imported from Phrygia is not known.21 In a context ca. 600bc at Ephesus we
have seen that Phrygian-type fibulae of gold and electrum were found. Here,
too, it is not possible to conclude whether these fibulae were locally made or
whether they were imported from Lydia or Phrygia.22 And other sites in the
area, such as Bayrakli, Larisa, Chios, and Lesbos, have yielded Phrygian-like
fibulae from the seventh to the fifth centuries which thus demonstrate that
these types were locally manufactured during those periods.23
Bronze bowls, with or without omphaloi, and often decorated with petals,
or with spools and elaborate handles, are found in all eighth century tombs
at Gordion and Ankara.24 At both sites, evidence exists that these bowls
were still being manufactured in the seventh and sixth centuries (or at
least were still available as heirlooms during this period). Tumulus J, of
the seventh century, and Tumuli F, M, and E, of the sixth century, all at
Gordion, contained in their tomb deposits various products among which
were bronze omphalos and other bronze bowl types. And one of the tumuli
excavated by Makridi at Ankara, of seventh or early sixth century date,
contained bronze bowl fragments.25
come from several of the Aegean islands and from Ephesus; see also N. Firatli’s report in
Annual of the Archaeological Museums of Istanbul (Istanbul, 1958), pp. 75f., Fig. 13, Nos. 1–4,
from an area near Izmir.
21 Phrygian Fibulae, pp. 25, 44, Figs. 75, 76, p. 33, n. 45.
22 Ibid., p. 44, for a discussion of some of these fibulae as Lydian objects.
23 Ibid., pp. 31 f., n. 31, 35, 36, 37, 39 f. Belt buckles in the form of fibula arcs occur in
the eighth century and continue through the seventh, J. Boardman, “Ionian Bronze Belts,”
Anatolia, 6 (1961–1962), 179 ff.
24 Gordion, p. 73, Nos. 60–72, Figs. 53, 54; p. 74, Nos. 73–80; R.S. Young, “Gordion 1956:
Preliminary Report,” AJA, 61, 4 (1957), p. 328, Pl. 93, Fig. 31; “The Gordion Campaign of 1957:
Preliminary Report,” AJA, 62, 2 (1958), p. 150; “The Gordion Campaign of 1959: Preliminary
Report,” AJA, 64, 3 (1960), p. 230; T. Özgüç, M. Akok, “Die Ausgrabungen an zwei Tumuli auf
dem Mausoleumshügel bei Ankara,” Belleten, XI, 41 (1947), pp. 57–85; H. Zubeyr, “Ankara Gazi
Orman Fidanliginda Buhman Eserler,” TTAED, I (1933), 10f., Nos. 1–3, p. 12, No. 5; see n. 18 for
Akurgal’s late dating of these tumuli.
25 Gordion: Univ. Mus. Bull., 17, 4 (1953), pp. 30 f., 34 f.; Univ. Mus. Bull., 16, 1 (1951), pp. 17f.;
E. Kohler, E.K. Ralph, “C-14 Dates for Sites in the Mediterranean,” AJA, 65, 4 (1961), p. 362;
Tumulus M is not yet published. Ankara: M. Schede in AA (1930), cols. 481f.; T. Özgüç,
“Untersuchungen über archäologische Funde aus Anatolien,” Belleten, 10, 40 (1946), p. 619;
E. Akurgal, Phrygische Kunst, pp. 84, 130 for dating. There seems to be some disagreement
about the provenience of a bronze vessel fragment: Özgüç, op. cit., p. 617, fig. 33, calls it an
object from Makridy’s second tumulus, whereas Akurgal in Bayrakli (Ankara, 1950), p. 89, says
it came from the third tumulus excavated by Makridy.
phrygian or lydian? 795
In Lydian territory there is rare but significant evidence that bronze bowls
of Phrygian types were in use. A plain bronze bowl with Phrygian-type
handles and spools was found in a grave at Manisa, near Sardis; its date is
stated to be seventh or sixth century bc.26
A small bronze omphalos bowl with a raised central groove was found
at Ephesus in a poor stratigraphical area.27 Its size is smaller than most
Phrygian bowls and it is possible that it was a local product and not imported
from Phrygia.
One of the best pieces of evidence that Phrygian-type bowls were in use
in the sixth century bc in the west is the apparently metal bowl carried by
the well-known ivory lady found at Ephesus.28 This bowl could certainly be
accepted as an eighth century product of Phrygia. Unfortunately, we are not
in a position to establish whether the bowl carried by the lady was an import
from Phrygia, or Lydia, or whether it was made in the Greek East.
One more example of a sixth century date for a Phrygian-type bowl is
a petalled specimen found at Trebenishte in Yugoslavia.29 There can be no
doubt about the date of the grave in which it was found, but we are not able
to make a specific statement about the bowl’s original home or whether it
was an heirloom.
With the omphalos bowls one is, therefore, faced with a problem parallel
to the one discussed above concerning our gold fibula. Thus, if a Phrygian-
type bowl is encountered on the antiquities market isolated from its original
context, it would be difficult to decide whether it is Phrygian or Lydian,
or whether it is eighth century, or later. A silver petaled omphalos bowl
allegedly from Anatolia now in the collection of the Metropolitan Museum
of Art (Pl. IV, Fig. 9), and another similar example in the Birmingham City
Museum are two such examples.30 Another silver omphalos bowl, plain and
26 Akurgal, Phrygische Kunst, p. 81, Pl. 60a, dates it to the sixth century: A.K. Knudsen,
“From a Sardis Tomb: A Lydian Pottery Imitation of a Phrygian Metal Bowl,” Berytus, XVI
(1964), 63 dates it to the seventh century. Compare the Manisa bowl to examples from the
Baumschule tumulus at Ankara, Phrygische Kunst, Pls. 58a b, 59a, b, and Gordion, R.S. Young,
“Gordion on the Royal Road,” PAPS, 107, 4 (1963), p. 362, Fig. 19.
27 Hogarth, op. cit., p. 152, Pl. XV, 13.
28 Ibid., Pls. XXI, XXII; Knudsen, op. cit., p. 64, n. 11; Akurgal, Phrygische Kunst, pp. 81f. I
agree with Akurgal and Knudsen that the bowl reflects Phrygian rather than Greek influ-
ences.
29 B. Filow, Die archaische Nekropole von Trebenishte (Berlin, 1927), pp. 75f., Fig. 93, 1;
with grooves around the omphalos, now in a private collection and said
to come from Anatolia, fits into this category (Pl. IV, Fig. 10). The size of
these three bowls conforms to that of the bronze examples excavated at
Gordion and Ankara, and in typology they appear to be eighth century
Phrygian products. Yet, in Anatolia, aside from a silver omphalos bowl found
at Boǧazköy of uncertain date, all of the many bowls excavated at Gordion
and Ankara are made of bronze (two bronze examples from Tumulus 1 at
the Mausoleum Hill in Ankara are said to be gold plated).31 And none of the
petaled bronze bowls found at Gordion has the incised lozenge decoration
around the omphalos that exists on the silver bowl shown in Fig. 9. It is,
therefore, tempting—but certainly premature at present—to conclude that
these silver bowls might be Lydian and sixth century bc in date. Certainly
one must be prepared to entertain such a possibility. What is needed, of
course, is a detailed study of the techniques and styles of the many excavated
bowls from the Phrygian tombs before one generalizes about unexcavated
material appearing on the market.32
Imitations of Phrygian metal bowls made in terracotta were manufac-
tured in the eighth century at Gordion and Boǧazköy.33 This practice contin-
ued at Gordion into the later centuries, for pottery vessels imitating metal
shapes of typical eighth century types have been excavated on the City
Mound in the archaic levels.34 One late example from Gordion, a spouted
pitcher in a local black-polished ware, a masterpiece of craftsmanship, imi-
tates in all details—bolsters, studs, and separate handle—a metal prototype
(Pl. V, Fig. 11). The excavator made these pertinent remarks: “… the Phry-
gians must have kept right on producing elaborate vessels of metal down
into the fifth century, for the shape of our jug is typically Phrygian, known at
31 Boǧazköy: H. Otto, “Die Funde von Büyükkale,” MDOG, 78 (1940), 48f., Fig. 9, 1, “aus der
ersten Schicht”; the bowl does not appear to me to be a typical eighth century type and seems
to be later. T. Özgüç, M. Akok, op. cit., Pl. X, Fig. 18 for one of the two allegedly gold-plated
examples. Also see note 36 below for a comment about silver vessels in the sixth century bc.
32 See for example M.-L. Buhl, Skatte fra Det Gamle Persien, Nationalmuseet (Copenhagen,
1968), pp. 86 f., No. 199 with figure on p. 87: a petaled bronze omphalos. The caption says
the bowl is Greek, was found in Iran (presumably it was bought from a dealer), and is sixth
century in date. Judging from the photograph the bowl appears to be eighth century Phrygian!
33 Gordion, pp. 66 f., Fig. 42; R.S. Young, AJA, 61, 4 (1957), pp. 328f., Pl. 95, Fig. 33; T. Beran,
“Eine Kultstätte phrygischer Zeit in Boǧazköy,” MDOG, 94 (1963), 45ff., Figs 13, 14; Knudsen,
op. cit., pp. 65 ff., discusses this object as well as the use of bronze handles associated with
wooden bowls.
34 Gordion, pp. 196 f., Fig. 195; Knudsen, op. cit., p. 67, n. 22, and infra. Cf. Eva-Marie Bossert,
Plate IV
Gordion as early as the eighth century, and the potter must have been using
as his model a nearly contemporary metal vessel of local manufacture.”35
Another example from Gordion, an omphalos fragment in black polished
ware, was found in a reused Phrygian house and “could belong to either
Phrygian or Persian times”; the excavator also remarks: “… the omphalos
fragment is like that shown by many of the bronze bowls from the Phrygian
tombs.”36
One should call attention in this context to the glass omphalos found in
the eighth century tomb of Tumulus P at Gordion. It, too, imitates a standard
metal prototype, similar to those excavated at contemporary Gordion, and
it neatly demonstrates that imitations of metal vessels were not confined to
terracotta.37
We may conclude from these brief comments that, albeit on a limited
scale, Phrygian craftsmen in the sixth and fifth centuries bc manufactured
pottery of the same types as did their eighth century ancestors.
Turning our investigation now to Lydia, we see a similar situation obtain-
ing. Lydian craftsmen of the sixth century imitate Phrygian metal vessels
in pottery and at the same time they also copy Phrygian pottery vessels.
Pottery of these types, found both in metal and pottery at Gordion and
elsewhere, have been excavated at Sardis: side-spouted pitchers with strain-
ers,38 bowls with spool handles and bolsters (Pl. V, Fig. 12), and horizontal
35 R.S. Young, “The 1963 Campaign at Gordion,” AJA, 68, 3 (1964), pp. 282f., Pl. 84, Fig. 12.
I would like to thank Professor Young for allowing me to republish the vessel here for the
convenience of the reader. We would like to know more about the fabric of this vessel: is the
fabric different from that of eighth century examples of black polished wares?
36 R.S. Young, “The Gordion Campaign of 1967,” AJA, 72, 2 (1968), p. 235, Pl. 76, Fig. 15. See
also R.S. Young, “The 1961 Campaign at Gordion,” AJA, 66, 2 (1962), pp. 154f., Pl. 41, 1, 2 for
a silver Achaemenid vessel and its counterpart in black polished ware found in the archaic
level. Incidentally, this silver example informs us that in the sixth century at Gordion silver
vessels were in use.
37 R.S. Young in AJA, 61, 4 (1957), p. 328, Pl. 94, Fig. 32. Young considers the bowl to be an
import but cf. Axel von Saldern in “Glass Finds at Gordion,” Journal of Glass Studies, I (1959),
22, 25 ff., Figs. 1, 2; Von Saldern accepts the possibility that the bowl was a local product but
prefers to see it as an import.
38 Gordion: AJA, 64, 3 (1960), Pl. 56, Figs. 6, 8 (the latter is bronze); AJA, 68, 3 (1964). Pl. 84,
Fig. 12 (here Fig. 11); Gordion, pp. 57 f., Figs. 21–23; pp. 62ff., Figs. 28, 29, 31–34; Akurgal, Phry-
gische Kunst, Pls. 11, 12, 14b. Cilicia: H. Goldman, Excavations at Gözlü Kule, Tarsus (Princeton,
1963), Pl. 88, 1242. Alishar: H.H. von der Osten, The Alishar Huyuk, Seasons 1930–32, II (Chicago,
1937), OIP, XXIX, 365 f., Figs. 414–415. Boǧzaköy: K. Bittel, H.G. Güterbock, Boǧzaköy (Berlin,
1935), p. 57, Pl. 13a, b (cf. Fig. 11 in this study). Sardis: Phrygian Fibulae, p. 44, Fig. 104, p. 47, n. 30
(= Andrew Oliver, Jr., “Lydia,” BMMA, XXV, 5 (1968), p. 199, Fig. 8 upper left): see also Phrygian
Fibulae, pp. 66, 74 f., n. 45.
phrygian or lydian? 799
Plate V
39 Gordion: Gordion, pp. 72 ff., Figs 50–54; AJA, 62, 2 (1958), Pl. 25, Figs. 15, 21; Pl. 27, Figs. 22,
24; R.S. Young, “The Gordion Tomb,” Expedition, I, 1 (1958); AJA, 61, 4 (1957), Pl. 93, Fig. 31.
Ankara: Akurgal, Phygische Kunst, Pls. 57–59. Sardis: Oliver, op. cit., p. 199, Fig. 8 lower right;
H.C. Crosby, Sardis, I (Leyden, 1922), 119, Fig. 126 center (= Fig. 12 here); Knudsen, op. cit., Pls.
VIII, IX.
40 The Gordion metal example from the eighth century Tumulus MM has not yet been
published. Sardis: Phrygian Fibulae, Fig. 103 (= Oliver, loc. cit., upper right). See Von der Osten,
op. cit., p. 150, Fig. 191, for a Hittite version from Alishar.
41 Knudsen, op. cit., p. 67.
42 Compare the low foot or lack of a foot on the side-spouted vessels from eighth century
Gordion referred to in note 36 to the foot on the sixth century examples from Gordion (Fig. 11
here) and Sardis. The example from Bogazkoy has a low foot, the one from Tarsus has no
foot. G.F. Swift, Jr., “The Ninth Campaign at Sardis (1966),” BASOR, 186 (1967), 34, refers to
p. 24, Fig. 8 as a Lydian Geometrio pottery fragment that imitates Greek Geometric styles. It
seems to me that a Phrygian prototype could actually have served as a model, e.g. Gordian,
pp. 58 ff., Figs. 22, 25, 26. For a discussion of the relationships between Greek and Phrygian
Geometric pottery see Akurgal, Phrygische Kunst, pp. 33ff., 121, 126 and supporting comments
by this writer in Phrygian Fibulae, pp. 60 ff.
43 Phrygian Fibulae, pp. 6 f., 8, 10, n. 36 and 37, 44; R.S. Young, “Making History at Gordion.”
overlying tumulus. This practice began in the eighth century and continued
into the seventh and sixth centuries. The Lydians also built tumuli over their
tombs, and these tombs were usually placed off center from the middle
of the tumulus. It seems almost certain that the idea was borrowed from
the Phrygians. The very same type of tumulus constructon with the grave
off center was known in northwest Iran and in Cyprus during the first
millennium bc, and the idea may have spread to these areas from Phrygia.44
Future research will help us solve some of the issues and problems raised
in this study. We know very little about stylistic differences between eighth
and sixth centuries bc Phrygian pottery, especially of the black-polished
varieties, because no comprehensive study of the Gordion or Boğzaköy
pottery has yet appeared. Likewise, we have no published analysis of the
bronze bowls from the eighth century tombs comparing them to the few
sixth century examples found in the later tombs. We would like to know
something about the types of metals and techniques used to manufacture
them, as well as a stylistic discussion. From the Lydian side we would also
want a study of Lydian pottery so that we may better be able to discuss
the extent of Phrygian influences. And knowing this, we would be able to
compare the relative values of both Greek and Phrygian (Near Eastern)
influences on Lydian culture.
Abstract: In two recent publications of material from tombs excavated at two Near
Eastern sites, Marlik Tepe in NW Iran and Assur in Assyria, the occurrence of
fibulae among the grave goods and the concomitant chronological conclusions to
be obtained from their presence have been either ignored or distorted. In this paper
it is argued that a correct analysis of the chronology of the advent of fibulae in
the Near East leads to significant results, results other than those perceived by the
scholars responsible for the publications.
“Still, it is an error to argue in front of your
data, you find yourself impossibly twisting
them round to fit your theories.”
—Sherlock Holmes
In their discussion of, respectively, Iranian and Urartian fibulae (the term
for ancient safety pins), Roman Ghirshman and Baki Öğün both thought
it appropriate to quote Solomon Reinach’s comment made in 1896, that
the fibula is “un principe de classification chronologique … et que la fibule
devait être considéréé comme ce ‘fossile directeur’ de l’ antiquaire ….”1 Rei-
nach’s old statement remains cogent and meaningful, all the more so as
subsequent research has confirmed his conclusion in general terms and,
with more refined determination of stratified finds, has allowed scholars to
consider the fibula to be one of the most important chronological indicators
available in archaeological research. This situation obtains for Europe and
the Aegean, where fibulae first appear in the late 2nd millennium bc, as well
as in the Near East. One might say the fibula, wherever found in an Old World
context, is a precise chronological indicator, a Leitfossil, at least in regard to
post and ante quem dating.
Speaking only with concern for the Near East (fig. 1), it may be asserted
with some certainty that except for a relatively small number of fibulae
* This article originally appeared as “Fibulae and Chronology, Marlik and Assur,” Journal
2 David Stronach, “The Development of the Fibula in the Near East,” Iraq 21:2 (1959)
181–206. For the East Mediterranean fibulae, see 182–185, fig. 1, Map 2. For the Near Eastern
examples, see 185–206, especially 185 and 193 for chronology. The dating of the East Mediter-
ranean Type II, with round or arched bow and with mouldings or beads at the terminals,
is somewhat problematic. It may have come into existence before the late 8th century but
by no means with certainty. Part of the problem concerns the occurrence of Type II fibu-
lae at Megiddo and the chronology of the excavated levels there. The excavators, R.S. Lamon
and G.M. Shipton, Megiddo I (University of Chicago Press: 1939) xxvii, give a chart with a
tentative chronology of Levels I, II, III, and IV, but warn that it must be used with “utmost
fibulae and chronology, marlik and assur 805
yielded fibulae in Syria, Anatolia, Urartu, Assyria, and Iran has charted the
chronological position of these artifacts. Initially two scholars, Ch. Blinken-
berg and D. Stronach, documented the occurrences and stratigraphy of hun-
dreds of Near Eastern fibulae and classified them into various types.3 They
exist in numerous varieties and forms. Some are semicircular or arched,
either with plain or beaded arms (Stronach’s Types I and II). Others, by far
the most common and ubiquitous, are triangular in shape (often referred
to as knee or elbow fibulae), the great majority decorated on the arms with
discs, ribs, and beads (Stronach’s Types III and IV); the catch is flat, bent to
receive the pin, and in a large number of examples is cut in the form of a
human hand. Exactly when the hand catch first appears is uncertain, but it
can be documented nowhere before the late 8th century, nor does it appear
earlier on Cypriote and East Mediterranean fibulae.
More recent scholarship has but reinforced and further clarified the
fibula chronology proposed by Stronach for Syria and Mesopotamia, and
the interested student is invited to study the evidence independently.4
reservations,” hardly a basis for discussions of absolute chronology for the fibulae. On pls. 78
and 79 are examples of East Mediterranean and Near Eastern fibulae from Levels I, II, and III,
which they date to the 8th century and later; only one fibula, pl. 78:20 (= Stronach’s Type II, 4,
fig. 6:4), derives from Level “IV filling,” i.e., presumably pre-8th century (1000–800 bc). In any
event, it derives from the fill, not from a locus in Level IV. In G. Loud, Megiddo II (University
of Chicago Press: 1948) only one fibula, pI. 223:79, is listed as from Level IV; it is isolated in
the text and presented with no discussion. I therefore suggest that Stronach (op. cit., in this
note, 191) misstates the evidence that his fig. 6:4 emerges “about 900 bc.” There is no evidence
that compels one to accept as fact that Type II fibulae are earlier at Megiddo than the typical
Near Eastern examples; they coexist there together.
3 Ch. Blinkenberg, Fibules grecques et orientales (Copenhagen 1926), Type XII, 24–25, 204–
225; Type XIII, 236–238, 239 247. This monograph is still the handbook and the initial study
for all Fibelforschers; Stronach, op. cit. (in note 2).
4 M.V. Seton-Williams, “Preliminary Report on the Excavations at Tell Rifa"at,” Iraq 23:1
(1961) 86, pl. XLI, 11, 12; Judy Birmingham, “The Development of the Fibula in Cyprus and the
Levant,” PalExplorQuar 95 (1963) 80–112, 91, 95 (but with imprecise chronologies and strati-
graphical references); D.E. McGown et al., Nippur I (University of Chicago Press: 1967) 69–71,
99, pl. 151:19–21; P.R.S. Moorey, Cemeteries of the First Millennium bc at Deve Hüyük (BAR Inter-
national Series 87: 1980) 85–86; Arndt Haller, Die Gräber und Grüfte von Assur (Verlag Gebr.
Mann: Berlin 1954) 12–161 (for a total of 99 fibulae from 65 graves and tombs, all neo-Assyrian).
This last publication is impossible to use with regard to a diachronic understanding of fibula
types recovered in the earlier and later neo-Assyrian burials. Haller makes it quite clear on
pages 5 and 12 that his division of the burials into a “Frühe Zeit 1100–824 v. Chr.” and a “Späte
Zeit 824–612 v. Chr.” is not firm and absolute. The division is based in part on a few inscrip-
tions and on an interpretation of the finds, for which he significantly states (page 5) “In vielen
Fällen werden in der Datierungsfrage Zweifel entstehen ….” Of the objects used to make the
division, he further notes that they are “mehr oder weniger charakteristisch.” Haller posits
that “Armfibeln … in runder Form” occur in the earlier burials, triangular ones in the later,
but of the 99 fibulae mentioned he illustrates only one, a triangular example from Gruft 30
806 chapter twenty-seven
Furthermore, the most recent studies targeted on the distribution and chro-
nology of the fibulae from Anatolia,5 Iran,6 and Urartu7 have also reinforced
the previously held conclusions. These studies, based “… surtout sur les
témoignages tirés des circonstances de trouvaille,” to quote Blinkenberg’s
criterion for “la determination chronologique des fibules,”8 have indicated
that no fibula was manufactured or in use in any of these areas before the
late 8th century bc. Before that time straight pins served to fasten clothing.
(pl. 22), which he dates to the earlier part of his putative later period, late 9th century bc (110;
see infra for a discussion of this fibula). Haller’s terminology in his description of the fibulae
is unclear and inconsistent: “Fibula,” “Bronzefibula,” “Armfibula,” “eckige Armfibula;” in only
three cases does he refer specifically to a “rundgebogene Fibula” (Grab 681), “Arm fibula … in
rundlicher Form” (Grab 758), “gerundete Armfibula” (Grab 809), none of which is illustrated.
Grave 758 had only anklets along with the fibula as burial contents. All we can say for the
Assur fibulae is that they occur only in the neo-Assyrian period, and suggest that none has
been demonstrated to occur before the late 8th century bc.
5 Oscar White Muscarella, Phrygian Fibulae from Gordion (Bernard Quaritch Ltd: Lon-
don 1967); R.M. Boehmer, Die Kleinfunde von Boğazköy (Gebr. Mann Verlag: Berlin 1972)
46–66; idem, “Phrygische Prunkgewänder des 8. Jahrhunderts v. Chr.” AA (1973) 151, 166, 172.
Recent discussions of Phrygian fibulae imported into the Greek spheres reinforce the Ana-
tolian chronology: E. Sapouna-Sakellarakis, Die Fibeln der griechische Inseln. Prähistorische
Bronzefunde XIV, Band 41 (Munich 1978) 120–129; Klaus Killian, Fibeln in Thessalien von der
mykenischen bis zur archaischen Zeit. Prähistorische Bronzefunde XIV, Band 2 (Munich 1975)
151–154. An earlier date for the advent of fibulae at Gordion is suggested by R.S. Young, Three
Great Early Tumuli (University Museum: Philadelphia 1981) 9, based on his belief that Tumu-
lus W, which contained fibulae, was to be dated to the end of the 9th century bc or “slightly
later.” For a criticism of this chronology see Oscar White Muscarella, “King Midas’ Tumulus
at Gordion,” The Quarterly Review of Archaeology (December 1982) 8–9.
6 Roman Ghirshman, op. cit. (in note 1) 92; Oscar White Muscarella, “A Fibula from
Hasanlu,” AJA 69:3 (1965) 233–240; P.R.S. Moorey, A Catalogue of the Persian Bronzes in the
Ashmolean Museum (Oxford University Press: Oxford 1971) 124, 127; Louis vanden Berghe,
“Les Fibules provenant des Fouilles au Pusht-i-Kuh, Luristan,” IranAntiq 13 (1978) 35–74 (a
comprehensive and most valuable summary of all the fibulae found to date in Iran; for
chronology see pages 49, 67, 69; add W. Kleiss et al., Bastam I [Gebr. Mann Verlag: Berlin
1979] 171:7, 9, 172:3; and Marlik, see infra). The evidence of the excavations at Hasanlu in
NW Iran is telling: no fibulae were recovered in the settlement or cemetery of Period IV,
which terminated ca. 800 bc, but they occur in Periods III and II, the former, Iron Age III,
commencing in the late 8th century bc or later. Note that all the spring fibulae excavated
in Iran are indistinguishable from those excavated in Syria and Assyria. A great number
of stray, Near Eastern type fibulae published by many scholars (including Ghirshman and
Stronach, and many others too numerous to cite) as deriving from Iran in fact derive from
the antiquities market, i.e., not from Iran, archaeologically speaking. For one special group of
misattributed fibulae see Oscar White Muscarella, “Five Additions to the Norbert Schimmel
Collection,” APA 7/8 (1976–1977) 316, no. 3.
7 Muscarella, 1965 op. cit. (in note 6) 235–237; for a comprehensive review of fibulae from
Urartu see Öğün, op. cit. (in note 1) 178–188. On page 178 Öğün dates the advent of fibulae in
Urartu “vor dem letzten viertel des VIII. Jahrhunderts nicht.”
8 Blinkenberg, op. cit. (in note 3) 26.
fibulae and chronology, marlik and assur 807
In summary, and to the point, after much research and review by a large
number of scholars of all the information available from excavations, the
evidence is certain—as certain as is possible in archaeological activity—
that the presence of a fibula in a stratified or closed context in the Near East,
at least in areas east of the eastern Mediterranean zone, signifies an ante
quem non date of late 8th century bc or later. And only the occurrence of a
Near Eastern fibula in an unequivocally dated closed context from an earlier
period will alter these conclusions; but I submit that such an anticipated
occurrence seems highly unlikely given the evidence to hand. I know of few
other classes of artifacts aside from pottery that, by all the copy book rules
and concepts of archaeological conclusion formation, can be accepted with
some ease as a reliable “fossile directeur,” The fibula in the Near East is a
classic, paradigmatic, ante quem non artifact.
not a single tomb group with complete inventory and illustrations has been
published (Negahban has informed me that the full publication of all the
tomb groups is forthcoming). In his 1964 and 1977 reports Negahban gave
in the texts only the excavation grid numbers (in Roman numerals) for the
locus of the objects he illustrated (in 1977 in a somewhat confusing manner,
e.g., “tomb VI B+,” “tomb XVIII C,” or “grid XX F”); in 1979, 1981, and 1983
he gave for each object illustrated both a grid (“Trench”) Roman numeral
and a tomb number (in Arabic numerals), thereby clarifying for the first
time the important distinction between grid number and specific tomb
number.11 More importantly, in the 1979 and 1981 reports a complete plan
was published with each tomb drawn in its respective grid and identified
by number (Tombs 1 to 53). Working through all the tomb and artifact
references, matching and sorting out grids and tombs using the grid-tomb
plan as a guide, and based on what has been published with illustrations,
I have been able to isolate the partial contents of about 31 of the 53 Marlik
tombs. Here I will discuss the contents of only one of the tombs isolated. It
is, as will become evident, one of the most important tombs in the Marlik
cemetery.
In his most recent publication of the Marlik finds (1983) Negahban illus-
trates and discusses 63 metal vessels, 15 of gold, six of silver, and 42 of bronze,
excavated from 32 identified tombs. On pages 89–90 we are given a list (but
without previous publication references, except for the vessels) of all the
artifacts found “in one of the richest [tombs] in Marlik both in terms of
quantity and quality of the burial gifts.” This is Tomb 36, which is in grid
XVIII C, the only tomb in that square. I checked all the objects published
over the years that were identified either as from grid XVIII C or Tomb 36, in
many cases matching them with the 1983 list, and present the results here
with Negahban’s publication date references. Tomb 36 contained at least:
four gold vessels (1964, figs. 111, 112, pl. VIII A; 1983, G6, 10, 11, 12); one gold
male bust (1964, fig. 101, pl. X); one gold bracelet (1964, fig. 77); a gold needle
(1964, fig. 42); a gold ring (1964, fig. 81); a gold seal (1977, 94, fig. 9); a gold
chain (1964, fig. 127); two bronze female figurines (1964, fig. 99; 1979,170–171,
pls. 33: 2, 3; 34: 1, 2); three bronze seals (1977, 99, 100–102, figs. 19, 20, 25, 26);
a bronze vessel (1964, fig. 28; 1983, B28); a bronze cymbal (1964, fig. 54); a
11 This conflation of grid and actual tomb number has caused confusion among scholars:
e.g., Peter Calmeyer, Datierbare Bronzen aus Luristan und Kirmanshah (Walter de Gruyter
& Co.: Berlin 1969) 56, 101, 104 cites “Grab XIII B,” “Grab XXIIIc [misprint for XXIII G],”
inadvertently mixing grid numbers with tombs.
fibulae and chronology, marlik and assur 809
bronze tweezer (1964, fig. 37); a bronze stag figurine (1964, fig. 96 A); seven
male and female terracotta figurines (1964, pl. XI; 1979, 159–161, pls. 29–31); a
terra-cotta charioteer (1964, fig. 94); a number of terra-cotta animal figurines
(1964, fig. 93, pl. IX).
Found with these objects in Tomb 36 was a Near Eastern bronze elbow
fibula, mentioned by Negahban in the aforementioned list of 1983 (page 90),
and which through his exemplary generosity and concern for scholarship
(and being aware of the intent of this paper) is published here for the first
time (fig. 2). The fibula has a wide triangular bow with a flattened bead
flanked by two discs or ribs, and a catch in the form of a human hand with
the thumb separate. It easily fits within Stronach’s Near Eastern Type III, 7
group, where many parallels are forthcoming both in form and hand catch
with an isolated thumb. Close and relevant parallels are to be found in
Syria and Assyria, and, significantly, at sites in Iran.12 All, without exception,
12 Type III, 7, has, as Stronach pointed out, op. cit. (in note 2) 197, “many different varieties,”
and there are many comparisons readily made with the Marlik fibula. I call attention here
only to a few, which are closest in form: Deve Hüyük, Moorey, op. cit. (in note 4) fig. 14:350,
dated to the 8th or to the 5th century bc; Nimrud, Stronach, op. cit. (in note 2) pls. L, 7, 10
(note the catch), LI, 1–3, and 9–12 for the catch; Gerar, Stronach, ibid. fig. 9:5, 6; from Iran
(excavated) see Godin Tepe, Bisitun, Susa, and sites in Luristan, vanden Berghe, op. cit. (in
note 6) fig. 11:5, 6, 7, 24; and fig. 9:9, 12, 15, 17, 18, 22 for the catch.
810 chapter twenty-seven
are dated in the late 8th or 7th centuries bc, or even later. Apparently no
other tomb at Marlik contained a fibula, but I maintain that because of the
presence of one in Tomb 36 there can be no doubt that at least this tomb was
closed not earlier than the late 8th century, and possibly in the 7th century,
i.e., in the Iron Age III period of NW Iranian chronology.
In all the publications concerning Marlik, Negahban has maintained that
the chronology of the cemetery is to be perceived as beginning sometime in
the 14th or 13th centuries bc and terminating in the 10th century, presum-
ably early in that century.13 In this posited chronology he has been followed
by a large number of scholars (too many to cite here), who over the years
have accordingly dated either the excavated materials from Marlik or many
related, but plundered or recently forged, objects from the antiquities mar-
ket to the late 2nd or, more rarely, to the early 1st millennium bc.14 It is not my
intention, considering the limited aims of this paper, to comment on the full
chronological range of all the Marlik tombs, except to note in passing that
some contained unbridged spouted vessels of Iron Age I, 2nd-millennium
13 Op. cit. (notes 9 and 10 for references): 1964, 37–38; 1965, 312, 326; 1979, 157, 173; 1981, 369;
1983, viii, 95. In 1965, 326 Negahban posited that the Assyrians “about the ninth century bc
or … the Urartians a little later, brought Marlik to an end.” In 1983, viii, 95, it is the Assyrians
who do this in the 10th century bc. There was, however, no Urartian penetration, neither as
settlers nor as raiders, into western Iran beyond Azerbaijan, many miles from the Caspian
Sea, as W. Kleiss has painstakingly demonstrated: see for example his Topographische Karte
von Urartu (AMI Ergänzungsband 3: Berlin 1976). Nor did the Assyrians at any time extend
their control, let alone penetration, so far east in Iran. What is more, the late date of Tomb 36
makes such a high date for the end of Marlik totally untenable.
Note, in the context of Urartian-Caspian relations, that Guitty Azarpay, Urartian Art and
Artifacts (The University of California Press: 1968) 12, 16, 82, note 28, pl. 1, refers to two Urartian
horse blinkers in the Feroughi collection, which she casually informs us derived from a
“recent discovery … in northwestern Iran …”, but which in fact derive from uncontrolled,
unwitnessed plunderers. This uncritical, and incorrect, attribution follows from a fictional
creation of Roman Ghirshman, “Deux Oeillères en bronze des Rois d’Urartu,” ArAsiae 27:
1/2 (1964–1965) 57–58, that the blinkers, inter alia, were discovered “dans une des tombes
des montagnards de la région du sudouest de la mer Caspienne,” and that the discovery
vouchsafes hitherto unknown information on Urartian-Caspian relations. The objects, of
course, do no such thing: they have no provenience, nor do we know what was found together
in situ or assembled in a dealer’s shop; see G. Gropp, “Ein Pferdegeschirr und Streitwagen aus
Urartu,” IranAntiq 16 (1981) 95, note 1.
14 The only exception of interest to this acceptance of chronology is Ekrem Akurgal,
Urartäische und Altiranische Kunstzentren (Ankara: 1968) 67, 103–106, who dates all the Marlik
material and that from Hasanlu and elsewhere in Iran to the 7th century bc, on the basis of
a preconceived notion that most of Iranian art is dependent on that of Urartu. Paradoxically,
Negahban cites modern forgeries in his lists of alleged parallels for his excavated pieces: 1983,
14, note 33; 19, note 71; 29, note 141; 58, note 12; 61, note 29.
fibulae and chronology, marlik and assur 811
form, which indicates that these tombs may date to that period.15 Nor is it
the time and place to initiate discussion with regard to the absolute date
of manufacture of all the objects deposited in Tomb 36, some of which
could conceivably have been curated heirlooms. Creative and intelligent
commentary cannot occur until all the tombs are fully published in discrete
groups—and even then all these problems will not be resolved. Neverthe-
less, the recognition that an Iron Age III, 8th–7th-century bc tomb exists
in the Marlik cemetery, an occurrence hitherto unperceived, will inevitably
profoundly alter our conception of the chronology of a number of Marlik
and Iranian artifacts of similar or related styles, primarily vessels and fig-
urines. And because of the existence of Tomb 36, archaeologists will have to
consider the cultural and political implications of an unplundered (in antiq-
uity) cemetery, in which many rich tombs were built for a time spanning
apparently four or five hundred years in a peaceful and protected environ-
ment.
One example of a downward shift in chronology that results from the
lowering of the Marlik terminal date is of special interest and warrants
brief review here. In 1972 I published a superb bronze vase that had been
purchased on the antiquities market, and which I dated to a time between
1000–800bc; my conclusions were in part based on the traditional Marlik
chronology.16 The vessel has six horizontal friezes or rows of animals that, as
I stated, strikingly parallel in form and style those represented on some of
the Marlik vessels. Since 1972 two scholars have suggested alternative dates
for the vase. Calmeyer17 believes that it was made in the 2nd millennium
bc, whereas Moorey18 suggested (but without discussion) that it should be
placed in the Iron Age III b period; in a subsequent publication he dated it
between 1000 and 650 bc.19 Four separate dates have thus been suggested for
the time within which the vase was manufactured, a clear indication that
15 Tombs 27, 45, 50, 52; Negahban, 1983 op. cit. (in note 10) S21, B43–47. There is as yet no
evidence that the unbridged spouted vessel, the classic Iron Age I form, continued into the
Iron Age II period.
16 Oscar White Muscarella, “A Bronze Vase from Iran and its Greek Connections,” Met-
MusJour 5 (1972) 25–50, figs. 1–11. Note that on pages 43 and 44 I expressed even at that time
some feeling that a 2nd-millennium date for all the Marlik material was not necessarily cor-
rect; Negahban (1983 op. cit. in note 10) neatly confirms that anticipation.
17 Op. cit. (in note 11) 205, note 442, c.
18 P.R.S. Moorey, “Some Elaborately Decorated Bronze Quiver Plaques made in Luristan,
we are less secure about the chronology, and the archaeology, of NW Iranian
material than is usually recognized.
Now, Tomb 36, the fibula tomb, contained two gold vessels20 decorated
with two rows of animal friezes, and they were mentioned along with oth-
ers in my 1972 list of comparanda for the bronze vase. Whatever the date
one may assume for the manufacture of the vessels, they were in use and
culturally available up to the time of the tomb’s closure. This pertinent infor-
mation, this lowering of the lower limit for the presence of the Marlik ani-
mal frieze on vessels, signifies that the Metropolitan Museum’s bronze vase
could have been made at a time later than the early 1st millennium bc. This
date will have been, I suggest, sometime between, to give the upper limits,
the 10-9th century and the years close to ca. 700bc.21 In the original publica-
tion of the bronze vase22 I argued that the form of horizontal friezes that it
shares with the Marlik vessels may have been the ultimate prototype for the
origin of certain East Greek animal friezes represented on pottery vessels,
those of the “Wild Goat” style. The problem outstanding at that time with
this argument was that the Greek examples did not appear before the mid-
7th century bc, at least a century and a half later than the lowest date I had
then assigned to the vase. But now, given the extended chronology at Mar-
lik that manifestly indicates a longer life for the animal frieze in Iran than
hitherto recognized, the chronological gap between the advent of the Wild
Goat style and the Iranian parallels is significantly narrowed, if not indeed
closed. A single fibula has taken us on a long and promising journey, one not
yet ended.
We now turn to the second publication to be reviewed, this one concern-
ing the contents and date of closure of but one of the hundreds of tombs
excavated many decades ago at Assur in Assyria. In this instance a fibula
recovered in a tomb was not so much ignored, as was the case at Mar-
lik; rather it has been unappreciated for the correct chronological message
it bears. And if the aberrant chronological conclusion presented there is
20 Negahban, 1964 op. cit. (in note 9) figs. 111, 140, 144, pls. VIII A; XII; idem, 1983 op. cit. (in
note 10) GI0, 12; see also the Tomb 36 listing, supra.
21 The upper limits for the time of manufacture still obtain because of the presumably
earlier dates of the other Marlik vessels with animal friezes. Also relevant is the existence at
9th-century bc Hasanlu of a small group of ivory plaques that depict winged bulls executed
in the same style and body decoration as found on the bronze vase, Oscar White Muscarella,
The Catalogue of Ivories from Hasanlu, Iran (University Museum: Philadelphia 1980) 108–115,
190–191, numbers 214–220. I still think it probable that the Metropolitan Museum’s vase was
made before 700bc.
22 Op. cit. (in note 16) 43–49.
fibulae and chronology, marlik and assur 813
29 Ibid. 116, Abb. 11, 124; Haller, op. cit. (in note 4) plate 22, c-e.
30 Sürenhagen and Renger, 103–108, Abb. 1–5.
31 Ibid. 128.
32 Ibid.
33 Stronach, op. cit. (in note 2) 196–197, fig. 8:2; actually, the fibula could be considered as
fitting into Type III, 7, as the central bead seems to be flanked by two smaller ones: but this
is not relevant to chronology.
34 Loc. cit. (in note 26). “Ich vermag nun keinen gravierenden Unterschied zwischen
dieser Fibelgruppe [Type II, 4] und D. Stronachs Gruppe III.6 … festzustellen, zu der auch
die Fibel II 190 al [from Gruft 30] gezählt wird, und schliesse ein älteres Entstehungsdatum
als das dart angegebene (7. Jhdt. v. Chr.) nicht aus.” In this context Sürenhagen refers to the II,
4 fibulae at Megiddo “(ca. 1000–800 v. Chr.),” for which see here note 2. He also accepts Haller’s
clearly tentative division date of 824 bc for the alleged time when round bowed fibulae were
replaced by triangular ones, for which see here note 4.
fibulae and chronology, marlik and assur 815
different form of bow than Type III, 6. Sürenhagen then goes on to cite only a
single excavated parallel of a triangular fibula. The sole parallel offered is not
without relevant interest in a chronological discussion, and its introduction
reveals far more than was intended. It is a Type III, 7, example from War
Kabud in Luristan, Iran. Without discussion and presentation of evidence
Sürenhagen suggests this site need not be 8th–7th century bc in date, but
could be 9th century. In fact, War Kabud, on the basis of all the information
available to Iranian archaeologists, must be an Iron Age III site, i.e., 8th–7th
century bc in date;35 and of all the other pertinent excavated parallels for the
Assur fibula, not a single one predates the late 8th century bc.36
Turning now briefly to the bronze lamp found with the earlier burials in
Gruft 30, a fine parallel was excavated at Baba Jan in western Iran from an
intrusive horse burial. Its date of deposition cannot be earlier than the late
8th century and is most probably in the 7th.37 Here, then, we have another
important piece of evidence for a post-9th-century bc date for the Assur
tomb, one not considered by Sürenhagen. And Sürenhagen’s chronological
discussions for other objects in the tomb, e.g., comparing a knobbed bronze
bowl to a different, lobed, example from War Kabud, or dating neo-Assyrian
cylinder seals only to their upper limits, do not in any objective sense
indicate a 9th-century date for the tomb or the fibula.38
What of Sürenhagens conclusion that the tomb was closed in the 9th
century bc and how secure is it? I am forced to suggest from an examina-
tion of the finds that his reconstruction of the stratigraphy is not accept-
able and warrants review. I also believe that it will develop that “Es besteht
somit kein Widerspruch zwischen baugeschichtlichem und archäologisch-
ern Befund.”39
Gruft 30 and its contents have not altered our opinion based on empirical
documentation that the Near Eastern fibula is a Leitfossil, an object of fun-
damental importance for dating material juxtaposed in tombs and secure
stratigraphical levels to the late 8th century bc or later.
35 Ibid. 124–125, note 63. See vanden Berghe, op. cit. (in note 6) for the chronology of War
Kabud and other Luristan sites; this important paper is not cited by Sürenhagen.
36 For references see note 12, and especially Moorey, 1980 op. cit. (in note 4), fig. 14, no. 350.
37 Clare Goff Meade, “Excavations at Baba Jan 1967: Second Preliminary Report,” Iran 7
(1969) 123–126, fig. 7; on page 126, notes 29, 30, the Gruft 30 lamp and fibula are cited. I plan
to write elsewhere on the chronology of Baba Jan but will note here that I firmly hold the
view that the site did not come into existence before the 8th century bc (Period III), and that
Period II is not earlier than the 7th century bc. If correct, this view supports the dating of the
horse burial to the 7th century bc.
38 Op. cit. (in note 26) 125–126, 128.
39 Ibid. 128.
chapter twenty-eight
* This article originally appeared as “Parasols in the Ancient Near East,” Source 18, no. 1
(1999): 1–7.
1 See, for a review and illustrations, H.G. Fischer, “Sunshades of the Marketplace,” Jour-
nal of the Metropolitan Museum of Art 6 (1972):151–156; id., “Sonnenschirm,” in Lexikon der
Ägyptologie 5 (Wiesbaden: 1984).
2 The stele was recovered in Susa, Elam, whither it had been brought as booty in the
twelfth century bc; Anton Moortgat, The Art of Ancient Mesopotamia (London: Phaidon,
1969), pl. 125.
818 chapter twenty-eight
above the head of a seated king (Fig. 2).3 A body-length shaft shown with
upper struts supports the shade from one side, not from the center, and a
long, narrow extension curves down from the rear of the flat shade, enclos-
ing the attendant.4 Perhaps this flat shade reflects the Egyptian form. The
latter, however, has its shaft centered, not placed to one side, and there is no
long rear element.
3 Foroughi Collection. E. Porada, “Aspects of Elamite Art and Archaeology,” Expedition 13,
cle “Sonnenschirm,” in Reallexikon der Vorgeschichte 12 (Berlin: 1928), pp. 308–310, esp. 309;
to describe an Assyrian curved form of sunshade, see below. The curved unit seems to form
a canopy-baldachin over the king’s space.
parasols in the ancient near east 819
5 See M. Miller, “The Parasol: An Oriental Status-Symbol in Late Archaic and Classical
Athens,” Journal of Hellenic Studies 112 (1992):95 and pl. 1a, and bibliography in nn. 8–11.
820 chapter twenty-eight
6 For example, R.D. Barnett, Assyrian Sculpture in the British Museum (London: 1975), pls.
27, 35, 37, 45, 65; id., Assyrian Palace Reliefs (London: Batchworth Press Ltd., n.d.), pls. 141, 158,
159, 163; J. Reade, Assyrian Sculptures (Cambridge: Harvard University Press, 1983), figs. 44, 62,
98; J. Bär, Der assyrische Tribut und seine Darstellung (Neukirchen Verlag, 1996), p. 215, n. 1677,
Abb. 53, 54.
P. Calmeyer in Altiranische Bronzen der Sammlung Bröckelschen (Berlin: 1964), pp. 50f.,
Abb. 8, published as ancient Assyrian an unexcavated bronze vessel decorated with a frieze
of motley, allegedly Assyrian troops attacking a fortress. One of the figures casually car-
ries a parasol over his shoulder, another at an angle behind a human-yolked vehicle; no
king is manifestly represented. Such depictions are impossible compositions in known—
excavated(!)—Assyrian iconography. I suggest that the scene is surely a modern addition to
an ancient vessel, a modern forgery, not an ancient Assyrian scene. The corrosion should be
tested.
7 S. Herbordt, Neuassyrische Glyptik des 8.-7. Jh. v Chr., The Neo-Assyrian Text Corpus
but it seems to be a Sonnendach held above the king in the same manner as
a parasol.8
Assyrianizing scenes decorating late eighth-seventh-century-bc Phoeni-
cian metal vessels depict a king with a parasol, emulating Assyrian royal
practice.9 Absorption of canonical Assyrian iconography and ideology is
also evident on mid-seventh-century-bc Urartian royal seals depicting King
Rusa II attended by a parasol bearer.10 These Urartian parasols were either
plain or had tassels (or bells?) at the rim (like those from contemporary
Assyria). While no earlier Urartian representations are extant, it is proba-
ble that royal-parasol ideology was practiced there earlier.
A little-known representation of a parasol exists in an eighth-century-bc
North Syrian context; it is the only known example in the art of that area.11
A stone funerary stele from Marash depicts a facing seated couple. Rather
than held by an attendant, the parasol extending over the male figure’s head
is shown as if its shaft is a curved extension of the chair’s back—a unique
depiction. This Assyrianizing iconography, locally modified, surely informs
viewers that the male is a king (or member of the royal family?).
A wooden knob with the struts of a collapsible parasol was excavated
at Gordion, the Phrygian capital in central Anatolia,12 from a child’s tomb
(Tumulus P), dated to the late eighth century bc. We do not know whether
this example was locally made or came to Gordion as a gift from an Assyrian
king (Sargon II). Nor do we know the extent to which Phrygians employed
parasols since no Phrygian representations of them exist. Given, however,
the gender-determined function of the parasol, Tumulus P surely was the
burial place of a prince (or a prematurely deceased king?).
The sliding knob section with eight rectangular cuttings belonging to
a wooden parasol was excavated at the Hera temple on Samos, where it
had been dedicated (probably by a non-Greek) sometime in the seventh
century bc.13
8 Barnett, Assyrian Palace Reliefs, pls. 158, 162; Unger, p. 309. Note that the typical parasol
Stellerinden Üç Ömek ve Gözlemler,” in Light on Top of the Black Hill, Studies Presented to
Halet Çambel (Istanbul: 1998), pp. 243, 249, figs. 2b, 8.
12 R.S. Young et al., Three Great Early Tumuli (Philadelphia: University of Pennsylvania,
The evidence manifests that from its incipience, the ancient Near Eastern
parasol was gender- and socially oriented. The parasol and its bearer were
restricted to the king’s person, his royal prerogative. It was not in common
use socially; it was not used in religious ceremonies; it functioned solely as
a marker of the king’s rule and authority.14
This prerogative continued without change into the Achaemenian period
(late sixth-fourth centuries bc), and at Persepolis several kings, accom-
panied by a parasol attendant, are depicted in doorways (that is, mobile;
Fig. 4).15 When the king was portrayed enthroned (that is, static), however,
a baldachin was fixed in place and covered the full throne area.16 Probably
the baldachin, which had already been represented in some royal scenes
in Assyrian times (i.e., on the Balawat Gates),17 became associated with the
charged meaning of the parasol, developing into its fixed counterpart, one
that shaded, protected, and defined the space around the king, not merely
his body. Parts of an iron parasol are reported to have been excavated at
Persepolis but remain unpublished.18
In Lycia, at Trysa and Xanthos, parasols were depicted on grave reliefs;19
and at Kizilbel, a tomb painting depicts a parasol bearer on a ship, this
parasol held by an attendant facing the protected figure.20 In this western
region of the Achaemenian Empire, a local, not canonical, Achaemenian
concept obtained, for the representations are manifestly not those of kings,
but of high-ranked males, possibly from satrap families,21 and a funeral role is
Aspects,” in The Civilizations of the Aegean and Their Diffusion in Cyprus and the Eastern
Mediterranean, 2000–600 B.C., ed. V. Karageorghis (Larnaca: 1991), p. 131, pl. XXX: 2; and
Kyrieleis’s forthcoming article in Anadolu 23. See also Miller, 95, n. 24.
14 See U. Magen, Assyrische Königsdarstellungen Aspekte der Herrschaft, Baghdader For-
no parasol scenes represented on seals. Note that the use of a parasol by Persian kings is
another example of iconographic and ideological continuity from Mesopotamian culture of
the previous era.
16 Ibid., pls. 77–78, 104–107.
17 Barnett, Assyrian Palace Reliefs, pl. 142; see also n. 8, above.
18 Personal communication from Judith Lerner, from a third source. It is also mentioned
from Cyprus, J.L. Myres, Handbook of the Cesnola Collection of Antiquities from Cyprus (New
York: 1914), pp. 228, 230, pl. 1365A, which seems Assyrianizing.
parasols in the ancient near east 823
22 For a good summary of the classical evidence of parasol function, see Miller (n. 5,
above).
Section Three
Abstract: The paper contrasts the professional activities of Professor Ezat Negah-
ban with that of Arthur Upham Pope. Beginning as a young man and continuing
throughout his career as a field archaeologist, Negahban combated the plundering
of archaeological sites in Iran, which he knew resulted in the destruction of ancient
Iranian—and the world’s—cultural history. In this endeavor, he was opposed by
many forces, dealers, collectors, members of the royal family, and even alleged schol-
ars, among them Roman Ghirshman. But he persevered.
One of Negahban’s opponents was Pope, an extraordinary, and complex individual
who played a large role in bringing forth information about ancient and Islamic Iran
through his involvement in exhibitions, institutions, and publishing. Paralleling
these activities, however, was that of Pope the dealer in Iranian antiquities, a role
which he deliberately kept secret from scholars. It is argued that selling antiquities
was the main force in his life, that it informed and manipulated his other, better
known, activities. The paper concludes that the beliefs and activities of the two
men represent the two opposing forces confronting archaeology today, excavation
or destruction.
“In addressing Negahban you walked right into the tiger’s den. He is a bitter
fanatic playing the artificial role of protecting, for the Persian people, their
rightful heritage of Iranian antiquities of which they are being shamefully
robbed by the antiquarians (all criminal thieves) and their criminal accom-
plices, the greedy and unscrupulous museum directors. He had prepared a
rather violent resolution to be passed, he hoped, by the congress … Ghirsh-
man and I voted against it and … 300 sheep voted for it.”
This animadversion was written by Arthur Upham Pope to one of his dis-
ciples, Rexford Stead (Gluck and Siver 1996: 537). The congress mentioned
was the Vth International Congress of Iranian Art and Archaeology meeting
in Iran in 1968, the year the letter was written, and the “violent resolution”
was in fact a moderate appeal: given “the growth of clandestine and ille-
gal traffic in antiquities” which “destroys both the cultural and historical
* This chapter originally appeared as “The Pope and the Bitter Fanatic,” in The Iranian
World: Essays on Iranian Art and Archaeology Presented to Ezat O. Negahban, eds. A. Alizadeh,
Y. Majidzadeh, and S. Shahmirzadi (Teheran: Iran University Press, 1999), 5–12.
828 chapter twenty-nine
heritage of mankind,” the congress appeals “to all countries and UNESCO
to devise and adopt effective ways and means to prevent the export from
and the import to one country of antiquities from another except through
official and authorized channels.”
Pope’s sarcastic evaluation of Negahban and his disdain for the purpose
of the congress resolution, which Negahban played a large role in formulat-
ing,1 immediately set in my mind the theme for the Festschrift article I had
been seeking—the vital difference between the work and goals of Negahban
and those of Pope. One could not think of two more appropriate individuals
whose achievements better epitomize the two opposed systems that prevail
in the archaeological world at large. While one worked all his life to save
the past for the future, to preserve the planet’s history, the other incessantly
worked to destroy it, to reduce it to exhibitions of “antiquities” on museum
and collector’s shelves, and to ridicule those who opposed these objectives.
The great virtue of Negahban was that he took his role as archaeologist seri-
ously, he respected archaeology’s needs and protected the artifacts within
its care. Early in his career he recognized that he was engaged in a battle
with powerful forces, and it was a battle he has fought all his life, in the posi-
tions he has held, and in his writings and lectures; and Pope wasn’t his only
enemy, others included archaeologists, such as Roman Ghirshman. Negah-
ban was one of the first archaeologists I had ever encountered who spoke
openly about the tragedy wrought by plunder and the role of museums and
collectors in this activity (not once in all my years as a student did anyone
ever mention, let alone explain, how objects reached museums and collec-
tions). And, to speak personally and appropriately, it was a lecture of his that
first turned me on to the problem (this and a lecture by Machteld Mellink
and the early—and isolated—writings of Clemency Coggins helped me to
understand the nature of collecting).
To this day many students and colleagues remain unaware of Negahban’s
other important contribution to archaeology, that he was not only the exca-
vator of Marlik and other sites. With this in mind, I write this article as a
dedication to Negahban as an expression of my respect for his fighting the
1 In the Vth International Congress of Iranian Art & Archaeology, Teheran, 1968: xxxiv,
Negahban claimed that there were “only two negative votes” on the resolution; he refrained
from mentioning names. Pope says it was he and Ghirshman. This is contradicted by Gluck
who proudly informs us that he too, along with another individual, also voted against the
resolution (534, 540). What editor Gluck never informs the reader, thereby paying the best
compliment to his mentor by imitating him, is that he is, in addition to being a publisher, an
antiquities dealer operating from Japan.
the pope and the bitter fanatic 829
good fight—and for being a decent man. I trust the offering will alert stu-
dents (I am not optimistic about many of their teachers) to the two opposing
forces exemplified by Negahban and Pope. Alas, they should realize that at
present the Negahbans are on the losing side.
Negahban’s concern on these matters began to generate in the 1950s
before he excavated at Marlik (Negahban 1996: 5 f.), and accelerated there-
after, for in all his subsequent international lectures he never ceased to
contrast archaeology’s work with that of its enemies. He went to Marlik in
the first place because he became aware that plundering was in progress
there, that plunderers, dealers, and smugglers were active. It turned out that
there was to be only one campaign at Marlik. For, after commencing exca-
vation he was forced to remain at the site for fourteen continuous months
(1961–1962) in order to save the site from destruction. This perseverance
was unprecedented, and I suggest it was heroic. From day one, he had to
confront intrigues by dealers, corrupt officials, and members of the royal
family, who collectively attempted to terminate his work so that they could
begin their work (which would not have taken fourteen months, merely a
few days). Only an appeal to the Shah allowed him to continue. Few schol-
ars are aware that he suffered a further, terrifying, ordeal, that his team was
physically attacked and his camp destroyed, and that soldiers were sent to
protect him—from the Popes of the world. As he succinctly noted “… we
had to be continually on the alert to protect the excavation and even our
own lives” (Negahban 1996: 10). Here we have a real life drama—with none
of the protagonists playing an “artificial role”—enacted by the two oppos-
ing forces; the performance is a paradigm of archaeology versus the plun-
derer/collectors. Negahban has written a professional and personal autobi-
ography, which, inshallah, we will soon read.
Arthur Upham Pope is a name known to all scholars of Iranian art. He
played a role in the organization of several international congresses on Ira-
nian art and archaeology, in Philadelphia (1926), London (1931), Leningrad
(1935), New York (1940), Iran (1968). He traveled to Iran many times, record-
ing countless records and photographs of Iranian monuments. He was the
founder (1928) of the Asia Institute (later adopting different names), which
had a faculty of several European refugees, and which sponsored archaeo-
logical excavations. He was an “advisor” (for purchasing antiquities) to sev-
eral museums. And, his best and most lasting work, he organized, edited,
and brought to fruition the first major publication of Iranian art ever pro-
duced in the west, The Survey of Persian Art (SPA 1938–1939) in six large
volumes. He was beloved by the Iranian royal family for his love of Iran
and for his contributions to disseminating knowledge of its rich heritage. In
830 chapter twenty-nine
1966, the royal family gave him and his wife a home in Shiraz, and after their
deaths (1969; 1977) a mausoleum and a park named for him in Esphahan.
This summary does not do justice to the activities and life of this truly
extraordinary, complex individual. To a large degree the full range of the
activities and life of Pope emerge in a biography published in 1996, Surveyors
of Persian Art, edited and compiled by Jay Gluck and Noël Siver. A large vol-
ume of 658 pages, it is a collection of, inter alia, selections from biographical
and anecdotal records, and selected letters, many written by or to Pope and
his wife, Phyllis Ackerman. Two prominent collaborators in the venture are
Robert Payne and Rexford Stead, both of whom were cooperating on writ-
ing a biography of Pope, but died before its completion; included are also
extracts from Pope’s unfinished autobiography which he called Nine Lines.
The collaborative biography is a labor of love, assembled and written
by Pope’s students and friends, all of whom “loved, no, adored him” (xix)
utterly—a sentiment in evidence throughout. This love compelled the pub-
lication, it weighs heavily on all the collaborators’ and editors’ assessments
of Pope’s deeds and attainments, and determined many of their positive
evaluations that is often expressed against the evidence they themselves
present; many critical evaluations of Pope are perceived as an “attack.”
Nonetheless, there remains a positive, indeed, an honest, quality, which
enriches the value of the work considerably, inasmuch as the editors have
chosen (after much wrestling with themselves, i.e., 574 f.), to give us the warts
too. This openness now allows us to discern that, far more intensely and for
a much longer time than hitherto realized, Pope was involved in the antiq-
uities market. He was in fact an active and aggressive antiquities dealer.
And the published evidence is there to demonstrate that this role was the
core of his life, was the motivating factor that informed most of his other,
apparently scholarly, activities. To appreciate the full thrust of this charge,
one must read all the published evidence; here I present appropriate high-
lights.
Like Negahban, Pope began his life-long career early. The earliest evi-
dence of Pope as a dealer seems to be 1908, when he began to sell rugs in
Providence, Rhode Island (49); he still listed himself as a rug dealer as late
as 1921 (as seen on a letterhead: personal knowledge). That he (along with
his wife) was dealing in antiquities at least from the mid-1920s is manifested
from Pope’s correspondence (154ff., in a chapter labeled “Some Correspon-
dence on Dealing”—the “some” indicating that only a sampling of the avail-
able information is presented). Included are letters written in the 1920s and
1930s about sales, prices, “professional fees,” etc., to the dealer Rabenou, with
whom he maintained a life-long business relationship—and whose family
the pope and the bitter fanatic 831
called him Uncle Arthur: (155–156, 170, 179f., 204, 207), and to many museum
directors, to whom he invariably bragged that his prices were considerably
cheaper than those of other dealers (adjusted to today’s prices, his figures
would be multiplied considerably). Mentioned inter alia is his “archaeolog-
ical collection” (159—written by Ackerman); that he has “purchased over
three-quarters of a million dollars worth of Persian things” (167); that “I have
… supplied more than $800,000 worth to more than twenty museums” (413);
and a boast that “… I have contacts with eighty percent of the sources of Per-
sian art in Persia …”
In Nine Lives, Pope lists as one of his accomplishments that he was a
“Purveyor of Persian art2 to several famous collectors and finally exclusively
to nearly a score of museums in both this country and abroad.” Gluck (xv)
reports that from their Paris home Pope and Ackerman “did art business
…” (notice the editorial euphemism slipped in here and elsewhere in the
biography—Pope was in the art business, not the antiquities business); and
Stead reports (583) that one of Pope’s aides told him that Pope “purchased
objects every trip …” he made to Iran. In 1968, just before he died, the Tehran
Journal reported that he was advising a young college student, “one of my
[Pope] very newest students,” how to purchase antiquities, and he called her
attention to one particular Tehran shop where he had “purchased more than
a quarter million dollars worth of things …” (527).
While I had known for some time that Pope was a clandestine antiquities
dealer (Muscarella 1979: 5f., 13), the newly released information fleshes out
details of this activity significantly and blatantly.
Pope of course never revealed his role as a dealer in his publications and
within his institutional activities. But there was explicit public notice that he
and his wife Phyllis Ackerman owned antiquities, for they identified them-
selves (sometimes using one of their names, sometimes both) as owners of
a number they illustrated and recorded in allegedly disinterested scholarly
publications they controlled: SPA Vol. 1: Figs. 38, 65, 71; Vol. 7: Pls. 10, 33, 41, 59,
60, 62, 74, 111, 124, 192, 255, 256; in the 1931 London Catalogue of the Interna-
tional Exhibition of Persian Art: pp. 2–3, 8, 12, 16–17; in Ackerman’s 1940 Guide
to the Exhibition of Persian Art: listing on page i, etc.; and in Pope’s Master-
pieces of Persian Art, London, 1945, Pls. 6–7, 27; and let it be noted that many
of the antiquities he published in all these publications had already been
sold by Pope (see also Illustrated London News (ILN) below). There is also at
2 Which title would have been most appropriate for the biography: did the editors not
realize that the title they chose would inevitably be smiled at?
832 chapter twenty-nine
least one published scholarly record that Pope sold antiquities to a museum,
here the University Museum, University of Pennsylvania (Legrain 1934: 8,
19); we can now cite the letters on this transaction published in the biog-
raphy (157f.). Worth relating in this context is that although Rexford Stead
acknowledges in a 1982 letter discussing my 1979 paper (169) that he knew
the SPA included “objects … in dealer possession and … a great many others
owned by collectors who probably obtained them through Arthur when, in
the 1920s, he was actively involved in selling,” he does not mention, what he
must have known, that one of the dealers was Pope.
Just as he manipulated readers in his major publications, Pope mani-
pulated—brilliantly, one is inclined to observe—the ILN. The editors and
the magazine’s readers, which included students and scholars, believed that
Pope was a disinterested scholar publishing recently “discovered” archaeo-
logical material (viz. ILN 5/6/39: 790, where the editor writes that Pope is
“known … for his discoveries of ancient Iranian bronzes”). In fact, he was
publishing objects being offered for sale, or that he had sold, thus using
the ILN as a sales catalogue (see Pope’s ILN bibliography on pp. 613 f.; Mus-
carella 1979: 13). A clue to this activity exists in ILN 9/13/30:444, lower right,
where he specifically identifies objects illustrated as “in the collection of
Mr. A. Upham Pope.” A 1930 letter published in the biography (160) reveals
Pope bragging to the Boston Museum that a number of objects he sold them
will be illustrated in a forthcoming ILN—just part of his sales service. These
camouflaged sales catalogues were accorded authority by Pope’s presenting
himself as Director of the International Exhibition of Persian Art (viz. ILN
9/6/30: 292; 9/6/30: 388; 9/13/30: 444), and as Director of his Institute (ILN
5/31/41: 718; 3/1/41: 292).
Pope further used the magazine to enlighten us on archaeological activ-
ity in western Iran, to share his professional knowledge concerning what
archaeology and its techniques were. It is the equivalent of a semester’s
education to compare Pope’s straight faced and casual description of how
Lurestan bronzes were “discovered,” how he reported what in reality was
the plundering and total destruction of the Lurestan culture that he was
financing, with Negahban’s fourteen months work at Marlik and the subse-
quent activities he observed after he left the area. In the ILN of 10/22/32: 613
Pope introduces us to a Mr. Rahim, who with his brother was “closely associ-
ated with the finding of the bronzes …” The brother spent “ten consecutive
months, collecting bronzes …” Rahim also travelled to Piravand, “where a
great quantity of fine and typical Lurestan metal was found, about a thou-
sand pieces in less than two months.” Rahim was also “present at the opening
of many graves …” When Nwgahban (1996: 11) returned to the Marlik area
the pope and the bitter fanatic 833
after his excavations terminated, he counted “more than 2000” ditches cut
by the plunderers; and that as many as “400 persons had been employed in
this clandestine digging, with the local gendarmes receiving a bribe of 5 rials
per worker to look the other way.”
Pope used his close involvement in the international exhibitions he spon-
sored to sell his wares. This was already indicated by the illustration of
objects he owned in the publications of those exhibitions mentioned above.
More explicit information that these expositions were used by Pope as a
venue for his sales is now clearly divulged in the biography. In a letter to
Robert Payne in 1982 Rexford Stead wrote: “Much of what was shown at the
Exposition [Philadelphia 1926] was actually for sale … Much of the objects
on view had been lent by dealers, and even AUP and PA had works for sale
…” (577). In ano ther letter to Payne (578f.) written the same year, he states
that Pope arranged in 1926 on his own authority to organize a parallel expo-
sition at the Pennsylvania Museum (later the Philadelphia Museum of Art)
because Pope questioned whether “the Exposition was the best vehicle for
such a presentation, whether it would attract the wealthy collectors AUP
wanted to reach.”3
In a letter to a Museum Director (163) written in 1932, Pope offers for
sale an object that “was one of the knockouts of the show”—the show was
the 1931 London Exhibition. Pope also slipped in the sales pitch that the
Hermitage was negotiating for the piece, so this museum director had better
act fast to purchase. Pope himself, in his biography, refers to his exhibiting
in the 1931 Exhibition his own wares and those he had just sold (189).
As for Pope’s manipulation of his Institute to serve his antiquity dealing,
one need only refer to his use of his title Director of the Institute in his
ILN catalogues (above), and to what amounts to a confession by Stead from
a 1982 letter already cited above (169): “Although his Institute association
doubtless [sic] enhanced his later [not earlier?] sales, I’m confident that
3 In fact, there were individuals involved with the 1931 Exhibition, who charged that Pope
controlled the committee selecting objects for exhibition, that he was selling antiquities as
well as acting for dealers, and that he knowingly sold forgeries. A committee held hearings,
which disbanded when accusers refused to press their charges in public. Pope was exoner-
ated: see the chapter “The ‘Trial’ ”, pp. 205 ff. Payne (206) exclaims in his unfinished biography
“… Arthur was handsomely vindicated.”
Although Pope was often accused of selling forgeries, I do not discuss that issue here
because I have no objective information that he did so knowingly. He certainly defended
himself openly and clandestinely against the charge (Muscarella 1979: 5f.). There is no doubt
that he sold forgeries, including the well-known Andarz Nameh manuscript.
834 chapter twenty-nine
he never used the nonprofit Institute illegally, yet he did personally get
involved, and this presents an ethical problem.” No more need be said on
this issue.
Although the contributors to the biography often defend and rationalize
Pope “the art dealer” as one who only took commissions (see 19, 154 ff., 169,
206f., 216, 316, 574, 577), they deserve praise for sharing with us contrary
evidence as well as their misgivings about the man, some of which we have
already seen. Some years before 1981 Payne is quoted by a third party to have
said of Pope “What a magnificent charlatan” (408).4 Stead wrote to Payne
in 1979 “… is the picture [of Pope’s accomplishments] compromised by his
dealer aspect? This is troubling my conscience” (574). Three years later he
wrote to Payne a cri de coeur (579): “Robert, I’m beginning to think that
all of us who were drawn to Arthur had some personality gaps or quirks
or weaknesses. Perhaps AUP sought out such people! We made him into a
father figure, possibly out of need.” This says much about Pope, his devotees,
and the impulses behind the present biography.
Gluck (himself an antiquities dealer) and Stead feebly raise the issue of
anachronism, judging Pope by today’s standards, Pope the “scholar” who
long ago also in normative fashion sold antiquities (xix, 584; as for Payne,
he was fully blind to archaeological research and goals). This ploy is disin-
genuous and a feint. For it is not so much that Pope was a dealer, like others,
such as E. Herzfeld (Ghirshman is mentioned in this role, 584), but the piv-
otal fact that selling antiquities was the force that drove his career, that his
public scholarly activities, Institute, Exhibitions, and publications were wil-
fully employed to screen and serve this dealing. Further, that he was both
aware of and utterly indifferent to the destruction he caused, transmogri-
fying what he knew was going on in Iran, what his dealing entailed, into
archaeological activity, is evident by his covering and protecting these facets
of his life from the scholarly community, and his invective against those who
suspected or knew. There is nothing anachronistic about reacting to this
abhorrent behavior, about removing the transparent fig leaf.
An overview of the work of Gluck and Siver have neatly allowed us to
place Pope’s negative evaluation of Negahban into proper perspective: a
Purveyor of Persian Art was defending his bazaar shop, unaware that at the
same time he was also bestowing a badge of honor.
4 The editors also include other negative comments received in their biographical re-
search: see pp. 170, 541, 585; Ackerman’s nephew thought Pope was “a congenital liar.”
the pope and the bitter fanatic 835
Bibliography
Gluck, J. and Noël Siver 1996: Surveyors of Persian Art California, Mazda Publishers.
Legrain, L. 1934: Luristan Bronzes in the University Museum, Philadelphia, University
Museum.
Muscarella, Oscar White 1979: Unexcavated Objects and Ancient Near Art: Addenda,
Malibu, Undena.
Neghaban, Ezat 1996: Marlik: The Complete Excavation Report, Philadelphia, Univer-
sity Museum.
chapter thirty
* This chapter originally appeared as “The Antiquities Trade and the Destruction of
Ancient Near Eastern Cultures,” in A Companion to the Archaeology of the Ancient Near East,
ed. D.T. Pott (Oxford: Wiley-Blackwell, 2012), 107–124.
838 chapter thirty
The plunder of sites encased within a mound (Persian tepe, Turkish hö-
yük, Arabic tell) formed by successive settlement constructions, and burials
and tombs, has a long history. The archaeological record reveals that the
practice occurred throughout antiquity. Numerous tombs in Egypt were
plundered millennia ago, the most spectacular being that of Tutankhamun,
which was looted after the king’s burial but soon thereafter resealed. Royal
tombs built within a contemporary, inhabited palace at Nimrud (Iraq) were
partially plundered and then repaired while the site was still inhabited.
And numerous tombs buried under mounds of earth, called tumuli, and
visible to all, were totally or partially destroyed in antiquity and thereafter.
Examples include Pazyryk in the Altai (but much was recovered), Sé Girdan
in northwestern Iran, Kerkenes Dağ in Anatolia, where scores of tumuli
there have been obliterated, and the Sardis area in western Turkey, where
90 percent of the tumuli have been plundered (Luke and Kersel 2006: 185–
186; Roosevelt and Luke 2006: 173–187).
The prevalence of ancient tomb plundering across the centuries within
Near Eastern cultural regions is unknown. But a good number of burial sites
have been excavated in modern times. Examples include tombs at Nimrud,
Kish, and Ur (Iraq); Umm-el Marra (Syria); Alaça Höyük and Arslantepe
(Turkey); Susa, Hasanlu, Dinkha Tepe, Marlik, and many in Luristan (Iran);
and Tillya Tepe (Afghanistan), where two burials were looted but six were
excavated, containing thousands of artifacts of gold, silver, and ivory (Hie-
bert and Cambon 2008: 210–293). Artifact contexts of undisturbed burials
are not merely of inestimable value for knowledge of the ancient cultures
involved; in some cases, they are our only source of cultural data. They
also vividly inform us of the information forever lost from the countless
plundered tombs.
Ancient plundering was presumably conducted both as desecration and
to acquire loot. Looting is the basis for all current plundering, evidenced by
the vast number of destroyed cemeteries throughout the Near East. These
activities increased in the 19th century, a result of the renewed interest in
antiquity and fueled by a fulfillment of social ambitions exemplified by the
increased collecting of antiquities by museums and private collectors every-
where (Meyer 1973: 46–47, 191–197). Egypt, the Holy Land, Greece, Cyprus,
and Italy were the primary victims, never vacating that status. In the 1870s
plundering occurred somewhere near the wide-ranging Oxus river, and a
quantity of gold and silver objects (including modern forgeries) labeled the
“Oxus Treasure” was acquired by the British Museum. But no archaeologist
can identify its find-spot(s) (Muscarella 2003a). In the late 19th century Luigi
Palma di Cesnola looted countless sites in Cyprus. He sold thousands of
the destruction of ancient near eastern cultures 839
purchased them, and this continued for years as more “Ziwiye” material
surfaced. A number of the bronze, gold, and silver objects had been cut
into pieces and partitioned among the plunderers, an action resulting in
scores of fragments sold all over the world—a common practice (Hiebert
and Cambon 2008: 67–79). The partition required years of work by scholars
to sort out and match the scattered fragments. Moreover, it was impossible
to know how many of the hundreds of artifacts purported to have come from
Ziwiye were actually recovered there or in fact came from elsewhere (other
sites, e.g. Qaplantu, have been proffered by dealers). Excavations at Ziwiye
by American and Iranian archaeologists recovered not a single comparable
artifact: but an historically important Urartian 7th century bc seal was
excavated there. The Ziwiye episode epitomizes the utter destruction of
a complex polity’s integrity and culture, and led to increased plundering
across Iran. Thus, following excavation in the southwest Caspian region at
Marlik, sites in the area were subsequently attacked. The Kalmakarra Cave
in Luristan in the early 1990s yielded scores of hitherto stylistically unknown
artifacts that have surfaced in the antiquities market (Muscarella 2000b: 30
n. 6). Another Iranian polity destroyed is that of the Sasanian kingdom in the
south. Hundreds of gold and silver artifacts labeled Sasanian have surfaced
over the years, many of which are genuine, but far more of those exhibited
in museums or collectors’ homes are modern forgeries (Muscarella 2000a:
203–204, 528–535 nn. 68–70). Only some Sasanian sites in Iran and Iraq and
rock carvings have survived.
Plundering in Turkey also increased in the 1930s, accelerating after World
War II (Meyer 1973: 57–64). The earlier examples include Açemhüyük, Ho-
roztepe, and, allegedly, “Ordu” (Muscarella 1988: 394–411). More recent cases
include Perge (Rose and Acar 1996: 77–78; Renfrew 2000: 32–34); possible
Hittite royal tombs, not one of which has ever been recovered; Elmali (Grae-
pler and Mazzei 1994: 92; Rose and Acar 1996: 80–82), and the dynamiting-
mutilation of Phrygian rock-cut façades north of Afyon (which I have seen).
The 1965 destruction of several tumuli at Uşak, east of Sardis, is a woeful
example of tumuli plunder (Özgen and Öztürk 1996; Greenfield 2007: 420–
423; Waxman 2008: 135–137, 144–154). Here, hundreds of Greek, Lydian, and
Achaemenian artifacts, including painted frescoes torn from tomb cham-
ber walls, were soon thereafter purchased by the MMA, fully knowing their
geographical origin. The museum immediately assumed the role of their
“guardian” by hiding them in a storeroom for decades (I was secretly allowed
to see them). A small number were revealed in 1970, all deceitfully labeled
as Greek antiquities, a number of which were not. Not until 1993 did all
become public, when, after years of legal costs of millions of US taxpayers’
the destruction of ancient near eastern cultures 841
distant foreign ports, where dealers might employ schemes that camouflage
the original derivation, thereby allowing them to claim that the antiquities
belonged to a collector in Australia or Hong Kong, etc.—in other words, a
forged provenance.
Clarification of the expressions plunder/looting, theft, illicit, and “spoils
of war” are essential, inasmuch as each may reflect a different episode and
background or a combination of more than one. Determination of illicit
acts is but a legal component of the plunder problem, one of the many
concerns involved (Bator 1982). To determine if a past or contemporary
acquisition involved legal or illegal acts, each requiring distinct courses of
action, all antiquities must be evaluated with regard to their acquisition
histories. A paradigmatic example, although not per se a Near Eastern mat-
ter, is determining which terminology is suitable in the ongoing discussions
between England and Greece regarding the repatriation of the “Elgin Mar-
bles” removed from the Parthenon in 1801 by Lord Elgin’s agent Philip Hunt,
and purchased from Elgin by the British Museum in 1816 (thereafter, con-
tinuous removal of sculptures occurred, acquired by European museums;
see Waxman 2008: 81). A number of metopes and sculptures were taken,
based, it was argued, on the firman, a government authorization. Elgin’s
role in acquiring the marbles, and past and ongoing negotiations for their
return, are fully documented by St Clair (1998) and Greenfield (2007: 41–96).
Based on contemporary standards at the time of the removal, the issue is
partially one of looting, given that unauthorized removal was involved, and
bribes were paid to local workers, and threats were used. What remains to be
resolved legally (leaving aside national and historical issues) is the distinc-
tion between the metopes illegally removed, contra the firman, and those
legally removed. This action is an issue of modern law and is of equal con-
cern to archaeologists and the public. The objects that Elgin removed legally
do not come into the category of plunder.
The modern, legal ownership of the Rosetta Stone, discovered in Egypt
by Napoleon, is an example germane to spoils of war issues, not to those of
plunder or theft. It was captured from the French by the British and taken
to England, and then given to the British Museum. Egypt has demanded its
return. In another example, Egypt’s demand for the return of the head of
Nefertiti, now in Berlin, is different. This was certainly acquired by theft and
thus constitutes a modern legal matter: the head was stolen from Amarna
by the German archaeologist Ludwig Borchardt, who excavated it in 1912.
Borchardt deceptively hid the head in a crate beneath layers of sherds for
export to Berlin, subverting the legal division regulations for excavated arti-
facts. The Nefertiti head was kidnapped and smuggled into Germany: an act
the destruction of ancient near eastern cultures 845
deceptive provenance by claiming that their antiquity derived from “an old
private collection” recently discovered in a basement in Italy or Germany, or
derived from a “noble European family” or from the “Collection of Monsieur
R” (Atwood 2004; Simpson 2005: 29–30, 32; Muscarella 2007: 610; Christie’s,
London, 10/25/07: 83). To document a purportedly old provenance, dealers
will supply forged letters, eagerly embraced by their customers, as documen-
tation that the purchase was legitimate (Atwood 2004: 84). A prominent
example is the purchase by the Getty Museum of a Greek kouros in 1984,
for which acquisition a letter dated 1952 was presented to confirm that it
had belonged to a Swiss collector for decades, therefore acquired “in good
faith”. Some years later, the museum’s director announced that he had only
recently discovered that the letter’s envelope had an anachronistic postal
zip code and was a forgery (Elia 1997: 95; Lapatin 2000b: 43–53; Renfrew
2000: 41); some scholars think the kouros itself is a forgery.
Dealers often cite an earlier auction sale as a provenance for their antiq-
uity, which is merely a record of yet another modern provenance. They
ship their antiquities for sale to foreign auction houses, enabling a pur-
chaser to claim a provenance in France or England. Dealers also utilize
auctions to sell their merchandise anonymously, especially when they sus-
pect it is a forgery. Another antiquity-selling market is the internet—in par-
ticular, eBay—where, alongside genuine artifacts, forgeries are offered for
sale (Stanish 2008). Countless postings offer objects alleged to derive from
Africa, Mexico, and South America (Kelker and Bruhns 2010). Such behavior
is classic “bazaar archaeology” (Muscarella 1995; 1999; 2003a: 264–265; 2006:
151–152, 157, 161–165).
It is evident that museums worldwide have been and remain the fore-
most purchasers of plundered antiquities (Koczka 1989: 192–193; Muscarella
2000a: 23–25; 2007: 611–612). Curators, some of whom are archaeologists,
initiate their museum acquisitions, seeking out and proposing purchases
(Muscarella 1974; 2007: 612–613; 2009a: 400–401; Cook 1995: 181, 185; Grae-
pler 2004), but ultimately directors and trustees make the final purchase
decisions. Unknown to most scholars and the public is that they make pur-
chases (and accept donations) knowing that they were plundered and smug-
gled abroad, an activity rarely reported in the press (for rare examples, see
E. Wyatt, The New York Times, 1/26/08: 1, 13; 1/30/08: All). Trustees include not
only wealthy and powerful citizens but also national and local government
officials and owners of important newspapers, all functioning in conflict-of-
interest roles (Silver 2006: 3; Muscarella 2009a: 399; 2009b: 7, 11–12). Some
trustees collect antiquities, in part for eventual tax-deductible donations to
their museums (Nagin 1986: 24; Renfrew 2000: 27–35; Atwood 2004: 141–142;
the destruction of ancient near eastern cultures 847
Silver 2006: 1; Wald 2008; Muscarella in press b). Private collectors are also
wealthy individuals of social importance, exemplifying these roles by their
purchases. These are exhibited in, or donated to, museums, for which they
have galleries named after them, and receive tax-deduction benefits based
on the alleged increase in value since the original purchase (Brodie and Ren-
frew 2005: 353–356; Silver 2006: 1,7–8; Greenfield 2007: 259). And noteworthy
is the fact that it is self-serving antiquities dealers who furnish the museum
appraisals. Collectors are cited by dealers and museum personnel as “promi-
nent” or “serious” (read “serial”) collectors, as having a “lust” or passion for
art, thus revealing their infatuation (Graepler and Mazzei 1994: 81–84,87;
Muscarella 2000a: 9, 11–13, 23 n. 5).
Consequently, pivotal to comprehending the nature of the plunder cul-
ture is full awareness that, worldwide, museums and private collectors are
the financers and sponsors, the beginning of the long chain of the process
(Muscarella 1974; Elia 1994: 20; Brodie and Renfrew: 2005: 349). An Iraqi
official addressing the value of antiquities succinctly articulated this in the
following words: “For me, for you, it is all priceless, but for them [the plun-
derers] it is useless if they can’t sell it in the market” (S.L. Myers, The New
York Times, 06/26/02: 6).
Museums and collectors identify themselves as protectors of the world’s
culture, stewards of antiquities, “Guardians of the Past,” fulfilling a “public
responsibility to collect” (Muscarella 2000a: 1–29; Renfrew 2000: 30; Macken-
zie 2005: 158–162; for an accurate elucidation of stewardship, see Lynott
and Wylie 2000: 35–39). Curators will lie about the actually known site of
their museum’s antiquities, sabotaging archaeology, as well as mocking the
museum’s educational mission (e.g., the MMA’s Uşak purchase: see above).
To justify their deeds, museums and collectors identify plundered countries
in classic imperialistic language as “source nations” (Cuno 2008: 89); they
proclaim that plundered artifacts are merely the “self-proclaimed cultural
property” of these nations’ chauvinistic, nationalistic, indeed “racist” atti-
tudes (Cuno 2008: xxxii–xxxv, 13–15, 26, 124; Waxman 2008: 176). They insist
that antiquities are not the property of anyone nation, and to state other-
wise is a political construction, for they are the common property of a world
society, composed of “encyclopaedic” (read “non-Near Eastern”) museums
(Cuno 2008: 129, 139; Muscarella in press a); and plunderers (a word never
used by Cuno), sellers, and buyers are engaged in normal, licit business and
positive cultural transactions. Underlining this decree, which consistently
refrains from discussing how antiquities are obtained, is that once muse-
ums and collectors have acquired property from a source nation, ipso facto
it becomes their legal, non-racist, non-nationalistic property (Muscarella
848 chapter thirty
Lynott and Wylie 2000; Brodie et al. 2001, 2006). A non-professional, lay orga-
nization in the United States, Saving Antiquities For Everyone (SAFE), has
been vigorously active in exposing and fighting the plunder culture, reveal-
ing its existence everywhere and naming those responsible.
Although under-appreciated by most scholars (and thus their students
too), the connection between the antiquities trade and the existence of forg-
eries is obvious. As the archaeologist E. Unger succinctly stated, “As long
as there are people who collect antiquities, there will also be people who
forge antiquities” (Muscarella 2000a: 12). Forgeries are created to be sold as
ancient antiquities. Virtually all museums and collectors worldwide have
purchased and labeled forgeries as ancient antiquities; several museums in
the United States possess nothing but forgeries. Archaeologists incorporate
forgeries into their courses, innocently integrating them into their teachings
of ancient cultures. Museum-employed curators and archaeologists exhibit
and publish forgeries, sometimes knowingly, obeying museum orders for
fear of offending rich collectors or their colleagues (Muscarella 1980: 117
n. 3; 2000a: 4–5, 13–14, 37–39). Archaeologists also, innocently or not, pub-
lish forgeries as ancient artifacts. A prominent example is the archaeolo-
gist Roman Ghirshman, whose many publications, both monographs and
exhibition catalogues, are standard scholarly texts. He was the most prolific
publisher of Iranian forgeries, baptizing them as ancient productions and
providing them with forged (by him) proveniences; many of them were in
the antiquities market (Muscarella 1977b: 164 n. 42a, 182 n. 83; 2000a: 28 n. 11,
34, 205 nn. 2&4). On one occasion, Ghirshman published details of antiqui-
ties he asserted had been “recently discovered” in Iran in the respected Illus-
trated London News (ILN, April 2, 1960: 550), then an important outlet for the
publication of actual archaeological discoveries. An unnamed archaeologist
(Ghirshman?) employed the same tactic in the same venue (December 1967:
54–55) concerning two antiquities, alleging them to have been recovered
together with Parthian coins in Iran. On both occasions all the antiquities
presented were forgeries, their provenance a forger’s workshop (Muscarella
2000a: 8). Since Ghirshman had published reports on his excavations in the
ILN, the ILN assumed this ruse was but another such report. Arthur Upham
Pope also deployed the ILN to sell his antiquities (Muscarella 2000a: 210
n. 38).
The detection of forgeries takes years of studying excavated artifacts, their
specific styles and motifs, as well as the structuring technologies and mate-
rials employed by individual cultures, i.e. to employ connoisseurship (a
word now condemned by some). Connoisseurship, like all heuristic inves-
tigations, is fallible but is absolutely essential for the study of artifacts and
852 chapter thirty
antiquities, with the caveat that, aside from scholarly mistakes and igno-
rance, it has its manipulators (Muscarella 1977b: 165–169 n. 68; 1980: 118–119;
Löw 1993: 39–41; Simpson 2005: 28–34; Grann 2010). While archaeologists are
becoming capable of recognizing forgeries, some (most?) who are anthropo-
logically trained are not. Brazenly rejecting these skills, they assert, “Archae-
ology is anthropology or it is nothing” and they scorn traditional archae-
ologists as “object oriented self serving … antiquarians” (Muscarella 2000a:
10–11; Wylie 2000: 139, 144). Accordingly, those who proclaim this off-the-wall
dogma theory ignore stylistic evaluations of artifacts they encounter (except
pottery), and lack both knowledge and interest in evaluating forgeries.
Some forgeries exhibit good workmanship and artistic skills (e.g., Wax-
man 2008: 153–162), others reveal unskilled hands, incorporating stylistic
errors or anachronistic details (Löw 1993; 1998: 533–562; Lawergren 2000;
Muscarella 2000a: 31–215). Both categories are sold and published as ancient
objects. Forgers copy both forgeries and excavated artifacts. They also create
pastiches, utilizing a genuine core with the addition either of non-related
ancient or modern-made elements, or add engraved scenes to genuine
unadorned plaques or vessels (Muscarella 1999). Forgers often attempt to
create a unikum, a hitherto unrecorded type of artifact and therefore all the
more valuable to customers and scholars (Butcher and Gill 1993: 386; Mus-
carella 2000a: 17–19, 209 n. 31; 1999; 2006: 166–167). Forgers have lifetime jobs,
although some are now working part time.
Scholarly awareness of forgeries of ancient Near Eastern artifacts began
in the late 19th and early 20th centuries but then declined (Muscarella 1977b:
154–155, 169 n. 68; 2000a: 9). Early discussions were primarily concerned
with a specific object or inscription, many of them alleged to be Hebrew
and Christian texts. A noteworthy example of such a forgery occurred in
Italy with the appearance of a gold fibula inscribed with the name Manios
that was presented by an archaeologist in 1887. The fibula was enthusias-
tically accepted because its inscription was considered to be the earliest
Latin writing ever recorded, a prize indeed. Years later it was revealed to be
a forgery, commissioned by the archaeologist, who was seeking professional
fame (Muscarella 2000a: 11; for a description of another gold fibula that even-
tually turned out to be a forgery or pastiche, see Simpson 2005: 28–32). At
present, relatively few scholarly references to forgeries occur in archaeo-
logical literature, often through ignorance, but also because some scholars
deliberately suppress discussions to defend themselves or colleagues, which
deeds play a significant role regarding general ignorance of their existence
(Muscarella 1977b: 154–156, 161–163; 1980: 117–118 n. 3; 2000a: 2–5, 7–10, 12; in
press a; Butcher and Gill 1993: 387, 396, 399 n. 4, 396 n. 36).
the destruction of ancient near eastern cultures 853
Not surprisingly, Anatolia has experienced the same fate. Among the forg-
ers’ most successful productions are vessels and scores of nude terracotta
and stone female figurines alleged by dealers and their purchasers to have
been recovered at the Neolithic site of Haçilar. These figurines are readily
distinguished from excavated examples because of crucial misunderstand-
ings both of the ancient manufacturing method and of the originals’ thigh
and leg positions—joined, not separated (Muscarella 1980: 120; 2000a: 135–
157, 434–453). The use of stone for forged figurines has been described (accu-
rately!) by its museum purchaser as a “Novum” (Muscarella 1980: 120). A
number of plundered Urartian artifacts have been embellished with modern
engraved “Urartian” and non-Urartian decorative scenes, accepted by some
scholars and museums as data for enlarging our knowledge of ancient Urar-
tian art, history, and foreign contacts (Muscarella 2006: 153, 165–166, 174). A
small number of Hittite forgeries exist (Muscarella 2000a: 143–156, 454–461).
A “discovery” reported by an archaeologist involves objects allegedly de-
rived from two tombs at Dorak, in northwestern Turkey. They were first
introduced in the ILN (November 28, 1959: 754, Pl. i–iii) by the brilliant and
charismatic archaeologist James Mellaart. He later claimed that in 1958 he
met a woman on a train who took him to her home in Izmir, there show-
ing him objects from two tombs she said were discovered in 1922, which
is 36 years before Mellaart saw them. She allowed him only to sketch the
objects, one of which bore an Egyptian inscription, as well as the tomb plans
showing the objects in situ. Published in the ILN were painted drawings
created in Ankara from the sketches. If genuine, the corpus would attest to
the existence of a hitherto unknown, complex culture in western Turkey in
the 3rd millennium bc close to Troy. The precisely drawn tomb plans (first
published in 1967), with artifacts shown in place, explicitly suggested that
an archaeologist had painstakingly excavated them, and the drawn objects
are spectacular and unique. However, not a single object or photograph
has surfaced to date and many scholars believe that the “Dorak Treasure”
is a masterful (psychological?) fraud, the drawings and tomb plans modern
creations (Muscarella 1988; 397–398 n. 4; 2000a: 141; Greenfield 2007: 416–
417; S. Mazur in Scoop: 7/27/05, 10/4/05, 10/10/05 [www.scoop.co.nz]). The
very same problem applies to several drawings of fragmentary wall paint-
ings alleged by Mellaart to have been excavated at another of his sites, Çatal
Höyük. No archaeologist at the site had ever seen these fragments or the
paintings or their photographs (Muscarella 1988: 397–398 n. 5; 2000a: 141–
143). Forgeries implanted by archaeologists at their sites have also occurred
in Japan (Harvard Asia Quarterly, VI/3, 2002) and is alleged to have hap-
pened at a site in the United States (The New Yorker, 8/12/95: 66–83).
the destruction of ancient near eastern cultures 855
objects from the Getty Museum, Boston Museum of Fine Art, Princeton Art
Museum, Cleveland Museum of Art, and the Louvre have been returned
to Italy, Greece, and elsewhere (Brodie et al. 2000: 9, 47–48, 54; Waxman
2008: 63–64, 298–342, 356–364). Also returned were a few plundered antiq-
uities from some collectors, including MMA Visiting Committee member
Michael Steinhardt (to Italy) and the collector and MMA trustee Shelby
White (to Italy and Greece). But for more than two decades, White consis-
tently refused to return the head of Hercules plundered from Perge in Turkey
(Rose and Acar 1996: 77–78; Brodie at al. 2000: 32; Atwood 2004: 144–145;
Waxman 2008: 137; Muscarella 2010), until September 2011, when she and
her co-owner partner, The Boston Museum of Fine Arts, returned the head
(a gift from White) to Turkey, but with the caveat that they had acquired it
“in good faith” (Boston Globe, September 11, 2011).
As primary and essential reading for scholars, students, and the concerned
public interested in the subject matters discussed herein, the volumes edi-
ted by Vitelli (1996), Brodie et al. (2000, 2001, 2006), Lynott and Wylie (2000),
Brodie and Tubb (2002), Brodie and Renfrew (2005), and Vitelli and Colwell-
Chanthaphonh (2006) are especially recommended, as are the publica-
tions of R. Elia, J. Greenfield, U. Löw, C. Renfrew, E. Simpson, K.E. Meyer,
S.R.M. Mackenzie, and O.W. Muscarella. Bibliographies in all these volumes
contain many more relevant works. Also strongly recommended are Atwood
(2004) and Watson and Todeschini (2006). Together, these last two books are
exemplars of the problem, documenting succinctly and fully the machina-
tions and worldwide connections of the plunderers and their facilitators.
For legal matters, a complex issue dealing with old and contemporary laws,
see Bator (1982), Simpson (1997b), St Clair (1998), Majd (2003), Gerstenblith
(2006), Prott (2006), and Sokal (2006). For scientific matters, Muscarella
(2008a) and Grann (2010) are recommended.
Bibliography
Anthes, R.J. (1954), The Head of Queen Nofretete, Verlag Gebr. Mann. Berlin.
Atwood, Roger (2004), Stealing History, St. Martin’s Press.
Bator, P.M. (1982), ‘An Essay on the International Trade in Art’, Stanford Law Review
34, 2: 275–384.
Bogdanos, M. (2005), ‘The Casualties of War: The Truth About the Iraqi Museum’,
AJA 109: 477–526.
858 chapter thirty
———. (2008a), ‘See No Evil, Museums, Art Collectors and the Black Markets They
Adore’, in Rothfield 2008: 57–61.
———. (2008b), ‘Thieves of Baghdad’, in The Destruction of Cultural Heritage in Iraq,
P.G. Stone, J.F. Bajjaly, eds., The Boydell Press: 109–134.
Brodie et al. (2000), Stealing History. Cambridge.
Brodie, N., et al. (2001), Trade in Illicit Antiquities: The Destruction of the World’s
Archaeological Heritage. Cambridge.
Brodie, N. and Renfrew, C. (2005), ‘Looting and the World’s Archeological Heritage:
The Inadequate Response’, ARA: 343–361.
Brodie, N., et al. (2006), Archaeology, Cultural Heritage and the Antiquities Trade.
University of Florida.
Butcher, K., Gill, D.W.J. (1993), ‘The Director, the Dealer, the Goddess and Her
Champions: The Acquisition of the Fitzwilliam Goddess,’ AJA 97: 383–491.
Cook, B. (1995), ‘The Trade in Antiquities: a Curator’s View,’ in Tubb 1995: 181–192.
Cuno, J., (2008), Who Owns Antiquity? Princeton.
Elia, R.I. (1994), ‘The world cannot afford many more collectors with a passion for
Antiquities’, The Art Newspaper, No. 41, October: 19–20.
———. (1995), ‘Conservators and Unprovenanced Objects: Preserving the Cultural
Heritage or Servicing the Antiquities Trade’, in Tubb 1995: 244–255.
———. (1997), ‘Looting, Collecting, and the Destruction of Archaeological Resour-
ces’, Nonrenewable Resources 6, 2: 85–98.
Gerstenblith, P. (2006), ‘Recent Developments in the Legal Protection of Cultural
Heritage,’ in Brodie et al 2006: 68–92.
Graepler, D., et al (1993), Fundort: Unbekannt. Munich.
Graepler, D. (2004), ‘Archaologie und illegaler Antikenhandel: Die Rolle der Univer-
sitatssamlungen’, in Illegale Archaologie? Weissensee Verlag: 116–130.
Grann, D. (2010), ‘The Mark of a Masterpiece’, The New Yorker, July 12 & 16: 50–71.
Greenfield, J. (2007), The Return of Cultural Property. Third Edition. Cambridge
University Press.
Hiebert, F. and Canibon, P., eds., (2008), Afghanistan: Hidden Treasures from the
National Museum, Kabul. National Geographic.
Jarvis, H., The Looting Question Bibliography. Internet: Documents/lootbib.
Kaye, L.M. and Main, C.T. (1995), ‘The Saga of the Lydian Hoard antiquities: from
Usak to New York and back and some related observations on the law of cultural
repatriation’, in Tubb, ed.: 150–162.
Kelker, N.L. and Bruhns, K.O. (2010), Faking Ancient Mesoamerica. Left Coast Press.
Kletter, R. (2003), ‘A Very General Archaeologist-Moshe Dayan and Israeli Archaeol-
ogy’, Article 5 in The Journal of Hebrew Scriptures 4.
Koczka, C.S. (1989), ‘The Need for Enforcing Regulations on the International Art,
Trade’, in Messenger 1989: 185–208.
Lapatin, K.D.S. (2000), ‘Boy Gods, Bull Leapers, and Mother Goddesses’, SOURCE XX,
1: 18–28.
Low, U. (1993) ‘Kunsthandel und Falschungsproblamatik,’ MDA-Ve v. Heft 1: 36–41.
———. (1998), Figurlich verzierte Metalalgefasse aus Nord- und Nordwestiran, Mun-
ster (reviewed by O.W. Muscarella in BibOr LVII, No. 1/2 188–195).
Luke, C. and Kersel, M. (2006), ‘Mysterious Shepherds and Hidden Treasure: The
Culture of Looting in Lydia, Western Turkey’, JFA 31: 185–198.
the destruction of ancient near eastern cultures 859
Lynott. M.I. and Wylie, A. (2000), Ethics in American Archeology. Society of American
Archeology.
Mackenzie. S.R.M. (2005), Going, Going, Gone: Regulating the Market in Illicit An-
tiques. Institute of Art and Law.
Majd, Mohammad Gholi (2003), The Great American Plunder of Persia’s Antiquities.
1925–1941. University Press of America, Inc.
Mallampati, H. (2005), ‘Archaeology and Collecting: Law, Ethics, Politics’, in M.C.
Root, ed., The Fertile Land Signs and Symbols in the Early Arts of Iran and Iraq.
Messenger, M. (1989), The Ethics of Collecting Cultural Property. University of New
Mexico Press (reviewed by O.W. Muscarella, AJA, 95, 1991: 342–344).
Meyer. K.E. (1973), The Plundered Past. Atheneum.
Muscarella, O.W. (1973), ‘Antiquities and Collections: A Curator’s Viewpoint,’ JFA 1:
2–5.
———. (1974), ‘The Antiquities Market: A Colloquium,’ Paper Two, JFA I: 221–222.
———. (1976). ‘Antiquities Legislation Pending in Congress’, Archaeology, 29, 275–
276.
———. (1977a), “‘Ziwiye” and Ziwiye: The Forgery of a Provenience’, JFA 4: 197–219.
———. (1977b), ‘Unexcavated Objects and Ancient Near Eastern Art,’ in Mountains
and Lowlands Essays in the Archaeology of Greater Mesopotamia, L. Levine, T.C.
Young, eds. Undena Publications: 153–205.
———. (1988), Bronze and Iron, Ancient Metal Artifacts in the Metropolitan Museum
of Art, New York.
———. (1980), “Die gefalschte Kunstgechichte-Ein Hausputz in der Vorderasiatis-
chen Altertumskunde”: A Reply, APA 11/12 117–120.
———. (1995), ‘Bazaar Archaeology’, in Beitrage zur Kultgeschichte Vorderasiens,
Festschrift for Michael Boehmer. Mainz: 449–453.
———. (2000a), The Lie Became Great. Styx.
———. (2000b), Editor, Special Issue on Forgeries of Ancient Art, SOURCE, XX, 1.
———. (2000c), ‘Excavated in the Bazaar: Ashurbanipal’s Beaker,’ in Muscarella
2000b: 29–37.
———. (2000d), ‘The Pope and the Bitter Fanatic,’ in The Iranian World, Essays on
Iranian Art and Archaeology, eds. A. Alizadeh et al, Iran University Press: 5–13.
———. (2001), ‘Jiroftand “Jiroft”-Aratta,’ BA1 15: 173–198.
———. (2003), ‘Museum Constructions of the Oxus Treasure: Forgeries of Prove-
nience and Ancient Culture’, ACSS 93–94: 259–275.
———. (2005a), Review of Ernst Herzfeld and the Development of Near Eastern Stud-
ies, 1900–1959. A. Gunter and S. Hauser, eds. Leiden. 200S, JAOS, 125: 431–432.
———. (2005b), ‘Some thoughts on fakes, addendum,’ N.A.B.U. no. 4: 97–98.
———. (2006), ‘Urartian Metal Artifacts: An Archaeological review.’ ACSS 12: 147–175.
———. (2007), ‘Archaeology and the Plunder Culture,’ IJCT: 602–618.
———. (2008), ‘The Veracity of Scientific Testing by Conservators’, in Original-Copy-
Fake?, E. Pernicka, Silke von Berswordt-Wallrabe, eds. Mainz am Rhein: 9–18.
———. (2009a), ‘The Fifth Column in the Archaeological Realm: The Great Divide’,
in Studies in Honor of Altan Cilingiroglu, eds. E. Gens, D. Celik. Istanbul: 395–406.
———. (2009b), ‘A Review and Addenda: Michael Gross’ Rogues’ Gallery, SCOOP,
Internet, September 16, 2009: 1–17.
———. (2010a), ‘Archaeologists and Acquisitionists,’ IJCT 18, 2: 449–463.
860 chapter thirty
———. (2010b), ‘An Unholy Quartet: Museum Trustees, Antiquity Dealers, Scientific
Experts, and Government Agents’, in Memorial Volume for Massoud Azarnoush;
in press.
———. (2010c), ‘Shelby White’s Foundation Expansion,’ SAFE Corner (Internet),
June 1.
Nagin, C. (1986), ‘Patrons of Plunder,’ Boston Review, August: 5–6, 23–25.
Özgen, L. and Öztürk, J. (1996), Heritage Recovered The Lydian Treasure. Ankara.
Renfrew, C. (2000), Loot, Legitimacy and Ownership: The Ethical Crises in Archaeol-
ogy. Duckworth.
———. (2006), ‘Museum Acquisitions: Responsibilities for the Illicit Traffic in Antiq-
uities,’ in Brodie 2006: 245–257.
Rothfield, 1., ed., (2008), Antiquities under Siege. Altimara Press.
Silver, V. (2006), ‘The role of museum trustees,’ Bloomberg News, February 25: 1–13.
Simpson, E., ed. (1997), The Spoils of War, Harry M. Abrams.
Simpson, E. (2005), ‘Tall Tales: Celts, Connoisseurs, and The Fabrication of Archae-
ological Context’, SOURCE XXIV, 2: 28–41.
St. Clair, W. (1998), Lord Elgin and the Marbles. Oxford.
Stanish, C. (2008), ‘Forging Ahead’, Archaeology (May/June): 18: 58–66.
Tubb, K.W. (1995), ‘The Antiquities Trade: an archaeological conservator’s perspec-
tive,’ in Antiquities Trade or Betrayed Legal, ethical and conservation issues, Kath-
ryn Walker Tubb, ed. Archetype Publication: 256–263.
———. (2002), ‘Point, Counterpoint,’ in Brodie, N. and Tubb, K.W., eds., Illicit Antiq-
uities, The Theft of culture and the extinction of archaeology. Routledge: 286–292.
Vitelli, K.D. (1984), ‘The International Traffic in Antiquities: Archaeological Ethics
and the Archaeologist’s Responsibility,’ in E. 1. Green, Ethics and Values in Archae-
ology. The Free Press.
Vitelli, K.D., ed. (1996), Archaeological Ethics. Altimara Press.
Vitelli, K.D. and C. Colwell-Chanthaphonh, eds. (2006), Archaeological Ethics. Sec-
ond Edition Altimara Press.
Waxman, S., Loot (2008). Times Books. New York.
chapter thirty-one
Abstract: The issues discussed here have been on my mind for a long time, and
some have surfaced in earlier publications.1 Here I expand on them, believing they
are appropriate for a Festschrift for Altan Çilingiroğlu, an excellent archaeologist, a
fine and honest scbolar and teacher; and a dear friend with whom I excavated for
seven years at Ayanis on Lake Van. In his honor I address a significant archaeological
condition, one regularly avoided or covered up: the contra-archaeological behavior
of some archaeologists. Their professional and moral behavior has resulted in the
multi-faceted, fractured, and unanchored nature of much of the discipline today.
Most archaeologists are aware of the on-going plunder of sites and concomi-
tant destruction of this planet’s history, yet relatively few are actively con-
cerned with terminating the destructions. Gradually more and more infor-
mation is being published by archaeologists, and by non-professionals in the
press or the Internet and in investigative reports and books, thus increas-
ing attention to the topic. Archaeologists who over the years have consis-
tently fought the fight include C. Coggins, E. Negahban—these two were
the first archaeologists to call attention to the problem (they converted me),
K.D. Vitelli, N. Brodie, R. Elia, E. Herscher, C. Renfrew, D. Graepler, D. Gill,
and C. Chippendale. And the number of archaeologists who undertake sim-
ilar actions in classrooms is increasing. Nevertheless, many archaeologists
remain reticent about engaging in vigorous and long-term efforts to stop
plundering and collecting activities that are contrary to the nature of their
discipline. Their indifference is matched by archaeological organizations in
Europe and the United States—viz. the Archeological Institute of America:
manifested by the paucity of sponsored lectures and on-going professional
and public discussions; the absence of active lobbying in the press and with
* This chapter originally appeared as “The Fifth Column within the Archaeological
Realm: The Great Divide,” in Studies in Honour of Altan Çilingiroğlu, eds. H. Sağlamtimur,
Z. Derin, and E. Abay (Istanbul: Arkeoloji ve Sanat Yayınları, 2009), 395–406.
1 viz. Muscarella 1991; 2000: 22–29; 2008a: 613–614.
862 chapter thirty-one
reputable), who have standing orders from the level above to supply them
with antiquities, and forged documents manifesting tbe antiquity derived
from an Italian or German family that possessed it for generations.
The dealers sell their plunder to the upper level, the fourth column. It
consists of wealthy private collectors, university, private and public museum
directors and their curatorial staffs and trustees, and is supported by news-
paper owners and by elected government officials. Plundered artifacts are
referred to as “antiquities,” or “Art”.6
Collectively, Plunder Culture members are the raison d’ etre for looting
and destruction. The upper membership level, museums and collectors,
purchase antiquities as an on going class-power-identity activity. They ini-
tiate the process by sponsoring and financing the purchase orders, which
pay the on-site looters to destroy (the power, the power!) ancient sites, the
smugglers, the bribes, so that they can possess/save “Art.” For museum cura-
tors, some are archaeologists, others art bistorians see Muscarella 2000: 2–3,
13, 23–25, 27, footnotes 5, 7; the “acquisition” of antiquities is a major compo-
nent of their job description, for which raises and promotions reward them. I
once heard one of the very rich, and thereby very powerful curator-sponsors
of plunder, Dietrich von Bothmer of the Metropolitan Museum of Art, sneer
at corrupt Italian officials—all lower class of course, for accepting bribes:
which he himself had paid!7 Courtesy of their social and financial status,
upper-level members also have the power to conceal their roles effortlessly,
for example keeping their activities out of newspapers and other media (this
is gradually changing).
Although the Plunder Culture’s activities are becoming better known, pri-
marily from non-archaeologist writers and reporters,8 and increasingly from
professional archaeologists, the reports rarely present full details about the
wealthy and upper-social class perpetrators. A recent volume with essays
by scholars on plunder by scholars is Brodie et al. 2006;9 earlier examples
6 For newspaper owners and staffs suppressing information about plundering museums,
see Meyer 1974: 11, Horsley 1997, Atwood 2007 (also below), and Muscarella in Mazur 2006: 2–
3.
7 In an Internet interview I referred to DvB as a “son of a bitch.” I wish to make an apology
for that stupid remark. I have been blessed by living with many female dogs, all loving,
truthful, and loyal bitches, and I apologize to them all for my solecism.
8 Two of the best recent books on this issue are works of Atwood 2004, and Watson
2006 (see also Watson 1977); also Nagin 1986. To date the only substantial public reports
on the archaeological issues raised here occur on the Internet: see Mazur 2006. A work that
presents—in a long and rambling manner—the arguments and positions of dealers, auction
houses, museums, and legal issues (but little on archaeologists’ positions), is Mackenzie 1988.
9 Brodie et al. 2006; see also Muscarella 2008a.
864 chapter thirty-one
of collected articles on plunder include Brodie et al. 2000 and 2001. David
Gill’s internet blog is an important example of professional engagement;
equally the McDonald Institute for Archaeological Research at Cambridge
University (now non-functioning), via its journal Culture Without Context
and monographs.10 In the past the Journal of Field Archaeology had a unique
“Antiquities Market” section edited by K.D. Vitelli, and later Ellen Herscher;
recently it has been revived. Very important is the public organization, Sav-
ing Antiquities For Everyone (SAFE), founded by lay people (Cindy Cho and
Paul Kunkel), ordinary citizens. They do work that not many archaeologists
and their organizations attempt: bring individuals of lay and professional
backgrounds together, to monitor, expose and fight plunder and those who
entail it, wherever encountered.11 Maybe eventually (inşallah) professionals
will wake up and use them as a model.12
continuously fought to prevent illegal antiquities arriving in the United States. He rarely
succeeded, but he did play an important role in getting back to Turkey the Metropolitan
Museum of Art’s plunder, “The Lydian Treasure” (see below; see also his article in Messenger
1989: 185–198; Muscarella 1991: 344).
13 For discussions and documentation of this ideological/professional divide see for
example my articles: Muscarella 1977; 2003; 2006; also Simpson 2005. I disagree with Winter’s
1992: 31 challenge to what she calls the “oppositional stereotypes … such as museum curators
vs. academics.”
the fifth column within the archaeological realm 865
14 Rose 2007: 6.
15 In Brodie 2006: 289–290; see Muscarella 2008a: 615–616 for discussion.
16 Griffin 1986: 465.
17 See examples below; also Muscarella 1991: 343.
18 Griffin 1986, 1989.
19 Griffin 1989: 106.
866 chapter thirty-one
“Purists would try to stop collecting …” for “Collecting can be positive and
creative.” Further, he is a “serious collector”.20 To him, it is archaeologists—
not collectors—who are “self righteous”.21
For Harvard’s (one time) curator James Cuno’s sophisticated homage to
orphaned artifacts see Muscarella 2008a: 612. His museum purchases—
182 Greek vessels in 1995 alone—prove he can deceive by transmogrifying
plundered artifacts and destroyed sites into “Art” belonging to everyone.
For similar beliefs we have a director of the St. Louis Museum in Missouri,
James Burke, claiming “the world is a better place when world cultures are
shared …”22 (also when the world’s women are shared in the same manner?).
Planned dissimulation was orchestrated by the Metropolitan Museum of
Art curator Dietrich von Bothmer, the museum’s director Le Comte Guy-
Philippe Lannes de Montebello, and Arthur Ochs Sulzberger, the owner of
the New York Times and (conflict-of interest) museum Trustee, serving for
decades on its Purchasing Committee. They purchased, exhibited and pub-
lished purchased artifacts plundered from Turkey, and baptized “A Greek
and Roman Treasury”.23 The Trustees and curatorial staff knew that the
objects had been plundered from now-destroyed tombs in western Turkey,
which was revealed many years later when the artifacts were returned to
their correct archaeological and geographical locus: in western Turkey, not
Greece.
The deceit has never been discussed openly within the museum, nor ever
been reported in the New York Times; only one employee in the museum
objected to the purchase.24 Karl Meyer obliquely reported the conflict of
interest behavior of Sulzberger when discussing the purchase of the plun-
dered Euphronios vase: “for the paper to editorially reproach the Met would
involve criticism of the publisher’s friends.”25 The conflict of interest role was
confirmed by Ken Auletta in The New Yorker (June 28, 1993: 4). He reported
that Sulzberger’s son Arthur refused to become a Trustee at the Metropoli-
tan Museum “in order to avoid any appearance of conflict of interest with
his duties as publisher.” Another disturbing revelation of the behavior of
New York Times’ staff and editors is that of C.B. Horsley who documents
their involvement in suppressing of any information of the Metropolitan
26 Horsley 1997.
27 Atwood 2007: 3–7.
28 See Muscarella on the Internet, December 12, 2005, February 9, 2006, via S. Mazur
(2006), also Brodie–Renfrew 2005: 349–350; on the MMA Trustees, see Nagin 1986: 6.
29 Muscarella 2000: 4.
30 For the objects see Muscarella 2008b: 13–14.
31 For specific details of this still unrecognized problem see Muscarella 2008b.
868 chapter thirty-one
torn from its now-destroyed locus, objected to its purchase. The retort was:
“If you don’t like what we do here, you can leave”.32 The Trustees purchased
the plundered artifact. And the archaeologist-curator was awarded a profes-
sorship at a major university (recommended by archaeologists associated
with the awardee’s museum). The protesting curator was subsequently dis-
missed.
Witness also the 180-degree professional and moral turnaround by an
archaeologist who applied for the position of an art museum director. He
was appointed and immediately accepted his job description duty, to
actively purchase plundered artifacts, and employ all the museum-ritual
language to justify such actions.33 The metamorphosis from archaeologist
to Fifth Column citizenship was accomplished with ease. As a reward for
“building up” his art museum’s antiquities collection, he was subsequently
hired to direct a university archaeological museum.
Another serious problem is that many academics (and thus their stu-
dents) unhesitatingly cite as reality the certification of curator/archaeolo-
gists that purchased or donated artifacts are genuine and archaeological
evidence for cultural history. They also declare as archaeological fact a spe-
cific geographical provenience, a locus, for the purchased object, which
information was furnished by a dealer (although never mentioned). Mu-
seum-ritual demands that one eliminate the crucial empirical distinctions
between excavated and plundered artifacts; wbich behavior falsifies the his-
torical record. Museum curator/archaeologists also are “bound by the rules
of confidentiality,” rules that control all information about acquisitions.34
Museum confidentiality loyalties take precedence over archaeological loy-
alties, for which the operating word is omertà.
Years ago an ex-museum curator, archaeologist and Director asked me
to look at a number of photographs of “antiquities” that a collector wanted
to donate to his museum: were they genuine or forgeries? My colleague
argued that if they were forgeries it would be bad for the museum, the public
and scholars to exhibit them. I agreed. To me all the objects were forgeries.
He told me he would inform his curator/archaeologist successor, and ask
him not to accept the gifts. Sometime later he informed me that his advice
was ignored. The gift was accepted because the rich collector could not be
offended. Decades ago I was asked by Teddy Kolleck, Mayor of Jerusalem,
to vet a donation of antiquities to the Israel Museum by a rich collector. I
32 See also Winter 1992: 34–35; Watson 2006: xviii–xix, 102, 331.
33 viz. Muscarella 2000: 24–25.
34 See Muscarella 2008a: 613.
the fifth column within the archaeological realm 869
reported that all but one object were forgeries. Against the wishes of the
museum’s curator Kolleck correctly rejected the gift.
The examples cited here depict but one component of the Great Divide
crippling the discipline, and I turn to another group of archaeologists who
operate comfortably and successfully within the Plunder Culture. Function-
ing for decades, they operate successfully as a Fifth Column within the
archaeological realm, essentially covertly and unrecognized by many col-
leagues and the public. Most scholars aware of their roles are either indiffer-
ent or remain silent—the latter because they fear their powerful academic
connections in the awarding or denying academic positions and fellowships.
I give some examples:
University employed archaeologists accept invitations from the Metro-
politan Museum’s Department of Ancient Near Eastern Art to become mem-
bers of their Visiting Committee, which meets several times a year for
socials, lunch, private discussions, lectures, and private viewings of antiq-
uity purchases and gifts. The Committee is composed of museum curators,
dealers and antiquities collectors, and a good number of archaeologists.35
They are invited to become members by the museum’s Trustees, after which
they receive a letter from the museum’s legal Counsel announcing that the
“Museum’s Board of Trustees” has appointed them members. One promi-
nent archaeologist wrote in his acceptance letter: “I certainly appreciate this
connection with the Museum … [and] I am looking forward to the contin-
uing association … [and] hope to be of some help in a visiting, admiring
and advisory capacity” (italics mine). Another wrote that membership was
“rewarding … both on a personal and professional level …” (italics mine):
the precise nature of this professional engagement was manifested when in
a letter to an archaeologist colleague he referred to his “official connections”
with the museum that “obliged” him not to discuss an archaeological matter.
Other archaeologists wrote the more typical “I accept with pleasure” (… to
sit alongside my fellow-members, powerful and rich antiquity dealers and
collectors).
Noted for years in the invitation letters is that the Committee Chair is
Shelby White, possessor of hundreds of plundered antiquities. Other mem-
bers include the antiquity collectors Jonathan Rosen (also an antiquities
dealer), Sheldon and Barbara Breitbart, and Michael Steinhardt. In Forbes
Internet report (12/29/06: 4) the latter, defending his purchasing activities,
35 See Muscarella 2008a: 613, 616–617; for a full listing of members see the Museum’s
Annual Reports.
870 chapter thirty-one
boasted that he is not intimidated (what billionaire is?), that collecting was
“dangerous, but that makes it exciting.”
The archaeologist members serve their colleagues loyally: omertà again.
They provide their dealer and collector colleagues with lectures on their
excavations and research. They visit their homes for social gatherings, and
offer advice on their purchases. An archaeologist member was also asked
by the museum’s Counsel to contact colleagues and officials of Turkey as
the Museum’s representative (another “official” function) to present their
position in a matter that concerned a large group of plundered artifacts, the
so-called Lydian Treasure, which the museum had purchased and refused
to return.
A few years ago I learned that the President of the NYC chapter of the
Archaeological Institute of America had voluntarily accepted the Trustees’
invitation to become a Committee member. I informed her of its dealer and
collector membership, supplying their names, and asked why she would
associate with them. This archaeologist shouted that the archaeologist X
was also on the Committee, and stalked away. Years later she piously wrote
in the Fall 2007 AIA New York Society News: “The AIA has worked hard to
stop the trafficking of antiquities and established itself at the forefront of
preserving the world’s archaeological resources and cultural heritage.” No
footnote disassociated her from this claim (after all, she knew that a past
AIA President has been a member of the Visiting Committee for more than
three decades).
Over many years non-member archaeologists have been invited by the
Committee to give lectures on their fieldwork. With ease these archaeolo-
gists accept the honor to address White, Steinhardt, Rosen, etc. Once when
Joan Oates accepted an invitation to lecture about her site Nimrud before
the Committee, I was told by the head of the ANE Department not to come (I
had no intention to do so): because Shelby White would leave if I appeared,
and that could not be tolerated. Another time when Ian Hodder was speak-
ing at a ticketed meeting at another institution for which the Committee had
reserved 25 places, the same authority informed me (through an intermedi-
ary) that I could not attend; no ticket was available for me. In both cases,
an archaeologist was denied permission to hear an archaeologist speak, but
the dealers and collectors were not denied: on orders from the plunderers.
Why did these archaeologists consider it appropriate to give archaeological
lectures to plunderers?
Collaborationist scholars also openly support collectors and facilitate
their plunder entering the United States. A cuneiformist from Cornell Uni-
versity helped Jonathan Rosen import recently plundered tablets from Iraq
the fifth column within the archaeological realm 871
(in 2000) by informing Customs that they were “educational material with
no commercial value”.36 Rosen took a $1,000,000 tax deduction when he sub-
sequently gave the tablets to Cornell; and the collaborating cuneiformist will
publish them.
Archaeologists also support plundering and the selling of the orphaned
antiquities to dealers, museums and collectors, by opposing all legislation
attempting to stop them. This was evidenced in Washington D.C. during
the congressional hearings on antiquity legislation in the 1970s, where over
some days I encountered archaeologists who vigorously opposed any anti-
plunder testimony before Congress. In Israel the Antiquities Traders Associ-
ation campaign against stopping plunder and the existence of dealers’ shops
was joined by a number of prominent local archaeologists, and (needless to
add) museum curators.37
Others openly serving the plunderers are the many archaeologists who
write articles for the plunder/dealer-supporting magazine owned by an
antiquities dealer (Minerva), or those functioning as fronts for antiquity
dealers and collectors (Biblical Archaeological Review, Archaeological Odys-
sey). Known to these contributing archaeologists is that the dealer-owned
magazine, aside from adversary behavior toward archaeologists, in each
issue displays many advertisements from antiquities dealers all over the
world, and informs its readers about antiquity auction sales worldwide. The
front-magazines (one has ceased publication) contain advertisements from
antiquities dealers (reduced in recent years), consistently defend dealer
and collecting activities and attack anti-plunder archaeologists and their
professional organizations. Not one issue of the three magazines is without
articles by archaeologists, who report (for a fee) on their excavations or other
research matters.38 Archaeologists also serve on their Editorial Boards.39
or participated in the purchase of one post 1970/1 (pace the self-serving and false claims by
an antiquities dealer), the year when I first learned from the lectures of Clemency Coggins
and Ezat Negahban what purchasing antiquities meant; and when I was instantly converted.
the fifth column within the archaeological realm 873
historians are its high priests,” and in which temples plundered artifacts
become “art.” It is not a coincidence indeed that one of the owners of Har-
vard University, James R. Houghton (serving on its Board of Overseers) is also
one of the owners of the Metropolitan Museum of Art (as member and chair
of its Board of Trustees). For years in both institutions he has encouraged
the purchase of “art.”45 In the early 1900s Yale University’s museum “bor-
rowed” many artifacts from Peru, from Yale’s excavations at Machu Picchu.
Yale never returned them to Peru and as of 2008 continues to refuse repatri-
ation.46
Other plunder-active United States university museums include Prince-
ton, Indiana University, University of Missouri, Emory, etc …47 University
trustees, presidents, faculties, including archaeologists, support these pur-
chases, actively or by their silence: they do not write letters to their univer-
sity presidents demanding an end to their participation in the destruction
of the archaeological record. For European University museum participa-
tion in the Plunder Culture, see Graepler 2004.
Conclusions: The archaeological Fifth Column is a significant component
of the forces that sponsor and thrive on cultural destruction. Its members
are as equally culpable as the advancing columns, tombaroli, antiquity deal-
ers, collectors and museum staffs. They continue to flourish, their activities
proceed successfully and unabated, they get awarded—revealing that the
discipline of archeology has no comprehensive sense of itself, no unclouded
self-knowledge, no awareness of its moral and academic weakness.48
There remains a parallel conclusion: follow the archaeological work, the
integrity and accomplishments of archaeologists like Altan Bey, whose
actions manifest proper archaeological behavior.
45 In a classic conflict of interest action Houghton wearing his Harvard hat awarded an
honorary Ph.D. to his Metropolitan Museum employee de Montebello, a degree this director
could/would never have earned.
46 E. Karp-Toledo, “The Lost Treasure of Machu Picchu,” New York Times, 2/23/08: A17.
47 Muscarella 2008a: 612–614; Winter’s (1992: 35–36) comment that university museums
ences—hence the often usage of the pronoun I. There are surely many other examples of
which I am unaware.
the fifth column within the archaeological realm 875
Bibliography
Atwood 2007: R. Atwood, “A critical look at U.S. media coverage of antiquities issues,”
SAFE (Internet), 1–7.
Brodie–Renfrew 2005: N. Brodie, C. Renfrew, “Looting and the World’s Archaeolog-
ical Heritage: The Inadequate Response,” Annual Review of Anthropology, 343–
428.
Brodie et al. 2000: N. Brodie, J. Doole, P. Watson, Stealing History: The Illicit Trade in
Cultural Material, Cambridge.
Brodie et al. 2001: N. Brodie, J. Doole, C. Renfrew, Trade in Illicit Antiquities: The
Destruction of the World’s Archaeological Heritage, Cambridge.
Brodie et al. 2006: N. Brodie, M. Kersel, C. Luke, K.W. Tubb, Archaeology, Cultural
Heritage and the Antiquities Trade, University Press of Florida.
Coggins 1972: C. Coggins, “Archaeology and the Art Market,” Science 175, 263–266.
Graepler 2004: D. Graepler, “Archäologie und illegaler Antikenhandel: Die rolle der
Universitätssammlungen,” W. Heilmeyer, J.C. Eule (Eds), Illegale Archäologie?,
Berlin, 116–127.
Griffin 1986: G.G. Griffin, “In defense of Collectors,” National Geographic 169, 4, 462–
465.
Griffin 1989: “Collecting Pre-Columbian Art,” in Messenger 1989, 103–115.
Horsley 1997: C.B. Horsley, “Chinagate,” The City Review: 1–16.
Ilan et al. 1989: D. Ilan, U. Dhari, G. Avni, “Plundered!,” Biblical Archaeological Review
XV, 2, 38–42.
Mackenzie 1988: S.R.M. Mackenzie, Going, Going, Gone: Regulating the Market in
Illicit Antiquities, Institute of Art and Law, Leister.
Mazur 2006: S. Mazur, “Italy-Met Euphronios Accord?” Scoop Internet, February 22:
1–12.
Meyer 1974: K. Meyer, “Behind the Damask Curtain,” MORE: 11–14.
Messenger 1989: P.M. Messenger, The Ethics of Collecting Cultural Property: Whose
Culture? Whose Property?, Albuquerque.
Muscarella 1977: O.W. Muscarella, “Ziwiye and Ziwiye: The Forgery of a Prove-
nience,” Journal of Field Archaeology 4, 197–219.
Muscarella 1991: O.W. Muscarella, Review of Messenger 1989, American Journal of
Archaeology 95, 2, 342–344.
Muscarella 2000: O.W. Muscarella, The Lie Became Great, STYX.
Muscarella 2003: O.W. Muscarella, “Museum Constructions of the Oxus Treasure:
Forgeries of Provenience and Ancient Culture,” Ancient Civilizations from Scythia
to Siberia, 8, 3–4, 259–275.
Muscarella 2006: O.W. Muscarella, “Urartian Metal Artifacts: An Archaeological
Review,” Ancient Civilizations from Scythia to Siberia 12, 1–2, 147–177.
Muscarella 2008a: O.W. Muscarella, “Archaeology and the Plunder Culture,” Inter-
national Journal of the Classical Tradition, 14, 3/4, 602–618.
Muscarella 2008b: O.W. Muscarella, “The Veracity of ‘Scientific’ Testing by Conser-
vators”, Original, Copy Fake?, Mainz, 9–18.
Nagin 1986: C. Nagin, “Patrons of Plunder,” Boston Review, 5–25.
Rose–Burke 2004: J.D. Rose, D. Burke, “Making Money from buried treasure,” Culture
Without Context, 14, 4–8.
876 chapter thirty-one
Rose 2007: C.B. Rose, “From the President: Common Ground, Can Archaeologists
and Museums Work Together?”, Archaeology 60, 5.
Simpson 2005: E. Simpson, “Tall Tales: Celts, Connoisseurs, and the Fabrication of
Archaeological Context,” SOURCE XXIV, 2, 28–41.
Watson 1977: P. Watson, Sotheby’s: The Inside Story, London.
Watson 2006: P. Watson, The Medici Conspiracy, New York.
Winter 1992: I. Winter, “Changes in the American Art Museum: The (An) Art His-
torian’s Voice,” in Different voices: a Social, Cultural, and Historical Framework for
Change in the American Art Museum, Association of Art Museum Directors, 30–
57.
Section Four
Forgeries
chapter thirty-two
BAZAAR ARCHAEOLOGY*
(PLATE 34 A.B)
schichte Vorderasiens, Festschrift für Rainer Michael Boehmer, eds. U. Finkbeiner, R. Dittmann,
and H. Hauptmann (Mainz: von Zabern, 1995), 449–453.
880 chapter thirty-two
The first disc (pl. 34 a; diameter 2.6cm., thickness .8 cm.) surfaced in 1976
and has been published in two dealers’ sales catalogues and subsequently
in two scholarly venues; it now rests in a New York private collection.1 In its
first non-dealer publication (see note 1: Porada 1992; the second publication,
Porada 1993: 50ff., is less detailed), the iconography and style of the carved
scenes are concluded to date to the Early Dynastic period. On one side is
depicted a seated male figure wearing a tufted skirt with his right hand clasp-
ing the left; before him is a cuneiform inscription that several cuneiformists
read as “Rimush LUGAL KISH,” that is, the Akkadian king who ruled about
ca. 2270 bc. The other side depicts an Anzu-Imdugud bird monster grasping
in each claw a horned animal. The rim is pierced for suspension.
Concerning the seated figure it is noted (Porada 1992: 69), with Early
Dynastic style in mind, that the head is “very large”, in proportion to the
body, the eye is “huge”, and, “more unusual” the ear is “very large”, as is also
the nose; and “only one foot is indicated …” With regard to the bird monster
1 Collectors’ Journal of Ancient Art, Vol. 5, no. 2, 1985, front and rear cover; Hesperia
Arts Auction Ltd. Antiquities, November 27, 1990, no. 58; E. Porada, A Lapis Lazuli Disc with
Relief Carving Inscribed for King Rimush, in La circulation des Biens, des Personnes et des
idées dans le proche-Orient ancien: Acres de la XXXVIIIe Rencontre Internationale (1992)
69 ff.; idem, Seals and Related Objects from Early Mesopotamia and Iran, in J. Curtis (Ed.),
Early Mesopotamia and Iran (1993) 44 ff.; neither article mentions the other. The disc is also
listed in D. Frayne, Sargonic and Gutian Periods (2234–2113bc), RIM volume 2 (1993) 71, no. 51
(incorrectly said to be in the collection of the P. Morgan Library—it is only on loan there).
Note that no. 52 in the RIM volume is the second disc discussed by me in this article, infra.
I also wish to thank J. Rosen for allowing me to examine the disc, and David Wright of the
Morgan Library for facilitating the examination.
According to Porada (1992: 69) the disc came to its present provenance from the dealer
(“the former owner”) Joel Malter, who “got it from another dealer who had acquired it from an
old European collection.” Aside from the problem of the last undocumented claim, which in
most cases—as every plunderer, smuggler, dealer, collector, curator, knows—is either gossip
or outright dissimulation, usually the latter, the given trade sequence is incorrect and much
telescoped. Indeed, the actual modem trickle-down trade dynamics is more complex. In 1976
a dealer, X, purchased it from another dealer. X then re-sold it to another dealer, who claimed
to be acting as agent for a client. At sometime after X had purchased and sold the disc, he was
told that it was once in the Gillet collection, but no evidence for this claim, or information
about how long the disc allegedly existed in the collection, was given to him. Nine years
after X sold it, in 1985, the disc was put up for sale (Malter), with no information given about
previous owners. Five years later the disc surfaced again, in an auction catalogue (Hesperia:
which is not mentioned by Porada); it then went to its present home in the J. Rosen collection.
Thus, before its final sale at least five, six, possibly seven (if we count the “old family”
collection) transactions, and possibly more not recorded (between 1976 and 1985, and 1985
and 1990?), had occurred; and one assumes that the price had risen as many times. Do we
have here an ethnoarchaeological model for ancient trade?
I also wish to thank Elizabeth Simpson for reading this paper, for making good sugges-
tions. and for demanding that I remove many preaching sentences.
bazaar archaeology 881
scene, it is noted (ibid.: 70) that the grasped goats with long, plain horns
“are not paralleled in Mesopotamian cylinders …” and that the depicted
monster-bird “… differs from all Mesopotamian representations … by having
the head of a bearded man.” Also noted is that the dot circles on the rim
were made “with a mechanical tool.” Following these correctly observed
deviations from that which is known to be Early Dynastic (which is the
existence of a fairly large amount of excavated material),2 formal parallels
that allow the disc to be dated to a post-Urnanshe period are drawn (Porada
1992: 70; 1993: 50). The date suggested for the disc is the period of either
Eannatum I or II, ca. 2450bc, despite the author’s opinion that the presence
of one row of tufts at the base of the gown is an indication of a pre-Urnanshe
date (ca. 2470bc; 1992: 70).
In fact, this list of deviations from that which is known to be Early
Dynastic (see note 2) does not reveal the full extent of the problem. The
observant viewer will readily add the following (some are more visible in
autopsy than in photographs): not only are the nose, ears, and eyes too large,
they are crudely executed, and in forms unparalleled and unexpected in the
extant corpus: note the eye outline that curves back and then loops forward
to incompletely enclose the pupil, and also the awkward juxtaposition of
the eye and ear. The large head lies heavily on the shoulders, and the curved
mouth in the projecting chin is clamped, not merely closed. The body, single
foot (a blob) and arms, and the bottom row of skirt tufts are not executed
in typical Early Dynastic forms. The hands have no structure, the fingers
are not discernible, and instead of a thumb extending from each of the two
hands, a thumb (?) and second finger awkwardly project from the under, left
hand. (Actually, all one need do is compare this figure to Porada 1992, Fig. 2
to appreciate the significant differences.) Also, the figure sits not on a stool
with a curved top but unsurely inside a chair. The dotted circles along the
exterior rim of the disc are unevenly spaced and some overlap.
And there is more. It is unusual for a seated figure represented on Early
Dynastic reliefs to be depicted with hands clasped; this gesture seems to
be reserved for standing figures in relief and in the round; and the holes
indicating the nipples do not to my knowledge occur elsewhere on relief
2 For purposes of examining the references for the appropriate excavated Early Dynastic
material, see for example H. Frankfort, Sculpture of the Third Millennium B.C from Tell Asmar
and Khafajah, (Chicago, 1939): pls. 105–114; idem, More Sculpture from the Diyula Region
(Chicago, 43): pls. 63–67; A. Moortgat, The Art of Ancient Mesopotamia (London, NY, 1969):
pls. 35, 42–49, 109–121; I. Fuhr-Jaeppelt, Materialien zur Ikonographie des Löwenadlers Anzu-
Imdugud (München, 1972), passim.
882 chapter thirty-two
figures, rather they seem to be confined to statues in the round to hold inlays.
As for the deviations mentioned above for the bird monster, aside from the
presence of the goats and the beard (it is definitely undercut to indicate a
beard and is a misunderstanding of the hatching directly below the feline’s
nose-mouth area noted on most representations of the creature), note the
crudely drawn wings, the truncated bodies of the grasped animals that sink
below the edge, and their indistinct heads.
The absences of parallels in style, iconography, and execution are all the
more telling when one considers the essential—but never mentioned—
fact that no lapis lazuli disc with an Early Dynastic scene has ever been
excavated, anywhere.
Even keeping in mind only the short list of deviations mentioned in
the two articles, one wonders why it did not trigger questions or doubts,
especially considering the object’s lack of an archaeological history. Several
answers may be suggested for the omission, but one is explicit in the pub-
lications: it was uncritically accepted as a given, as an inherent truth, and
hence assserted, that the dealer-derived, unexcavated disc originated (i.e.
was manufactured) in the Kerman area of southeastern Iran (Porada 1992:
69f., 72; 1993: 50ff.). The evidence offered for the assertion that an object
with an Early Dynastic design was manufactured outside of its expected
homeland was that the disc resembles another, also unexcavated, lapis lazuli
disc (Louvre, ex-Foroughi Collection; Porada 1993: Fig. 19). This example
depicts what seems to be a scene more at home in Iran than elsewhere, and,
following P. Amiet, is assigned to the Kerman area. In this mode of analy-
sis, the “provenience” of one unexcavated disc is taken to be the same as
the “provenience” of another unexcavated disc, differences in style notwith-
standing. Further, the deviations in style and execution from the corpus
of Early Dynastic artifacts have no authority other than as obvious (after
the fact) evidence and support for a perceived non-homeland source. Thus,
even before an Iranian origin was proclaimed (Porada 1992: 70), it was
stated without evidence and matter-of-factly in defense of one deviation
that it is “probably a type at home in southeastern Iran.” And the histori-
cally significant conclusion that the disc was made in a non-Mesopotamian,
provincial area was posited before stylistic discussion commenced, provok-
ing the recognition of a “close relation to the Mesopotamian works and
their unquestioned combination with foreign elements …,” the latter being
the deviations noted (ibid.: 72). From the beginning, the inscription was
accepted without question as a genuine component of the ancient disc.
The ethnic and geographical backgrounds of the disc having been estab-
lished, a suitable historical scenario was then devised to support them: the
bazaar archaeology 883
3 Opinions concerning the whereabouts of Marhashi are still being expressed, and schol-
ars disagree about its geographical position: see for example the article by P.R.S. Moorey in
the Curtis 1993 volume mentioned in note 1, p. 38; wherever it existed, it has never been
demonstrated that any Mesopotamian army ever penetrated into Iran so far as, or even close
to, Kerman. In both sales catalogues (see note 1) it is stated (with scholarly help?) that the
disc was manufactured in 2800bc and that Rimush “may have” or “probably” acquired the
disc as booty when he conquered Kish more than 400 years later. Speaking only to these
problems, first, there is no evidence that Rimush conquered Kish—Sargon probably accom-
plished that feat. And, second, the translation of the inscription is probably: Rimush, King of
the World/King of the Universe/König des Alls (as Frayne and others have noted).
4 Three cuneiformists examined the inscription at my request; because if the disc could
be a forgery then perhaps the inscriptions might reveal some mistake. Two of the cuneiform-
ists had previously seen it and thought it genuine (one was D. Frayne, see note 1), but
after reexamination they changed their minds. The third—a tenured professor in a major
university—refused information because he said he did not want to offend the disc’s owner
884 chapter thirty-two
Each feather of the bird monster on the other side is delineated by crude
oblique lines; the wings are truncated and unevenly curved on each side; the
ears are pointed and cat-like. The felines held in the bird’s claws are depicted
like those on the Dudu plaque from Tello, with the full face seen from the top
facing the bird, but here not so satisfactorily accomplished: on the disc the
heads are not a pair and are crudely executed; the same is observable for
the asymmetrical front and rear legs and truncated feet; and the tails hang
straight down.
Disregarding that many of the disc’s most striking features, details, and
iconography, are unparalleled in Early Dynastic (or any other ancient
period’s) art, an interpretation has been vigorously offered (in communica-
tions and dialogues) for the iconography is that it is a major example of polit-
ical propaganda created in the Akkadian period (argued thusly because of
the inscription): a bearded Rimush stands equal in height to and facing the
unbearded deity Ningirsu (on the unproven assumption that the Anzu bird
is the symbol primarily of this deity). The grasping of the Anzu-Imdugud
bird monster by both of the figures represents their sharing of earthly and
divine power, truly an extraordinary political event and an unprecedented
scene in ancient Near Eastern art. The execution of an Akkadian period
object in what is normally called Early Dynastic style is perceived in this
interpretation as an overlap of the new political-cultural authority with the
older, but still present, style—and the disc is cited as the very evidence
demonstrating this phenomenon. Questions raised by concern for the lack
of parallels for style and execution of the unique iconography produce the
response that no forger could have conceived the ideologically exquisite
message of Machtkunst—which argument automatically absolves the stylis-
tic and chronological issues from consideration. Indeed, the argument is
precisely that the lack of stylistic and iconographical parallels—in both the
Early Dynastic and Akkadian periods—is the very evidence that indicates
the disc’s uniqueness and therefore its singular importance as an ancient
object.
If one accepts the disc as an Akkadian product, one must confront both
the presence of the apparent Early Dynastic style and the absence of any
other evidence of the continuation of that style into the Rimush period (not
to mention the earlier ca. 60-year reign of Sargon). An Akkadian attribution
can be maintained only if the inscription can be demonstrated to be both
ancient and contemporary with the depiction—that is, not to have been
added two centuries after the carving. On the other hand, if it is assumed
that the disc was made in the Early Dynastic period, then both the aberrant,
unparalleled, style and the unique iconography and ideology, although not
886 chapter thirty-two
the inscription, would have to be accounted for, just as with the first disc
discussed. One thing is certain, however: the proponents of either view will
need to invoke the ubiquitous Deus ex machina known as the Provincial
Craftsman.
The history of both lapis lazuli discs begins in the bazaar, and all evidence
indicates that their style and iconography are unparalleled. Why then, and
by what rules of archaeology/art history, are we required to accept them
as genuine, ancient artifacts? Can they be demonstrated to be ancient?
Clearly no. And it is not so much that their antiquity has decidedly not
been demonstrated—let alone proven—but that no verifiable evidence for
their antiquity has been presented. Given the lack of real evidence for their
antiquity—in this case evidence that would have been provided had the
discs been excavated, the possibility, indeed probability, that the objects are
not ancient must be considered.6
Correctly understood, the dilemma of the lack of verification takes us
beyond specific arguments that the discs may be or are forgeries. For an
explicit consequence of their condition is that although some scholars may
honestly argue that the discs are genuine, or that they are “not sure” the discs
are forgeries, the fact remains that the discs have ceased to be, and cannot
be cited as, artifacts known to have been made in an ancient culture. For,
even if it is argued that it cannot be proven that the discs are forgeries, it
equally cannot be proven that they are genuine. This is significant, and it
manifests that the discs are lost to archaeology—even if in fact they are
genuine! Such is the fate of all similarly situated objects that surface in the
bazaar and exhibit unique, anomalous characteristics that are unparalleled
6 Taking into consideration the possibility that the history of the discs exists solely in the
modern world, some inferences emerge. The discs were probably made in the early 1970s
in the same workshop (in Iran?) but apparently by two artisans. The forgers either knew
of decorated lapis lazuli discs (two are known: one already mentioned in the Louvre; the
other excavated in Egypt—Porada 1993: figs. 18, 19) and used them as models for the new
creations, or they acquired genuine (plundered) blanks that begged for decoration (labo-
ratory analysis may clarify this issue; also needed is an analysis to verify that the material
is lapis lazuli). The forgers attempted to replicate generic Early Dynastic scenes. They bor-
rowed, adapted and enhanced features from the often published Early Dynastic iconography,
viz. decorated stone plaques and the Eannatum Vulture stele. Somehow (through cooper-
ation with a cuneiformist or by library research) the forgers came across the many exam-
ples of Rimush inscriptions and copied them. The Jerusalem disc’s inscription—if not the
other—has been vigorously defended as authentic, but this authenticity may be in a formal
sense, that is, it could have been expertly copied. The forgers may not have been aware of
the chronological/cultural discrepancy involved between Early Dynastic style and Akkadian
inscriptions: or perhaps they did and planned to be clever.
bazaar archaeology 887
in the ancient corpus. This fate is the gift of the bazaar to modern culture,
the claim that all material offered for sale is genuine plundered artifacts,
gifts from the broken earth.
The bazaar model requires that we accept a methodology that incor-
porates unanchored intuition and a priori faith, and rejection of the facts
within archaeology’s care. This methodology instructs us that stylistically
un-canonical, idiosyncratic, and poorly executed artifacts that have their
origin not in mounds and site reports but in dealers’ shops and catalogues,
are ancient artifacts, and that scholars have a duty to study the bazaars,
not dig in excavations, to acquire from these aberrant objects meaningful
knowledge about the history and culture of the ancient world. This is a
model that does not believe archaeological culture is worth preserving, that
it has—or deserves—a future as a serious scholarly discipline.
888 chapter thirty-two
Boehmer, ed. U. Finkbeiner et al. (Mainz: 1995), pp. 449–453. For a fine documentation of
its adepts defending unicums (forgeries), see K. Butcher and David W.J. Gill, “The Director,
the Dealer, the Goddess, and Her Champions: The Acquisition of the Fitzwilliam Goddess,”
American Journal of Archaeology 97 (1993): 383–401.
890 chapter thirty-three
coming The Lie Became Great: The Forgery of Ancient Near Eastern Cultures. Many scholars,
including me, have published forgeries as genuine. In some cases, it was inadvertent—e.g., a
mistake in interpretation—but not necessarily resulting from innocence of bazaar behavior.
My mistakes appear in The Lie Became Great (see above).
3 Of course, excavations provide variations of known iconographies and styles—e.g., a
bowl excavated at Arjan in southwestern Iran: Y. Majidzadeh, “The Arjan Bowl,” Iran 30 (1992):
131–144 (John Curtis reminded me of this vessel decorated with unique scenes). Leaving aside
the fact of excavation, its uniqueness does not raise eyebrows, and every line, scene, etc. is
what we expect in an ancient work (here an Assyrianizing Iranian production).
See also the neo-Neo-Assyrian cylinder seal offered in Sotheby’s Antiquities from the
Schuster Collection (London: 10 July 1989), no. 45, depicting a siege scene; it was recently
made to be sold (like other modern seals exhibited in the same said-to-be antiquities sales
catalogue). E. Bleibtreu also published it (“Pestungen auf neuassyrischen Rollsiegeln und
Siegrlabrollungen”) in Beschreiben und Deuten in der Archäologie des alten Orients: Festschrift
für Ruth Mayer-Opificlus, ed. M. Dietrich and O. Loretz (Münster: Ugarit-Verlag, 1994), p. 10,
no. 4, Abb. 5; see my The Lie Became Great, Assyrian, no. 40, correctly noting “kein Beispel
bekannt … keine Parallelen” but not indicting it. For another neo-Neo-Assyrian seal with a
siege scene published as ancient, see J. Eisenberg, “Glyptic Art of the Ancient Near East—
Part II,” Minerva 9, no. 4 (1998): 11, fig. 32.
excavated in the bazaar: ashurbanipal’s beaker 891
the Miho Museum, who knew that I had different views regarding the vessel’s age. I have been
informed that a monograph on the vessel has recently been published by Bleibtreu—“Ein
Vergoldeter Silberbecher des Zeit Assurbanipals in Miho Museum: Historische Darstellung
des 7. Jahrhunderts v. Chr.”-in Archiv für Orientforschung Beiheft 28 (1999). See also “Museum
Topics,” Miho Museum News 4 (1999): 4.
5 An examination of small-scale figures depicted on excavated cylinder seals or on royal
nide,” Revue du Louvre 4 (1995): 81; V. Donbaz, “A Median (?) Votive Inscription on a Silver
Vessel,” Nouvelles Assyriologiques Brèves et Utilitaires (N.A.B.U.) 2 (1996): 37–39. A fanciful,
undocumented history of the find is given by the antiquity dealer H. Mahboubian, Treasure
of the Mountain (London: 1995). It resulted from an “excavation” conducted by an “archae-
ologist” showing “diligence”—i.e., a dealer in antiquities. Others, including scholars, repeat
892 chapter thirty-three
the dealer’s report of an excavation from (we are generously given two choices) a “cave in the
Zagros mountains, near Khorramabad” or “between Hamadan and Kermanshah” (ibid., pp. 11,
18). The dealer date of discovery is pre-1934 (see the first unnumbered page and pp. 10ff.). In
“Le royaume élamite de SAMATI,” N.A.B.U. 1 (1996): 21f., F. Vallat concludes that the state
mentioned is Samati and that the dates are c. 585–539bc.
An apparent example of a plundered “cave” vessel with an Elamite inscription is illustrated
in Christie’s sales catalogue Antiquities (New York: 4 June 1999), no. 204; it sold for $43,700.
A leaf pattern is incised in its in-curved sides. There is no longer any knowledge of the
provenience of the “cave” vessels.
7 Either way, the author commands us to believe that the (decorated) vessel rested in the
earth until found “durch Zufall”—no doubt by poor ignorant peasants working in their cave
garden but canny enough to help get it to Japan.
8 Bleibtreu cites the appropriate relief literature but betrays no awareness that the depic-
tions differ essentially in all details, including iconography, from the scenes on the Miho
beaker.
excavated in the bazaar: ashurbanipal’s beaker 893
differently and crudely depicted. All hair styles are wrongly or badly exe-
cuted; some figures seem barefoot, others not. In the second row from the
top in Figure 2 is depicted an Elamite driving an Assyrian chariot (note
the wheel studs), outside of which is a warrior (nationality?) standing on
the axle (?), clubbing an Elamite passenger: this anomaly is set amidst a
nonbattle—victorious—scene (note the pipest serenading the slaughter);
the scene is described without comment (p. 28). Note the full body-covering
shields shown in front view.
In the lower scenes, bowmen fold their hands while also carrying bows;
one bowman is beardless, the others bearded. The groveling Elamites have
unique headbands. In the lowest, fourth, scene (Fig. 1), there is an empty odd
chariot driven by an Elamite (p. 28). To the right, the Elamite king’s garment
is wrongly depicted, the side fringe should slope left to right, depictions of
his hat are misunderstood and not uniform and his leg floats in the air; he is
also shown again isolated by the king’s chariot.
The king’s chariot is depicted twice (Figs. 1 and 4), large and overlapping
two zones, as on the palace reliefs. The royal crown is inconsistently drawn.
The parasol forms are misunderstood. Likewise the fan bearers and fans,
which in one case strike the back of the head of the parasol bearer, in the
other cross behind his head and touch the king’s crown; the number of
fan units varies in the two scenes, and the heads of the fan bearers are
inconsistently drawn. The parasol bearers hold neither a parasol handle nor
a fly whisk. The figure walking alongside the chariot wheel in one case grasps
spokes with both hands, politely ignoring the floating pair of bodiless hands
grasping a spoke; on the other chariot, this figure holds a spoke in one hand
while the other hangs at his side; here also in one case one foot is booted,
the other bare. In excavated Assyrian art, this figure holds one spoke, never
two, sometimes also the hub, and in the former position, he carries a baton;
also, his posture is rendered upright, not bent. Behind the lowest chariot
are two figures holding very long, thin, plain poles. Bleibtreu notes (p. 25)
that, “bisher” (wonderful!—in a word disclosing how data are acquired in
bazaar archaeological scholarship), they have no parallels in Assyrian art.
But inasmuch as they are said-to-be ancient artifacts, they are labeled rulers’
staffs. Not recognized is that the arrangement of the scenes above bands
of (poorly executed) palmettes is also (“bisher”?) not Assyrian; nor are the
palmettes Assyrian-they were copied from Phoenician patterns.9
9 See H. Frankfort’s The Art and Architecture of the Ancient Near East (Harmondsworth
and Baltimore: Penguin, 1954, 1963), fig. 94, and S. Moscati, The World of the Phoenicians
(London: 1968), figs. 10, 19, 20, and pls. 1, 7, 92 (the Moscati reference is from Glenn Markoe).
894 chapter thirty-three
Fig. 1. Gilt silver beaker, view 1. Height: 24.5cm. Miho Museum, Japan.
excavated in the bazaar: ashurbanipal’s beaker 895
Fig. 2. View 2.
896 chapter thirty-three
Fig. 3. View 3.
excavated in the bazaar: ashurbanipal’s beaker 897
Fig. 4. View 4.
898 chapter thirty-three
Something is very wrong, but not in ancient Assyria. The scenes were
chased,10 consistent with ancient practice; but not mentioned is a corro-
sion analysis of the chasing or inscriptions, or an investigation of the gilding
process. Absent a sense of archaeology and scholarly insight, the question
“Why is the beaker ancient?” was never asked. To the (anticipated) out-
raged bazaarist defense that one cannot prove by stylistic and iconograph-
ical analyses that the decoration is not ancient, the spontaneous rebound
shouts—and more accurately: one cannot prove that it is ancient. Thus,
whether one accepts modern fraud or ponders an unsure posture, the said-
to-be Assyrian beaker ceases to function in archaeological, cultural, and
art-historical discourse. (Of course, museum and other bazaarist personnel
have their own protocols and will continue to exhibit, publish, and educate
the public on the beaker’s beauty and uniqueness.)
Consider these scenarios. The beaker is probably ancient (is the base
added?), seemingly one of many plundered from western Iran in the early
1990s and smuggled to western bazaars. It bears a Neo-Elamite inscription,
but, being plain, its sales value is restricted. A bazaar-wise entrepreneur
(known in the bazaar and museum world as “an honest and reliable dealer”)
employed a—yes—provincial artisan (we may call him the Miho Mas-
ter) to embellish the begging-to-be-decorated plain vessel with an Assyrian
inscription and scenes loosely copied from (photographs of?) Assyrian wall
reliefs. As expected (from past experiences), the scheme succeeded: the
Beaker of Ashurbanipal, created to be sold, was sold (of course, “in good
faith”), exhibited and praised by several museum directors and curators,
baptized (born again) authentic, and published by a believer. Bazaar archae-
ology, sublime and elegant.11 Or both inscriptions as well as the decoration
10 Richard Stone examined the published photo and determined that the incisions were
chased.
11 The Vienna Museum director Wilfried Seipel also proudly exhibited and published as
ancient other Miho forgeries, equally discovered in bazaar stalls by alert directors and schol-
ars. The catalogue entries written by Trudy Kawami (with “technische Angabe von Pieter
Meyers”) assert as a given—but never demonstrate—their ancient age: no. 6, “Bactrian,” and
nos. 12, 13, 14, 15 (utterly perverse). These repeat what Kawami earlier asserted in Ancient Art
from the Shumei Family Collection (no editor!) (New York: Metropolitan Museum of Art, 1996),
nos. 3, 11, 12, 13; no. 10 here is also a forgery, and no. 7 is a modern pastiche. As a welcome anti-
dote to all this, see U. Löw, Figurliche verzierte Metallgefässe aus Nord- und Nordwestiran
(Münster: Ugarit-Verlag, 1998), pp. 325 ff., 450 ff., 593 (reviewed by me in Archiv für Orient-
forschung 57, nos. 112 [2000]: 187–195), and the Marlik section of my The Lie Became Great
(see n. 2, above).
Forgery factories recruit fairly good (“provincial”) artisans by guaranteeing them lifetime
jobs.
excavated in the bazaar: ashurbanipal’s beaker 899
The tracking of cultural transfers begins with evidence of objects that stand
out from the norm. Exotic materials, new technologies and deviations in
form, style or decoration are the tell-tale signs of imports or foreign-inspired
products of local manufacture. Their presence alerts us to the existence of
outside stimuli which raises questions about the nature of contact and its
impact on the recipient culture. Knowing the provenience of these objects,
their archaeological context, is fundamental to the investigation which they
inspire into cultural exchange and interrelations. Without this information,
research is severely hampered but unexcavated objects can still yield valu-
able, if restricted, insights into the nature of outside stimuli, if evaluated in
an appropriate fashion.
In his 1987 article on ‘tracking cultural transfers’, Moorey explores the
question of whether writing and other aspects of early civilization reached
Egypt from Mesopotamia along with the import of Uruk-type seals, pottery
and iconography that were found in predynastic contexts. Moorey focused
his research on excavated material, but he was also aware that insights into
the character of cultural exchanges and influences on early state forma-
tion in Egypt stem in part from the evidence of unprovenienced products.
Numerous ivory and embossed sheet-gold knife handles and large carved
slate palettes decorated with figural and floral motifs lack archaeological
context. Deemed authentic and assigned to predynastic Egypt on stylistic
grounds, they bear witness to the syncretistic tastes of an Egyptian elite.
Their selection of certain Near Eastern motifs to decorate native ceremonial
and votive objects supports the general conclusion, based on a wider range
* This chapter originally appeared as “Von Bissing’s Memphis Stela: A Product of Cultural
Transfer?” in Culture through Objects: Ancient Near Eastern Studies in Honour of P.R.S. Moorey,
eds. T. Potts, M. Roaf, and D. Stein (Oxford: Griffith Institute, 2003), 109–121.
1 It is a personal pleasure and honour for me to dedicate an article to Roger Moorey about
whom many good things can be said, all true. He is a learned and intelligent scholar, he is
utterly honest and generous without hesitation to all, and he is a paradigm of what a scholar
should be in and outside of scholarship.
902 chapter thirty-four
of evidence, that at the end of the fourth millennium bc, elements of foreign
imagery were adopted by an already complex society and adapted to suit its
own ends. Lack of context may restrict a full analysis of these unexcavated
artefacts but provided they are genuine, they contribute to the reconstruc-
tion of intercultural contact.
Different by definition, objects of foreign or multicultural inspiration
sometimes stretch the boundaries of our existing frame of reference. Exca-
vated material is beyond suspicion. But we are obliged to question the
authenticity of unconventional objects that lack provenience and find
viable criteria that will serve to increase our knowledge of the genuine and
eliminate from the record the false information derived from the fake. Fail-
ure to query and investigate the background of such artefacts can lead to a
skewed understanding of the culture in which they allegedly originated. The
present paper is concerned with one of the many examples where scholars
failed to question their sources. To some extent our understanding of Egypt’s
interaction with areas to the East during the fifth century bc has been embel-
lished by the oversight.
In 1930 Freiherr F.W. von Bissing, an Egyptologist as well as a collector
and vendor of antiquities (Bissing 1930; Anthes 1934: 90), published a stone
relief (called a ‘Totenstele’) of yellow limestone (Fig. 1), which he had heard
about from a dealer named Cassira based in Egypt. According to Cassira’s
testimony, presented by von Bissing with conviction, the ‘stela’ was discov-
ered at Mithrine, the site of ancient Memphis, in 1909. Von Bissing neglects
to mention in his 1930 publication that he owned the stela or when he pur-
chased it from Cassira (cf. Parlasca 1972, below).2 The important fact of the
stela’s purchase and its implications were therefore consistently overlooked
by scholars who cited the relief in subsequent years, accepting it as having
derived from Memphis. Indeed, some took for granted that it was excavated
there, resulting in the name the Memphis Stela by which the object has come
to be known in literature (viz. Anthes 1934; C. Picard 1935; Culican 1965: 154;
Kyrieleis 1969: 39, 148; Parlasca 1972: 76; Gallo 1993: 273; Martin and Nicholls
1978: 66–67; Ray 1988: 273; Briant 1992: 90–91; 1996: 974; Mathieson et al.
1995).
2 When did he purchase the relief? In the photo caption for the piece (Abb. 1), its
hat.
904 chapter thirty-four
gesture according to von Bissing, while the right figure tears at her hair. At
the foot of the bed stand two men, apparently without shoes, each wearing
a knee-length garment over trousers as well as a hat with an appendage very
much like that on the hat of the deceased except that here it clearly hangs
from the back. The two figures visible in the upper corners are described as
sirens without further explanation (ibid. 229), possibly because the one at
the right seems to have a feathered tail and a clawed front foot. The smaller
one to the left overlaps the bed cover to leave room for an open-mouthed
horse that is led by a servant with Egyptian hair. He wears a jacket but no
trousers, hat or shoes.
Von Bissing noted that his relief was strange and its style coarse (ibid.
229), but as he never doubted its authenticity, these observations did not
deter him from attempting to find its place within the artistic traditions
of the ancient world. As a relevant comparison, in particular for the pres-
ence of mourners in the death scene, he cites a fifth-century bc Egyptian
stela with an Aramaic inscription in the Vatican (ibid. 229–230, Fig. 3). The
only parallel for the funeral bed with its upturned head section as well as
for the stool comes from Neo-Assyrian art; namely, Assurbanipal’s garden
relief (ibid. 231). Also considered Assyrian in inspiration are the hair and
beard of the deceased. Von Bissing refers to a statue of Assurnasirpal II and
reliefs of Sargon II from the ninth and eighth centuries bc although he also
notes deviations. The typically Assyrian horizontal division of the beard, for
example, does not occur on the Memphis relief; a deviation which von Biss-
ing explains as a misunderstanding on the part of the ancient craftsman and
construes as evidence for its Egyptian manufacture. Another comparison for
the hair and the dead man’s hat with what he saw as side flaps leads him to
a vessel from the Crimea (ibid. 231, Fig. 4) which appears to show Persians
wearing hats with side flaps—unlike those worn by the Persian tribes on the
Persepolis reliefs that have only one appendage hanging from the back. The
hair of these Persians also compares favourably with that of the deceased
and his attendants on the Memphis relief.
The funeral bed, the stool as well as the dead man’s hair and beard all
suggested an Assyrian background. However, the hats with flaps pointed to
a Persian connection, as did the trousers worn by the male mourners on the
Memphis Stela and the presence of a horse which played an important role
in Persia (ibid. 234). The Memphis Stela also included Greek and Egyptian
elements. The sirens were deemed to be Greek, although von Bissing knew of
no parallels (ibid. 235), and the wailing women recalled Egyptian practices,
although it appears that they were not unknown among Persians (ibid. 231–
232).
von bissing’s memphis stela: cultural transfer? 905
Faced with a ‘stela’ that was said to come from Memphis and had obvious
Assyro-Persian comparanda, von Bissing concluded that the dead man was
an important Persian resident in Egypt who died during the reign of Darius
the Great, at the end of the sixth century bc, when Achaemenid art still
closely adhered to Assyrian prototypes. To him, such works incorporating
Persian, Greek and Egyptian elements were not uncommon in Memphis
and he presented as evidence other examples, a number of hybrid seal
impressions and terracotta figurines of foreigners, all of which were also
unprovenienced and privately owned, indeed, some by von Bissing himself
(ibid. 234–235).5
R. Anthes published the stela in 1934, recording its purchase that year by
the Berlin Museum (no. 23721) from von Bissing. He did not doubt that it
derived from Memphis (1934: 99). This findspot, the material, the coarse
technical execution and the overall style were all indications, he claimed,
of its Egyptian origin, although the rest appeared to him to be strange. The
shape has no Egyptian parallels, although he calls it a stela, and he notes
that the representation of the dead man’s right arm rising from behind the
body is also un-Egyptian. Anthes refers to the rectangular area over the dead
man’s head as a ‘Bahrtuch’ and offers only Greek parallels (ibid. 93, n. 2). Of
the figures, only the mourning females could be considered Egyptian and
perhaps also the servant beside the horse, unless he wears the same clothes
as the male figures below. Anthes does not cite the Vatican stela (ibid. 100)
but agrees with von Bissing that the bed and table and the hair and beard are
Assyrian in style, and that the clothing worn by the dead man and his two
male attendants recall the attire of the Persians and Scythians represented
at Persepolis. Their hats, however, he would not compare with those on the
Crimean vessel although they could fit into an Iranian or Scythian sphere;
to support this view he cites the horse. He too cannot find parallels for the
sirens. Thus, non-Egyptian (that is, Assyrian and vaguely Iranian) elements
5 Von Bissing also published four other photographs at the end of his text, only two of
which he mentions in the text (1930: 236). One is his Abb. 6, a clay head, which he says is
Achaemenid from Memphis. He compares it with a clay head (Abb. 8) which he says is a
Semite (no provenience is given). Both would appear to be forgeries. In the photo credits
on p. 238 it transpires that Abb. 5, a clay horseman, Abb. 7, a clay head, and Abb. 8 are in
his private collection. The last two are listed in the photo captions as from Memphis and,
again, both are surely forgeries. The photo credits fail to mention that Abb. 1, the relief under
discussion here, also belonged to von Bissing. I thank Professor Wildung of the Egyptian
Museum in Berlin for granting me permission to publish this relief, knowing that I question
its authenticity.
906 chapter thirty-four
parallel for the funeral bed was that on Assurbanipal’s garden relief. To
Kyrieleis, the rectangular area over the funeral bed is a cloth characteristic of
Persian custom although in support of this claim, he cited not a Persian bed
but the cloth covering Persian throne seats (ibid. 148). K. Parlasca accorded
the relief only one sentence and a footnote to observe that it derived from
Memphis and depicted the laying-out of a Persian from the time around
500bc (1972: 76 with n. 22). He repeated that the relief was found in 1909
and that von Bissing purchased it from Cassira in Cairo, not previously
mentioned by von Bissing. When he next referred to the relief, he described
it as among the ‘most famous archaeological testimonies of the first Persian
period …’ (1979: 318).
In more recent times von Bissing’s stela continued to be accepted as an
important Egyptian monument from Memphis that merited serious dis-
cussion. According to G.T. Martin and R.V. Nicholls (1978: 66–67), it is a
stela that represents ‘a Persian dignitary in “Median” dress from Memphis’.
They note comparisons with the Persepolis reliefs but also bring in
Graeco-Persian grave reliefs from Daskylion and elsewhere in Asia Minor
and conclude that it ‘shows an essentially Greek concept rendered in a
Graeco-Persian style.’ The relief is considered late sixth century bc in date
because in their view the deceased’s ‘beard and Median dress … and the
curving profile of the horse’s head’ reflect the ‘conventions developed by
Darius at Persepolis.’ The authors relate the relief’s style and technique to
the Vatican stela which, although unexcavated, they also believed originated
from Memphis.
In 1980 (36) I too cited the piece, not as a forgery, for I barely looked at it,
but as a forgery of provenience because there was no evidence that it derived
from an Egyptian site.
In the Cambridge Ancient History, J.D. Ray refers to the stela to demon-
strate that Egypt was a ‘richly cosmopolitan state’ (1988: 273). Again, the
scene is described as showing ‘a foreigner in “Median” costume’ but Ray
ventures further to propose an historical interpretation, suggesting that ‘the
representation could be taken as a likeness of Atiyawahi or his brother’
(Egyptian children of a recorded Persian-Egyptian marriage) and that, more-
over, ‘the artist may have been East Greek.’
P. Briant accepts that the subject of the Memphis relief is a Persian noble,
but following Picard, he compares the scene with that of a Greek prothesis
(1992: 90–91). Later, he refers to the same relief as an ‘intéressant document
funéraire’ that was found at Memphis and portrays a man clothed as a
Persian or Mede (1996: 974). Gallo, in his article on stelae, merely mentions
the relief as being from Memphis and depicting the prothesis of a Persian
908 chapter thirty-four
8 For the predisposition of scholars to equate dealers’ claims about proveniences with
archaeological reports, see Muscarella 2000a: 13–14, passim. See also idem (2000b: 29–37).
von bissing’s memphis stela: cultural transfer? 909
9 I became interested in the Memphis relief after my involvement with the so-called Per-
sian Princess, which I determined was a forgery. Because some individuals, including several
Zoroastrians in California and several newspapers, asserted that the mummy documented
Persian practice of mummification after the Egyptian fashion, I decided to investigate. It was
then that I recalled the von Bissing relief in Berlin and began to look into its background
(more fully than I did for my 1980 paper!). For the true history and uncovering of the forgery
see Romey and Rose (2001). The creators of a TV documentary about the mummy give an
inaccurate account.
10 The Saqqara stela is made of Tura limestone. That of the Memphis Stela is described a
circlet around the head of the enthroned Persian, the drape and mattress
on the chair on which he sits, the torque or ring presented to him,
the second offering table and the finger position which seems closer to the
Assyrian manner (Mathieson et al. 1995: 30–31). All other elements of
the funerary scene can be traced in Late Period Egyptian or Achaemenid art
while some, like the winged disk and the lion bier, combine characteristics
of both (ibid. 28, 31).
On von Bissing’s ‘stela’ viable and recognizable elements and compar-
isons are few and imprecisely rendered. The two ‘sirens’ and the rectan-
gular bed cover have no parallels. The stool/table, although considered
Assyrian, is also without analogues and like other sections of the relief
seems to have been consciously mutilated to suggest age or conceal poor
912 chapter thirty-four
iconoclasm directed at key figures of authority. The funerary monument of a nameless foreign
resident in Egypt is unlikely to have aroused such feelings of aggression and, in any case, the
damage to the Memphis relief is randomly distributed, affecting not only the face and left
hand of the deceased but two of his attendant mourners and a harmless stool/table as well.
von bissing’s memphis stela: cultural transfer? 913
References Cited
Anthes, R.
1934 Fünf Neuerwerbungen in der ägyptischen Abteilung. Berliner Museen Be-
richte aus den Preussischen Sammlungen 55: 90–101.
Briant, P.
1992 Darius. Les Perses et l’empire. Gallimard.
1996 Histoire de l’empire perse. De Cyrus à Alexandre. Paris: Fayard.
Canby, J.V.
1979 A note on some Susa bricks. Archäologische Mitteilungen aus Iran 12: 315–
320.
Culican, W.
1965 The Medes and the Persians. New York.
Gallo, P.
1993 Une stèle ‘Hellènomemphite’ de l’ex-collection Nahhman. Bulletin de l’in-
stitut français d’archéologie orientale 93: 265–274.
Kyrieleis, H.
1969 Throne und Kline. Berlin.
12 Although not funerary scenes, the Petosiris reliefs from Egypt (of post-Achaemenid
date) are also relevant for such discussions. Here we have classic Egyptian-made tomb reliefs
with scenes of ethnic Egyptian craftsmen accurately making Achaemenid-style vessels. See
Muscarella (1980: 28–29, Pls. VIII–IX).
914 chapter thirty-four
Mathieson, I. et al.
1995 A stela of the Persian period from Saqqara. Journal of Egyptian Archaeology
81: 23–41.
Moorey, P.R.S.
1987 On tracking cultural transfers in prehistory: the case of Egypt and lower
Mesopotamia in the fourth millennium bc. In Centre and Periphery in the
Ancient World (M. Rowlands, M. Larsen and K. Kristiansen, eds.). Cam-
bridge: Cambridge University Press, 36–46.
Muscarella, O.W.
1980 Excavated and unexcavated Achaemenian art. In Ancient Persia: the Art of
an Empire (D. Schmandt-Besserat, ed.). Malibu, California: Undena Press,
23–42.
1992 Fragment of a royal head. In The Royal City of Susa (P.O. Harper et al., eds.).
New York: Metropolitan Museum of Art, 219–221.
2000a The Lie Became Great. Groningen: STYX Publications.
2000b Excavated in the Bazaar: Ashurnasirpal’s beaker. Source 20/1: 29–37.
Parlasca, K.
1972 Eine Gruppe römische Sepulkralreliefs aus Ägypten. Forshungen und Be-
richte 14: 72–78.
1979 Persische Elemente in der Frühptolemäischen Kunst, Akten des VII. inter-
nationalen Kongresses für iranische Kunst und Archäologie, Munich, Sep-
tember 7–10, 1976. Archäologische Mitteilungen aus Iran, Ergänzungsband
6: 317–323.
Picard, C.
1935 La satrape au tombeau, relief de Memphis. Revue Archéologique 6: 92–94.
Ray, J.D.
1988 Egypt 525–404bc. In Cambridge Ancient History 4/2, 254–286.
Root, M.
1979 The King and Kingship in Achaemenid Art. Acta Iranica 19, Series 3, Leiden.
Seidl, U.
1976 Ein Relief Darius’ I. in Babylon. Arcäologische Mitteilungen aus Iran 9: 125–
130.
1999 Ein Monument Darius’ I. aus Babylon. Zeitschrift für Assyriologie und Archä-
ologie 89/1: 101–114.
chapter thirty-five
* This article originally appeared as “Gudea or Not Gudea in New York and Detroit:
Fig. 1a. Gudea. Diorite. The Metropolitan Museum of Art, Harris Brisbane
Dick Fund, 1959 (59.2). Image © The Metropolitan Museum of Art.
gudea or not gudea in new york and detroit 917
The Background
1 Some of the issues mentioned here have been discussed in less detail in Oscar White
Muscarella, “Seated Statue of Gudea,” in Art of the First Cities: The Third Millennium bc from the
Mediterranean to the Indus, ed. J. Aruz and R. Wallenfels (New York: Metropolitan Museum
of Art, 2003), pp. 428–430, no. 305.
2 Questions had been raised about whether the statues had actually been excavated
in the temple or purchased from local Arabs who had previously plundered the site. See
A. Parrot, Tello: Vingt Campagnes de Fouilles (1877–1933) (Paris: 1948), pp. 15–16 and n. 9,
p. 18, and p. 152; F. Johansen, Statues of Gudea, Ancient and Modern (Copenhagen: 1978),
p. 29 and n. 104; and D.P. Hansen, “A Sculpture of Gudea, Govemor of Lagash,” Bulletin of
the Detroit Institute of Arts 64, no. 1 (1988):18, n. 1. Parrot expresses his suspicions by writing
(pp. 160–164) “Trouvée par de Sarzec au Palais (7),” whereas in his reference to a statue
excavated in 1903 (p. 165, no. 13) there is no question mark. I. Winter, “‘Idols of the King’:
Royal Images as Recipients of Ritual Action in Ancient Mesopotamia,” Journal of Ritual
Studies 6, no. 1 (1992):18, and H. Vogel, “Statuen, die sichtbar machen zur Bedeutung der
Statuen des Gudea,” Baghdader Mitteilungen 31 (2000):17, n. 7, accept the Seleucid-palace
locus.
3 Gudea statues were first offered for sale in 1925. For the 1924 plunder, see, among
others, Parrot, pp. 27, 167; Johansen, pp. 5–6, 18–19, 32, who doubts the date because it was a
dealer who reported it (see n. 8, below); E. Møller, “On the Gudea Statue in the NY Carlsberg
Glyptotek,” Assyriological Miscellanies (1980):52–53; also nn. 8–10, below.
4 L.A. Heuzey, Catalogue des antiquités chaldéennes: Sculpture et gravure à la pointe (Paris:
1902), pp. 167–188; Parrot, pp. 160–172; E. Strommenger, “Gudea,” Reallexikon der Assyriolo-
gie und vorderasiatischen Archäologie (Berlin: 1971), pp. 681–682; and E.A. Braun-Holzinger,
Mesopotamische Weihgaben der frühdynastischen bis altbabylonischen Zeit (Heidelberg: 1991),
pp. 158–159, all provide lists of the excavated and not excavated statues. See also Muscarella,
p. 429 and n. 4.
920 chapter thirty-five
Louvre.5 There are also five seated figures in the Louvre (for instance, Fig. 4,
statue I), four of them headless.6 Two isolated heads in the Louvre also derive
from Tello, one comes from Nippur, and three come from Ur.7
All the dealer-derived, unexcavated Gudea statues are in museum collec-
tions; seven are standing,8 and three (possibly four) are seated.9 There are
also about sixteen isolated Gudea heads, more than twice the number of
excavated examples.10
From ancient texts, we know that Gudea built many temples and was
engaged in trade with many distant lands from which he acquired precious
pp. 7–13, pls. 1–16; Muscarella “Seated Statue,” p. 429, n. 5. Statue L is rejected as that of Gudea
by Braun-Holzinger (p. 265), H. Steible (“Versuch einer Chronologie der Statuen von Lagash,”
Mitteilungen der deutschen Orient-Gesellschaft 126 [1994]:83), and C.E. Suter (Gudea’s Temple
Building: The Representation of an Early Mesopotamian Ruler in Text and Image [Groningen:
2000], pp. 330, 335) and is not counted here. U is a fragment in the British Museum that was
excavated in 1850 by W. Loftus at Tell Hamman in southern Mesopotamia.
6 For statues B, D, F, H, and I, see Parrot, pp. 161–165, nos. 2, 4, 6, 8, 13; Johansen, pp. 10–12,
pls. 19–43; and Strommenger, p. 682, 1b 1–5. Statue I is completely preserved (see statue P,
below); but note that most of the excavated figures are headless.
7 Parrot, pp. 168–170, nos. 21, 22 (the latter is bald), 25, and pp. 165, 170–171, nos. 12, 27–
32; C.L. Woolley, The Early Periods: A Report on the Sites and Objects prior in Date to the Third
Dynasty of Ur Discovered in the Course of the Excavations, Ur Excavations, vol. 4 (Philadelphia:
1955), pp. 51, 189, pl. 42; and Suter, pp. 332–333. Most of the excavated figures are headless.
8 Copenhagen’s Nationalmuseet (O), two in the Louvre (N), British Museum (V), the
Cleveland Museum of Art, the Detroit Institute of Arts (M); and possibly an example in the
Golénisev Collection (Braun-Holzinger, p. 267, no. 124, T [I have not seen it]). See Muscarella,
p. 429 and n. 9; most of these have heads. For scholarly discussions about possible forgeries
among the unexcavated corpus of Gudea statues and heads, see Oscar White Muscarella, The
Lie Became Great: The Forgery of Ancient Near Eastern Cultures (Groningen: 2000), pp. 172–173.
9 The Metropolitan Museum of Art (P), Baghdad’s National Museum (Q), and the British
Museum (acquired in the mid-nineteenth century from Mesopotamia) may not be Gudea;
of these, only the Metropolitan Museum statue preserves its head (see Muscarella, “Seated
Statue,” p. 429 and n. 10). A damaged statue in Harvard’s Semitic Museum (8825; new No.
1936.5.1), called R, may not be a statue of Gudea. E. Sollberger, “Selected Texts from American
Collections,” Journal of Cuneiform Studies 11 (1956):11–13, fig. 1, claims the inscription was not
made at Gudea’s order, although it may depict him; he also calls attention to the “queer
attitude” of the torso leaning to its right. Johansen, pp. 25, 40, and Steible, 83, also do not
consider it to be a statue of Gudea. Parrot, p. 171, published it as Gudea, no. 31, pl. XVII, as
does Strommenger, pp. 682–683, d, 1.
10 They are housed in New York’s Metropolitan Museum of Art; Boston’s Museum of Fine
Arts; the Cleveland Museum of Art; Harvard’s Fogg Art Museum; Harvard’s Semitic Museum;
University Museum of the University of Pennsylvania; Norfolk, Virginia, Chrysler Museum
of Art; Paris’s Louvre; Berlin’s Staatliche Museen; Leiden’s Rijksmuseum; Jerusalem’s Israel
Museum; the National Gallery of Victoria, Australia, and Madrid’s Museo Nacional del Prado.
See Muscarella, “Seated Statue,” p. 429, n. 11.
922 chapter thirty-five
11 See the inscription on statue B in D. O, Edzard, Gudea and His Dynasty, The Royal
Inscriptions of Mesopotamia, Early Periods, vol. 3/1 (Toronto: 1997), pp. 31–36. See also Suter
for a good discussion of the political, textual, and material cultural history and records of the
Lagash II and Ur III periods, especially of Gudea’s time, and J. Evans, “Approaching the Divine
…,” in Aruz and Wallenfels, pp. 417–424, for a neat, concise summary of the period.
12 For a bibliography on the Sumerian and Akkadian background to clothing, hand posi-
tion, named statues, etc., see Muscarella, “Seated Statue,” p. 249 and n. 12.
13 Parrot, p. 167, pl. XVI a, listed statue P as being in a private collection (read antiquity
dealer) and as “Fouilles clandestine de 1924.” It was first offered for sale to the Metropolitan
Museum of Art in 1934 by a dealer (Feuardent Fréres), in 1937 by another dealer (J.E. Géjou),
and in 1950 by the latter’s daughter; each time it was rejected. In 1958, the figure surfaced
again, in the possession of yet another dealer (E.S. David), and this time it was purchased
(1959). In 1950, the curator (C.K. Wilkinson) who purchased the statue in 1959 thought that
the head was genuine but that the body was not.
14 Thus, statue P shares most of its inscription with statue I but its name with statue C.
“a copy”: the inscription’s ductus (writing style) is aberrant, the stone is not
diorite, the throne’s feet are flat and not organically rendered, and the fur of
the hat is clumsy. In a chapter in the same volume, the cuneiformist A. Alster
accepted Johansen’s condemnation. He believed that the “minimal” change
in the text between statues P and I could easily have been construed by 1930.
Oddly, neither he nor Johansen mentioned the sharing of the statue’s name
with statue C.15
More or less based on stylistic determinations, some scholars supported
statue P’s authenticity,16 others followed Johansen and Alster, and still oth-
ers were unsure whether it was modern or ancient. G. Colbow thought that
statue P could have been modeled and copied from statue I and the inscrip-
tion copied from both statues I and C. To her, this might indicate that statue
P is a forgery but also that it is “wahrscheinlich echt” (probably ancient).17
Others were more certain. E. Braun-Holzinger condemned it on the basis of
the inscription as well as style: poor execution of the eyes, a too fleshy nose,
overly modeled cheeks, wrong ears and hat, and incorrectly rendered cloth-
ing; and both H. Steible and C. Suter, concentrating on the coinciding of the
inscriptions, condemned it as a modem work.18
I find it difficult to reject statue P as a forgery—to accept the negative
stylistic observations expressed.19 I perceive the style and execution of the
face, hands, muscled arms, and feet to be in accord with ancient Gudea
sculptures, and the stone indeed is diorite or gabbro, consistent with the
others. If statue P is a forgery, one has to confront the following possible
scenario: a skilled forger in the 1920s had to have had hands-on access to
statue I in the Louvre, obtained accurate measurements—perhaps made a
mold. Furthermore, he or someone else (see note 25, below) had to have
15 For both, see Johansen, pp. 12, 24–25, 34, 39, 57–58. To me, Johansen’s attempt at a few
stylistic criticisms of statue P has no manifest verity and seems to be not substantive; he
spends no time on the crucial shared-inscription issue; and he is wrong about the stone (see
below). And Alster ignores the stylistic issues raised.
16 Aside from its purchasers, see Strommenger, p. 682, 1b 6; B. Schlossman, “Portraiture
in Mesopotamia in the Late Third and Early Second Millennium B.C.,” Archiv für Orient-
forschung 26 (1978–1979):60, who sees statues P and I as possibly deriving “from the same
workshop”; Møller, 51; A. Spycket, La statuaire du Proche-Orient ancien (Leiden: 1981), p. 190
and n. 29.
17 G. Colbow, Zur Rundplastik des Gudea von Lagas (Munich: 1987), pp. 88–89.
18 Braun-Holzinger, p. 266; Steible, 96–97; Suter, p. 36, n. 26, 33: “an obvious fake” (for the
I thought it a clear-cut forgery, I would readily proclaim it as such loudly and clearly.
gudea or not gudea in new york and detroit 925
24 Quoted from one of Claudia Suter’s personal and generous communications with me.
See also Vogel, 67. Vogel does not mention the specific text problem concerning statue P.
25 This signifies—if this Gudea is a forgery—that a Sumerian scholar must have informed
and guided the forger about copying the inscription of statue I up to the point where the
statue’s name appeared, and then directed him to another statue in the Louvre, statue C, and
the position where that statue’s name existed!
26 V. Scheil, “Une nouvelle statue de Gudéa,” Revue d’Assyriologie et Archéologie orientale
20, no. 2 (1925):41–43. Little of the substantive issues I raise here appears in the entry for this
statue in the Metropolitan Museum catalogue mentioned in note 1, above: E. Peck, “Standing
Statue of Gudea,” pp. 430–431, no. 306.
27 G. Contenau, Manuel d’ Archeologie orientale …, 4 vols. (Paris: 1927–1947), II, pp. 721–722;
Parrot, pp. 165–166; H. Frankfort, The Art and Architecture of the Ancient Orient (Baltimore:
1954), pp. 47–48. For the otherwise common use of diorite, see above.
gudea or not gudea in new york and detroit 927
is in the Department of Ancient Near Eastern Art of the Metropolitan Museum of Art and is
dated 19 February 1973. Edzard’s criticism of the inscription has not hitherto been mentioned.
32 Their opinions were given in communications, dated 3 February 1977, on file in the
Metropolitan Museum of Art. Also mentioned by them is the name of a Sumerologist who
was considered by scholars in Copenhagen to have been the one who aided the forger in
928 chapter thirty-five
creating the inscription on the Gudea statue in their museum, a statue that they and others
think is a modern creation. See Johansen, pp. 14–24, 40.
33 Ibid., pp. 52–53.
34 Spycket, p. 190; Colbow, pp. 85–86, 88–89; Møller, 51; Hansen, 14–17. Hansen also men-
tions that there is a “curious aspect” of the statue, a misalignment of the vertical axis, which
is a “difficult trait to explain.” And I find it difficult to understand—whether or not the statue
is ancient or modern—how one could claim that the statue is “masterpiece”: to my eyes, the
piece is grossly sculpted and inferior in aesthetic aspects to Gudea statues, not to mention
other ancient productions. It would be self-serving to label the statue a provincial ancient
work, ancient but uniquely uncanonical: no scholar can demonstrate this for an unexcavated
piece, and it is an old argument used to defend the indefensible.
35 P. Amiet, “La Statue d’ Ur-Ningirsu reconstituée,” La Revue du Louvre 24 (1974):243;
Braun-Holzinger, pp, 158, 265; Steible, 84, 100–102; Suter, p. 36, n. 26.
gudea or not gudea in new york and detroit 929
plundered in 1924, and the inscription added in modern times. All the Tello seated statues
have inscriptions; a seated statue in the British Museum, perhaps of Gudea (see n. 9, above),
bears no inscription (Johansen, p. 14, pls. 53–54).
37 See Muscarella, The Lie Became Great, p. 173, and above.
chapter thirty-six
It is essentially the same paper delivered in Bochum, with additions and bibliography.
(The Bochum Conference was one of the most vigorous and open-minded I have ever
attended. Frankness and a willingness to share different, even opposing views, frankly and
as colleagues prevailed throughout: thanks to our hosts, Alexander and Silke von Berswordt,
and their helpful assistant, Hilke Wagner). I wish to thank two conservators and a scholar
for reading this paper and making valuable suggestions—and for correcting and modifying
some scientific misunderstandings on my part.
1 Muscarella 2000, and bibliography on pages: 223–224; for use of the words artifact and
causes for its existence; and the degree to which forgeries have corrupted
our knowledge of the world’s ancient material and spiritual past.
I submit also that among scholars and the public at large there is another
crucial related problem about which they remain quite uninformed. This
concerns the methods employed to detect forgeries other than the hands-
on, direct, visual investigations anchored in the empirical knowledge of
archaeologists and art historians, an undertaking usually labeled (some-
times pejoratively) “connoisseurship.” I address here the contentions of
some scientists and conservators that their technical analyses and conse-
quent conclusions are based on “objective” scientific analysis, and thus are
more accurate than connoisseurship observations, and they are complete,
and (in some instances) final rulings.2 Scholars, students, and the public
are conditioned to believe such claims, that a scholar’s “subjective” inves-
tigation must yield to a conservator’s “objective” determination, which is
invoked as impartial, regarding whether an alleged antiquity under inves-
tigation is ancient or modern, even if the conservator is a paid employee of
the object’s owner (viz. McDonald 2006; Spier 1990: 627; and infra). And it
took me some time to comprehend this perception, to recognize that con-
clusions submitted in scientific or conservator reports can be as incorrect, or
in conflict with conclusions of their colleagues, as are empirical, “connois-
seurship” determinations by archaeologists and art historians. Further, some
scientists/conservators casually segue from an apparent technical/scientific
analysis into an interpretation or explanation about technical or stylistic
oddities or aberrations, issues not strictly within their purview (viz. Mus-
carella 2000: 7–38, nos. 40, 41 for examples). The assumption that “subjec-
tive” stylistic analysis is distinct from “objective” scientific analysis is a false
dichotomy—which a number of honest conservator/scientists recognize.
The individuals involved in both forms of analyses are subject to fallibility,
ignorance, and error—as well as personal considerations, which includes
fear, intimidation, and dishonesty.
Conservators, like archaeologists and art historians, have no magic
machines; they are not infallible engineers. J. Spier (1990: 623) put it directly
and accurately: “Many technical and scientific studies, however, are not con-
clusive, especially in determining authenticity, and often appear to
be invoked by archaeologists as a desperate appeal to the unattainable,
2 Some individuals involved in testing antiquities are primarily scientists, and call them-
selves scientists or scientist-conservators; others, more involved in an history but who also
have some scientific training call themselves objects conservators. I use both forms here.
the veracity of “scientific” testing by conservators 933
‘objective’ result rather than as a proper study.” Spier 1990 is one of the few
scholars who raise this significant issue in print.3
In addressing this problem, four interrelated matters must be considered.
First, and culturally significant, the issue of forgeries of ancient artifacts
involves only objects that have not been excavated by archaeologists at a
known site.4 Involved are exclusively objects that have surfaced in bazaars
and in dealers’ shops (all over the world) or offered as alleged family heir-
looms, all objects that are subsequently sold and exhibited by collectors
and museums. Second, all antiquity dealers and auction houses insist that
they never sell forgeries (but they say others do). And every antiquity pur-
chased by museum curators or collectors is ipso facto claimed by them to
be ancient. Third, the skills and the scholarly and scientific knowledge of
forgers functioning all over the world cannot be overestimated. Nonethe-
less, every owner of a challenged antiquity will shout that no forger could
have made the antiquity they purchased, as it is too complex to be conceived
and then made by a modern artisan. The owner will assert that the antiq-
uity is formally correct (two eyes, a nose, mouth …), and that if there are
anomalies, the object is all the more valuable, a rare and thus (financially)
a precious original, an Unikum, made in a provincial workshop (Muscarella
2000: 18). The reality is that many forgers (and their families) have lifetime
jobs for a good reason: museums and private collections are continuously
acquiring their productions. In addition to being good artisans—and per-
tinent to the thrust of the Bochum Conference’s discussion, forgers also
monitor scientific reports on age-determination techniques. And they and
their sponsors employ scholars and conservators both for alleged stylistic
and scientific analyses and advice, and sometimes to help with creating
verisimilitude.
And fourth, pertaining to common perceptions of the academic training
and skills of scientists and conservators contracted or employed by private
collectors and museums, two issues must be emphasized. As with archae-
ologist and art historian connoisseurship studies [Muscarella 2001 (2005) is
3 His article contains a balanced discussion of the merits both of traditional connoisseur-
ship and scientific analyses and disagreements and discrepancies between them, recognizing
the problems with both methods. He supports the former more than the latter with some
reserve—as I myself have done, and he correctly notes that too much unexamined reliance
has been assumed for the latter; see also pp. 626, 628. His report on a fragmentary forged
torso, alleged through a connoisseurship study, to be close in style to the Getty kouros, is less
satisfactory.
4 For exceptions see Muscarella 2000: 181, no. 24, and page 215, note 61, where I refer to
the “salting of objects into sites” by excavation workmen, also mentioned by Unger 1957:6.
934 chapter thirty-six
this, inferring I was speaking the obvious. But indeed, non-scientists/conservators do not
know this: because few conservators openly write or speak about such matters.
6 Suzan Mazur “The Odyssey or Stuart Pivar’s Roman Bronze Boy,” on the Internet:
SCOOP, December 22, 2006: 7–9; see again Spier 1990: 627.
the veracity of “scientific” testing by conservators 935
object his/her museum has purchased. The same situation also exists for
objects donated by a rich donor. I give an example where a museum scien-
tist suffered when disagreeing with a museum curator. This curator believed
that one of his (many curators believe what they curate belongs to them)
antiquities was ancient, but when someone (not a scholar) publicly claimed
it to be a forgery, he publicly accepted the claim. However, when it was sub-
sequently demonstrated by a museum conservator that the object was in
fact ancient, and that the curator had been caught making a mistake in pub-
lic, he actively campaigned to have the conservator fired. The curator then
protected himself by creating a special exhibition for the object, stressing
how and why he always thought it to be genuine.
I turn to a critique of the various technologies that scientists and con-
servators employ to determine the authenticity of various materials by pre-
senting a few examples. I begin with an appropriate model, a 3-dollar United
States bill in my possession (Fig. 1). Everyone, including foreigners, is aware
that no such U.S. denomination exists. But I propose that I could employ
a “specialist” censervator who would write a report stating that the paper
looks good, the number 3 is in all four corners, there is a president’s face in
the center; and, to explain the oddities, conclude that the bill was surely
printed in a provincial mint. Therefore he sees “no reason to doubt its
authenticity.” (I am of course being sarcastic, but not formally inaccurate.)
Terracotta: Thermoluminesence (TL) testing is normally conducted by
scientists to determine the age of fired terracotta artifacts, For many years
it has been acclaimed as a definitive “scientific” technique: “… thermo-
luminesence testing … has become the standard method of testing the
firing dates of ceramic objects. Nearly all important ancient [sic] ceram-
ics are routinely subjected to thermoluminescence testing to remove any
936 chapter thirty-six
doubts regarding their authenticity”; and one of the “best scientific test-
ing techniques for dating …” (Meyers 1997: 314); see also Spier (1990: 628)
who accepts TL unhesitatingly. The reality is otherwise: TL is by no means
as accurate as proclaimed. Some scientists and scholars have known for
decades that TL dates are not per se precise, that the technique is very com-
plex; and there is a statistical error of +/- %8.5. Further, at least recently
some forgery laboratories have engineered a system of falsifying TL results,
so that a forgery baked last week can be treated to reflect an age of millennia.
Forgers also make pastiches by bonding pieces of ancient terracotta artifacts
together with modern units, or by adding ground-up ancient terracotta frag-
ments to objects fashioned from locally gathered clay (see also below). TL
testing will thereby authenticate the modern creation as an ancient artifact
(Brent 2001 a, and b). Forgers all over the world know such techniques, as
also do scientist-conservators, some of whom consciously ignore them for
reasons given above. But not all archaeologists know this.
A said-to-be Hittite vessel in the Cleveland Museum of Art (Fig. 2) is a
manifest, embarrassing forgery; it was purchased in 1985, and published
the veracity of “scientific” testing by conservators 937
as a small photo in 1986 (Muscarella 2000: 143–144, no. 6). I wrote to the
museum’s curator Arielle Kozloff, who purchased it, requesting that the
vessel be TL tested. She responded briefly on a post card that the vessel
had been “TLd” before the purchase by Oxford University. After a further
request, she sent me the Oxford report, which stated that the vessel was
made between 1600 and 300bc.
I contacted the Oxford Laboratory, whose scientist-conservator re-
sponded to my questions that the vessel was tested not for the museum
but for a dealer, who had sent in the sample, and that the sample had been
taken from the base. I should say allegedly taken, for the dealer could have
taken a sample from a genuine Hittite vessel, claiming it to have derived
from the one subsequently sold to Cleveland. I was also informed by the
Oxford conservator that unwitnessed testings and sent-in samples were no
longer accepted (compare, however, her statement to the contrary in Brent
2001b: 31), but the sampling from the vessel’s base was defended: “I doubt
very much that the base is genuine and that the rest of the vase fake but
it is of course a possibility” (on this wrong assumption relating to another
ancient-modern pastiche, see Brent 2001a: 29). My doubts were thus chal-
lenged. I sent Kozloff a copy of the Oxford report requesting that she check
the base-body construction and take a TL sample from the body; there was
no response. And this neo-Hittite vessel is still exhibited in Cleveland to edu-
cate students and the public about ancient Hittite culture—because the
museum’s curator and a scientist-conservator said that it is ancient, than
which there can be no higher authority. I had assumed immediately that the
vessel’s base was the only part of the vessel TLd: because soon after TL was
developed, forgers began to construct vessels built upon an ancient base.
Every forger is cognizant of the curator’s code, to never, ever, damage the pre-
cious plundered vessel you have purchased—take samples for testing only
from the base (some curators are becoming aware of this).
A figurine accorded a Hacilar birthplace by its vendor (Fig. 3) was man-
ufactured to be all the more valuable because it depicts neolithic period
lovemaking, a rarity (none has yet been excavated). It was tested for a Frank-
furt dealer in 1967 by a conservator at the Römisch-Germanisches Zentral-
museum in Mainz (Muscarella 2000: 139–140, 449, no. 58), whose statement
(written in English) is presented as a scientific report—which was shown to
potential purchasers by the dealer. The conservator reported that the loving
couple (and another Hacilar forgery made by the same artisan: Muscarella
2000: 139, 450, no. 59) was “tested … by several methods, some of them chem-
ical.” Not one of the tests is described, although one would assume that TL
would have been the primary test. I quote the conservator’s conclusion: “… I
938 chapter thirty-six
do not see any possibility to doubt that both idols are genuine Hacilar works
of art … [and] are extremely valuable … [and] absolutely unique”—a perfect
example of a conservator obeying and serving his patron’s wishes. Indeed,
the figurine is “unique”; it is an authentic modern Hacilar villager’s creation
made to be sold, which the conservator’s paid-for advertisement is intended
to facilitate. These are but two of many scores of Hacilar forgeries housed in
museums and private collections in Europe, the United States, and Turkey
(Muscarella 2000: 135–141, 434–453).7
A terracotta plaque (Fig. 4) recently offered for sale at Sotheby’s, New
York,8 is labeled Assyrian, dated, we are informed, on the basis of TL testing
to the “1st half of 1st millennium bc, or earlier.” (!) No details are given,
such as who tested it, where, and what specific results were revealed; the
announcement itself suffices. It is to my eyes a badly made forgery, a modern
$ 3 bill forgery (was it radiated?).
A clay plaque (49.5× 37cm.) depicting a frontally nude goddess in high
relief was recently purchased (2003) by the British Museum (i.e. by British
taxpayers; Fig. 5). That it was born in the bazaar, and not excavated, is never
mentioned in the museum’s publications, as such information is considered
to be unimportant by museum staff; it most certainly has not been “correctly
identified as coming from ancient Iraq” (Collon 2005: 5). First published in
940 chapter thirty-six
The Illustrated London News (June 13, 1936) it was soon named The Burney
Relief, after its vendor, but following its purchase the museum’s director bap-
tized it “The Queen of the Night” (Cellon 2005 11). It has recently (casually)
been asserted—“soweit ich sehe”—to be ancient (W. von der Osten-Sacken
2002); also by Cellon (2005). However, from all analysis of all its cultural
and stylistic elements in depth, Pauline Albenda (2005) has argued that the
plaque is most probably a forgery. She discusses the plaque’s modern his-
tory, its many defenders, and an earlier challenger, Dietrich Opitz (in 1937).
She has painstakingly demonstrated that The Queen is utterly unique in
motif, composition, and style, for example no owls are known in ancient
the veracity of “scientific” testing by conservators 941
Near Eastern art. The British Museum curators cite TL testing (from 1975),
which gave a range across 2000 years, 1725–25bc, as conclusive evidence
manifesting ancient manufacture (such a date range per se is not a nega-
tive). Tests were taken from the rear and front of the plaque: but not on the
lady’s body, which may be a crucial omission. I contend that the plaque is not
manifestly ancient, and whether it is a forgery or genuine remains an issue
for scholarly discussion (soweit ich sehe); More refinements have developed
in TL testing, and new testing is therefore mandatory, and also on the body
(for one wonders if the body is a modern addition to an ancient terra cotta
plaque).
In 1953 The City Art Museum of St. Louis, Missouri, purchased a frag-
mentary four-foot high statuette and labeled it as an Etruscan Diana with a
stag (Fig. 6a), and ecstatically proclaimed it to be one of the finest Etruscan
works in existence. When purchased it consisted of many broken pieces,
considered by the purchaser (and intended by the forger) to manifest its
ancient age; subsequently it was assembled by a well-known restorer (see
below, Fig. 6b). Published three times as a masterpiece of Etruscan statuary
942 chapter thirty-six
9 T. Kawami published them as genuine and ancient in Ancient Art 1996: 26–31, and Miho
Museum Southwing, Miho Museum 1997: 51–52, nos. 10, 11; she mentioned but ignored the
lack of excavated parallels.
the veracity of “scientific” testing by conservators 943
Silver: A cast silver plate, once in the Schimmel collection (Fig. 8), can
be considered a forgery on the basis of the animal’s execution and style
(Muscarella 2000: 60, no. 4, 336). Schimmel sent it for authenticity analy-
sis to the laboratory in Mainz (the same mentioned above regarding the
Hacilar figurine, although a different conservator wrote the same mean-
ingless type of report). When the report was shown to me, I was surprised
(I was innocent then) that there was no mention of inter-granulation or
corrosion, etc. But the conservator claimed ipse dixit: “Es gibt keinerlei
[sic] Argumente gegen die Echtheit der Silberschale …” Notwithstanding
this authoritative, “objective” conclusion, another scientist examined it
subsequently and presented definitive evidence that the plate is a modern
construction. Two conservators, two analyses, and two utterly different con-
clusions.
944 chapter thirty-six
10 The object remains unpublished, although photographs are available on the Internet.
11 For a record of one conservator involved see Jean D. Portell, “Art Crimes and Conserva-
tion …,” American Institute of Conservation News, 330: 6: 8.
the veracity of “scientific” testing by conservators 945
12 A. Steinberg et al., in Rodney S. Young, Three Great Early Tumuli, Philadelphia. 1981:
286–288.
13 Compare the conclusions of J. Riederer, “Die Erkennung von Fälschungen kunst- und
15 As for example, the eager acceptance by W.G. Lambert concerning methods employed
at the. University of Georgia, U.S.A. that allegedly authenticate alleged ancient stone artifacts
(N.A.B.U, 2004, 3: 61).
948 chapter thirty-six
16 P. Kuniholm “Dendrochronology and Radiocarbon Dates for Gordion and other Phry-
gian Sites,” Source VII, 3/4, 1988: 6, 8. Unfortunately for archaeological accuracy. Kuniholm
and his Gordion colleagues have ignored this crucial reuse documentation in their later con-
clusions about Gordion chronology: by using the recovered beams in the destruction level as
well as one in the Early Phrygian fortification wall as evidence. Keenan and James present
a strong critique (Mielke less so) of Kuniholm’s analyses and conclusions regarding the dat-
ing of sites from recovered wood. Note also Keenan 2006 (13–16) for a rejection or the recent
dendrochronological dates announced for Assiros in Macedonia by Kuniholm et al. These
dates demand a re-dating of the Proto-Geometric period to the early 1st millennium bc, a
major upward chronological adjustment for that period. Sec also J. Yakar’s acceptance of the
Assiros dendrochronological analysis and its early chronology proclaimed in the same vol-
ume as Mielke 2006 (1–19).
17 See for example its use in determining the age of European furniture in “The Getty, A
Bibliography
Albenda, Pauline 2005: “The ‘Queen of the Night’ Plaque—A Revisit,” Journal of the
American Oriental Society 125.2: 171–190.
Ancient Art from the Shumei Family 1996, Metropolitan Museum of Art (abbreviated
here as Ancient Art).
Born, H. 1987: “Die Bedeutung antiker Herstellungstechniken zur Beurteilung fäl-
schungsverdächtiger Bronzen”, Deutsche Gesellschalt für zerstörungsfreie Prüfung
13: 146–155.
Brent, Michael 2001a: “Faking in African Art,” Archaeology, January/February: 27–30
2001b: “The Limits of TL,” Archaeology, January/February: 31.
Colon, Dominique 2005: The Queen of the Night, The British Museum Press.
James, Peter 2002: “The Dendrochronology Debate,” Minerva July/August: 18.
Keenan, Douglas J. 2002: “Why Early-Historical Radiocarbon Dates Downwind From
the Mediterranean Are Too Early,” Radiocarbon 44, 1: 225–237, “Anatolian
Treering Studies Are Untrustworthy,” available at http://www.informath.org/
ATSUO4a.pdf.
Lapatin, Kenneth 2000: “Proof: The Case of the Getty Kouros,” Source XX, 1: 43–53.
Löw, Ulrike 1998: Figürliche verzierte Metallgefaße, Ugarit-Verlag, Münster.
McDonald, Robert 2006: “Authenticity of Tseng Artifacts Questioned,” Daily Sundial,
Internet, May 4, 2006.
Meyers, Pieter 1997: “Acquisition, Technical Study and Examination,” in Miho
Museum, South Wing, Miho Museum: 313–320.
Mielke, Dirk Paul 2006: “Dendrochronologie und hethitische Archäologie—einige
kritische Anmerkungen,” BYZANZ 4: 77–94.
Muscarella, Oscar White 2000: The Lie Became Great, Styx, Gronigen, 2001 (pub-
lished in 2005): “Jiroft and Jiroft-Aratta,” Bulletin of the Asia Institute, 15: 173–
198.
Osten-Sacken, E. von der 2002: “Überlegungen zur Göttin auf dem Burney-relief,”
in Sex and Gender in the Ancient Near East, eds. S. Parpola and R.M. Whiting,
Helsinki: 479–487.
952 chapter thirty-six
Abstract: For almost 30 years many hundreds of objects in gold, silver, bronze, and
terracotta have been accepted and published by scores of scholars as having derived
from the ancient site of Ziwiye in NW Iran. All of these objects in fact came from
antiquity dealers, none was excavated by archaeologists. The “Ziwiye problem” is
discussed from an archaeological perspective that examines the background for
this acceptance. It is argued that a study of the publications of the alleged find
results in the conclusion that there are no objective sources of information that
any of the attributed objects actually were found at Ziwiye, although it is probable
that some were; that dealers and uncritical scholars together are responsible for
treating the objects as an actual archaeological find from one find-spot; and that the
objects have no historical and archaeological value as a group. Further, it is argued
that the methodology employed by many scholars concerned with the material has
been defective and improper and that many art historians and archaeologists have
generally tended to ignore the serious implications involved in seeking firm cultural
and archaeological conclusions when working with objects claimed to derive from
sites that were not excavated by archaeologists.
The study does not attempt to offer its own conclusions regarding the alleged
nature, chronology, and artistic elements of the objects allegedly derived from
Ziwiye; it is only concerned with an analysis of the scholarly treatment of a mass
of material offered for sale by dealers and its subsequent integration into a putative
reconstruction of the past.
The first published reference to the alleged discovery of the “Ziwiye Trea-
sure” was made by André Godard and appeared in the Musée Cernuschi
Catalogue of the exhibition of Iranian art in 1948 (Godard 1948, 9–14). The
name of the site was not mentioned but was referred to as “une montagne
voisine de Sakkiz.” Godard cited in passing a gold gorget (“pectoral”), and
he cited in the catalogue, without photographs, three objects “trouvés aussi
dans la region de Sakkiz.” Number 5 in the catalogue was an incised red ter-
racotta vessel with a duck’s head protome that served as a spout; No. 7 was a
* This article originally appeared as “ ‘Ziwiye’ and Ziwiye: The Forgery of a Provenience,”
pair of silver bracelets, one with lion, the other with ram-head terminals; and
No. 8 was a greyware terracotta ram-header vessel. The first two objects were
listed without explanation as belonging to the private collection of Yedda
and André Godard, the third to the antiquities dealer, Rabenou. Neither
Godard nor anyone else ever mentioned that Godard collected antiquities;1
nor did Godard state in this 1948 work—or elsewhere—that Rabenou was
both the dealer most intimately involved with the sale of objects said to
come from Ziwiye and the commercial dealer working at the site itself (this
was not mentioned in print until Dyson 1963, 34). The two silver bracelets
and the duck’s head spouted vessel were mentioned again in Busagli 1956,
Nos. 204 (pl. XXVI) and 208, as still in the Godard Collection;2 and the ram-
headed vessel seems to be the example published in Godard 1950, fig. 57, but
there said to be in the Teheran Museum.3
The first scholarly report on the find was made by Godard before the
Académie des Inscriptions et Belles Lettres, which was published as a com-
munication in their 1949 report (Godard 1949). In this report Godard gave
no specific details regarding the find except to point out that it was made by
local villagers who divided the spoils, which were subsequently dispersed
(presumably to dealers). He mentioned certain objects said to have been
found in a bronze container by the villagers but gave no illustrations, and
he presented his opinions about the artistic attributions of the objects, their
chronology, and importance. This information was repeated in Godard 1950
and need not be further discussed here.
On April 8, 1950, in the French magazine France Illustration, Yedda
Godard published for the first time the alleged account of how the find was
made:
1 Thus Godard 1950, fig. 49 is obviously the same vessel as Godard 1948, No. 6, although
no mention is made of this fact: see here also note 8. Dussaud 1951, 287–288, stated that
Godard got the Teheran Museum to purchase “l’ ensemble;” see also Samedi 1960, 18; Bussagli
1956, 119 implies that Teheran recovered all the finds (even though he includes two objects
from Godard’s private collection!); Ghirshman 1950, 181 stated that the Teheran Museum
acquired “les pièces capitales,” white Ghirshrnan 1961, 81, and 1964b, 6, claim the major part
was acquired by Teheran. Note also the less than candid comment of Y. Godard 1950b, 714
(quoted in the text). Furthermore, Wilkinson’s statement (1975, 7) that Godard “did his best
to conserve the treasure in the country of its origin,” does not seem to be supported by the
facts.
The need to refer repeatedly to publications in the body of the text in order to document
most conveniently the history of the Ziwiye affair required a deviation from the usual JFA
format for references. A full bibliography is appended at the end of the article.
2 Ghirshman 1964a, 113, mentions silver bracelets but does not give their location; Ghirsh-
man 1973, No. 110, refers to a silver bracelet with a ram’s head terminal in the Teheran Museum.
3 It is of course possible that the vessel passed from Rabenou to the Museum.
“ziwiye” and ziwiye: the forgery of a provenience 957
II y a pres de trois ans, une jeune berger qui gardait ses moutons sur une
haute colline du Kurdistan vit briller un fragment d’or parmi des débris de
constructions et de fortifications anciennes. Il se ramassa et pensa qu’il prove-
nait d’une fissure apparue à la suite d’un précédent orage. Il appela un de
ses camarades occupé un peu plus haut à déterrer de ces racines d’astragale
qui fournaissent le gomme adragante, et tout deux se mirent à élargir la
crevasse. Ils en sortirent quelques objects qui leur parurent être un don de
Dieu. Craignant pour leur nouvelle fortune, les deux enfants l’enveloppèrent
dans quelques guenilles et songeaient à la cacher de nouveau lorsque survint
un Israélite qui, de son côté, etait venu sur la colline pour y recueiller une
certaine terre propice au nettoyage rituel de ses casseroles. Surpris de la trou-
vaille, il offrit de la troquer contre quelques pièces de monnaie, ce qui fut
accepté.
La nuit suivante le Juif revint sur le site de la découverte, trouva d’autes objets
d’or et les enfouit dans le jardin de sa maison, mais, les enfants ayant raconté
l’aventure, l’affaire s’ébruita, et l’Israélite fut menacé, molesté, pillé. Les habi-
tants du village voisin se rendirent sur le lieu de la fouille et creusèrent, l’un
surveillant l’ autre. D’autres objets d’or furent exhumés, et d’auprès discus-
sions eurent lieu au sujet de ce qui en reviendrait à chacun, à la suite de quoi
tout fut coupé en morceaux et partagé. (Y. Godard 1950a, 331).
Almost three years ago, a young shepherd who was keeping his sheep on a
high hill in Kurdistan noticed a gold fragment glittering among the debris
of ancient buildings and fortifications. He picked it up and thought that it
came from a cleft in the rock that had appeared as a result of an earlier storm.
He called over one of his comrades, who was occupied a little further up
in digging around the roots of an astragalus tree, which supplies tragacanth
gum, and the two devoted themselves to enlarging the crack. They brought
out several objects—to them, a gift from God. Apprehensive about their new
fortune, the two boys wrapped it up in some rags and thought of hiding it
again, when there came upon them an Israelite, who, for his part, had come
up the hill to gather some variety of mould necessary for a ritual cleaning of his
saucepans. Surprised at the find, he offered some money, which was accepted
in exchange for several pieces.
The following night the Jew returned to the site of the discovery, found some
other gold objects, and buried them in the garden of his house; but, since
the boys had recounted their adventure, rumor of the affair spread, and the
Israelite was threatened, harassed, and robbed. The people of the neighboring
village went to the place of the find and dug, each keeping an eye on the
others. Other gold objects were dug up, and, after some discussion about the
allotment of the finds, everything was cut into pieces and distributed.
(Translated by Lesley K. Cafarelli.)
She gave no source for this story, presented as historical fact—a leitmotif
that characterized almost all future discussions of the Ziwiye treasure. Nor
did she explain why the “Juif/Israélite” was so designated, inasmuch as he
958 chapter thirty-seven
presumably was one of the local villagers; perhaps the designation was
added to document the intrigue involved. What is more, she hinted at
incidents of intrigue, jealousy, and revenge, but refrained from discussing
them for fear that “Les lecteurs … croiraient que j’ ai entrepris d’ écrire un
roman policier.”
The same story was retold in English in a slightly abridged version in the
May 6, 1950 issue of the ILN. Here an additional sentence was added, one
of interest: “this priceless find was broken up into small pieces and shared
among all the contestants; and has only recently been reassembled—
though whether in full is doubtful” (Y. Godard 1950b, 714).
For reasons never offered by their authors, the details of Y. Godard’s story
were metamorphosed in subsequent retellings. Thus van Ufford (1962, 25)
claimed that “des laboureurs ont trouvé” the treasure; Samedi (1960, 17)
said it was found by “trois jeunes bergers;” and Kantor (1957, 10) claimed,
although not sure it was true, that “some women … needed fresh earth with
which to repave the floors of their houses. When they dug into the slope of
the nearby mound (sic) …, they came upon a bronze sarcophagus …; it was
crammed with objects of precious metal and ivory. Although the women
hastily concealed their find and their men-folk furtively carried it into their
houses at night, its existence could not be kept secret.” Y. Godard’s “petit
berger” out with his sheep becomes “trois jeunes bergers,” “des laboureurs,”
and “some women,” and the Jew disappears. Even the alleged date of the find,
stated by Y. and A. Godard to be 1947, was not universally accepted: Ghirsh-
man 1950 said it occurred “pendant la deniere grande guerre” (later, however,
1964a, 98, hegave the 1947 date), followed by Schefold (1954, 428); Sulimirski
(1954, 298) gave the date as 1946; and Carter (1957, 107, 112) casually claimed
that it had been made in 1928. Moreover, the alleged specific location of the
find-spot moved about in subsequent publications. In Godard 1949 (168) the
treasure was discovered “au sommet de cette colline;” in Godard 1950 (7–8)
it was changed to “quelque distance du sommet de la colline;” according
to Ghirshman 1950 (181) it was found “dans le flanc d’ une colline;” Porada
(1965, 123) said it was found “below the walls of a citadel;” and Dyson 1963
(34) reported that the villagers claimed it was found “about halfway down
the slope … in a deep gully.”
The first time the scholarly world saw in print any of the objects said
to come from the treasure was in the two Y. Godard publications of 1950.
She published only four gold objects, but let it be known that there were
numerous fragments. Later in the same year the first major description of
the objects was published (Godard 1950) and it is this volume that is usu-
ally cited by scholars as the starting point and as the standard work on the
“ziwiye” and ziwiye: the forgery of a provenience 959
treasure. Godard 1948 and 1949 and Y. Godard 1950a and b are rarely cited
and appear to be generally unknown. Paradoxically, Godard 1950 did not
mention the details of the find discussed by Y. Godard; we are given only
a brief undocumented statement that the objects were found in 1947 in a
bronze container, that they were cut into pieces by the villagers and sub-
sequently dispersed (Godard 1950, 7–8), information already published in
his 1949 report.4 In addition to this information, Godard made a distinction
between objects he claimed derived from the treasure, from the receptacle,
and those objects he claimed derived from the site at large, “provenant de
Ziwiyè,” “sur la colline de Ziwiyè mais ne font pas partie du trésor,” “des envi-
rons de Ziwiyè,” and so forth (Godard 1950, 55, 56, 65, figs. 46, 47, 49, 50, 51,
53, 57, 58; see also Busagli 1956, 122). What evidence, if any, that was avail-
able to Godard for these distinctions was not vouchsafed to us. Moreover, in
a number of cases he did not state the modern provenience of the objects
published.
One receives from this publication the implicit, if not explicit, impression
that the corpus of the Ziwiye material—gold, silver, bronze, ivory, stone,
and terracotta—is being presented. In some instances objects are simply
cited without description and photographs: on page 7 “nornbreuses armes,
poignards …, pointes de flèches et têtes de lances en fer, des fragments de
sièges luxueux, en bois de cypres recouvert de bronze, des bases de toutes
formes …” are mentioned; and on page 9 a gold vase, a gold cordelière, gold
earrings, necklaces, and pins are cited, none of which objects is further dis-
cussed (one of the gold necklaces might be Dimand 1950, 145, left). Other
objects, however, are more fully discussed and are illustrated. Godard pub-
lished the following 14 gold objects:5 a complete gorget (figs. 10–24, 33; also
published by Y. Godard); fragments of a trapezoidal plaque with three dec-
orated registers extant (fig. 25; for the lower part see Dimand 1950, 145); two
joining fragments of a plaque decorated with a hero fighting a (missing)
lion (fig. 27); a fragment of a belt (?) decorated with Scythian style animals
and birds’ heads (fig. 29); two griffin heads (fig. 30) and two lion heads,
4 The omission by Godard is all the more striking because Y. Godard’s two. articles
appeared on April 18 and May 6, 1950, while the manuscript of A. Godard’s book was com-
pleted months earlier, in January of that year (Godard 1950, 126).
5 After some thought I decided to list in the text, rather than place in footnotes, all the
objects with their descriptions and references published by Godard and other scholars (but
see note 13). This choice was made so that the reader could immediately recognize what
object was being discussed and so that he would receive the full impact of the number of
objects introduced over the years.
960 chapter thirty-seven
6 The same bracelet was subsequently published many times, but that of Anonymous
1956, 107, lower right corner, is the most constructive for illuminating how the provenience of
unexcavated objects changes in publication: the bracelet is here said to come from Hamadan
and it is conveniently dated to the Achaemenid period.
“ziwiye” and ziwiye: the forgery of a provenience 961
not, as Ghirshman claimed, from the treasure; his source was “le fouilleur
lui-même” (of which more later). Godard did not discuss why he himself had
not mentioned this plaque in his earlier publication; moreover, he tacitly
accepted as deriving from Ziwiye both the silver plate with gold inlays (“qui
semble bien provenir de Ziwiyè, peut-être même du Trésor,” Godard 1951, 242;
italics mine), and the fibulae, but again not mentioning that he had not
originally published or cited them.
The first publication of material said to derive from Ziwiye by a scholar
other than Godard or Ghirshman appeared in 1952. Wilkinson published
a number of ivories acquired by the Metropolitan Museum of Art, all of
which are of the same type of style as those published by Godard 1950. Some
are obviously part of the same panels, and one was the very same ivory as
published by Godard (Wilkson 1952, fig. on page 234; Godard 1950, fig. 85).
Wilkinson (1952, 239) stated that they “were found in a bronze receptacle
that was probably a coffin of Assyrian workmanship,” obviously following
Godard. He also published a new fragment of the bronze receptacle that had
been acquired by the museum and that was clearly related to the fragments
published by Godard. Wilkinson offered no discussions or opinions about
the nature of the discovery or other objects that allegedly may have been
found at Ziwiye.
A second article by Wilkinson in 1955 published more objects acquired
through purchase or loan by the Metropolitan Museum, “from Zawiyeh …
many of which were found in a bronze bath or coffin made for an Assyrian
prince” (Wilkinson 1955, 213; see also ILN April 16, 1955, 699). These included
the museum’s large fragment of a trapezoidal plaque, similar to the example
published by Godard. This article informed scholars for the first time that
there were in existence fragments of at least two similar plaques, that there
was another fragment of the ibex, stag, en face lion plaque, and that the gold
bracelet in the A.B. Martin Collection was the mate to the one published
by Godard (where only one had been mentioned; Samedi 1960, 18, 24, still
referred to one bracelet). Mentioned without illustration (220) was a gold
fragment in the Pomerance Collection that was part of Godard 1950, fig. 27.
In addition to these fragments Wilkinson also published as deriving from
Ziwiye a gold plaque acquired by the Metropolitan Museum in two parts
and decorated with the bodies of two rampant lions with one face placed
below a tree (figure on page 216), a piece never mentioned before as part
of the treasure (for a further discussion of this piece see below, Section
8).
The following year Wilkinson published two large silver ram-headed ves-
sels, one purchased by the University Museum in Philadelphia, the other
“ziwiye” and ziwiye: the forgery of a provenience 963
by the Metropolitan Museum. Wilkinson cited both Ziwiye and the appar-
ently clandestinely dug site of Kaplantu, three miles SE of Ziwiye (cf. Godard
1950, 6–7, fig. 1; below, Section 7): “That the two silver vessels came from the
area of these villages is practically certain” (Wilkinson 1956, 9). No docu-
mentation for this strongly worded assertion was offered, nor was it noted
that neither Godard nor Ghirshman had mentioned such objects in their
reports (see also Wilkinson 1967; Shepherd 1966; Ghirshman 1961). That this
type of strong but undocumented certification has itself generated other
equally strong statements with the same thrust is neatly illustrated by the
comments in Mallowan and Herrmann (1974, 55): “We may also feel some
assurance in accepting as genuinely belonging to the collection a number
of objects which reached the Metropolitan Museum, New York, during or
immediately after the early stages of licensed digging.”
In 1960 and 1963 two more articles on material said to come from Ziwiye
were published by Wilkinson. Wilkinson’s 1960 article was primarily con-
cerned with dating the bronze container in which the treasure was allegedly
found, but a few more objects were also introduced (pl. XXX): an ivory from a
private collection (whose Ziwiye provenience was challenged by Mallowan
and Herrmann 1974, 56); a gold fibula with two lions on the arc, and a silver
example (see also Muscarella 1965, pl. 58, fig. 3; the fibula cited by Wilkin-
son in note 10 is gold, not silver as stated; and note that Ghirshman 1964a,
100 mentions 4 gold and 38 silver fibulae from the treasure); and a silver
pomegranate-head pin. Wilkinson 1963 published for the first time two more
fragments from the Metropolitan Museum’s trapezoidal plaque; a decorated
bronze bucket “reputedly found in the Ziwiyeh area” (Wilkinson 1963, figs. 14,
15); and another ivory fragment (fig. 16).
The most recent report by Wilkinson on Ziwiye was his 1975 publication
of the ivories and miscellaneous objects in the Abegg Collection. Thirty-
seven ivory fragments exist in the collection, twenty-two of which conform
in style to those published by Godard in 1950. There was no hesitation
evident in assigning not only these fragments to the treasure, but the others
as well. Thus, although Catalogue No. 21 was singled out as not belonging
“by style to the group hitherto described,” it is claimed for the treasure: “In
spite of this [the differences], there seems no reason to believe that it belongs
anywhere else than in the Ziwiye treasure” (italics mine; Wilkinson 1975, 53).
That statement implies that there is some reason, albeit not revealed, for
assigning it to the treasure. Likewise, the ivories nos. 24–37, all examples
hitherto unpublished and all of types not reported by Godard 1950, or by
Wilkinson in his previous publications, were incorporated into the Ziwiye
repertory.
964 chapter thirty-seven
Wilkinson 1975 was not to be the last publication of “Ziwiye ivories,” for
K. Ishiguro (1976, Nos. 120, 147) published two from his collection. No. 120 is
a fragment of a winged sphinx in the round (of “ink well” type?); No. 147 is
an ivory chape. The former is a type hitherto not attributed to Ziwiye, while
the latter is of Achaemenid style, and also not hitherto attributed to Ziwiye.7
In the meantime Ghirshman had published his book on Iran for the Pen-
guin series (1954), wherein he discussed as deriving from Ziwiye some of
the objects previously published by him in 1950, still referring to the col-
lection as “The Treasure of Sakkez,” and not mentioning the name Ziwiye.
He did not mention a few of the objects he published in 1950 (figs. 4, 18,
19, 22, 23), which had been challenged by Godard 1951, but he did include
the gold bracelet and belt that were published by him in 1950 and that also
had been challenged by Godard; and he said nothing about fibulae or more
important, about ivories. In passing he published a terracotta ram-headed
situla (112, pl. 13b), attributing it to Azerbaijan although it had been previ-
ously published by Godard (1950, 68, fig. 57), as “des environs de Ziwiye.”
Later (Ghirshman, 1961, 100, No. 610, 1962, 75, fig. 20), ignoring his previous
designation, he moved the vessel back to Ziwiye.8 Ghirshman also contin-
ued to maintain that Scythian writing existed on the silver plate, a position
first advanced in 1950, but dismissed by Godard 1951 (244–245, a reference
never cited by Ghirshman).
Godard had little to say regarding Ziwiye after 1951 (see Godard 1962,
95–96), but Ghirshman continued over the years to write on the subject
and to introduce new objects into the repertory. Many other scholars had
contributions to make regarding Ziwiye and objects said to derive from there
and they will be discussed shortly.
Inasmuch as Ghirshman has been involved in the publication of Ziwiye
material for 25 years, one may follow the basic development of the modern
history of Ziwiye in his works. In 1961 Ghirshman played the major role in
7 See R.D. Barnett, Catalogue of the Nimrud Ivories (London 1975) T9, for a parallel to the
Ishiguro ivory No. 120. I believe that the ivory chape should be tested before it is accepted as
ancient; the eyes and ears are not similar to those noted on excavated examples.
8 This situla is apparently now in the Teheran Museum; it was originally in the possession
of the dealer Rabenou: Godard 1948, No. 8. The other example published by Godard, 1950,
fig. 58, as in his own collection, a fact not mentioned: but see Amiet 1969, 335–336, figs. 19,
20, where it is stated that it was given to the Louvre by Mme. Godard. Note also that Amiet
says that the vessel came from Kaplantu, not Ziwiye, as was claimed by Godard (see note 19).
Tuchelt 1962, 58, No. 6, thought that the Godard situla was in Teheran. Keeping a record of
the modern proveniences of published “Ziwiye” material is not one of the easiest tasks for
those interested in this problem.
“ziwiye” and ziwiye: the forgery of a provenience 965
inconsistency, as should be clear from a reading of the text. Thus Ghirshman 1964a, 120, says
“only one element of psalia … has survived,” i.e. his fig. 540, and 1961 no. 584. He therefore
ignores his own 1961. No. 583 reference and also Godard 1950, fig. 46 (which he does not
otherwise challenge as deriving from Ziwiye); and note that Ghirshman 1973 cites an iron
example in Teheran; see also Porada 1965, fig. 73: and Jettmar 1967, 223!
966 chapter thirty-seven
piece (605); a bronze sphinx (606; same as Amandry 1966, 126, pl. XXVIII,
2a–c); a terra cotta gazelle vessel (608; apparently not the same as Samedi
1960, fig. 34); a terracotta boot goblet (611); other terracotta vessels (612–617,
620, 621); and fragments of terracotta sculptures (618—probably same as
Ghirshman 1964a, fig. 172, 619). Note that some of these objects are in the
Teheran Museum, while others were in the hands of antiquity dealers (“Coll.
particulière”) at the time of the publication.
The pattern established in 1950 of publishing material claimed to be from
Ziwiye did not change, rather it was reenforced: no documentation or expla-
nation to underpin the introduction of new objects was deemed necessary;
one was apparently expected to accept the attributions automatically as an
expression of faith in the implied secret knowledge of the publisher. And
that the knowledge of both the object’s existence and its alleged source came
from a dealer was not revealed.
The exhibition travelled to Essen in 1962 and a new catalogue was issued
(7000 Jahre Kunst in Iran). The Ziwiye section was reduced in size as some
objects were removed by their owners and no additions to the Ziwiye reper-
tory were made. The exhibition next moved to Zurich where a new catalogue
was issued (Kunstschätze aus Iran); many objects not published in the Paris
and Essen catalogues were added in a Nachtrag. In the Ziwiye section of this
Nachtrag 14 objects never hitherto published were introduced; several were
in the hands of dealers. They were as follows: a stone bull’s head (847); a gold
protome (848, fig. opposite page 40); two gold beakers (849, 851, pls. 53, 32);
a fragment of a gold beaker (850); a gold cow (852, pl. 14); two gold goats
(853, pl. 15); a gold stag (854); a gold chain (855); two gold plaques (856, 857);
an ivory plaque (858); an ivory head (859); and a gilt bronze diadem (860).
Only five of the objects were illustrated, and none of these, I believe, should
be considered to be the work of an ancient artist (see below); inasmuch as
the others are not illustrated no comments can be made.
Of special concern with regard to the Ziwiye problem is that following
upon the massive intrusion of new Ziwiye entries in the Paris catalogue,
within a few months 14 additional objects surfaced from dealers and col-
lectors and were rushed into print and placed on view in the Zurich exhibi-
tion.10
In 1964 Ghirshman published The Arts of Ancient Iran which in-
cluded a chapter on “The Scythians and the Royal Tombs of Ziwiyeh.” Here
10 On page 9 of the Zurich catalogue Ghirshman, K. Erdman and P. Amandry are listed as
Ghirshman presented his personal views concerning the nature of the trea-
sure, of which more later. He also introduced more objects into the Ziwiye
repertory, again without documentation: an ivory sword hilt (fig. 155; on
p. 425 it is claimed that a gold pommel with ivory lion heads belongs to this
piece); a second example of a sword hilt (fig. 161); a gold pommel (fig. 158;
cf. Ghirshman 1950, fig. 7; 1964, fig. 157); gold and silver weapons (100); seven
iron spear heads (100, 119, fig. 166); terracotta shields and lion-headed omboi
(100, 321, fig. 392), 11 bronze bells (120; cf. Ghirshman 1961, no. 582, where only
one is mentioned); stone vases (123); silver bracelets (113); terracotta animal
headed vessels (figs. 395, 396); and bronze and bone trilobate arrows (119,
fig. 165; see Dyson 1963, 37).
A number of photographs published in this 1964 work were of objects first
cited by Ghirshman in 1961 and earlier, but inexplicably no mention of this
fact occurs and one must painstakingly check each object against previously
published material.11 Moreover, both the descriptions and measurements
of what one believes to be the same object in separate publications vary,
which makes it difficult to be sure whether one has recognized correctly
the same object in its multiple publications, or there are two separate but
similar objects under discussion. It should also be noted that relatively
few of the countless objects in the Ziwiye repertory have been published
with a photograph. Therefore, it is possible that in some instances my
claim that an object is being cited of published for the first time may be
incorrect.
The problem regarding the ability to recognize whether or not an object
had been previously published, and the problem regarding the introduc-
tion and publication without documentation of material claimed for Ziwiye,
is succintly exhibited in Ghirshman 1973. There 341 objects in the Teheran
Museum, and claimed to be from Ziwiye, are published without illustra-
tions and in simple list form. No information concerning the dates and
sources of the acquisitions of the objects is given.12 Of the 341 objects I
11 In the text in my discussions of the 1961 material I add the 1964a references where I
recognize them.
12 Many scholars assume that if objects said to come from a given site are in the collection
of a Near Eastern national museum that the objects were scientifically excavated. But officials
of these museums often purchase objects from peasants or dealers, and, like their colleagues
in the West, supply provenience labels supplied by the vendor (Muscarella, 1977). Samedi,
1960, 18, for example, specifically states that the Teheran Museum purchased the Ziwiye
material; so also state Dussaud 1951, 287 f. and Bussagli 1956, 119; see also Melikian 1961, 67.
How many objects published by Ghirshman 1973 were acquired in 1950 and how many were
acquired in subsequent years is at present unknown outside of Teheran.
968 chapter thirty-seven
have been able to conclude tentatively that over 121 entries (some of which
represent more than one specimen) have apparently not been previously
published, that approximately 32 may have been previously published, and
that apparently the others have been published before.13 Given the meagre
published descriptions these figures may not be accurate. In 16 instances a
footnote indicates a reference to a previous publication but, strangely, in
other instances, although it can be determined that the object had been
previously mentioned in print, no citation or reference is presented. And
in the 32 cases where it is not clear that the objects have been previously
published, it is the apparent inconsistencies in comparison with the descrip-
tions and measurements of previously published pieces that cause one to be
confused.
Immediately after Godard 1950 and Ghirshman 1950 were published, the
scholarly world naturally became excited by the reports and over the years
often referred to the material and incorporated it into their studies. In
addition to the ongoing publications of Ghirshman and those of Wilkinson
cited above, more than a score of scholars had something to say about
the significance of the Ziwiye treasure in the history of 1st millennium bc
Near Eastern art, the various artistic strains recognizable, the chronology
of the deposit and of individual objects, the nature of the deposit, and
so forth. Some of these writings over the next decades were concerned in
varying degrees with some or all of these specific problems (viz. Falkner
1952; Frankfort 1955; Barnett 1956, 1962; Amandry 1958, 1964, 1965; vanden
Berghe 1954, 1959; Parrot 1961; von Ufford 1962, 1963; Porada 1965; Farkas
1970; Mallowan and Herrmann 1974), while others incorporated the alleged
Ziwiye material into their work on other aspects of ancient art (viz. Grousset
1951; Wiedner 1951; Benson 1960; Tuchelt 1962; Muscarella 1965; Culican 1965;
van Loon 1966; Jettmar 1967; Azarpay 1968; Moorey 1971; Hrouda 1971; Burney
13 I will spare the reader a listing of each of these objects except to give their numbers.
Those I suggest were not previously published include at least Nos. 10, 14, 18 (apparently not
the same as Ghirshman 1961, Nos. 534, 535), 19–22, 30, 31, 36–38, 43, 110 (cf. Godard 1948, No. 7;
Ghirshman 1964a, 113), 131, 132, 152–159, 170, 202–204, 252–286, 288–292, 300–303, 305–307,
312, 313, 320–322, 325–336, 355–366, 410, 501, 503, 505, 506, 513, 514, 531 (apparently not the
same as Ghirshman 1964a, fig. 395, which is not mentioned in the present listing), 538, 539,
603–608, and possibly also 507–512, 515–529. Those that may have been previously published
include Nos. 20 (cf. Ghirshman 1961, No. 504), 23, 25 (could be Dimand 1950, 145, left), 32 (cf.
Ghirshman 1961, No. 534), 161–168 (cf. Godard 1950, figs. 167, 168, 169, 207–219 (some of these
appear to have been published by Godard 1950 and Ghirshman 1964a), 314, 315, 348–351; 532
may be the same as Samedi 1960, fig. 32.
“ziwiye” and ziwiye: the forgery of a provenience 969
1972; Phillips 1972; Calmeyer 1973). And still others continued the accepted
pattern of introducing new material by publishing an object acquired from
a dealer as automatically deriving from Ziwiye.
For example, Kantor published, in addition to the Cincinnati Museum’s
fragment of a trapezoidal plaque (1957, fig. 3), four other objects hitherto
not known to scholars: two fragments of gold plaques both depicting heros
killing lions (figs. 1, 2; see also Goldman 1964, figs. 9, 10); a gold and stone
bead necklace (fig. 4); and a gold bracelet (fig. 5). While I can say nothing
with regard to the authenticity of the last two items, on the basis of stylistic
analysis I believe that there is little reason to accept the first two as ancient
(cf. infra). What is more, in 1960 Kantor published a hitherto unknown
rectangular gold plaque depicting winged creatures (pl. I) acquired by the
Oriental Institute of the University of Chicago. To my eyes the details of the
creatures deviate considerably from those on the Ziwiye plaques to which
the Chicago plaque is compared; it equally deviates in style from other,
genuine, well-established, works of ancient art (cf. infra).
In 1957 Dyson published the University Museum’s alleged Ziwiye mate-
rial: more fragments of the ibex, stag, en face lion head plaque (fig. 25); a
fragment that is part of Godard 1950, fig. 29 (fig. 26). Also, for the first time,
he published a gold belt (?) fragment (fig. 26 top; cf. Ghirshman 1961, no. 518,
pl. XLI, 2, there said to be the same as Dyson’s fig. 26, but it is clearly another
fragment); a fragment of a gold plaque (fig. 26 right; cf. Ghirshman 1961,
no. 512, pl. XL, 3, from the Kofler-Truniger collection); and a gold bracelet,
formerly in the Kevorkian collection (fig. 28). Another fragment of a gold
plaque in the possession of the University Museum, and said to be from
Ziwiye, was published on the cover of the Guide to the Collections The Univer-
sity Museum (1965); on the basis of style, I am not convinced that it is ancient
(cf. infra).
Vanden Berghe 1959 (115, pl. 144 a) cited a bronze votive axe that he
claimed had been recently discovered at Ziwiye, but did not inform his
readers who made the discovery.
In Porada 1964 a gold funnel was introduced as being part of the treasure
(no. 437; it is said to be in the Teheran Museum but was not mentioned by
Ghirshman 1973; v. Calmeyer 1973, 143–144). And Porada 1965 introduced as
from Ziwiye an ivory statuette of a male now in the Cincinnati Art Museum
(125–127, pl. 35), and a bone psalion (fig. 73).
In 1966 the Musée Rath in Geneva held an exhibition of what it called
Trésors de L’Ancien Iran. Among a number of objects on view listed as from
Ziwiye, at least eight appear not to have been previously published: a gold
whetstone (No. 598, pl. 50); a gold horned griffin (No. 599); gold lion masks
970 chapter thirty-seven
(No. 600); a gilt bronze band (No. 603); a bronze mirror (No. 604); ivory
pyxis fragments (No. 605); an ivory plaque (No. 606); and an ivory human
mask (No. 607). Only one of these objects is illustrated and all seem to have
belonged to dealers.14
Finally, Amandry 1966 (121–122, pl. XXV a, b, c) published three gold
fragments of plaques decorated with winged lions and human figures, which
he had seen on the art market. To my mind, all three appear not to be
products of ancient artists, a possibility hinted at by Amandry (see infra and
note 21).
2. Ziwiye/Zibia
Because of the phonemic similarity, Godard (1949, 168; 1950, 5–7) professed
that he had recognized in the name of the village of Ziwiye its ancient
name: “que l’on peut identifier avec l’ancienne Zibiè,” mentioned by Sar-
gon II (722–705 bc) as a Mannean fortress. Many scholars have either with-
out reservation or tentatively accepted this equation: Dunand 1957, 287;
Wilkinson 1955, 213; 1967, 5 (but doubts raised in 1975, 9); Barnett 1956,
111–112; Needler 1957, 9; von Ufford 1962, 38; 1963, 101; Boehmer 1964, 19–
20, 21; Culican 1965, 42; Hrouda 1970, 249; Burney and Lang 1972, 135; Mal-
lowan and Herrmann 1974, 55. Aside from Wilkinson 1975, who pointed out
that the name Ziwiye “is an Arabic word that could have been given as a
place name … long after the fortress … had been destroyed (sic) and for-
gotten,” only Levine (1974, note 108) expressed reservations about the equa-
tion.
Ghirshman, on the other hand, laid stress on what he considered to be
significant, namely that Ziwiye was near the town of Sakkiz (it is actu-
ally about 40km. to the east of that site), and he emphasized what was
important about this proximity. To Ghirshman (1950, 201–206; 1954, 106–107;
1964a, 98, here with some reservations expressed), Sakkiz is to be iden-
tified with the Scythian capital of Partatua and his son Madyes. (It will
be recalled that in his earlier writings Ghirshman referred to the finds
as “le Trésor de Sakkez,” only later citing them as from Ziwiye.) Frank-
fort (1955, 205) accepted Ghirshman’s identification, while Schefold (1954,
14 I can give no opinion here regarding the authenticity of these objects, as I saw them too
long ago. I am not sure if No. 598 is genuine or not (Muscarella 1977, note 81).
“ziwiye” and ziwiye: the forgery of a provenience 971
1965, 207; Muscarella 1966, 381). Amandry (1965a, 897; see also 1965b, 150),
although accepting the possibility that “le ‘trésor’ accompagnait un mort
dans la tombe,” pointed out (contra Ghirshman) “on ne connait ni des con-
ditions ni même Ie lieu de la trouvaille. La restitution de rites funéraires,
fondée sur les objets retrouvés (sacrifice de chevaux, d’ esclaves ou de con-
cubines), comporte une grande part d’hypothèse. Au demeurant, les Scythes
n’ étaient pas seuls à immoler des chevaux dans les tombes.” This provides
a neat summary of the situation, indeed, although the last sentence implies
that horse bones were found at Ziwiye, which is not the case, as Moorey
(1971, 260) pointed out.15
15 Moorey also stated that no weapons were reported, which seems to be an error as
Godard, Ghirshman and vanden Berghe have cited them; and Dyson actually excavated three-
flanged arrows there.
974 chapter thirty-seven
5. The Chronology
Because of what appeared to him as stylistic parallels with the art of 9th
century bc Assyria, Godard (1949, 169–172; 1950, 10, 36–44, 45, etc.) dated the
treasure to that time, singling out for special discussion the gold objects,
the gorget, trapezoidal plaque, the gold bracelet and some of the ivories;
other ivories he dated to the 8th century bc. The date of the deposition of
the treasure, as distinct from the date of the objects, he placed in the late
7th century bc (1950, 123).
Ghirshman’s conclusions, published the same year as Godard’s book, and
citing some of the same objects singled out by Godard, were that on the basis
of style and because of the presence of fibulae, the treasure could not be pre-
8th century bc in date, and that it was probably to be placed in the second
quarter of the 7th century bc (Ghirshman 1950, 197–198, 202); the deposition
was given as ca. 625bc, at the time when the Scythians were forced out
of Iran by the Medes. Ghirshman maintained these dates in subsequent
publications. Godard (1951, 242) reacted immediately to Ghirshman’s low
dating for the treasure, rejecting the chronological conclusions based on the
fibulae (but, as noted above, neglecting to explain why he did not mention
their existence: see Muscarella 1965, 238, note 39).
In the many subsequent discussions of chronology, only two scholars to
my knowledge accepted Godard’s high dates for the manufacture of some
of the Ziwiye material, Parrot (1961, 142, 144), and von der Osten (1956, 57),
while a large number accepted Ghirshman’s low dating: Sulimirski (1954,
313–314), Wilkinson (1955, 218; 1960, 219–220), Barnett (1956, 112; 1962, 94),
Needler (1957, 9), Amandry (1958, 16), van Ufford (1962, 38), van Loon (1966,
177). Other scholars opted for a general 8th–7th century date (Kantor 1957,
14; 1960, 1), Huot (1965, 140), or a more specific late 8th century designation,
Hrouda (1970, 222, 249), Moorey (1971, 212, 260), Calmeyer (1973, 96, note 67),
and apparently also Mallowan and Herrmann (1974, 56). At least five schol-
ars preferred to recognize one or more objects as having been made in the
6th century, or preferred 7th–6th century general dates, Falkner (1952, 131–
132), von der Osten (1956, 57), Azarpay (1968, 46), Schefold (1954, 428), and
Benson (1960, 64, note 32); at least one of these scholars, Schefold, believed
that some of the material represented an early stage of Acnaemenid art.
recumbent stags and ibexes and en face lion heads, the gold ibex, several gold
strips and buckles, the silver inlaid plate, the stone seal and the bronze and
bone horse psalia. And it is in this particular area, as much as any involved
in the Ziwiye problem, that the true provenience and chronology of the
alleged Ziwiye material is of great significance in the history of ancient art.
For if the objects mentioned above in fact derived from one site in Western
Iran, and if they were deposited together before ca. 625–600bc, then they
legitimately could—should—be brought into a discussion involved with
the origins of Scythian art, which many authorities maintain manifested
itself in the Scythian homeland only during the 6th century bc, i.e. after the
alleged date of the deposition of the “Ziwiye Treasure.”
The theory that Scythian motifs and art were unformed and non-existent
until the Scythians came into direct contact with Iranian art during their
sojourn in Iran was first introduced by Godard (1949, 169–170; 1950, 24, 44–
49, 59–64). Given the high 9th century dating of the Scythian material, and
given the much later appearance of Scythian art in Russia, Godard’s conclu-
sion was inevitable. In the same year that Godard’s statements were enun-
ciated, Ghirshman (1950, 202) came to the opposite conclusion. Because he
attributed the material to a lower date, one closer to the traditional dates
for Scythian art, he concluded that the Scythians actually brought their art
with them from their homeland and that the Ziwiye treasure was the earliest
extant examples of that art. Later, however (1961, 82; 1964a, 327), he reversed
himself (without calling attention to his previous opinion) and accepted
Godard’s theory (without citing him); his low chronology still gave priority
to the Scythian elements at Ziwiye.
Whereas certain scholars involved with the Ziwiye material avoided com-
ing to any conclusion regarding the origins of Scythian art, others expressed
their opinions, some vigorously. Sulimirski (1954, 313–316), Potratz (1963,
97–107, 111, 116), Amandry (1965a, 895–896, 903–904; 1965b, 150–152, 159–
160), and Phillips (1972, 136), supported the Godard theory; Burney and Lang
(1972, 175) simply stated that the earliest Scythian art is to be found at Ziwiye,
but drew no further conclusions. One of the first to oppose the Godard the-
ory was Falkner (1952, 132) who stated that Ziwiye alone in Iran produced
Scythian elements,16 and that it is strange that Mannaean elements appear
at the same time as the Scythian incursion. Barnett (1955, 114) also reacted
16 Scythian style bone psalia have since been excavated at Hasanlu from Period III: R.H.
Dyson, Jr., ILN, Sept. 12, 1964, 372, fig. 3 (incorrectly labelled as from period IV: see Muscarella
1974, 79, note 16); see also note 18.
“ziwiye” and ziwiye: the forgery of a provenience 977
6. Doubts
At this point in our itinerary through the Ziwiye maze the reader may feel
overwhelmed both by the investigation of various scholars’ opinions, embel-
lishments, conclusions, omissions, disagreements, and so forth, and by the
listing of the many hundreds of objects that have surfaced over the years
and that have been assigned to Ziwiye. It is now pertinent to ask if anyone
17 Viz. Artamanov, which reference I know from Farkas 1970 and van Loon 1970, 69, note 23.
(I belatedly came across S. Sorokin, “The Curled-up Animal from Ziwiye,” [in Russian] Soob-
schenia Gosudarsvennovo Hermitazha 34 [1972] 75–78, where the gold pommel published by
Ghirshman 1964a, fig. 158, is used in an argument that the animal style of the Scythians had
its origins in the Near East.)
18 Unfortunately van Loon assumed that bone psalia of Scythian style had been excavated
challenged the assumption that all the ivories attributed to Ziwiye derived
from that site: “these ivories came not from one, but from many findplaces”
(57); but at the same time they accepted a number of examples, especially
those published by Godard and most of those published by Wilkinson, as
in fact coming from Ziwiye. Moreover, as mentioned above, they accepted
with “some assurance” a Ziwiye provenience for objects acquired by the
Metropolitan Museum. And this writer has also occasionally called atten-
tion to the problems of attribution or provenance (Muscarella 1965, 233,
note 6; 1966, 380–381; 1976, 210).
It must be stressed again, however, that all of the scholars who had
reservations of various degrees nevertheless at one time or another either
accepted some or most of the published material as coming from Ziwiye, or
in some cases themselves introduced new material as coming from there.
Thus, their doubts were primarily ad hoc ones, reservations against individ-
ual objects, or they were general ones, left isolated and not developed as a
problem probably effecting the whole Ziwiye corpus.
7. Kaplantu/Qaplantu/Ghaflantu
(No. 621), and eight stone ram heads (No. 623); also published here was the
gold rhyton of Ghirshman 1961, and another gold disc with a hero hold-
ing lions (No. 622). A gold finial in the Los Angeles County Art Museum,
formerly in the Berg collection, is listed as “Found at Ghaflantou, Iran” (Mus-
carella 1977, no. 168). And finally, in 1968 the Reallexikon der Assyriologie
und vorderasiatischen Archäologie published its entry on “Ghaflantou” (Seidl
1968, 344–346) wherein it is stated that “wenigstens 13 figürlich und orna-
mental verzierte Fundstücke bester Qualität aus dem I. Jt. sollen dorther
stammen.” As for “bester Qualität” I shall speak below; the “sollen” appar-
ently is meant to supply a clue that there is a lack of any facts regarding
provenience. But this latter intimation is unfortunately not pursued, and
one has the distinct impression that the reader is being informed about
material that came from Kaplantu.
8. Suspicious Objects
A. Ziwiye
We have seen how since 1948 a continuous stream of objects of various
materials said to be from Ziwiye has flowed into the literature and into the
exhibition cases. If a dealer, whether in Teheran, Switzerland, London, Paris,
or New York, said an object came from Ziwiye, this usually was considered
sufficient archaeological evidence by some members of the scholarly com-
munity to warrant an attribution. Given this abdication of scholarly respon-
sibility, it could almost have been assumed (as by Dussaud 1951) that objects
from other sites or from other periods would be inadvertantly included
within the Ziwiye corpus. And surely in the present case it is hardly cyni-
cal to note that given the total corpus attributed to the site, the single cuve
de bronze, accepted by most scholars as originally containing the treasure,
could never have held all the many hundreds of objects in existence; at least
three or four such containers would have to be posited!
Some scholars, as already noted, did call attention to the possibility of a
false Ziwiye provenience for stray objects, but very few discussed in print the
possibility that a false provenience might also provide a pedigree for forg-
eries or suspicious objects, objects that should never have been presented
as deriving from any ancient site, Ziwiye or elsewhere. To my knowledge
only the writer (Muscarella 1977, nos. 155–165) and Mallowan and Herr-
mann (1974, 55) specifically mentioned the possibility that forgeries may be
present in the corpus. Melikian (1961, 71) has presented the problem in gen-
eral, but his statement aptly applies to the Ziwiye material: “Rien n’ est plus
“ziwiye” and ziwiye: the forgery of a provenience 983
20 After publishing these plaques Amandry (122) suprisingly and without elaboration adds
Muscarella 1977, No. 165; 12, 13, 14, 15, 16) a gold ram, two gold vessels, a
gold cow, gold goats, Ghirshman, et al., (Kunstschätze aus Iran), Nos. 848,
849, 851–853; Muscarella 1977, Nos. 157–161; 17) silver horse frontlet, British
Museum, Barnett 1973, 125, pl. LIII a (which may be the same piece men-
tioned by Amandry 1966, 125).21 In addition, there are two other objects
whose authenticity or lack of same is not clear to me: a gold decorated strip,
Ghirshman 1950, figs. 18, 19 (cf. Muscarella 1977, No. 54); and a gold bracelet,
Cincinnati, Kantor 1957, fig. 4).
B. Kaplantu
From the relatively small group of objects listed as having derived from
Kaplantu, a good number, to my mind, do not compel one to accept them as
ancient, and they should be considered to be suspicious until proven other-
wise. As with the Ziwiye material singled out above as suspicious, analysis
of style and artistic motifs has been the criterion for judgement, not merely
a lack of apparent artistic relationships with material alleged to be from
Ziwiye or its neighborhood; all the objects have been mentioned above: 1)
gold rhyton, art market, Ghirshman 1961, No. 490; Seidl 1968, a; Muscarella
1977, No. 166; 2) gold vase, Geneva, Musée Rath No. 619; Seidl 1968, 1; Mus-
carella 1977, No. 170; 3) two gold goat handles, Freer Gallery of Art, Ghirsh-
man 1960, fig. 8; Seidl 1968, k; Muscarella 1977, No. 105 (to my mind actually
meant to be Achaemenian); 4) silver ram-headed vessel, Abegg collection,
Wilkinson 1967, 12–14; Muscarella 1977, No. 167; 5) gold whetstone, Nelson
Gallery of Art, Ghirshman 1960, fig. 4; Seidl 1968, e; Muscarella 1977, No. 169;
6) gold whetstone, Los Angeles, ex Berg collection (same as Ghirshman 1961,
493 and Seidl 1968, f?); Muscarella 1977, No. 168; I also believe that the gold
discs depicting a hero holding lions deserve more technical and visual exam-
ination before they are all accepted as ancient, Muscarella 1977, note 86; cf.
Seidl 1968, g-i.
21 This object was not mentioned in Muscarella 1977 because I only recently learned that it
was in the British Museum and had been published; previously, I knew it as an unpublished
object in a dealer’s possession, and therefore could not mention it. It is very clear that the
British Museum example differs in execution and in details from the published frontlet of
Godard 1950, fig. 109: compare the spacing and size of the punched dots; the trunk, flowers,
and design of the trees, as well as their respective qualities of execution; and compare the
obvious differences in the nature of the body decoration, genitals, tail ends, mouth, eyes, and
feet of the two lions. I believe that a laboratory analysis on the British Museum example is
warranted. Note also for the record that Melikian 1961 and Lorenz 1970 should be added to
the bibliography of Muscarella 1977.
“ziwiye” and ziwiye: the forgery of a provenience 985
22 And here. of course, we are concerned with Ziwiye, without the mandatory quotation
10. Conclusions
The Ziwiye problem had its beginning with the first publications by Y. and
A. Godard in 1950. The research involved in digging through and record-
ing each successive level deposited in the ensuing mound of bibliography
has been very similar to the activity carried out on an archaeological field
campaign—but a campaign conducted on a mound where each level is dis-
torted by the existence of many pits, disturbed and missing areas, and com-
plex, non-uniform structural additions, alterations, and repairs; a campaign
which yields not a controlled horizontal sequence of events, but rather often
a convoluted, spiral-form stratigraphy. And a section drawing of the multiple
scholarly views that refer to specific find spots and dates, chronology, artis-
tic elements, the nature of the deposition, and which objects may or may
not have been found, and so forth, would form a series of fragmented loops
and spirals, crudely interlocking and overlapping one another.
One of the aims of this study has been to isolate these spirals and to
attempt to unwind them, to make them horizontal, so to speak, so that
23 In the course of writing this paper I learned that Nosratollah Motamedi of the Iranian
Center for Archaeological Research conducted an excavation at Ziwiye in 1976. A brief report
on the finds said to be from two cemeteries is given in the Exposition des Dernières Décou-
vertes Archéologiques 1975–1976 of the Fifth Annual Symposium of Archaeological Research
in Iran, November 2- December 2. 1976, 27–29. with references to the material in vitrines 18a
and b. Among the objects mentioned being excavated at Ziwiye are terracotta vessels, a terra-
cotta lamp, bronze pins, bronze earrings, a bronze tweezer, an ivory fragment (!), frit cylinder
seals, a frit scarab, a spurred arrowhead, and a fragment of a terracotta with an enamelled
motif.
988 chapter thirty-seven
others may be able to form their own conclusions, to draw their own sec-
tions, if they choose to reject those offered here. Another aim has been to
set forth as much of the working data that is presently available in order
to assist in attempts to answer several important questions. Was anything
found at Ziwiye in 1947? What was found? Was anything found there after
1947 by commercial diggers? What historical and archaeological value does
the “Ziwiye” material have? To those questions we now turn.
Was anything found at Ziwiye in 1947? The alleged discovery of the trea-
sure is not challenged by anyone as having occurred “fortuitement.” Inas-
much as the first reported objects surfaced via the antiquities market, they
certainly resulted from a clandestine dig (one of the few undisputed facts
available in the Ziwiye problem!). However that came about, and whenever
it occurred, this fact must be the starting point of any investigation and dis-
cussion. Only the individuals involved in the clandestine find know where
the site was and what were the circumstances of the discovery. Therefore,
it follows that the Godards’ statements regarding the story of the Ziwiye
find, the list of objects recovered from the bronze container, and those
objects found outside it, did not derive from personal observation but from
second- and third-hand sources: from dealers and possibly—but on this
issue one can only guess—from the Ziwiye villagers themselves. As such,
all the Godards’ statements and claims are hearsay, and are not what is tra-
ditionally considered to be sound sources of information in archaeological
research, scientifically controlled and recorded observation by an archaeol-
ogist and his staff.24
It is equally certain that Ghirshman knew nothing objective or first-
hand concerning which objects were actually found at Ziwiye, nor even
the date when the find was said to have occurred; he too knew only what
was revealed by dealers—obviously in some cases dealers other than those
who talked to Godard! In 1950 the scholarly community was confronted by
the publications of Godard and Ghirshman, both works written apparently
independently and unaware of each other, both implying that a corpus of
the Ziwiye material was being presented, and yet each publishing material
not mentioned by the other, and omitting material given great publicity by
the other. Godard’s 1951 retort did not resolve these obvious contradictions,
rather it created more; the spiralling had begun. As reported above, Godard’s
24 In this context one need only recall the confusion, the bitterness, and the different
opinions held by the archaeological community regarding whether or not the so-called
Dorak treasure exists, merely because it was reported by an archaeologist working alone who
claimed to have seen it.
“ziwiye” and ziwiye: the forgery of a provenience 989
warning that dealers gratuitously assign labels to stray objects they acquire
was thrust forth solely contra Ghirshman, to score points in a specific argu-
ment, claiming that some of the “Ziwiye” material did not come from
Ziwiye.25 And because of the narrow intent of this warning, Godard did not
recognize it for what it actually was, a warning applicable to all material
offered for sale by dealers, which embraces all the material allegedly from
Ziwiye—including that published by Godard himself. Ghirshman’s warning
regarding the attribution of dealers was also ad hoc, against other scholars,
and was not meant to reflect on his own enormous list of Ziwiye material, all
of which he maintained “provenant d’un seul et même trésor” (Ghirshman
1950, 198).
If the earliest writers on Ziwiye established a pattern concerning the
treatment of the material attributed there, most scholars in the follow-
ing decades uncritically accepted it, indeed refined and institutionalized it.
They too assumed the right to publish any object purchased from a dealer
and assign a Ziwiye label. Generally, without any critical evaluation, they
too expected automatic acceptance of their assertion, masking the undocu-
mented provenience with phrases such as “said to come from Ziwiye,” or
more emphatically stating that it is “practically certain,” or “there seems
no reason” to deny, that the object came from Ziwiye. Museums also con-
tributed in large part to the dissimulation and confusion concerning prove-
nience by allowing exhibitions of “Iranian Art” to serve as a vehicle for the
exhibition and publishing of newly introduced material alleged to be from
Ziwiye. Moreover, much of this material was in the hands of dealers at the
time of the exhibition and the museums thereby institutionally stamped the
objects both as from Ziwiye and as genuine. That endorsement helped to sell
them, courtesy of the documentation acquired.26
Are we able to learn anything about the alleged find from the villagers
at Ziwiye itself? Dyson (1963, 34) stated that in 1956 the local peasants “vol-
untarily pointed out the spot where the chest had allegedly been found,”
and that he believed they had no reason to lie to him. Further, commenting
on the commercial digging he witnessed that same year, he thought that it
was doubtful “that the dealer would have been spending money excavating
25 There appears to have been a personal battle in progress between the protagonists
being worked out in print, and with archaeological knowledge as the victim.
26 Many of the objects in dealers’ hands on exhibition were subsequently sold to museums
and private collections: and always the vendor cited as evidence of authenticity and prove-
nience that it was exhibited in a major museum. These exhibiting museums thus functioned
both as bazaars and archaeological documentation centers.
990 chapter thirty-seven
unless some discoveries had been made there.” With regard to the first state-
ment, it could be argued that if nothing had been found at Ziwiye in 1947 the
peasants might not want to reveal such information because they would lose
their jobs and their much-needed wages. Furthermore, one wonders why
they would voluntarily point out the alleged find spot, since if they actually
knew it they surely would want to protect it, to exploit it for themselves, and
not turn it over to strangers. The second statement could be countered by
noting that a dealer would dig at Ziwiye, or any other site, if he believed on
the basis of reports that something of value had been found there at some
time. Thus I suggest that the statements of the local villagers and the pres-
ence of a commercial dig at Ziwiye cannot by themselves be accepted as
sufficient evidence that something of importance was found at Ziwiye.
If we were cognizant of something concrete concerning the manner of
the transfer of the material from its alleged find spot to Teheran, then we
would perhaps be in a better position to feel more secure that something
may indeed have been found there in 1947. Did, for example, dealers pur-
chase the objects at the village of Ziwiye itself? Or did villagers known to
live at Ziwiye approach dealers in Sakkiz or Hamadan (Samedi 1960, 18), or
Teheran? But if Godard and Ghirshman had specific information with regard
to these matters, they did not offer it in print. And although I do not doubt
that they believed in 1950 that the objects they presented as from Ziwiye
actually came from there, I argue that they actually could not have known.
Thus, if one believes on the basis of the published information that
something was found at Ziwiye in 1947, it must be understood that such
belief is based ultimately on faith, on the acceptance of circumstantial and
hearsay evidence, and not on data objectively derived. This conclusion, it
must be stressed, does not deny the possibility, even the probability, that a
major find occurred at Ziwiye in 1947; it merely places the problem into a
perspective and defines its parameters.
If, then, one tentatively accepts the possibility/probability that “some-
thing” was found at Ziwiye in 1947, the next question follows inevitably:
what was found at that time? It is probable that most scholars would no
doubt answer by citing the objects published by Godard 1950, and it has
already been noted how those who expressed some doubts about attri-
bution, nevertheless accepted Godard’s authority. It might be argued that
surely the gold gorget, trapezoidal plaques, one or more other plaques and
strips, all stylistically related, “must have come from Ziwiye.”27 Wilkinson
27 Indeed, it might be argued that if the stylistically related objects did not come from
“ziwiye” and ziwiye: the forgery of a provenience 991
Ziwiye they surely collectively came from another site, X. But this hypothesis only begs the
question: we still would not know where the site was located, whether or not in fact the
objects were found together, or whether they were found at one time. Personally, I accept
the possibility that they came from one find; but the acceptance is conditioned by the
reservations expressed in the text.
28 If the silver plaque had been correctly excavated at Ziwiye we would have in hand the
southernmost existing example of Urartian art in Iran. As it stands now, we merely have
a fine example of Urartian art that probably came from somewhere in Iran. For a state-
ment about two Urartian helmets in the British Museum that are said to come from Azer-
baijan and Luristan respectively, see my review of P. Calmeyer’s Reliefbronzen in babylonis-
chem Stil in a forthcoming issue of JAOS. And with regard to the citation of unexcavated
objects from unexcavated sites in historical reconstructions, see my comments concern-
ing the gold Achaemenid bracelet in the Karlsruhe Museum in Appendix B of Muscarella
1977.
992 chapter thirty-seven
is that the same conclusions presented above for the material published
by Godard and Ghirshman apply equally to the post-1950 material. But
these objects are many stages removed in time and place from the material
published in 1950 and as such offer even less authority (and require even
more faith) for acceptance into the corpus. In this context it will be recalled
that the first suspicious objects began to appear on the art market in the mid
1950s, after the publication of Godard 1950; these objects were attributed
to “Ziwiye” by the same vendors who had labelled, and continued to label,
other objects as from “Ziwiye.”
Is information volunteered by the Ziwiye villagers of any help in resolving
this question? Dyson (1957, 33; 1963, 34) reported that although commercial
digging continued at Ziwiye from 1947 to 1956, nothing of significance had
been found. This information came from the local villagers working for
the dealer licensed to dig there. It might be argued that this information
does derive from first-hand sources and indicates that in fact nothing of
importance was found at Ziwiye, at least after 1947, that all the material
said by dealers to have come from there after this date must have derived
from other sites. However, it could equally well be argued that perhaps
the workmen may have been under orders to suppress any information
regarding actual finds for several reasons, one of which could have been to
keep eager archaeologists away. Furthermore, it could be argued that the
enormous amount of material found in 1947 was hoarded by dealers and
released over the years as a sound business practice. In the final analysis,
statements by local villagers cannot be verified and have no objective value
as an argument either to support or to reject any theory concerned with
what was or was not found at Ziwiye after 1947. And any theories regarding
dealers’ inventories and sales techniques can ultimately only be based on
the claims of the dealers themselves, who, being parti pris, can hardly be
depended upon to solve our problem.
The answer to the final question—what historical and archaeological
value does the Ziwiye material have?—has been anticipated by the con-
clusions already suggested. If it cannot be treated meaningfully as a unit
representing the archaeological finds from one site, it therefore must fol-
low that any historical or archaeological theories or conclusions based on
the existence of a collective treasure or hoard from Ziwiye will be of limited
value, if of any value at all. To treat all, or a subjectively selected group, of the
published objects as a unit-find is to misrepresent, and therefore to abuse,
the few facts available to us.
But the objects exist, and the genuine ones obviously have a significant
value. Indeed, not only are they exquisite works of art, they are of great
“ziwiye” and ziwiye: the forgery of a provenience 993
importance for the history of ancient Iranian art.29 As such, they must be
integrated into our study of Iranian and Near Eastern art in general. The only
caveat that should be ever present, indeed that should underlie any inclu-
sion of the “Ziwiye” objects into discussions of ancient art and culture, is
that they are to be considered as individual pieces, with no known prove-
nience, and without any firm evidence that they were found juxtaposed to
any other objects. They can tell us nothing about a hoard/burial deposited
at a particular place by a particular people; they surely can tell us nothing
historical about the site of Ziwiye. Historical and archaeological hypothe-
ses, for example those regarding the artistic and chronological relationship
of Iranian or Near Eastern art to Scythian art, can be maintained only with
this information in the forefront of the discussion. And, not parenthetically,
it will be obvious that the same constraints logically apply for the site of
Kaplantu and the objects said by dealers to have come from there.
The failure of scholars to critically examine the data and to question
them, the failure to recognize the situation for what it actually was, not for
what was claimed for it, have led to the creation of the “Ziwiye problem.”
This problem laid out before us forms a paradigm of a larger problem of
Iranian archaeology. For if one examines the corpus of unexcavated mate-
rial claimed by dealers and then published by scholars as deriving from
Hamadan, Marlik, Luristan, Amlash, etc., one soon realizes that there is
a “Marlik” and a Marlik, a “Hamadan” and a Hamadan, and so forth. The
objects “said to come from” these sites are sold by the same dealers, the sto-
ries concerning their “excavation” are told by the same story tellers to the
same receptive audience, an audience of credulous individuals whose desire
for objects qua objects is so insatiable, they will destroy their own discipline
in order to acquire and publish them.
Acknowledgements
29 I do believe that the available information allows us to conclude that the objects derived
Bibliography
1948
Godard, André. Comments in Musée Cernuschi, Iran: Pièces du Musée de Teheran, du
Musée du Louvre, et de Collections Particulières (Paris) 9, 14.
1949
Godard. André. “Le Trésor de Ziwiyè,” Compte Rendus de L’Académie des Inscriptions
et Belles Lettres, 168–172.
1950
Dimand, Maurice S. “Treasures of Iranian Art,” BMMA 8, 5, 145–146.
Godard, Yedda. a) “Histoire d’un Trésor,” France Illustration, April 8, 331–333.
——— b) “The Arabian Nights Story of the Gold Treasure of Ziwiye,” ILN, May 8, 714–
715.
Godard, André. Le Trésor de Ziwiye (Haarlem).
Ghirshman, Roman. “Le Trésor de Sakkez, les Origins de l’Art Mède et des Bronzes
du Luristan,” Artibus Asiae 13, 181–206.
1951
Crousset, René. “Dernières Vues sur l’Art des Steppes,” La Revue des Arts 2, 107–110.
Dussaud, René. Revue of Godard 1950, Syria 28, 287–289.
——— Revue of Ghirshman 1950, Syria 28, 289–291.
Godard, André. “À Propos du Trésor de Ziwiyè,” Artibus Asiae 14, 240–245.
Wiedner, E.F. Note in AFO XV, 151–152.
1952
Falkner, Margarete. “Der Schatz von Ziwije,” AfO XVI, 129–132.
Wilkinson, C.K. “Some New Contacts with Nimrud and Assyria,” BMMA 10, 8, 233–
240.
1954
Ghirshman, Roman. Iran from the Earliest Times to the Islamic Conquest (Penguin).
Schefold, K. “Die Iranische Kunst des Pontusländer,” in Handbuch des Alterwis-
senschaft (Munich) 423–453.
Sulimirski, T. “Scythian Antiquities in Western Asia,” Artibus Asiae XVII, 282–318.
vanden Berghe, L. “Iran, De Stand van de Archaeologische Onderzoekingen … VIII
De Kurdistan,” Jaarbericht Ex Orient Lux 13, 382–384.
1955
Anonymous. “Assyrian and Persian Gold and Silver: Masterpieces from Recent Dis-
coveries in the Metropolitan Museum of Art, New York,” ILN, April 16, 699.
“ziwiye” and ziwiye: the forgery of a provenience 995
Frankfort, Henri. The Art and Architecture of the Ancient Near East (Penguin).
Wilkinson, C.K. “Assyrian and Persian Art,” BMMA 13, 8, 213–222.
1956
Anonymous. “New Treasures of Teheran Museum: Achaemenian Gold and Later
Silver,” ILN, July 21, 107.
Barnett, R.D. “The Treasure of Ziwiye,” Iraq XVIII, 111–116.
Bussagli, M. “Objects from the Ziwiyeh Treasure,” in Mostra d’Arte Iranica (Milan)
119–124.
von der Osten, H.H. Die Welt der Perser (Stuttgart).
Wilkinson, C.K. “Two Ancient Silver Vessels,” BMMA 15, 1, 9–15.
1957
Carter, Dagny. The Symbol of the Beast (New York).
Dyson, R.H., Jr. “Iran 1956,” UPMB 21, 1, 27–39.
Kantor, Helene, J. “Goldworks and Ornaments from Iran,” Cincinnati Art Museum
Bulletin 5, 2, 9–20.
Needler, Winifred. “Four Near Eastern Antiquities Lent by Mr. Joseph H. Hirsch-
horn,” ROM Bulletin, 25, 7–11.
1958
Amandry, Pierre. “Orfèverie achémenide,” Antike Kunst 1, 9–23.
1959
vanden Berghe, L. Archéologie de l’Iran Ancien (Leiden), 111–115.
1960
Benson, Jack L. “Unpublished Griffin Protomes in American Private Collections,”
Antike Kunst 2, 58–70.
Ghirshman, Roman. “Rich Treasures of Persian Animal Art—Recently Discovered,”
ILN, April 4, 550–551.
Kantor, Helene J. “A Fragment of a Gold Appliqué from Ziwiye and Some Remarks
on the Artistic Traditions of Armenia and Iran during the Early First Millennium
B.C.,” JNES 19, 1, 1–4.
Samedi, H. Les Découverts Fortuites … (Teheran).
Wilkinson, C.K. “More Details on Ziwiye,” Iraq XXII, 213–220.
1961
Ghirshman, Roman. “Le Trésor de Ziwiye,” in 7000 Ans d’Art en Iran (Paris) 81–102.
Hoffmann, Herbert. “The Persian Origin of Attic Rhyta,” Antike Kunst 4, 1, 21–26.
Melikian, Souren. “Les Fouilles clandestines de l’Iran,” Connaissance des Arts 117,
November, 66–73.
Parrot, André. The Arts of Assyria (New York) 138–146.
1962
Barnett. R.D. “Median Art,” Iranica Antiqua II, 1, 77–95.
Ghirshman, Roman. a) “Le Rhyton en Iran,” Artibus Asiae XXV, 57–80.
996 chapter thirty-seven
——— b) “Der Schatz von Ziwije,” in Kunstschätze aus Iran (Zurich) pp. 71–86; see
also same entry in 7000 Jahre Kunst in Iran (Essen) 93–114.
Godard, André. L’Art de l’Iran (Paris).
Tuchelt, K. Tiergefässe in Kopf- und Protomengestalt (Berlin).
van Ufford. L. Byvanck-Quarles. “Le Trésor de Ziwiyè,” BABesch 25–39.
1963
Dyson, R.H. Jr. “Archaeological Scrap; Glimpses of History at Ziwiye,” Expedition 5,
3, 32–37.
Potratz, J. Die Skythen in Südrussland (Basle).
van Ufford, L. Byvanck-Quarles. “À Propos du Trésor de Ziwiyè,” BABesch 100–105.
Wilkinson, C.K. “Treasure from the Mannaean Land,” BMMA 21, 5, 274–284.
1964
Boehmer, R.M. “Volkstum und Städte der Mannäer,” Baghdader Mitteilungen 3, 11–
24.
Goldman, Bernard. “Early Iranian Art in the Cincinnati Art Museum,” Art Quarterly
27, 324–341.
Ghirshman, Roman. a) The Arts of Ancient Iran (New York) 98–125.
——— b) “Invasions des Nomads,” in Dark Ages and Nomads, ed. Machteld Mellink
(Istanbul 1964) 3–8.
Porada, Edith. “The Treasure of Ziwiye,” in 7000 Years of Iranian Art (Smithsonian
Institute) 25–27, 85–86.
1965
Amandry, Pierre. a) “L’Art Scythe archaique,” AA 891–913.
——— b) “Un Motif ‘Scythe’ en Iran et en Grèce,” JNES 24, 3, 149–160.
Culican, William. The Medes and Persians (New York).
Dyson, Robert H., Jr. “Protohistoric Iran as Seen from Hasanlu,” JNES 24, 3, 205–207.
Huot, Jean-Louis. Persia I (Archaeologia Mundi, World Publishing Co., New York)
139–140.
Muscarella, Oscar White. “A Fibula from Hasanlu,” AJA 69, 233–240.
Porada, Edith. “The Treasure of Ziwiye,” in The Art of Ancient Iran (New York) 123–
136.
Terrace, E.L.B. “The Ziwiye Treasure,” in BullMFA LXIII, 13–15.
Young, T.C. Jr. “A Comparative Ceramic Chronology for Western Iran, 1500–500B.C.,”
Iran III, 53–85.
1966
Amandry, Pierre. “A Propos du Trésor du Ziwiyè,” Iranica Antiqua, VI, 109–129.
Muscarella, Oscar White. Review of Porada 1965 in AJA 70, 3, 380–381.
Shepherd, Dorothy. “Four Early Silver Objects from Iran,” Bulletin of the Cleveland
Museum of Art LIII, 2, 38–49.
Terrace, E.L.B. “Ancient Near Eastern Art,” in The Pomerance Collection of Ancient Art
(Brooklyn) 42–48.
Musée Rath. Trésors de l’Ancien Iran.
van Loon, M. Urartian Art (Istanbul).
“ziwiye” and ziwiye: the forgery of a provenience 997
1967
Jettmar, Karl. Art of the Steppes (Baden-Baden).
van Loon, M. Review of M. Mellink (ed.) Dark Ages and Nomads c. 1000 B.C. (Istanbul,
1964) in Bibliotheca Orientalis XXIV, 1/2, 21–26.
Wilkinson, C.K. Two Ram-Headed Vessels from Iran (Bern).
1968
Azarpay, G. Urartian Art and Artifacts (Univ. of California).
1969
Amiet, Pierre. “Notes d’Archéologie Iranienne,” La Revue du Louvre 19, 6, 325–338.
1970
Farkas, Ann. “The Near East,” in ‘Animal Style’ Art (Asia House, New York) 19–27.
Lorenz, Eva. “Raubgräberei—Nicht Aktenkundig,” Antike Welt 1, 35–43.
van Loon, M. Review of Potratz 1963 in JNES 29, 66–72.
1971
Hrouda, B. Vorderasien I (Munich).
Moorey, P.R.S. A Catalogue of the Persian Bronzes in the Ashmolean Museum (Ox-
ford).
Muscarella, Oscar White. “Hasanlu in the Ninth Century bc and its Relations with
other Centers of the Near East,” AJA 75, 263–266.
1972
Burney, Charles, Lang, D.M. The Peoples of The Hills (New York).
Phillips, E.D. “The Scythian Domination in Western Asia …,” WA 129–138.
1973
Barnett, R.D. “A Review of Acquisitions 1963–1970 of Western Asiatic Antiquities (2),”
BMQ 37, 3–4, 119–137.
Calmeyer, Peter. Reliefbronzen in babylonischem Stil (Munich).
Ghirshman, Roman. “A Propos du Trésor de Ziwiyè,” JNES 32, 4, 445–452.
Muscarella, Oscar White. “Excavations at Agrab Tepe, Iran,” MMAJour 8, 47–76.
1974
Ghirshman, Roman. “Une Mède sur les Bas-Reliefs de Nimrud,” Iraq 36, 1/2, 37–38.
Mallowan, Max and G. Herrmann. Ivories from Nimrud Fascicle III Furniture from
SW.7 Fort Shalmaneser (London) 55–57.
Muscarella, Oscar White. “The Iron Age at Dinkha Tepe, Iran,” MMAJour 9, 35–90.
1975
Orthmann, W. “Die Kunst des Iran,” in Der Alte Orient, Propyläen Kunstgeschichte,
93.
Porada, Edith. “Iranische Kunst,” in Ibid. 363–398.
Wilkinson, C.K. Ivories from Ziwiye (Bern).
998 chapter thirty-seven
1976
K. Ishiguro. The Mr. and Mrs. Ishiguro Collection of Ancient Art I (Tokyo).
Muscarella, Oscar White. Review of Mallowan, Herrmann 1974 in JNES 35, 3, 208–210.
1977
Muscarella, Oscar White. “Unexcavated Objects and Ancient Near Eastern Art,” in
press.
chapter thirty-eight
More than two decades have passed since Géza de Francovitch wrote that
Median art “is a myth,” that “not a single work exists of proven Median
origin,” and, further, that its discussion involves one in “the realm of myth
and fable ….”1 That this strongly presented view still obtains, is still forcibly
valid, even more so after the excavation of two apparent Median sites in
western Iran, and after the appearance of more scholarly writings that claim
to have unraveled and perceived the characteristics of Median art, will be the
conclusion of this paper.
The political and social history of the Medes is difficult to chart and
reconstruct in any of its specifics and at any point in time, and here we
can only deal with it in the broadest terms.2 The sources are primarily the
contemporary, often terse references in the Assyrian annals and records that
begin in the ninth century bc (836 bc) and continue into the seventh, the
brief but significant references in the seventh-sixth century sections of the
Babylonian Chronicle, and the later fifth-century bc writings of Herodotus
(l. 95–106). The Assyrian annals indicate that the Medes were powerful and
occupied many towns and cities, each under a chieftain, they were attacked
periodically by the Assyrians, and they fought with their neighbors. By the
last two or three decades of the seventh century, some political events
unknown to us had occurred, for by this time the Medes (or some Medes)
became united under one king. This unification strengthened them to the
degree that in the years between 614 and 609 bc under Cyaxares, “the king of
the Medes,” they invaded Mesopotamia and participated in the destruction
of the Assyrian state. By 585bc it seems that Cyaxares had secured control
* This article originally appeared as “Median Art and Medizing Scholarship,” Journal of
257.
2 I present here only the barest outline. Recent important research makes it clear that we
know far less about Median history than hitherto assumed; see, in particular, Payton R. Helm,
“Herodotus’ Medikos Logos and Median History,” Iran 19 (1981): 85–90, with bibliography.
1000 chapter thirty-eight
over most of the peoples of western Iran including the Persians, and he
or his son Astyages had advanced to the Halys river in Anatolia, where he
maintained a border with the powerful Lydians. Within a generation all
was lost. Cyrus II, the king of the Persians, rebelled against his father-in-law
Astyages and in turn subjugated the Medes (553–550bc). With this event
began a new and, from a modern perspective, decidedly clearer phase of
Iranian and Near Eastern history.
The heartland of the Medes in western Iran centered around Hamadan
(Ecbatana) and extended for some distance to the southwest, northeast, and
east in the Alvand mountain range.3 The Assyrian annals record wealthy
Median cities, but the only kinds of booty mentioned are weapons, cattle,
donkeys, horses, camels, and, rarely, lapis lazuli,4 the latter without doubt
acquired by Median trade further east. By the time of their unification or
shortly thereafter, they seem to have acquired the means to supply them-
selves with more substantial wealth. This may be inferred from the sixth-
century bc section of the Babylonian Chronicle (7, ii, 3–4)5 that records
Cyrus’s defeat of Astyages: “Cyrus (marched) to Ecbatana, the royal city. The
silver, gold, goods, property … which he carried off as booty (from) Ecbatana,
he took to Anshan.” The Chronicle does not define the nature and char-
acteristics of the precious metals or of the goods and property, except to
record that portable material is being discussed. Thus, whether the gold and
silver was in the form of bullion or crafted artifacts, and whether Median
Kleinkunst or state or religious objects were included in the spoils, is not
vouchsafed. But something valuable was taken by Cyrus from Ecbatana, and
I do not believe one can exclude the possibility that these valuables included
products of Median craftsmen.
There is yet another possible interpretation that may be brought forth
regarding the nature of the spoils taken from Ecbatana, an interpretation
3 A number of scholars have argued that the Median area of control extended as far
northeast as Tehran on the assumption that the Mt. Bikni associated with the Medes in the
Assyrian records is Mt. Damavand. However, L.D. Levine, “Geographical Studies in the Neo-
Assyrian Zagros-II,” Iran 12 (1974): 118 f., nn. 167, 168, following König, correctly demonstrated
that Mt. Bikni is most probably a mountain in the Alvand range, near Hamadan. J. Reade,
“Kassites and Assyrians in Iran,” Iran 16 (1978): 38, continues to support the Bikni-Damavand
connection.
4 D.D. Luckenbill, Ancient Records of Assyria and Babylonia (Chicago, 1926), vol. 1, pars.
text, see A.K. Grayson, Assyrian and Babylonian Chronicles (Locust Valley, New York, 1975),
pp. 91, 106.
median art and medizing scholarship 1001
suggested to me by T. Cuyler Young, Jr.: the spoils could have been booty
that the Medes themselves took from the Assyrians more than a half-century
earlier. Explicitly argued in this view (assuming it means that only Assyrian
valuables were kept at Ecbatana) is the conclusion that, for whatever rea-
sons, the Medes chose not to or could not employ their own craftsmen to
create works. Indeed, this conclusion is one that could follow after the fact,
so to speak, after reviewing the evidence presented in the main body of this
paper. But that evidence allows one only to recognize what does or does not
exist as a result of modern discovery, not what may have actually existed in
the sixth century bc. I am of course aware of the limitations in the use of
the Babylonian Chronicle to prove the existence of Median art and artifacts,
whether we posit them to be mundane objects or sumptuous products of a
Staatskunst (infra), just as I am aware of its limitations for disproving such
existence. Here I am merely exploring the range of possible interpretations,
rather than attempting to make definitive statements.
That the gold and silver recorded in the Babylonian Chronicle may not
have been solely in the form of bullion or Assyrian booty and may have
included objects crafted by Median artisans is perhaps suggested, albeit
tenuously, by the later inscription of Darius at Susa (DSf). Here we read that
“the goldsmiths who wrought the gold, those were Medes and Egyptians”;
the very same people are also said to be the ones who “adorned the walls.”6
Whatever the meaning of “adorned,”7 the Medes were working together with
Egyptians as goldsmiths, which suggests, even if it does not prove, that some
tradition of Median gold working existed prior to the early fifth century bc.
Whether this tradition existed in pre-Cyrus times or began during his or
Darius’s reign is not vouchsafed, but—again, perhaps—the evidence of the
Ecbatana spoils would support the former position.8
Given then both the historical background of a powerful, united Median
state in the late seventh–early sixth century bc and the references to silver
and gold objects, scholars have posited that a Median art and iconography
probably existed by the late seventh century and perhaps earlier, and that
objects of sumptuous and courtly quality, inter alia, were crafted by Medes.
To give one example of the thought process concerning what might be
but from the reference she supplies, it is far from certain that the Medes in Babylon were
craftsmen; the word ba-ak-tu is not a craft designation (Ira Spar, private communication).
1002 chapter thirty-eight
expected of the Medes, M. Root states, “It is not unlikely that at important
centers such as Ecbatana buildings were adorned with sculptures in stone
or glazed brick relief. Also, a tradition of monumental representational art
may well have been maintained … such as wall painting and wooden relief.”9
This observation is viable, as are those postulating the existence of portable
Median art, Kleinkunst, as well as sumptuous works. Such assumptions
not only do not conflict with, but are based on our perceptions of the
interrelationships of ancient Near Eastern art and politics. In this context,
scholars might also anticipate a Median writing system for economic and
political transactions and major architecture to accommodate the royal
family as well as local officials.
Before proceeding further, brief, but pertinent, comments and a caveat
on the terms Mede and Median art are called for. A people called Medes,
distinct from other peoples in western Iran, are known in Iran from the
ninth century bc. By the last two decades of the seventh century there is
a “Median” army under a king, the implication of this being that a cen-
tral Median authority is present. Is it legitimate to assume that by the later
seventh century the Median kingdom incorporated other peoples in west-
ern Iran who were thereby considered to be Medes in a political sense and
who were components of the Median army? The Babylonian Chronicle (3.5)
records that the Manneans in 616 bc were still a distinct political entity, inas-
much as they were allies of the Assyrians against the Babylonians, which
signifies that not all western Iranians (meaning here peoples living in west-
ern Iran regardless of languages spoken) were included in the Median state
at this time. But what of the Manneans a few years later, and what of oth-
ers, Harhar, Missi, Ellipi, etc., who cease to be mentioned in the late Assyrian
texts? Was there a Media before and a Greater Media after the last decade or
two of the seventh century bc? Would Median material culture and iconog-
raphy in the earlier period basically reflect one form, purely Median, and
in the later another composed of several cultures? And if so, what part was
“ethnically” Median and what parts “Ellipian,” “Mannaean,” “Missian,” etc.?
Indeed, only Media is mentioned in 550bc when Cyrus defeats Astyages, but
what is Media? We do not know the answers to all the questions posed. As
we have no conception concerning the artistic and iconographic nature of
most of the peoples and political entities of western Iran (excepting Luris-
tan) either before or after the late seventh century, any attempt to isolate
Median art must take into account both the complex cultural and ethnic
9 Ibid., pp. 32 f.
median art and medizing scholarship 1003
background of the area and the question of precisely at what time and where
in Iran our attention is being focused.10
Now, having in mind from the above discussion an anticipation of what
might be reflected in the archaeological record with regard to the culture
of the “Medes,” what “must have existed,” what in reality do we have? First
of all there is no Median writing extant in Median or in Elamite script,
not a single sign on any object, artifact, or work of art.11 With regard to
architecture an apparently different situation exists. In the mid-1960s two
almost certainly Median sites were excavated, Godin Tepe (Period II) and
Nush-i Jan, southwest and south of Hamadan respectively.12 Both sites were
founded sometime in the late eighth century bc, that is, at a time before
the unification under one king, presumably Cyaxares. They both have a
large building with a multi-columned hall, a form known earlier at Hasanlu
in northwestern Iran and later at Pasargadae and Persepolis; Nush-i Jan
has, in addition, a fort and two fire temples, all remarkably well preserved.
Ecbatana remains unexcavated,13 hence these two sites will serve as the focus
of future research on Median architectural features and characteristics.
10 See D. Stronach, “Excavations at Nush-i Jan, 1967,” Iran 7 (1969): 4, and Helm, “Herodo-
tus,” p. 87. Some of the thoughts expressed here have been sharpened by conversations with
T. Cuyler Young, Jr. See also my brief cautionary comments in Journal of the Metropolitan
Museum of Art (MMJ) I (1968): 17 f.
11 Thus, John Curtis, Nush-i Jan III: The Small Finds (London, 1984), p. 14, correctly refrains
from identifying as Median a few broken cuneiform signs incised on a silver fragment (pl. 8,
100) from Nush-i Jan. The absence of Median writing, of course, does not mean that it
never existed; see I.M. Diakonov apud R. Ghirshman in Bibliotheca Orientalis 15, no. 6 (1958);
260. The pre-Achaemenian Persians used Elamite for their inscriptions, as is evidenced by
seatings from Persepolis with the inscription, “Cyrus of Anshan, son of Teispes.”
12 T. Cuyler Young, Jr., Excavations at Godin Tepe: First Progress Report (Toronto, 1967);
T. Cuyler Young, Jr. and L.D. Levine, Excavations at Godin Tepe: Second Progress Report (To-
ronto, 1974). D. Stronach’s reports on Nush-i Jan have been published in Iran 1969, 1971, 1973,
1974, 1975, 1978; see also Curtis, Nush-i Jan III. W. Kleiss, “Zwei Plätze des 6. Jahrhunderts
v. Chr. in iranisch Azerbaidjan,” Archäologische Mitteilungen aus Iran (AMI) 9 (1976): 114ff.,
emphatically considers a surveyed site, Chorbulag, northeast of Maku in northwestern Iran,
with sixth-century bc pottery, to be “… der Anlage um einen medischen Herrensitz” (p. 114)
and “… das einzige Gebäude im Nordteil von Iranisch-Azerbaidjan, das von den Medern
errichtet worden ist” (p. 116). Yet, while the site may date to the Median period in general
terms, it may not be called Median in the narrow, ethnic sense. There is no evidence that
Baba Jan (C. Goff, “Excavations at Baba Jan: The Pottery and Metal from Levels III and II,”
Iran 16 [1978]: 41 f.) is Median.
13 A few sondages have been made there. A large number of unexcavated objects, some
ancient, some modern forgeries (infra), have been assigned to Hamadan by dealers and
scholars with the assumption, implicitly or explicitly expressed, that they are Median; see
my “Excavated and Unexcavated Achaemenian Art,” in D. Schmandt-Besserat, ed., Ancient
Persia: The Art of an Empire (Malibu, 1980), pp. 31 ff.
1004 chapter thirty-eight
14 O.M. Dalton, The Treasure of the Oxus … (1926; London, 1964), nn. 2, 10: in the first,
1905, publication, only the sheath was singled out, pp. 18, xvi, xxii. For my comments on the
provenience of the treasure, see my 1980 paper, “Excavated and Unexcavated Achaemenian
Art,” p. 26. I wish to make two modifications to that paper: it was R.D. Barnett, not Dalton,
who wrote the preface to the 1964 edition quoted on p. 27; and it should have been made
clearer that the silver rhyton published by Dalton, lx (not XL), no. 178, was not claimed to be
part of the Oxus treasure (p. 30) but, rather, from an alleged separate find at Erzerum.
In two recent articles by B.A. Litvinskiy and I.R. Pichikyan, “The Temple of the Oxus,” JRAS
1981, pp. 134 ff.; and “Monuments of Art from the Sanctuary of Oxus,” Acta Antiqua 27 (1983):
25 ff., it is claimed as an archaeological fact (“indisputable”) that the site of Takht-i Sangin
median art and medizing scholarship 1005
were a gold bowl (no. 18) decorated with six pairs of rampant winged lions
separated by a lobe, and a gold akinakes sheath (no. 22) decorated with
a scene, repeated five times, of a horseman shooting arrows at a rampant
lion. Criteria used to establish the tentative Median attribution were the
stylization of the lions on the bowl, the apparent Assyrian influences of
the hunt scene, and the high tiara headdress on the horseman. In 1921
E. Herzfeld accepted the sheath as “medische Arbeit,” and in 1941 he added
the gold plaques from the Oxus group.15
R.D. Barnett in 1957, 1962, and 1968 transformed Dalton’s tentative attri-
bution into a certainty.16 Discussing only the sheath, he correctly compared
it in form and scalloped border decoration to the example represented on
the well-known Audience Hall (Apadana) relief, where it is worn by a Mede.
Because of its presence on this relief, the akinakes sheath is “a traditional
Median type of scabbard,” and the scalloped border on the Oxus sheath
“betrays it as Median,”17 Having thus concluded that the akinakes is peculiar
to Medes, Barnett then leaped from a culturally-ethnically defined object
to a chronological position, automatically perceiving the Oxus sheath as
Median in a pre-Achaemenian sense, a product of the seventh century bc.
And so positive was he of this attribution that he claimed forcefully and
without qualification that the horseman on the Oxus sheath “can only be
Astyages himself.”
on the right bank of the Oxus River is the site from which the Oxus treasure derived. Other
than nineteenth-century opinions, the sole piece of archaeological information offered to
support this claim is the discovery at the site of an extraordinary ivory akinakes sheath (JRAS,
1981, pl. 1; Acta Antiqua, 1983, fig. 2) decorated in relief with a rampant lion clutching a stag.
Because this sheath is the same size as the gold example from the treasure (26.7cm) and,
to the authors, of the same date, the ivory find “tends to confirm the traditional opinion
about the find spot of the Oxus Treasure ….” But on their own merits alone, the date of
the sheath, found in a third-century bc temple with third-century bc material, as well as its
find-spot cannot possibly support the fiat that the original site of the Oxus treasure is now
revealed. I refuse to accept the claim and find its absoluteness not proper in archaeological
discourse.
15 Dalton, Treasure of the Oxus, pls. 14 and 15; E. Herzfeld, “Kattische und khaldische
77–95, esp. 78–80; idem, “The Art of Bactria and the Treasure of the Oxus,” Iran. Antiqua 8
(1968): 36, 38.
17 Idem, “Median Art,” pp. 79 f. For the Persepolis relief sheath, see E. Schmidt, Persepolis,
18 E. Porada, The Art of Ancient Iran (New York, 1965); p. 140; W. Culican, The Medes and
Persians (New York, 1965), p. 140; M. van Loon, Urartfan Art (Iistanbul, 1966), p. 178; E. Akurgal,
Urartäische und altiranische Kunstzentren (Ankara, 1968), p. 124; A. Farkas, “The Horse and
Rider in Achaemenid Art,” Persica 4 (1969): 69 ff.
19 R. Ghirshman, “Le Trésor de l’ Oxus, les bronzes du Luristan et l’art mède,” in Vorderasi-
atische Archäologie (Moortgat Festschrift) (Berlin, 1964), pp. 88–94, esp, 88–90; idem, The Arts
of Ancient Iran (New York, 1964), pp. 90–94.
20 Root, King and Kingship, p. 98, n. 169; see also pp. 277; 279, n. 144; 281f.; and further
comments below.
median art and medizing scholarship 1007
those who wore the akinakes. Aside from the Medes (Delegation I), on the
Apadana reliefs they are worn or carried by Scythians (Delegation XI) and by
Delegation XVII, who are either Scythians, Sogdians, or Chorasmians. Fur-
ther, on the tomb reliefs they are worn by thirteen Völker in addition to the
Medes.21 In short, the akinakes is not and cannot be associated solely with
Medes (see also notes 14, 23, and the Kelermes and Litoy-Melgunov sheaths,
below).
Of course Barnett was right to introduce the Audience Hall relief sheath
as an appropriate parallel to the Oxus sheath, but he was in error when he
misread its chronological message. The scalloped border and animal frieze
scene on the body of the relief sheath are explicit and formal parallels,
parallels that indicate an Achaemenian attribution for the Oxus sheath.
The hunt scene on the latter is Achaemenian in style and spirit and occurs
on a number of Achaemenian period seals where royal and apparently
other figures are depicted hunting animals from horseback, on camels, from
chariots, and on foot.22 And the high headdress worn by the horseman,
as Moorey has noted in his argument for an Achaemenian date for the
sheath, “may be the turban [sic] worn by a number of the peoples within
the Achaemenid Empire ….”23 (The headdress seems not to be a turban, but
the general idea still obtains.) To identify, and with certainty, the horseman
as a specific Median king, of whom we know little historically and nothing
art-historically, even assuming the sheath to be pre-Achaemenian in date,
transgresses the boundaries of archaeological activity, all the more so when
we recognize the Achaemenian, probably fifth century bc, depiction before
us. And consequent to the fact that all horseman in the Achaemenian period
21 For identification of the Apadana relief figures and the tomb bearers, see G. Walser,
Die Völkerschaften auf den Reliefs von Persepolis (Berlin, 1966), pp. 53ff., Falttafel I, and my
comments in my review of Walser in JNES 28 (1969): 283 f. See also B. Hrouda, Vorderasien,
vol. I (Munich, 1971), p. 286; and P.R.S. Moorey, Cemeteries of the First Millennium bc at Deve
Hüyük, BAR International Series 87 (London, 1980), p. 55.
22 Viz., H. Frankfort, Cylinder Seals (London, 1939), pp. 221f., pl. 37d, f, h, i, m, n. See also
the silver disc in the Oxus treasure in Dalton, Treasure of the Oxus, p. 13, no. 24.
23 Moorey, Cemeteries, p. 57. In this context it is worth noting the silver rhyton from
Erebuni that depicts a heavy-set man riding a horse. The man bears a typical akinakes and
wears a tapering high headdress: B.N. Arakelian, “Treasure of Silver Artifacts from Erebuni,”
Sovietskaya Archeologica I (1971): figs. 1–4, dated to the fifth-fourth centuries bc. The man is by
no means necessarily a Mede. P. Calmeyer “Zu einigen vernachlässigten Aspekten medischer
Kunst,” in Proceedings of the IInd Annual Symposium of Archaeological Research in Iran, Iran
Bastan Museum, October–November, 1973 (Tehran, 1974), pp. 112, 118, n. 5, dates the Oxus sheath
to the second half of the fifth century bc on the basis of comparisons with Greek art; Farkas
“Horse and Rider in Achaemenid Art,” p. 70, also raised this issue but did not develop it further.
1008 chapter thirty-eight
24 See n. 20, above. On p. 281, n. 151, Root says that there are no known Achaemenian
representations of the king wearing trousers, but she notes that Darius III wears them on
the Alexander Mosaic; the Oxus sheath scene does not alter this observation. O. Kimball
Armayor, “Herodotus’ Catalogue of the Persian Empire …,” Transactions of the American
Philological Association 108 (1978): 5, takes a rigid view of trousers contra Herodotus (1.71, 135;
3.87; 5.49; 7.61, 62), arguing that they were never worn by Persians, on which view see Root.
There can be little doubt that in battle (as Herodotus stated) and surely in the hunt, Persians,
and other people, wore trousers.
25 See, for example, a sixth-fifth century Archaemenian silver bowl excavated at Ushak
in western Anatolia (unpublished, in the Ankara Museum) and an example of the same date
(and possibly same site) in the Metropolitan Museum of Art, which D. von Bothmer, “A Greek
and Roman Treasury,” Bulletin of the Metropolitan Museum of Art (EMMA) 42 (1984): 25, no. 18
and also 19, incorrectly labeled “Greek”! Other examples include a silver vessel from Kazbek;
see A.M. Tallgren in Eurasia Septentrionalis Antiqua 5 (1930): 117, fig. 4, and another from the
same workshop from Rhodes, Clara Rhodos 8 (1936): 179f., figs. 168, 169; also an example said
to have been excavated at Ünye on the Black Sea; see E. Akurgal in Antike Kunst 10 (1967):
32 ff., figs. 1–3, pl. 8. 9: 2–6, surely from the Achaemenian period and from a local workshop.
26 Dalton, Treasure of the Oxus, p. 10, noted that doubts about the authenticity of the
sheath had been raised, but he says nothing of the gold bowl. I know the vessel only from
photographs but see no compelling reasons to doubt its ancient date. Herzfeld, “Khattische
und khaldische Bronzen,” p. 154, considered the bowl to be Urartian.
27 Barnett, “Median Art,” pp. 80 ff., 86 ff.
median art and medizing scholarship 1009
28 Ibid., pp. 89 f., n. 2, fig. 5. For published challenges to Barnett’s Urartian attribution of
the British Museum mirror, see A. Greifenhagen, “Schmuck und Gerät einer lydischen Mäd-
chens,” Antike Kunst 8 (1965): 17 f.; A. Oliver, “A Bronze Mirror from Sardis,” in D.G. Mitten et al.,
eds., Studies Presented to G.M.A. Hanfmann (Mainz: 1971), pp. 117f., 120; Moorey, Cemeteries,
p. 57. The animals on the rim of the British Museum mirror exhibit Achaemenian features,
as Akurgal, “Urartäische und altiranische Kunstzentren,” p. 124, noted—albeit dating it to
the Median period. The Berlin example, published by Greifenhagen, pp. 16f., pl. 5, is, as that
author claims, East Greek in style; it has no addorsed heads as clamps.
29 Barnett, “Median Art,” p. 90; he was supported in the Urartian attribution of the Tell
Fara dipper by R. Amiran, “Two Luristan Fibulae and an Urartian Ladle from Old Excava-
tions in Palestine,” Iran. Antiqua 6 (1966): 90 f. For Achaemenian examples, see P. Amandry,
“Orfèvrerie achéménide,” Antike Kunst I (1958); 13 ff., 18; idem, Collection Hélène Stathatos,
vol. 3 (Strasbourg, 1963), pp. 262, 264 f., 267, figs. 158, 163–168; Oliver, “Bronze Mirror from
Sardis,” p. 117; E. Stern, Material Culture of the Land of the Bible in the Persian Period (Jerusalem,
1982), p. 74, fig 90; p. 147, figs. 241, 244; p. 152, fig. 254.
30 Dalton, Treasure of the Oxus, pp. 13, 18; H. Otto, “Ein achämenidischen Goldwidder,”
ZA 48 (1944): 10, n. 2; 13; Amandry, “Orfévrerie achéménide,” p. 20; Ghirshman, Ancient Iran,
p. 358; Farkas, “Horse and Rider in Achaemenid Art,” p. 67; Moorey, Cemeteries, p. 57. Culican,
Medes and Persians, pp. 137, 139, considers the dagger to be Median; Akurgal, Urartäische und
altiranische Kunstzentren, p. 124, sees it as a product of “medisch-achämenidischen Stiles.”
Note that its sheath is decorated in Greek style; M.I. Artamonov, The Splendor of Scythian Art
(New York, 1969), p. 55, pls. 183, 184.
1010 chapter thirty-eight
the Kelermes sheath the shooting creatures alternate with passive striding
ones. Also of Urartian style is the pair of winged human figures flanking
a tree at the top of the sheath. On the side projection, however, there is
a reclining stag bordered by stylized eagles’ beaks, both features of classic
Scythian style, and the chape is decorated with two confronted Scythian
style leonine creatures. The confusion concerning the cultural interpreta-
tion of the sheaths is increased by the material associated with them: at Kel-
ermes, a Scythian axe and at Litoy, Urartian furniture fittings were reported.
Further associated with the Kelermes finds were a gold petaled bowl and
another with repoussé animals, both of which seem to be Achaemenian
works.31
All scholars concerned with the sheaths agree on the above description
and analysis; they differ in their interpretation and perception of the artistic
mixture—actually, juxtaposition is the better term—and the nature of the
cultural-ethnic workshop that produced them. In 1937 Herzfeld dated the
sheaths to the first half of the sixth century bc, “sicher vorachaemenidisch,
dabei so urartaeisch,” but in 1941 he placed them in the “Median period,”
along with the Oxus sheath, and he did not refer to Urartu.32 Barnett devel-
oped and expanded this casual “Median period” attribution with force, and
again by fiat, in his cited 1962 article. He rightly recorded the Urartian and
Scythian elements but only to announce that the connection of these two
styles plus the akinakes shape signifies we are in the presence of Median
art.33 The Scythian relationship had already been introduced in a casual
manner when, in connection with the Oxus sheath, he referred to the chape
represented on the Audience Hall akinakes.
A close reading of Barnett’s conclusions suggests that in fact he began
his studies with the idea that the Kelermes and Litoy sheaths were Median
and then isolated the co-existence of Urartian and Scythian motifs, thereby
producing a formula to demonstrate what already had been determined.
But Scythian and Urartian and Median are discrete cultural entities, two of
which are known, the other unknown, and simple arithmetic cannot yield
an archaeological reality, Median. The unknown has not been revealed.
ambitious formula or equation for Median art: Urartian plus Assyrian plus Scythian plus
Phoenician elements. What happened to the Assyrian and Phoenician elements in the five-
year interval was not explained.
median art and medizing scholarship 1011
It has already been noted that the akinakes sheath cannot be considered
to be a Median artifact, an object characteristic of Medes in any period, and
therefore it may not be brought into the discussion as a factor in a Median
equation. Nor indeed may the chape on the Audience Hall relief, which
displays a curled animal in pure Scythian style, for this particular decoration
is an Achaemenian period artifact, as all scholars agree.34 In any event,
the chapes on the sheaths under review here are composed of a different
Scythian form. We have thus to concern ourselves with the iconography and
style of the decoration on the sheaths, and others who have studied them
have arrived at conclusions other than that suggested by Barnett. Aside from
Herzfeld’s 1937 note, the only other scholar known to me who concluded
that the sheaths were Urartian works, incorporating Scythian elements, was
H. Kantor.35 Others have taken the contrary position, that they are Scythian-
made artifacts incorporating Urartian elements, some singling out the non-
Urartian form of the sheath as crucial to the argument.36
The problem to be investigated here is truly complex, as it involves under-
standing why the motifs of two distinct cultures appear juxtaposed on one
artifact, and some detailed discussion is inevitable. First of all, I do not think
that geography, the location of the find-spots, should play a role in analy-
sis; the final resting place of an object ought not necessarily determine or
34 Idem, “Median Art,” p. 79. For discussions of these Achaemenian chapes, see the com-
ments of the following; B. Goldman, “Achaemenian Chapes,” Ars Orientalis 2 (1957); 43–
54; R. Stucky, “Achämenidische Ortbänder,” Archäologischer Anzeiger (AA) II (1976): 13–23;
P. Bernard, “A propos des bouterolles de forreaux achéménides,” Revue archeologique 2 (1976):
227–246; S. Sorokin, “The Curled-up Animal from Ziwiye” (in Russian), Soobscheniya Gosu-
darstvennaya Akademia, Hermitazha 34 (1972): 75–78; Litvinskiy and Pichikyan, in Acta Anti-
qua 27 (1983): 49 ff., 54. Note that a circular bone plaque excavated at Sardis (BASOR 192 [1966]:
14, fig. 9; BASOR 211 [1973]: 33, fig. 11; Biblical Archaeologist 29 [1966]: 45, fig. 5) seems to be a
solid pommel, not a chape as claimed.
35 H. Kantor, “A Fragment of a Gold Applique from Ziwiye …,” JNES 19 (1960): 13; in n. 49
she cites Herzfeld’s 1937 reference. G. Azarpay, Urartian Art and Artifacts (San Francisco,
1968), pp, 70 f., 73, believes on the one hand that Urartian models or “perhaps even Urartian
craftmanship” may have played a role; but then, because the motifs are of “mixed style,” and
not a “normal pattern of Urartian art,” they are not Urartian works. But, of course, there is no
mixed style, there are two separate styles. B. Goldman, “The Animal Style of Ziwiye,” IPEK 24
(1974–1977): 55, 61, says that the Litoy sheath “may be an Urartian import,” but on p. 67, he
supports Barnett’s views on Median art; a contradiction seems evident here.
36 Barnett, “Median Art,” p. 91, n. 2; also Ghirshman, Ancient Iran, pp. 28, 303f., 327; Akurgal,
Urartäische und altiranische Kunstzentren, pp. 123 f.; Stucky, “Achämenidische Ortbänder,”
p. 20; Moorey, Cemeteries, pp. 56 f.; B.B. Piotrovskii, Urartu (London, 1967), pp. 21, 97ff. Dalton,
Treasure of the Oxus, pp. xxxv f., considered the frieze to be Assyrian; see also J. Boardman,
The Greeks Overseas (London, 1980), p. 258.
1012 chapter thirty-eight
37 Thus Barnett, “Median Art,” p. 93, explains the presence of the sheaths in Scythian
tombs as booty or gifts from the Medes acquired when the Scythians were in Iran. This
conclusion follows from the view that the sheaths are Median, but if not Median, then
another explanation must be sought for their distribution.
median art and medizing scholarship 1013
king commissioning the sheaths as gifts from one king to the other, gifts
reflecting the art (message) of both the donor and the recipient.
Finally, a non-Scythian, non-Urartian third-party may have commis-
sioned the sheaths. On the one hand, a king of Land X, itself having or not
having an artistic tradition of its own, may have summoned to his court a
Scythian and an Urartian, or a Scythian/ Urartian trained in both schools,
to create bicultural artifacts for some particular reason (an exotic curio?).
On the other hand, we could conceive the specific thrust of Barnett’s con-
cept; a king from a land without artistic traditions may have summoned
the appropriate artists because he intellectually and consciously chose as
the artistic expression of his land (state) the juxtaposition of Scythian and
U rartian motifs, anticipating that the contemporary (not to mention the
future) world would recognize it as such. With the acceptance of either one
of these schemes, another question ineluctably surfaces: which of the many
pre-Achaemenian states in Iran or in the northern regions was the third
party?38
Concerning the workshop staffing, a neutral position recognizes that
archaeological and ethnographical evidence indicates that an artist trained
in one cultural tradition could reproduce the art of another and, equally,
that two artists could work on the same object.39
I have no decisive suggestions to offer concerning the intriguing problem
of sorting out the mental and physical activities involved in the creation
of the sheaths—although I am inclined to believe that two artists partici-
pated. Further, I find the third-party thesis, in particular that articulated by
38 Barnett (“The Treasure of Ziwiye,” Iraq 18 [1956]: 114) revealed the characteristics of
Mannacan art as “crude, if lively, but barbarous and provincial in the extreme,” and he even
presented the archaeological community with actual examples, all picked at random from
the vast Iranian scrapbasket.
39 For an object made in two parts and decorated by two separate artists, see my article
“A Bronze Vase from Iran and Its Greek Connections,” MMJ 5 (1972): 25–50, figs. I–II. To this,
compare a vessel very similar in form and style, also made in two parts but here apparently
decorated by one artist described by P.R.S. Moorey in G. Markoe, ed., Ancient Bronzes Ceram-
ics and Seals (Los Angeles, 1981), no. 415. Although in the first instance both artists shared the
same artistic tradition, the concept of two artists working on a single object is established
empirically. For ethnographic evidence of this same phenomenon, see M.A. Hardin, “The
Cognitive Basis of Productivity in a Decorative Art Style …,” in C. Kramer, ed., Ethnoarchae-
ology (New York, 1979), pp. 92 f. For artists of one ethnic, cultural background manufacturing
works of another, see the Tomb of Petosiris in Egypt, where Egyptian craftsmen are depicted
making Achaemenian style objects; see my article “Excavated and Unexcavated Ancient Near
Eastern Art,” in T.C. Young, Jr. and L.D. Levine, eds., Mountains and Lowlands (Malibu, 1977),
p. 194, n. 100. Note that M. Rostovtzeff, Ionians and Greeks in South Russia (Oxford, 1922), p. 50,
believed that one artist “fashioned” the separate parts of the Kelermes sheath (a Scythian?).
1014 chapter thirty-eight
40 See my “ ‘Ziwiye’ and Ziwiye: The Forgery of a Provenience,” Journal of Field Archaeology
4 (1977): 197–219, esp. pp. 205 ff., 207 f.; see also K.R. Maxwell-Hyslop, Western Asiatic Jewellery
(London, 1971), pp. 206 f.; T. Sulimirski, “The Background of the Ziwiye Find ….” Bulletin of the
Institute of Archaeology 15 (1978): 7–31; R. Ghirshman, Tombe princière de Ziwiyé (Paris, 1979);
and B. Hrouda, “Der ‘Schatzfund’ von Ziwiyah ….” Iran. Antiqua 18 (1983): 97–108.
41 See M. Falkner in “Der Schatz von Ziwiye,” AfO 16 (1952): 132.
42 Barnett, “Median Art,” pp. 84 f., 91 f., 94 pl. 4a–c.
43 Porada in 7000 Years of Iranian Art (Washington, D.C., 1964–1965), p. 26; idem, Art of
median art and medizing scholarship 1015
also shifted his position on the nature of the treasure but more dramati-
cally and in a different direction than Porada. At first he confessed not to
know what the treasure represented, then that it was a Scythian treasure
deriving from a Scythian prince’s tomb, and still later that it came from a
tomb of a “Scythe-Median” prince, reflecting here his view that there was
a cultural-political symbiosis of the two peoples (cf. his conclusions about
Luristan, supra). Concomitant with the latter conclusion, he maintained,
with no documentation and without calling attention to a specific object,
that Median art was formulated from “many different strains—Urartian,
Assyrian, local Mannaean, perhaps even Ionian ….” Twenty pages later in
the same publication, after noting that the Ziwiye material “has much to
teach us about Median art,” he added the oxymoronic non-sequitur “it seems
impossible to differentiate the [material culture] of the Medes from the
Scythians and Cimmerians.”44 In her book on ancient Near Eastern jewelry,
Maxwell-Hyslop also accepted the conclusion that Median elements are to
be recognized in the Ziwiye corpus. As evidence to support this claim, she
singled out two gold roundels of (to my eyes) Achaemenian style which, on
the basis of comparison with two forgeries (infra), enabled her to conclude
that they could be products of a Median goldsmith.45
We need not dwell any longer on the Ziwiye-Median equation, for the
same arguments, to be sure with more permutations, brought forth in the
discussion of the Kelermes-Litoy sheaths, obtain here, a resort to specula-
tion and singling out of motifs and objects without reference to verifiable
data. And whereas previously a modest equation of which Western Asi-
atic motifs or elements constitute Median art was presented, in the case of
Ziwiye the grouping on the left side of the equation has been unabashedly
modified and enlarged.
About a half-dozen other stray objects and types have from time to time
been labeled Median artifacts. G. Pudelko, writing in the 1930s and ignor-
ing what was even then known about Urartian art, cited a typical Urartian
Ancient Iran, p. 134. For the record, Culican, Medes and Persians, pp. 46f., repeats Barnett’s
views; and H.A. Potratz, “Die Skythen und Vorderasien,” Orientalia, n.s. 28 (1959): 67f., had
earlier connected the Ziwiye treasure to the Medes.
44 For the bibliography on Ghirshman’s views, see again my article “Ziwiye,” pp. 205f.; see
in the Cincinnati Art Museum, see infra; for the other, see my article “Bronze Vase from Iran,”
p. 184, no. 156, a gold plaque in the Metropolitan Museum.
1016 chapter thirty-eight
attachment is discussed by G.M.A. Hanfmann, “Four Urartian Bulls’ Heads,” Anatolian Studies
6 (1956): 205–213, esp. pp. 207 f., no. 2, Louvre AO 17.207.
47 R. Ghirshman, “Notes iraniennes XI: Le Rhyton en Iran,” Artibus Asiae 25 (1962): 75, 77,
fig. 19.
48 See supra, and n. 28. In The Art of Greece (New York, 1968), p. 217, Akurgal refers to the
of Bactria,” pp. 46 f.; P. Calmeyer, Datierbare Bronzen aus Luristan und Kirmanshah (Berlin,
1969), pp. 39 ff.; see also my “Excavated and Unexcavated Achaemenian Art,” p. 34.
50 R. Ghirshman, “La Fibule en Iran,” Iran. Antiqua 4 (1964): 97, nos. 13–15, pl, 25; idem,
“Le Pazuzu et les Fibules de Luristan,” Mélanges de l’Université Saint Joseph 46 (1970): 123ff.,
pl. 2; Calmeycr, “Zu einigen vernachlässigten Aspekten medischer Kunst,” pp. 114f., D, E, F.
Ghirshman’s nos. 13, 14 lack the pin, but cf. Calmeyer’s E, with the spring on the Pazuzu head.
Calmeyer’s F consists of only one “arc” and is odd: there seems to be only one ram/mouflon
head at the “apex”; a bird’s head “catch” is on top of the Pazuzu head (we would expect it on
the opposite arc); and there seems to be no apex. If this object is a fibula, presumably there
was a female at the opposite arc. Nos. 15, D, and F have the ram/mouflon apex heads, nos. 13,
14, and E have the bird.
median art and medizing scholarship 1017
51 Calmeyer, Datierbare Bronzen, p. 99, figs. 100, 101: note that fig. 100 is not, to my eyes,
Ghirshman’s “La Fibule en Iran,” (1964), no. 14, as is claimed; I do not know what this example
is (a female head and a Pazuzu head on opposite arcs); idem, “Zu einigen vernachlässigten
Aspekten medischer Kunst,” pp. 114 ff.; Curtis, Nush-i Jan III, pp. 23, 33.
52 See my article “Fibulae Represented on Sculpture,” JNES 26 (1967): 82ff. For fibulae
associated with chains, see M. Mallowan, Nimrud and Its Remains, vol. 1 (London, 1966),
pp. 114 f., fig. 58; C. Schaeffer in ILN, April 27, 1935, p. 686, fig. 2; for two examples from
Luristan (excavated!), see L. Vanden Berghe, “Les Fibules provenant des fouilles au Pusht-i
Kuh, Luristan,” Iran. Antiqua 13 (1978): 51, nos. 10, 11, fig. 9, pl. 2.
53 See my article “Five Additions to the Norbert Schimmel Collection,” Acta Praehistorica
et Archaeologica 7/8 (1976–1977): 316, no. 3. The Schimmel fibula has a female on each of the
two arcs; they are of the same style as those on the Feroughi fibulae.
54 Amiran, “Two Luristan Fibulae and an Urartian Ladle,” pp. 88ff., pl. 17, I; Amiran
recognizes the Megiddo fibula to be an import there, but uncritically following Ghirshman,
she claims the import came from Luristan.
1018 chapter thirty-eight
55 See again my article in Young and Levine, eds., Mountains and Lowlands, p. 184, no. 152;
and idem, Addenda to this article (Malibu, 1979), p. 4, no. 24, for the bibliography of scholars
who published this piece as genuine: now add Akurgal, Urartäisch und altiranische Kun-
stzentren, p. 123 (also for the quote in my text); Ghirshman, Tombe princière de Ziwiyé, p. 21;
G.G. Beloni, Iranian Art (London, 1969), pl. 40. Cf. P. Calmeyer, “Hamadan,” in RLA, vol, 4, p. 66
[note that in my article “Excavated and Unexcavated Achaemenian Art,” in nn. 12, 33, above,
median art and medizing scholarship 1019
and configuration of the vessel, for the sharp, stylized, and poorly—inaccu-
rately—executed decoration of winged lions in relief, there is no parallel
in Achaemenian or ancient Near Eastern art; and the feet of the animal
handles rest directly on the vessel, contrary to proper Achaemenian prac-
tice. Another gold vessel of similar form and relief decoration, this one
in a private collection, has lion handles, and its shape and execution are
equally unparalleled in style and execution in Achaemenian and ancient
Near Eastern art. Here there is an additional motif of two lions sharing
one en face head, a motif found on still another forgery.56 To defend these
two manifestly modern misunderstandings, to ignore the loud and clear
message of an unexcavated object supplied with a forged provenience, is
to mock proper archaeological and art historical analysis. A third vessel of
gold, an animal-headed cup now in Tehran, is an object that I also believe
is not ancient; Ghirshman considered it to be ancient and “sous réserve”
Median.57
A gold strip in the British Museum depicting a winged lion in relief is also
to my mind not the work of an ancient artist, although claimed as Median by
Barnett; nor is a gold handle with addorsed lapis lazuli lions at the top, also in
the British Museum, to be considered the work of an ancient artist.58 A silver
bowl in the Ashmolean Museum decorated in the interior with two crossed
lions was originally published as Median and from Hamadan; subsequently
one of its publishers has recognized it as a modern work.59
Finally, in the Louvre is a fragment of a white stone head executed in
barely recognizable Achaemenian style. A. Parrot and P. Amiet have identi-
fied it as most probably a Median work because its character is just not quite
Achaemenian in either style or execution: the hair, beard curls, mustache
I incorrectly stated that Calmeyer in this publication claimed that the gold Achaemenian
objects in the Metropolitan Museum derived from Hamadan. In fact, on p. 65, left, he raised
doubts about this attribution. I here correct my error and apologize; I apparently misread my
notes].
56 See again my article in Young and Levine, eds., Mountains and Lowlands, p. 184, no. 153;
cf. no. 156 (see n. 45 above); see again my Addenda, p. 4, no. 25. Note that the roundel with
the head shared by two lions discussed by Maxwell-Hyslop, Western Asiatic Jewellery, p. 210,
may be genuine; see another probably genuine example of this motif, this time on a pre-
Achaemenian, Luristan disc pin, Ghirshman, Ancient Iran, fig. 384.
57 Ghirshman, “Notes iraniennes XI,” pp. 75, 77, fig. 18; see again my article in Young and
Levine, eds., Mountains and Lowlands, p. 185, no. 166; and my Addenda, p. 4, no. 29.
58 For the gold strip, see Barnett, “Median Art,” p. 78, pl. 1a; and my article “Ziwiye,” p. 211.
The gold and lapis handle (135908) has been published as a postcard dated 1975: “Median or
Early Achaemenian.”
59 See my article in Young and Levine, eds., Mountains and Lowlands, p. 184, no. 154.
1020 chapter thirty-eight
positioning, eyes, ears, nose—in short, everything about the head is wrong.60
Furthermore, there is a debate between the two scholars involved in the
head’s presentation concerning whether, inasmuch as the provenience is
accepted as Hamadan (although purchased in Switzerland!), it is a portrait
of Cyaxares or Astyages, a thought that surely did not enter the mind of the
forger. The Louvre head joins the other forgeries of Achaemenian objects
discussed above in the repertory of new Median art.
We turn now to the final group of objects encountered in the Median
odyssey, works that have actually been excavated in Iran itself. At the pres-
ent time, perhaps the least controversial element in the alleged Median
corpus is the group of rock-cut tombs: Dukkan-i Daud, Farhad-u-Shirin,
Sakavand, Fakhrika in Iran, and Kizkapan in Iraq. Herzfeld gave them promi-
nence as Median tombs because of their geographical positions and because
of the Median dress on some of the figures in relief on the tombs at
Dukkan-i Daud and Kizkapan, to which he compared the gold plaques from
the Oxus treasure (supra); he dated Sakavand to the early Achaemenian
period, to the year 521 bc.61 A few scholars, including I.M. Diakanov, H.H. von
der Osten, and Ghirshman, also believed the tombs to be Median, but recent
scholarship has rejected this early chronology. Research based on more
refined dating of tool marks and stylistic analysis of the columns before
some of the tomb openings indicates that they are Achaemenian or later.
We still seek the tombs of the Medes.62
When Stronach originally published a bronze Pazuzu head excavated
at Nush-i Jan, he considered it “not at all unlikely to have been looted
from Assyria.”63 Now, whether it was looted from Assyria or thence derived
by more peaceful means is not of course known, but its Mesopotamian/
Assyrian origin is almost certain.64 A Mesopotamian attribution and origin
is quoted in the text. I regret not including this piece in my 1977 paper on forgeries (in
Mountains and Lowlands) but correct the omission here; one need only read Root, King and
Kingship, pp. 114ff., to be convinced of its modern birth.
61 Herzfeld, Iran in the Ancient Near East, pp. 200ff.
62 I.M. Diakonov apud R. Ghirshman in review in Bibliotheca Orientalis 15 (1958): 261; von
der Osten, Die Welt der Perser, pp. 56 f.; Ghirshman, Ancient Iran, pp. 87ff. For the lower
chronology see H. von Gall, “Zu den ‘medischen’ Felsgräbern in nordwestiran und iraqi
Kurdistan,” AA 1 (1966): 19 ff.; Porada, Art of Ancient Iran, pp. 138f.; D. Stronach, Pasargadae
(London, 1978), pp. 283 f.; idem in “Achaemenid Village I at Susa and the Persian Migration to
Fars,” Iraq 36 (1974): 246 f.; de Francovitch, “Achaemenid Architecture,” p. 257.
63 D. Stronach, “Tepe Nush-i Jan: A Mound in Media,” BMMA 27 (1968): 185, fig. 14.
64 For discussion and bibliography of Pazuzu heads see M. Noveck in Oscar White Musca-
rella, ed., Ladders to Heaven (Toronto, 1981), pp. 139 f., 318f.
median art and medizing scholarship 1021
p. 100, fig. 104, Cc, 102, Au; J. Waldbaum, “Luristan Bronzes,” Record of the Art Museum Princeton
University 32 (1978): 9, figs, 2, 3; P. Amiet, Les Antiquités du Luristan (Paris, 1976), no. 90. See
1022 chapter thirty-eight
with vertical-horizontonal spout and without the figured plaque have been
excavated in Iran at Marlik Tepe, Sialk B, and Tepe Guran; and many stray,
unexcavated examples, all attributed to Iran, exist in collections.69 There
is no doubt that these metal spouted vessels, both figured and plain, are
Iranian, that is, they were made somewhere in Iran.
Now, whereas in 1969 Calmeyer stressed a Luristan provenience for the
majority of the stray vessels, in 1974 he played down this alleged provenience,
stressed the importance of the Ramadan find, and concluded that the sites
where the vessels were found (“gesicherten Fundorte”) constitute a “Gebiet
das im 8. Jahrhundert nur medisch sein kann”; therefore he believed that
the vessels are Median.70 Notwithstanding this claim, the places where the
plain metal vessels were excavated encompass a wide area in Iran, and
it is archaeologically not possible to assign them to one manufacturing
or cultural center. As for the Hamadan vessel with its figured plaque, it
stands alone without reference to any fixed datum that would allow it to
be considered as either an import or a locally made work. The style of
the figured plaque is best described as “western Iranian,” a general term
indicating that there is insufficient data to realize cultural differentiation
among the many first millennium bc artifacts deriving from western Iran.71
Indeed, it is not impossible that the Hamadan vessel is a Median work of the
pre-unification period, but I would argue that the Median label should be
kept in parentheses, so to speak, and that at present the vessel be considered
simply a product of western Iran.
Calmeyer is aware that his identification of the Pazuzu fibulae, the Nush-i
Jan Pazuzu head, and the metal spouted vessels as artifacts made by Medes
does not thereby furnish scholars with a clear-cut, unified Median style;
they are not products of a Staatkunst,72 Inherent in this conception is its
also The World of Persian Pottery: Gluck Collection (Tokyo, 1980), no. 66, which (if genuine)
has some form of plaque at the rear.
69 See Calmeyer, Datierbare Bronzen, pp. 99 ff. for the references; unfortunately, he treats
notes that the figured plaques resemble the figures holding two animals at bay on a non-
Luristan cheekpiece, no. 37A.
72 Calmeyer, “Zu einigen vernachlässigten Aspekten medischer Kunst,” pp. 116f.
median art and medizing scholarship 1023
73 For a discussion of “Hurrian” art and the impossibility of recognizing and defining it, see
M.T. Barrelet et al., Méthodologie et critique, I: Problèmes concernant les Hurrites (Paris, 1977).
The Urkish lions (about which I plan to write elsewhere) have a bona fide Hurrian inscription,
but stylistically they are not distinguishable from Akkadian or Ur HI art of the second half of
the third millennium bc. As such, the lions epitomize the problem of identifying Hurrian art,
distinct from the art of other Near Eastern cultures.
74 Alas, the negative aspects of this study do not end, for not only are Median art and
artifacts unperceived in our time, so are the Medes themselves: in the sense that (leaving
Persepolis aside) not a single figure depicted in Near Eastern art can be recognized with
security as a Mede, a Mede strictu sensu, not a Zagrosbewohner. I make this strong statement
with ease, and in spite of the fact that a number of scholars think otherwise and have
presumed to recognize Medes represented in art at “Ziwiye,” on stray objects, on Assyrian
reliefs, on Nimrud ivories, viz., Barnett, Iraq 18 (1956): 116; idem, in SPA 14 (1967); 3,000f.,
figs. 1,061–1,063; von der Osten, Die Welt der Perser, p. 57; C.K. Wilkinson, “More Details
on Ziwiye,” Iraq 22 (1960): 217; Culican, Medes and Persians, p. 148; Mallowan, Nimrud and
Its Remains, p. 250, fig. 215; Ghirshman, “Un Mède sur les bas-reliefs de Nimrud,” Iraq 36
(1974): 37 f., pls. 3 and 4; Sulimirski, “Background of the Ziwiye Find. …,” p. 26; Root, King
and Kingship, pp. 115, 260; The Metropolitan Museum of Art Guide (1983): 57, no. 36, On some
of these alleged identifications, all achieved by guesswork, see the comments of M. Wäfler,
Nicht-Assyrer neuassyryscher Darstellungen (Neukirchen-Vluyn, 1975), pp. 118, n. 601; 210,
n. 1,089; 277, n. 1,408. Even the very appearance of the Medes is denied us!
My aim in writing this paper has been to share the results of an examination of the claims
made over the last eighty years that Median art is an entity capable of being recognized by
modern scholars. My conclusions are entirely negative with respect to these claims, but I
strongly believe that the results are positive in the sense démolir pour mieux bâtir. Perhaps
now that the table is cleared, we can start afresh, beginning with the questions of what is
Media, what is Median. Perhaps only then can we confront the other question, why is there
no Median art?
chapter thirty-nine
Forgeries of Provenience and Ancient Culture,” Ancient Civilizations from Scythia to Siberia
9, nos. 3–4 (2003): 259–275.
1 One is the chariot box occupant mentioned both by Cunningham 1881, 151, 154, pl. XII.
no. 8 (as first in the collection of a Louis Cavagnari) and Dalton 1964, v–vi, xl. fig. 21 as in
the collection of Lord Lytton; it was donated to the British Museum in 1953. The British
Museum example seems to be Cunningham page 154, pl. XII. no. 8; the gold horseman
statuette donated by Lord Lytton in 1931, Barnett-Dalton 1964, v–vi, the plate supplement,
is Cunningham’s 154, pl. XIII. no. I.
2 Dalton 1964, 1–39, pls. I–XXI; Curtis 1997, 230–233.
3 Dalton 1964, v; Barnett 1968, 34; Curtis 1997, 231.
1026 chapter thirty-nine
the fact that Cunningham had “gradually bought” his objects over time.9 Dal-
ton and Curtis also noted that Franks purchased “Oxus” material directly
from local vendors as well as from Cunningham.10 There was no single ven-
dor for the British Museum’s “Oxus Treasure,” nor, as Curtis has established,
only one account for the robbed merchant story, or evidence for the actual
existence of Burton’s deposition.11 No evidence exists that Dalton ever met
Burton, and we do not know where the former got his information. Cun-
ningham mentions a Major Burton, not with regard to any story concerning
the discovery of the treasure, but solely as someone who owned “Oxus”
objects.12 Barnett said of Burton’s story that it “suggests more readily the
atmosphere of romance rather than the sober world of archaeology;” and
Pichikyan excitedly and positively described the adventures as “bordering
on fiction,”13 Indeed.
Dalton also noted—but again avoided the logical archaeological implica-
tions of his statements—that in the bazaars of Rawalpindi (and Peshawar?)
there had been “addition and interpolation” to the original Oxus collec-
tion; that dealers in Northwestern India (Pakistan today) “frequently receive
antiquities of various periods discovered within and beyond the frontier [ital-
ics mine] and … incorporate miscellaneous objects of various origins in
some ‘find’ …”14 His concerns were specifically expressed with the 1500 coins
that arrived “with the Treasure,” calling attention to the fact that “there is no
absolute certainty that they all were found with the treasure,” and therefore
they could not function as chronological markers. Put another way, he was
plainly reporting here that the “Oxus” treasure the Museum purchased as
deriving from one locus had been salted. Dalton was admitting—however
obliquely—that the treasure curated by him as from a single find was not
from one find, was not what he was claiming it to be. He was also aware that
in 1883 Cunningham wrote about coins be had seen only recently, i.e. years
after those he saw and reported in 1881 and which he continued to recognize
as deriving from the very same locus as those found in 1877.15
Dalton could not allow himself to dissociate the coins from the treasure;
and he did not mention that a numismatist had earlier challenged their
inclusion in the treasure.16 Nearly all scholars working with Dalton’s account
did not perceive the implications of his caution about the multiple origins
(and thereby their obvious absence of unity) of the coins (not to mention the
other objects in the treasure), blithely accepting them as part of the treasure.
Collon expressed initial doubts about the coins belonging to the treasure
but dropped it.17 And more recently, Curtis stoutly defended them as bona
fide components of the treasure; he cited Cunningham’s claim/defense on
the matter, which was based on nothing more than the “statements of the
collectors themselves,” i.e. the parti pris plunderers and the bazaar vendors.18
Curtis understood Dalton’s abeyance—“the origins of the Treasure seem to
be fairly diverse,” and followed his lead, requesting that scholars accept the
coins as a component of the treasure. With more import than he realized
he wrote “… it does seem illogical to accept the Treasure itself as a group
and to throw out the coins.” Quite so. But the argument expressed here also
expresses the reverse: it seems illogical to throw out the coins and accept
the treasure as a group, as Dalton and Curtis most surely understood, for
the coins are in precisely the same unprovenienced situation as every other
object in the treasure. Both museum curators therefore “saved” the coins—
hoping to convince scholars thereby of the archaeological integrity of the
complete treasure housed in the British Museum.
In fact all the information assumed to be known about the derivation of
the Oxus Treasure, which “has obtained the recognition of archaeologists
and scholars,” or “general agreement”19 derives from Cunningham, Franks,
and others noted by Curtis, and also Dalton—the latter two making archae-
ological determinations about the unity, the Oxus locus, of the Museum’s
collection from London. But the reality remains to instruct us that no one
writing about the Oxus treasure, and this includes Burton and Cunning-
ham, had any personal experience or knowledge whatsoever concerning
the find spot(s) of the purchased objects about which they reported. Cun-
ningham who “gradually bought” objects in bazaar stalls situated hundreds
of miles distant from their alleged find spot, and years after the find date
of the recovery reported to and by him, could not have “confirmed” any-
thing about the origin of the treasures he collected.20 Nor can any modern
scholar “confirm” merely by citing others who themselves cannot confirm
anything they report. What remains completely unknown is (1) where and
when the alleged Bokhara merchants acquired their (mostly unidentified)
alleged treasure goods, (2) what role the Peshawar/Rawalpindi bazaars
played in the assemblage of these very goods, and (3) whether the objects
purchased by Cunningham and Franks were in fact those carried by the
Bokhara merchants.21 Porada put it quite neatly and correctly: Dalton’s story
was a “reconstruction”22—and an inadequate one at that.
A believer’s anticipated claim that “surely the objects must have come
from somewhere nearby” is certainly possible, but from an archaeological
perspective is culturally meaningless. I have visited shops in Van, Turkey,
that sell goods from all over Turkey, the United States, Europe, Iran and Iraq;
in Ankara shops I have seen offered for sale material from the Caucasus, Iran,
Uzbekistan and other Central Asian states. The genuine objects sold in the
Peshawar/Rawalpindi Pakistan bazaar could have derived from anywhere in
Pakistan, India, Iran, Tajikistan, Uzbekistan or Afghanistan, yes, including a
site or sites on the Oxus River. Cunningham (who was a learned scholar)
and Franks were not liars; they merely reported what they had heard from
dealers. They were not the last persons to do this. And it is well known that
to this day dealers—including those at Peshawar—sell forgeries and supply
findspot names for the objects they sell to eager curators and collectors—
who “have to have them.” Dalton did what he believed every curator was
supposed to do, defend the material in one’s prestigious institution against
all non-evidence or evidence to the contrary; he saw it as his responsibility.
In several publications I have argued against accepting the Oxus Trea-
sure as a bona fide find from one specific site or even one general area,
based on some of the reasons given above.23 This position is beginning to
gain support;24 but, generally, scholars have rejected it. In addition to Cur-
tis, I.R. Pichikyan mounted a vigorous defense of the integrity of an Oxus
locus.25 Pichikyan knew that the objects (all of them) derived from one site.
But more: he also knew their precise locus, which was at Takht-i Sangin, a
temple site he was excavating on the Oxus (Amu Darya) River, in Tadjik-
istan: notwithstanding the reality that none of the Takht-i-Sangin artifacts
relate to those in the British Museum—Curtis can only note that three gold
“Oxus Treasure.”
25 Curtis 1997—he does not mention my 1980 and 1987 articles; Pichikyan 1992, I, 31ff.,
plaques are “broadly comparable to the fifty or so examples in the Oxus Trea-
sure.”26 P. Bernard, who supported the integrity of the British Museum’s trea-
sure and the general area from which it supposedly derived, also supported
Cunningham’s locus, claiming contra Pichikyan that the treasure derived
from Takht-i Kuwad (spelled different ways), near Takht-i Sangin.27 E. Rehm
supports Pichikyan’s view; Curtis, citing also Zeymal’s support for Takht-i
Kuwad, reserves judgement.28
The “Oxus Treasure” conception was generated by bazaar archaeology
methodology and scholarship. The objects ascribed to it were found (most
certainly not all, see below part II) by a number of individuals and sold by an
unknown number of bazaar merchants to a number of Europeans over sev-
eral years. It is troubling to read Curtis inconsistently first acknowledging
Cunningham’s and Dalton’s reporting the different dates for the find, and
(correctly) concluding: “It seems certain, then, that the Treasure was not
all found at the same time, and neither was it found in exactly the same
place,” and then two pages later concluding the very opposite: that the “evi-
dence makes it quite clear that the Treasure [all the objects in the British
Museum] was found on the banks of the Oxus.”29 Something is not quite right
here: hearsay about a find spot (time and place, when and where) equals a
known find spot—this is an archaeological oxymoron (an oxusmoron). Pace
Curtis, the British Museum’s treasure cannot be accorded archaeological,
cultural or historical value for it is nothing more than a museum-culture
artifact, from time to time polished and glamorized with museum-speak
rhetoric and defenses. Only with irony can it be proclaimed that Cunning-
ham and Franks “saved the Oxus Treasure for posterity.”30 Historical reality
acknowledges only two facts: Cunningham and Franks purchased over time
individual related and unrelated objects, and, in collaboration with museum
curators, created an “Oxus Treasure.”
There may be an “Oxus Treasure” but there is no actual Oxus Treasure,
no hoard from one site wonderfully preserved by many extraordinary indi-
viduals. All discussion must thus commence from the understanding that
each museum provenanced “treasure” object exists as an unprovenienced
unit, isolated even from any other unprovenienced unit, those alleged (with-
out evidence) to be associated with it. Stripped of the rhetoric, the “Oxus
In a long article written in 1997, Pichikyan returned to Oxus issues again, but
with a renewed and heightened fervor, for his defense this time was in the
context of another significant “Oxus” manifesto. He first reaffirmed his belief
concerning the Oxus locus of the British Museum’s treasure, its authorized
“history,” and its obvious relationship to his site at Takht-i Sangin. Then, with
undisguised passion he shared with his colleagues an amazing discovery he
made. This was seeing with his own eyes in the Miho museum in Japan a
component, the second part, of the British Museum’s Oxus Treasure, which
he proclaims is “the lost and now recovered Oxus treasure (OT-2).” It is “the
greater part,” of that treasure, the lesser being in the British Museum, which,
because of priority of recovery, is now to be known as OT-1.32 The Miho
museum’s “find [purchase] stems precisely from the Oxus treasure and not
any other hoard [italics mine]” which, along with OT-1, is Pichikyan’s temple
at Takht-i Sangin.33 Using the examples of Dalton’s Oxus treasure story (and
Indiana Jones movies) as exemplars, Pichikyan purports to report on the
archaeological discovery of the OT-2 treasure. An individual had acquired
the major part of the treasure that was found in 1877. He buried his share
of the material soon thereafter, but died before revealing the burial spot to
his children. From that time, his sons and grandsons (great-grandsons are
not mentioned) continuously searched for it and after 120 years of patient
digging the grandsons found the buried hoard! Thereupon, they somehow
got the mass of material to Peshawar thence to England—the old OT-1
trail—and finally to its host country, Japan.
31 See for example a gift Wollostan Franks donated to the British Museum, an axe he said
was found at Hamadan in 1880, a provenience Barnett accepted and passed on as gospel,
Barnett 1968, 46–47, note I, pl. VII, 5; for a summary of some of the provenience issues raised
here see Muscarella 2000, 31–32, 53–61, 76, 81–83, 141, 146–147.
32 Pichikyan 1997, 306–307; see his note 2.
33 Pichikyan 1997, 308–310.
1032 chapter thirty-nine
We turn to the chronology of the individual objects in the two museum col-
lections and ask: are all or some of the objects in the OT-1 collection ancient
artifacts, made by ancient craftsmen? The answer begins with what one
may truly refer to as an official document: not one object in the collection
was excavated, not one has a known provenience; each is a phenomenon
wi th no ancestry beyond provenance in bazaar stalls. Resulting from this
situation, we naturally ask of each object “why is it genuine?” and seek a
response by evaluation against what is known from manifestly genuine—
that is excavated—artifacts. What results is that for a large number of the
objects in OT-1 (for OT-2, see below) no defensible evidence commands our
acceptance of them as ancient artifacts; in other words, they are most proba-
bly modern forgeries. The results from an archaeological and common sense
perspective are the same: if perchance they are in fact ancient we can not,
and may never, know.
As early as 1905 Dalton, along with his other reservations, raised the
issue of forgeries among the Oxus material. He claimed that the chalcedony
cylinder seal (no. 114) in the treasure had recently been copied in gold.41
The gold cylinder was seen and reported by Cunningham (it is not revealed
whether he owned it), but he did not mention the chalcedony seal.42 Dalton
and curated in the British Museum did derive from the same locus, and
belong to one Oxus Treasure? This conundrum has been evaded by those
defending the treasure’s integrity—perhaps because confronting it leads to
contradicting the locus and chronology, the archaeological integrity of “The
Oxus Treasure.”
I add to the list another “Oxus” silver statuette curated in the Lands of the
Bible Museum, Jerusalem representing a sad-looking man holding a lotus in
his left hand, published as Achaemenian. In fact, it is a sad man and a very
bad modern forgery.49
Also discussed by Dalton was the matter of the authenticity of the gold
plaques.50 Again he accurately stated the problem: suspicion has been
aroused by the “character of some of the figures, which are so rude and
grotesque that a child might have designed them …” a thought appropri-
ately echoed by Collon, but here employed by her as a defense of a forgery,
no. 63.51 And again Dalton turned away from such a view, because it “pre-
cludes detailed criticism”—a phrase left unexplained, but in fact uninten-
tionally communicating that he was aware that no objective analysis of their
birth dates would ensue. A feeble rationalization is given instead: gold was
common in the area in ancient times and crudeness need not condemn
a piece—thereby suggesting that crudeness guarantees their authenticity.
Curtis also allows that “some of the plaques are extremely crude,” but sin-
gles out only two as forgeries, nos. 63 and 64.52 R. Ghirshman picked up the
crude-defense theme to explain why he knew that the “unique” plaques of
the “Oxus” treasure and some plaques (which in fact are all forgeries) in
the Foroughi collection were genuine votive objects donated to a temple:
“La qualité de l’exécution est secondaire.” Look to the metal to determine
ancient authenticity: gold = authenticity, an assessment also accepted by
Rehm.53
Inasmuch as the gold plaques have been declared genuine by fiat, they are
readily invoked by all “Oxus Treasure” believers as prime evidence to doc-
ument archaeological and cultural conclusions about votive gifts derived
from a putative temple. The most recent publication claims “… the gold
plaques seem to have been votive in character … placed in a temple to rep-
resent the worshipper in the presence of the deity. The association with a
49 I. Ziffer in Seipel 1997, 119–120, no. 166; other forgeries of Achaemenian art in the same
temple is reinforced by the fact that many of the figures are priests … we
are dealing with a temple treasury …”54 (italics mine). Earlier, Ghirshman
explained the differences in the garments worn by the figures represented
on the plaques as indicating different ranks; the plaques reveal class differ-
ences.55 The plaques thus—merely because they exist—reveal not only their
ancient birth dates and functions, they also inform us about social differen-
tiation in Oxus religious society.
Tendentious, after the fact, defenses of the plaques notwithstanding, to
the contrary, crudeness of style and execution as well as obvious misun-
derstandings of form and details are essential issues that most certainly do
not preclude critical review. Crudeness and apparent aberrations in ancient
(excavated) artifacts occur rarely, but we are here discussing born-in-the-
bazaar phenomena, non-excavated material, and quite a lot of it: and no
comparably crude Achaemenian material has to my knowledge ever been
excavated. The unparalleled nature of the plaques in the excavated cor-
pus is to be confronted solely on archaeological terms, without resort to
assumptions, fiats, beliefs and ad hoc cultural interpretations. What ensues
from such an investigation is that all essential details on the plaques—
eyes, noses, hair, faces, hands, feet, hats, clothing, shoes, and sleeves—are
completely unparalleled in any style and execution within any sphere of
ancient art—most certainly including Achaemenian, which seems to be the
intended cultural designation of their modern, neo-Achaemenian, creators.
Why are they ancient?
Plaque no. 48, first described as a priest by Cunningham, is glamorized by
Curtis as “one particularly fine plaque …”, “the finest of the gold plaques.”56
The plaque is worked in low relief, apparently unique in the collection. The
plaque has other unique—i.e. unparalleled—features aside from a poorly
articulated “Achaemenian” style. Note the trousers, hat, skirt, mask, etc.; the
sword’s scabbard and hilt position are wrongly conceived and executed, and
the sword is not attached to the belt and thus floats in space; the mask shows
the chin and does not seem to cover the mouth; the garment either has short
sleeves, or the two lines at the wrists may have been meant to be the sleeve
ends, or “bracelets”—but in fact, this issue is but another indication that
the “provincial craftsman” had no idea what he was executing in the first
place. The figure also has two left feet, as does no. 49—I have not found
two left feet on any figure in excavated Achaemenian art. Further, would a
We now turn to the issue of the modern forgeries in OT-2, published in the
Miho museum Catalogue (2002). I shall not expend much time here, for we
have before us another museum-generated scenario. I do not claim that all
the many objects catalogued in the Miho museum under the rubric “Bac-
trian Treasure” or OT-2 are modern forgeries, but a large number clearly are.
Not unexpectedly, the most obvious of these are the many decorated OT-2
gold plaques, of which I argue not a single example can be accepted with-
out reservations to be ancient. All are crude and unparalleled in execution of
body forms and stylistic details; indeed many were spawned from OT-1 mod-
els, formally and as artifact models (bastards begetting bastards), especially
the “barsom”-bearing priests (Catalogue nos. 64–66), although none here is
armed (nor are the OT-2 “priest” statuettes). The valuable new information
to be derived from bazaar archaeology is manifold.61 The OT-2 figures have
many varieties of clothing and trousers, some embellished with incised pat-
terns, varieties of hats that include a concave top, all have two left feet, all
have differently and strangely executed eyes, noses, hands, and shoes. The
other plaques depict a multitude of crudely executed figures, holding a vari-
ety of objects, and some have crowned heads. Especially insulting to those
viewing this material is the sun disc with a figure holding up his hands surely
in despair (no. 71), a fortress populated with females (no. 72), the chatting
scene (no. 81), and the “agriculture scene” (no. 86).
Not one of the original five, eventually six, gold and silver statuettes of
“priests” holding “barsom,” here indicated by thin strips of gold, can be
assumed to have had an ancient existence. They are fabricated as thick,
thin, tall, short and stocky, with thin arms, thick arms, straight and bow
legged, with fingered hands or blobs, with different hats, different sized
and executed face masks, and with different shaped and formed eyes and
eyeballs—the latter depicting dead, non-ancient men. Not one of these
lifeless creations could, would, have been made in an ancient workshop—
the artisans would have been deported.
As for the human and animal statuettes (nos. 51–62), who among us
can claim with certainty to know their birth dates, inasmuch as they are
unparalleled anywhere in the ancient excavated corpus? And how many of
the Achaemenian style bracelets, torques, and earrings (nos. 145–186; surely
nos. 146, 170–173) are modern forgeries or ancient is impossible to determine
without disinterested, non-museum generated laboratory testing: there are
scores of modern jewelers who could make any object on view in the Miho
museum. This cautious view obtains also for the plain gold vessels and ladle
(nos. 123–144), all of which could easily have been produced in modern
61 For more information about ancient Zoroastrian religion revealed from forgeries see
goldsmith shops: the same hesitancy and caution obtains for the other
scores of trinkets from the OT-2 treasure. I have no idea what the many
blanks, body parts, and miniature bows and daggers signify, or whence
they derived. That there are a number of apparently genuine artifacts—
plundered from now destroyed sites—that have been collected, curated,
published and displayed under the OT-2 rubric is most probable.
Outside the walls of two museums in opposite corners of the world
there is no Oxus Treasure. Both OT-1 and OT-2 are modern constructions
documented by pastiches of stories and composed of ancient and modern
artifacts. Collectively. they neatly illuminate not the culture of the ancient
Oxus region but the cultural interests, ideologies and internal agendas of
modern artifact-acquiring museums—and these interests and agendas are
not those of archaeology, which seeks (should seek) to comprehend an
accurate context of material remains in order to begin to understand their
cultural history.
Bibliography
Barnett, R. 1968: The Art of Bactria and the Treasure of the Oxus. Iranica Antiqua
VIII, 34–53.
Bernard, P. 1994: Le temple du Dieu Oxus à Takht-Sangin en Bactriane: temple du
Feu ou pas? Studia Iranica, 81–112.
Bleibtreu, E. 1998: Ein Chalzedon-Rollsiegel aus dem Oxus-Schatz und seine Nach-
ahmungen in Gold. Archäologische Mitteilungen aus Iran und Turan 30, 145–154.
Cellon, D. 1995: Ancient Near Eastern Art (Berkeley).
Cunningham, A. 1881: Relics from Ancient Persia in Gold, Silver, and Copper. Journal
of the Asiatic Society of Bengal 51, 151–186.
———. 1883: Relics from Ancient Persia, in Gold. Silver, and Copper. Second, Third
Notice. Journal of the Asiatic Society of Bengal 52, 64–67, 258–260.
Curtis. J. 1997: Franks and the Oxus Treasure. In M. Caygill. J. Cherry (eds.), A.W.
Franks, Nineteenth-Century Collecting and the British Museum (London), 231–
249.
Dalton, O. 1964: The Treasure of the Oxus with other Examples of Early Oriental Metal-
Work (London).
Ghirshman, R. 1964a: The Arts of Ancient Iran (New York).
———. 1964b: Le trésor de l’Oxus, les bronzes du Luristan et l’art mède. In K. Bittel
et al., (eds.), Vorderasiatische Archäologie (Berlin), 88–94.
Green, A. 2002a: The ‘Treasure of Bactria’ in the Miho Museum. In Miho Museum
Catalogue, 2002, 220.
———. 2002b: The Gold Plaques of the Bactrian Treasure. In Miho Museum Cata-
logue, 2002, 225–227.
Miho Museum Catalague 2002: Catalogue of Treasures of Ancient Bactria (Miho
Museum).
forgeries of provenience and ancient culture 1041
Introduction
Ancient Persia: The Art of an Empire, ed. D. Schmandt-Besserat (Undena, 1988), 23–42.
1044 chapter forty
1 Few authors, including the present writer, have avoided this error in the past. What
is argued here is that in the future traditional attitudes that accept received information as
necessarily reflecting reality be eliminated, and that rigorous attention to sources be the rule.
excavated and unexcavated achaemenian art 1045
and terracotta rhyta, silver, bronze, glass and stone bowls, gold, silver and
bronze bracelets and other types of jewelry, decorated sheet metal, textiles,
ivories, and various small items. Moreover, that list knowingly presented
as incomplete, should be expanded considerably, and to this issue we now
turn.
Appendix B (Muscarella 1977a) was compiled with the view to supplying
a corpus of those Achaemenian objects concerning which one could be cer-
tain came from specific sites.2 Maintaining a strict documentation of prove-
niences yields two facts, both objective and not subject to personal opinion:
“it is not a question of believing, or judging, but of knowing” (Ridgway 1977,
20). First, and foremost, all the objects derived from excavated sites are gen-
uine products of the Achaemenian period, and as such should serve as the
prime data that inform us about the nature and characteristics of Achaeme-
nian art. And it is this ordered category that forms the base from which
one analyzes and controls the unexcavated material. Second, any attempt
to record accurately the find spots and alleged occurrences of Achaemenian
art throughout the Near East and surrounding areas ought to begin—and
end—with a strictly defined list of proveniences. Put another way, specific
objects from specific sites furnish the data for meaningful research. If the
data are not under control in a disinterested fashion, so to speak, if there is
no certainty that the objects are in fact what is claimed for them, the con-
clusions will be at the least misleading, at the worst, fictional.
At the same time, I have come to the conclusion that certain important
objects deserve a separate listing almost equal to that in Appendix B. In this
proposed parallel listing the situation is such that while no unbiased author-
ity witnessed the actual discovery of the objects, all the evidence compels
us to conclude that the find occurred: if not necessarily from the suggested
site (allowing for caution), certainly from a recognizable area, e.g. some-
where in Bulgaria (at least the Balkans), or in the Caucasus, and so forth. It
is understood that some subjectivity is involved—hence the separate list—
2 The claim made in Appendix B (p. 193) that it was compiled solely from material “exca-
vated under controlled conditions” was not intended to suggest that an archaeologist employ-
ing modern techniques supervised the excavation. Rather, the intent was to convey the
fact that some archaeologist or reliable authority—i.e. not merely alleged local peasants—
witnessed the find. Thus, for example, although the silver rhyton from Arinberd was listed, it
was found during a construction operation and turned over to the authorities.
1046 chapter forty
but it can be kept to a minimum and under control if each find is analyzed
on its own merits based on the particular circumstances recorded in each
instance. Accuracy, not pedantry, is the goal (cf. Hamadan below). Thus, I
now believe that sumptuous Achaemenian finds credited to Duvanli in Bul-
garia, Prokhorova in the Urals, north of the Caspian, and Achalgori, and with
reservations, Kelermes in the Caucasus, may be accepted as Achaemenian
finds from these respective areas (cf. Muscarella 1977a, 192 f.).
The Duvanli find apparently occurred in 1925 and was in part dispersed.
The well-known silver amphora (fig. 1) was presented for sale in Sofia but the
handles were acquired separately in Duvanli itself, where a plundered tomb
was observed. Still more material surfaced both at Duvanli and in Sofia; and
in other instances objects acquired at one place matched those from the
other (Filow 1934, 40f.). On the basis of these facts it seems that we take
no risk in accepting the Achaemenian amphora and a silver bowl (No. 15,
fig. 60) as found where claimed; at the very least we may accept them as
from Bulgaria.
Rostovzeff (1922, 123) vigorously defended the finds from Prokhorova as
definitely deriving from that site. Indeed, the objects of particular concern to
us, two Achaemenian silver bowls and a seal, were secured by the Oranburg
museum. Surely it cannot be denied that the objects were found somewhere
in this remote area (cf. Altai), for if found further south, they probably would
have found their way to one of the market cities there, Tiflis, for example.
The Achalgori finds—actually said to have occurred in the village of
Ssadseguri but sold to a dealer in Achalgori—were tracked down within
a short time of their sale in Tiflis (Smirnov 1934, IX). To my mind, the fact
the E.S. Takaischwili did a creditable job of tracing the find, documented
by Smirnov, and because the objects were offered for sale in Tiflis, suggests
that they were found somewhere in the Caucasus, which is sufficient for our
needs. Included among the objects that concern us are gold earrings, gold
plaques, silver bowls and a silver amphora without handles, a unique object
(Smirnov 1934, pls. III, IV, VII–XII).3
The alleged finds from Kelermes—two superb gold bowls, one deco-
rated with repoussé petals and lozenges, the other with repoussé birds and
3 Of some interest is a silver gourd-shaped vessel with two holes on the same side,
48 f., No. 68, pl. XII, as it is exactly paralleled by five terracotta examples in northwest Iran.
Three from Hasanlu and one from neighboring Dinkha Tepe came from Iron Age II, late 9th
century bc contexts, and one was seen in a dealer’s shop in Rezaiyeh. Oscar White Muscarella,
“The Iron Age at Dinkha Tepe,” MMA 9 (1974), 78. Whether the silver example is Achaemenian
or earlier is not clear from the available information.
excavated and unexcavated achaemenian art 1047
animals (Rostovzeff 1922, pl. VII; Schefold 1938, figs. 4, 5)—present spe-
cial problems given the unfortunate circumstances both of the find and its
recording. Rostovzeff (1931, 278ff.) published succinct information regard-
ing the uncontrolled nature of the finds attributed to Kelermes: aside from
the controlled excavations of Wesselowski, one Schulz, a licensed digger
(“an amateur,” Rostovzeff 1922, 5), claimed in a report to the Archaeolog-
ical Commission of Russia to have made important discoveries. Rostovz-
eff describes Schulz’ notes as “manchmal ungenauen, bisweilen absichtlich
erlogenen Berichten,” and claims that he also melted down some of the gold
and sold it for personal gain. Schefold (1938, 9) also noted that the impor-
tant finds came “nicht aus den beiden genau untersuchten Gräbern, sondern
aus Raubgrabungen in der Nachbarschaft dieser Kurgane.” Although accept-
ing the bowls as part of the Kelermes group he lists them as “Unbestimmter
Herkunft.”
From all this what can one conclude about the provenience of the bowls?
Given Schulz’ tendency to lie (if we follow Rostovzeff) we really do not know
whence they derived, Kelermes or elsewhere. At the same time, the cir-
cumstantial evidence surrounding the report suggests that the bowls came
from somewhere in the area of Schulz’ activities (could he have purchased
them and resold them to the authorities?). It seems to me that a caution-
ary attitude allows one to accept a Kelermes provenience only tentatively;
less tentatively, one may accept a general Caucasian provenience. Indeed,
the information available is far from decisive, and any use of the bowls as
Achaemenian objects from the Caucasus must be made with a caveat. (N.b.
that the bowl with the animal reliefs could be provincial Achaemenian or
perhaps even slightly earlier.)
The so-called Oxus Treasure, on the other hand, cannot by any stretch of
the imagination be accepted as a find from one specific site or even area;
to quote Dalton (1964, Preface), the story of its acquisition “suggests more
readily the atmosphere of romance than the sober world of archaeology.”
The story (stories!) attached to its discovery and recovery claims it occurred
over a period of years beginning in 1877 at more than one site, and that the
objects passed through many hands and traveled to many places, includ-
ing Afghanistan; some were even counterfeited. Thus each piece must ulti-
mately be treated as an individual object without a specific provenience,
rather than as a part of a single find (Dalton 1964 XIII ff., XVI; Barnett 1968,
34ff., Porada 1965, 168ff.).
Several other objects omitted in Appendix B have come to my attention
and should be noted. One is a bronze bracelet with animal head termi-
nals and the characteristic Achaemenian indentation opposite the opening,
1048 chapter forty
excavated at Khorsabad (Loud and Altman 1938, pl. 59, No. 127; Amandry
1963, 271, No. 14; Moorey 1971a, 220). Amandry, the source for other appar-
ent finds, suggested that another bracelet from Khorsabad (No. 123) is also
Achaemenian but this is not so certain. Amandry also refers to two gold
Achaemenian bracelets with lion-griffin terminals in the Izmir Museum
from the area of Magnesia (Amandry 1965, 585). The nature of the available
information suggests that the bracelets were acquired by the museum by
purchase or confiscation, rather than excavation. Be that as it may, if they are
Achaemenian as stated, we may assume that they came from somewhere in
western Turkey—with the reservation that “assume” is the governing word
here. Amandry (1963, 270f., No. 6) further reports a major find of Achaeme-
nian material that occurred in 1962 in Vani, Georgia, and which is now in
the Tiflis museum. To my knowledge, both the objects and the nature of
the find remain unpublished, but gold bracelets, earrings, a diadem, a pec-
toral (gorget?), and a bowl are mentioned. Until the objects are available
for examination we must reserve judgement, although given the widespread
dispersal of Achaemenian objects it would not be surprising if such a find of
Achaemenian material occurred in the Caucasus.4
4 Amandry (1963, 267, notes 2, 13) mentions a bronze ladle found in 1904 at Aydin, another
from Sardis, and a bracelet from Antioch, all in the Istanbul Museum. I have not been able to
get information on these objects up to the time this paper went to press (to meet a deadline).
Otto (1944, 10, n. 2) mentions two bracelets from Nippur, also in Istanbul. Lütfi Tuğrul informs
me he can find no record of their existence. Kantor (1957, 21) cited as Achaemenian a stone
relief fragment of a striding lion excavated in Byblos. Half of the head is missing but enough
of the lion remains to allow one to conclude that the attribution is not certain. One would
expect belly hair “wings” and two raised areas under the eye; further the depiction of the
mane is not typically Achaemenian in execution.
excavated and unexcavated achaemenian art 1049
5 This fact explains why A. Godard, Le Tréser de Ziwiye (Haarlem, 1950), 69ff., believed
that the same artists who carved the reliefs were also the gold workers; cf. Stucky 1976, 21f.
for the opinion that from the Achaemenian point of view two and three dimensional objects
were not considered to be separate entities.
6 For other fibulae represented at Persepolis and excavated there and at various sites in
Iran see Muscarella 1969, 283 (IX), 284 (XIX). For additional published references to fibulae
excavated in Iran see: Zendan AA 1965, 739, fig. 59 a, b; Baba Jan, Iran VIII (1970), 176; Bisitun,
AMI 3 (1970), 154 f., fig. 221, pl. 75:1; Ghalekuti, T. Sono, et al., Dailaman III (Tokyo, 1968), T5, T7;
Godin, T.C. Young, Jr., Excavations at Godin Tepe, First Progress Report (Toronto, 1969), pl. 123,
no. 2; Nush-i-Jan, Iran VII (1969), pl. X, b; Pasargadae, Iran I (1963), 41; War Kabud, Phoenix
XIV (1968), 123. fig. 42, left; Bastun, Iran VII (1970), 177; AMI 8 (1970), 51. pl. 33:1 (late). For the
“Ziwye” fibulae see Muscarella 1977b, passim. There are no fibulae from Sialk, cf. Moorey 1971,
174.
1050 chapter forty
7 Amandry in 1965, 583—see also 1958b, 1140—seems to have changed his mind and
accepted Delegation VI as Lydians. It is of interest that he did not compare the pair of
bracelets with lion-griffin terminals in the Izmir museum mentioned above with the same
bracelet-type carried by the Delegation VI tribute bearers. A pair of bracelets of the same
type is also part of the Oxus Treasure (Dalton 1964, pl. I, No. 116).
excavated and unexcavated achaemenian art 1051
erbringen.” The valuable objects carried by the Delegations could have come
from any part of the empire, not necessarily “auf Ateliers in diesen Gebieten,”
for these objects “waren begehrtes Kaufgut des achämenidischen Reich-
sadels, das man sich wohl von weither kommen liess ….” With these opinions
I am in agreement; the final resting place of the objects illustrated on the
reliefs is our sole concern, because that is all we know, and that place is
Persepolis in Iran.
At the same time it should be noted that a number of scholars have not
neglected to discuss the illustrated works of art on the reliefs in context
with similar, three dimensional works: viz. Amandry in his definitive arti-
cles (1958a, 1959, 1963, 1965) on bracelets and amphorae; Luschey (1938a
and 1939) on bowls; Filow (1934) on amphorae; Moorey (1971a, 219, 313) and
Pudelko (1933–1934, 85) on bracelets; Kantor (1957) on bracteates; Otto (1944,
9 f., n. 2) on vessels and bracelets; … Goldman (1951) and Stucky (1976) on
chapes; Walser (1966, Chapter V) for objects in general. But the literature is
also filled with missed opportunities. If the reliefs are considered at all it is
to cite them in passing as comparanda for three dimensional pieces, rather
than as the starting point of the discussion (see the chapters on Achaeme-
nian art in Ghirshman 1964, Culican 1965, Porada 1965).8 And there are puz-
zling examples where a scholar has corrupted his argument concerned with
an object or motif in Achaemenian art by citing unexcavated and spurious
objects rather than the reliefs where an appropriate parallel exists (Culican,
Chapter V, passim; Kantor 1977, 14, n. 16).
The horn rhyta9 and lion-griffin recognized on the reliefs of the Tomb of
Petosiris (Lefebvre 1923, pls. VII, VIII, IX), figs. 4, 5, a tomb of an official at
Hermopolis in Egypt, dated to ca. 310–300bc, may not be quoted in the same
manner as that suggested for the Persepolis reliefs. For although Achaeme-
nian objects are clearly depicted they cannot be said to be illustrated. For
some reason no attempt was made to create an exact copy of a rhyton or lion
griffin with all details “beinahe photographisch” represented. Their value,
speaking only with regard to the issue under discussion, is not so much one
8 That the style of the relief documents is sometimes ignored may be illustrated by one
example. No one, to my knowledge, has pointed out that on the reliefs the animals on the
amphorae do not rest their feet directly on the vessel and that their heads face out; these
features are also noted on the Duvanli amphora and on the one depicted on the tomb wall at
Karaburun, Mellink 1973, pl. 44 (see also Muscarella 1977a, 179, no. 102–105). Note that rhyta
are not represented on any Achaemenian relief, pace Ghirshman 1964, 252.
9 At least one—possibly more—specifically Achaemenian style rhyton is represented on
of style but one of provenience; the reliefs inform us about the manufacture
of Achaemenian objects in Egypt around 300 bc (Muscarella 1977a, 194,
n. 100; Adriani 1939).
The rhyta and bowls depicted on the walls of the tomb have not been
ignored by certain scholars: Luschey 1939, 74, no. 422c; Adriani 1939, 352, 357,
359ff.; Svobada 1956, 12f., 61ff.; and Culican (1965, 153) specifically pointed
out that the tomb yields “evidence for the manufacture of Achaemenian
objects in Egypt ….” But it is of some interest to note that Montet (1926),
who argued vigorously for evidence of Persian influence in the scenes on
the reliefs, as opposed to Greek influences suggested by Lefebvre, totally
ignored the rhyta and bowls; he stressed what he believed were Mithraic
elements in the bull sacrifice scene (Lefebvre 1923, Pl. XIX). Picard (1930)
correctly dismissed this opinion, but went out of his way to deny any Persian
influence, even for the characteristic lion-griffins on the bier. To Picard even
the rhyta could be Greek, and he made no mention of the Persian examples
available to him (cf. Adriani 1939, figs. 4–6).10
The same conclusions expressed for the Tomb of Petosiris objects must
equally obtain for the vessel with griffin headed handles depicted on the
wall painting at Karaburun in Lycia (Mellink 1973, pl. 44). Here too, judging
from the photograph, we have evidence of an Achaemenian amphora with
animal handles in an excavated context.11
To summarize the above conclusions, a map with excavated find spots
recorded must include: amphorae, swords and chapes, bracelets and vari-
eties of vessels at Persepolis in Iran; horn rhyta, vessels and lion-griffins in
Egypt; and an amphora in Lycia.
We turn now to the second question posed above, concerned with recog-
nizing whether or not the proveniences offered for unexcavated objects are
verifiable or just convenient. The position argued here is that if an object is
10 Note that Adriani (1939, 359 ff.) also noted a Persian background for the Petosiris rhyta,
but he believed (360) they “sono le stesse influenze che si riconoscona in altri prodotti di arte
egiziana contemporanea ….” He believed that the Persian examples are in fact to be dated to
the time of the Petosiris tomb, ca. 300bc, and were made in Egypt.
11 Because I was concentrating on objects that could be used stylistically both to define
Achaemenian art and to function as a guide for unexcavated objects. I left out the Petosiris
and Karaburun evidence from Appendix B.
excavated and unexcavated achaemenian art 1053
12 My own experience has been more positive than negative. In writing this paper curators
from several museums have graciously and promptly answered my queries. Not to put too fine
a point on it, the answer given to my queries concerning objects “said to come from X” was
the one expected; see Muscarella 1977a, 160 ff.
13 Inasmuch as the amphora was not excavated by an archaeologist, the source of the
anomaly, one object “excavated” in three separate areas: but, as we shall see,
not an uncommon phenomenon.
The pair of Graeco-Persian handles from a now lost amphora, one in
Berlin, one in the Louvre, have also been found, if we believe our sources,
in more than one place. For they “proviennent selon certaines informa-
tions, d’Amnisos (actuelle Samsoun), à l’est de Sinope et, selon d’ autres,
d’Arménie” (Amandry 1959, 53f.). Amandry himself had earlier tentatively
accepted the Armenian provenience (Amandry 1958b, 1141), and, indeed,
most scholars have elected Armenia as the place of origin (Luschey 1938a,
761, 764; 1939, 57, n. 330; Filow 1934, 203; Svoboda 1956, 47; Culican 1965, 125).
The ex-Borowski, Berlin amphora, on the other hand, has been allowed to
come only from somewhere in Iran (Amandry 1958b, 1141, fig. 8, “recently
found in southern Persia;” 1959, 43; Tuchelt 1962, 85); and the superb silver
rhyton-amphora in the Pomerance Collection was published as “found in
Iraq,” based on a dealer’s claim (Terrace 1966, 52).
Moving to other vessels, we find similar problems; in some instances an
example is accepted by consensus as deriving from one place, in others there
are plural proveniences. Thus, nearly all scholars have accepted as coming
from Syria the Berlin bronze rhyton (31158), a bronze amphora in a private
collection (Amandry?), and a cup with a bull protome purchased by Woolley
in Syria (Amandry 1958b, 1140, fig. 4; 1959, 44 ff.; n. 75; Culican 1965, 122, 125;
Moorey 1974, 159; Luschey 1938a, 762; 1939, 57, n. 330; Tuchelt 1962, 83, no. 1,
85; Zahn 1930, 148, fig. 1; Svoboda 46: with a slight reservation expressed for
this and other rhyton proveniences). The only one of these three vessels
that has a verified background in Syria is the one purchased by Woolley; but
where his Armenian vendor actually got it, and from whom (he claimed he
found it at Marash), must remain an open question.
All interested scholars have accepted the purchased Metropolitan
Museum silver horse-head rhyton (47.100.87, fig. 6) as coming from Mazan-
deran in Iran, probably on the dubious authority of Pope (1935, 1; Casson
1938a, 355; Otto 1944, 10, n. 2; Luschey 1938a, 763, n. 5; Tuchelt 1962, 59,
n. 13; Ghirshman 1964, 252). Nevertheless, all that is objectively known of
its history is that it belonged to the Jacks collection before the Metropolitan
Museum purchased it.
hardly qualifies as a “reliable person.” Compare the reliable people discussed by Ridgway
(1977), who prevented the truth about the Manios fibula’s true provenience—a dealer—
coming to light for almost a century.
excavated and unexcavated achaemenian art 1055
14 I wish to thank Pierre Amiet for sending me information on these objects and other
information.
15 For bowls see: Ackerman apud Casson 1938b, 371; Terrace 1962, no. 61; Amandry 1963,
260, 262, 271; Culican 1965, 120, 136: ILN, September 8, 1962, p. IV; Luschey 1938b, 78f.; 1969,
42, 49, 53; Silver: ILN, September 8, 1962, p. IV; Amandry 1959, 41, pl. 21: 4, 5; Stone vessels:
see Muscarella 1977a, n. 42a, p. 182, n. 83. In the Gazette Archeotogique 8 (1883), 237ff., pl. 41,
E. Babelon published a small stone relief from the Duc de Luynes collection, which was
“acquis en Syrie où il a peut-être été trouvé ….” See also Herzfeld 1941, fig. 362; Amandry 1959,
42, n. 39, “Achetée en Syrie.” The plaque is indeed Achaemenian and it is unique; its origin,
however, cannot be claimed with any certainty.
1056 chapter forty
various portable objects that he acquired. The small objects were illustrated
in seven plates, four (figs. 157–160) were labelled objects “trouvés dans les
ruines d’Ecbatane,” and three (figs. 161–163) as objects from “Ecbatane et
environs.” In no instance did de Morgan inform the reader how he acquired
the objects, that is, which, if any, had been personally excavated, and which
were purchased. The latter would have resulted from what de Morgan de-
scribed as “une exploitation régulière,” for “il n’ existe plus guére aujourd’ hui
de parties de l’antique Ecbatane qui n’aient été exploitées” (de Morgan 1896,
236). But with regard to his method of acquisition nothing more was given
than the ambiguous statement about objects “qu’ on recontre dans les ruines
de la ville …” (de Morgan 1896, 253). Therefore, although I do not think it too
strong to state that the objects published by de Morgan were acquired by
him at or near Hamadan, in the final analysis we do not know how many
were found, or acquired, in outlying villages, and how many were in fact
found within the city itself.
Calmeyer (1972, 65) accepts the de Morgan finds as bona fide material
from Hamadan, a position I find too secure given the inadequate informa-
tion available. From all the objects illustrated by de Morgan, and obviously
from different periods, Calmeyer singles out a bronze animal pendant, a
small gold figurine of a female, and a small bronze duck weight (figs. 160: 10,
157: 9, 159: 14) as Median or Achaemenian finds from the site. To my eyes, the
gold female is not Achaemenian, nor necessarily Median; nor is the pendant
easily characterized except as pre-Achaemenian; the duck weight, however,
is probably Achaemenian (cf. Schmidt 1957, pl. 82:4).
In 1914, a French expedition consisting of Fossey and Virolleaud con-
ducted excavations at Hamadan. Photographs of Le Breton’s restored draw-
ing of fragments of a pre-Achaemenian spouted vessel (figs. 7, 8),16 and a
number of small animal figurines (Ghirshman 1964, figs. 122, 124; Calmeyer
1969, 102f., fig. 106; 1974, figs. 3–5), are the only objects made available for
study. While both groups of objects are clearly pre-Achaemenian, there is
no reason to assume that they are Median (cf. Calmeyer 1972, 65; 1974, 114).
These objects alone of all the many published as deriving from Hamadan
16 It will have been observed that the cast curved spout with the animal head terminal
of Calmeyer 1974, fig. 3 (Figure 8 here) is not the same spout in the restored drawing.
Pierre Amiet, who has generously allowed me to republish the Hamadan vessel fragments,
has informed me that Le Breton did not believe the spout belonged to the same vessel as
represented by the other fragments, and therefore drew the spout he thought appropriate.
If this is true, and Amiet supports the idea, we then have parts of two vessels excavated at
Hamadan.
excavated and unexcavated achaemenian art 1057
17 Even in such an instance there are problems: see Muscarella 1977a, n. 26; and note again
18 In Herzfeld 1941, 195, it was stated that the object was lost but it was purchased by
the Metropolitan Museum of Art in 1952 (52.119.12; fig. 9). How it got lost by Herzfeld and
acquired by a dealer is not known (by sale?). With regard to objects purchased in Hamadan
it is pertinent to recall a reference in Aurel Stein’s Old Routes of Western lran (London, 1940),
291, to “antique seekers from Hamadan” digging in Luristan. Surely they brought back their
booty to Hamadan to sell: which would then be listed by the purchasers in publications as
“from Hamadan.”
19 A reading of Herzfeld 1926, 1928 and 1930 produces no statement that Herzfeld himself
witnessed the find he discusses as follows: “Beim Neubau einer Häuschens kam ein antikes
fundament zutage, in dem die Goldplatte zwischen zwei durch ihre gute Bearbeitung auffälli-
gen Quadern endeckt wurde” (1926, 2105), and “In the town of Hamadan, there was discovered
a few years ago …” (1928, 1). Who was the source for this story? Where did Herzfeld encounter
it? Where did he first see the tablets?
20 On p. 187 of Wilkinson 1949 is the statement that in 1947 the Metropolitan Museum “pur-
chased some of our finest pieces, but we had the opportunity of adding twenty-four antiqui-
ties from prehistoric and Achaemenian Persia, all from scientifically conducted excavations.
These last were acquired by exchange from the National Museum, Teheran ….” In the article
excavated and unexcavated achaemenian art 1059
right). Here again Herzfeld neglected to reveal where he had viewed the
objects, but at the same time he volunteered no information regarding the
alleged find spot of the bowls (even in AMI 1937, where he was defend-
ing their authenticity). Nevertheless, in an anonymous article in the ILN,
April 16, 1955, 699, the Metropolitan Museum bowl (47.100.84, fig. 10, ex-
Brummer) as well as other objects acquired on the market by the museum
(below), were presented as “gold and silver objects from Hamadan.” Earlier,
Olmstead (1948, 353) also believed that the bowls were made at Hamadan.
But this attribution did not go unchallenged and he was cautiously ques-
tioned by Schmidt (1953, 37, n. 356): “We must assume that Olmstead had
definite information with regard to the find-location of the bowls …. Herz-
feld … reported neither their find-spot nor their present location ….” Aman-
dry also challenged the Hamadan provenience for the bowls (1963, 271, n. 17)
but not for other objects.
By far the most important group of objects traditionally accepted as deriv-
ing from Hamadan is the collection of gold “rhyta,” bowls and daggers that
are now in the Metropolitan Museum and in Teheran. In the first publication
of the Teheran gold objects (ILN, July 21, 1956, 107) it was merely stated by
the anonymous author that “The provenance of these objects is not exactly
known, but they are believed to have come from Hamadan;” however, who
“believed” it, and why, is not revealed (cf. Mostafavi 1953, 145, 147; the bowl on
148 is listed by me as a probable forgery, Muscarella 1977a, no. 108). Of inter-
est to those who keep a record of shifting proveniences, is the gold bracelet
(ILN, July 21, 1956, 107 lower right corner) listed as “reputedly from Hamadan”
and conveniently dated to the “sixth-fifth centuries bc;” this bracelet and its
mate are usually assigned to Ziwiye and to the seventh century (as in ILN,
April 16, 1955, 699, center; see Muscarella 1977b, 200, n. 6).
Four gold Achaemenian objects acquired by purchase by the Metropoli-
tan Museum were categorically claimed to derive “from Hamadan” in the
anonymous ILN article of April 16, 1955 mentioned above: a gold bowl with a
Darius inscription (fig. 11), a buckle (fig. 12), a dagger (fig. 13), and a “rhyton”
(fig. 14) (54.3.1–4). In the same year Wilkinson (1955, 213, 220 ff.) published
these objects in more detail, describing them as “from Hamadan,” “found
in the ruins of Hamadan,” or with the modifications “reputedly found in
three Achaemenian objects, one the silver Artaxerxes bowl, are published with no comment
regarding their acquisition. Lest there be doubt, none of these objects formed part of the
exchange; all three were purchased from the Brummer collection. The only Achaemenian
object given in the exchange was a bronze horsebit (48.98.19) of the same type as Schmidt
1957, pl. 78, no. 4, pl. 79, no. 7.
1060 chapter forty
21 It is not clear why Calmeyer (p. 66) discusses a gold vessel in Cincinnati and a gold plate
in Teheran in an essay on Hamadan. Indeed, Calmeyer does note that the find spot of these
and other objects “werden oft bezweifelt,” but does not follow this thought to a conclusion
nor elucidate it; both pieces have been cited by me as probable forgeries (Muscarella 1977a,
nos. 109, 152). Cf. Luschey 1968, 117 who states of objects said to be from Hamadan: “Jedoch
sind die Fundumstände und die Herkunft nicht ganz gesichert.” Moorey 1971a, 315 accepts a
Hamadan attribution for a silver bowl in the Ashmolean Museum as possibly “correct in this
case,” and suggests that it may be “a rare example of ‘a Median’ silverwork.” It is now known
that the decoration (Achaemenian style) on the bowl is modern: Muscarella 1977a, no. 154.
excavated and unexcavated achaemenian art 1061
41, T—with the wrong acquisition number). The facts about the acquisi-
tion are as follows, and I thank J.E. Curtis for sending me the following
information: “The axehead was presented to the British Museum in 1880
by Sir A.W. Franks; it is said to have been found at Hamadan by a certain
C.P. Clarke.” The juxtaposition of these two statements, one published, the
other from the Museum files, is instructive: an unverified claim from one
party to a second is transmogrified into an historical fact when published
by a third; “Hamadan” becomes Hamadan.
A stone Achaemenian head, probably from a capital (cf. Schmidt 1953,
fig. 55), was published as acquired by the Kansas City Museum “from the
ancient Median and Achaemenid site of Ecbatana,” (ILN, April 20, 1957,
642; Culican 1965, pl. 76). Vanden Berghe (1959, 110, pl. 137d) accepted this
claim—and the implication that the head came directly from Hamadan to
Kansas City: “Tout récemment, on y [Hamadan] a exhumé un fragment de
sculpture acheménide.” To the casual reader the phrases “acquired … from
Hamadan” and “on y a exhumé,” might logically suggest that the tasks of
excavation and shipment were performed by an archaeologist rather than
a plunderer; they certainly imply that firm information was available to the
authors—information they could not have had, given that the piece was
sold by a dealer.22
Vanden Berghe further claimed that two bronze heads once in the Brum-
mer collection were found before the last war “aux environs de Hamadan”
(vanden Berghe 1959, 110, pl. 137a, b). The larger, and finer of the two heads
(fig. 15) was first published in the ILN, January 10, 1931, frontispiece, as “an
Achaemenian King” that was “found near Hamadan,” and is the source of
vanden Berghe’s attribution.23 Both heads were subsequently published by
22 Vanden Berghe’s volume could have been the most valuable work on Iranian archaeol-
ogy to date if only he had limited himself to excavated material. The unexcavated objects and
the hearsay evidence associated with them should have been placed in a separate section.
23 It will be obvious to anyone who examines the initial publication of objects in the
ILN that the magazine functioned as a laundry for the antiquities market. Inasmuch as the
magazinc has a well-deserved reputation as a publisher of material from recent excavations,
most scholars have assumed that everything published there has a legitimate pedigree, the
ILN seal of approval. ILN titles often suggest that actual excavation reports are to be discussed
when in fact plundered objects are being offered for sale: “Recently Found Masterpieces”
(December 17, 1958); “Recently Found Treasures” (July 17, 1948); “Recent Discoveries” (August
21, 1948); “Recently Discovered Gold, Silver …” (April 2, 1960); “Newly Found Masterworks”
(May 23, 1959); “New Discoveries in Iran …” (May 31, 1941), and so forth. And in some instances
the objects were apparently not recently found, but recently manufactured (see Muscarella
1977a, 160, n. 31, 163).
1062 chapter forty
Casson (1938a, 355f., pls. 105, 106, 107a, B) but with apparently two sepa-
rate proveniences, neither of them Hamadan. For the smaller head Casson
claimed in the text that “It comes from Salmas near Lake Van” but gave no
provenience for the larger. Puzzling, however, is the statement offered in the
previous footnote 3: “This [the smaller head] and the other head are said to
have been found together, in Ādharbāyjān.” These contradictory statements
may be understood, I suggest, by assuming that either Ackerman or Pope
added the footnote in his or her capacity as editor, for when Pope (1945,
17) published the heads he stated matter-of-factly that they were found “in
the vicinity of Lake Urmiya,” ignoring the earlier Hamadan and Lake Van
proveniences. Whatever the reasons for the two proveniences in his arti-
cle, Casson accepted both heads as products of the Achaemenian period
(not pre-Achaemenian as stated by Porada 1965, 233, n. 34), perhaps inad-
vertently echoing the original Hamadan reference.
The larger head was eventually purchased by the Metropolitan Museum
(47.100.80), where, following Pope, its provenience was given out confi-
dently as “found in northwest Persia” (Wilkinson 1949, 192 f.).24 This latter
attribution is the one presently accepted (Porada 1965, 62, 233, n. 34, fig. 38;
Calmeyer 1972, 65 also challenged the Hamadan provenience), as is the ear-
lier date suggested by Wilkinson. In fact the only certain provenience we
have for this mute masterpiece is that at one time it was in the Brummer
collection: nothing more or less.
24 Its shifting provenience does not end here. One unpublished source gave the prove-
nience of the Metropolitan head as “Tepe Tikhon,” another as “Chouchichi, near Lake Van.”
excavated and unexcavated achaemenian art 1063
Cairo have always attracted foreign merchants, especially given the number
of available westerners who seek antiquities. Therefore, when an Egyptian
museum catalogue cites an object with a provenience, but no documenta-
tion, or one with no provenience, it is difficult at this distance to establish
whether the object was excavated or purchased, or confiscated.
With these thoughts in mind, two bracelets in Egypt, and not mentioned
in Appendix B, should be considered as additional examples of Achaeme-
nian art in Egypt (unless local investigation proves otherwise). They were
published by Vernier (1927, 63, 188, pls. XVII, 52.148, XXI, 52.587; Amandry
1958, 14, n. 39, 20, n. 82) as deriving from Edfu and Mendes.25
Further, I believe it was an error on my part not to have cited as examples
of Achaemenian art from Egypt representations of jewelry on Egyptian
statuary, merely because their find spots are unknown. That the statues
were made in Egypt is beyond doubt. The first and best known example is
the torque on the statue of Ptah-hotep in the Brooklyn Museum (fig. 16),
(Cooney 1953a, figs. 1, 2, 5; Bothmer 1960, pl. 60, fig. 151; Amandry 1958a,
16, n. 55). This torque is a classic Achaemenian work and is paralleled by
a superb example in the Guennol collection (Cooney 1953a figs. 6, 7) and
by bracelets from the Oxus Treasure (Dalton 1964, nos. 136, 137). The second
example, less preserved, is a bracelet with animal head terminals on a statue
in the Vatican Museum (Botti and Romanelli 1951, 32 ff., no. 40, pls. XXVII,
XXVIII; Amandry 1958, 16, n. 55).26
But there are certain Achaemenian objects in western collections that
have been claimed to derive from Egypt and which were consciously omit-
ted from Appendix B. These include a limestone plaque in the Oriental
Institute depicting animal friezes (Frankfort 1950, 111 f., pl. III; Kantor 1957,
21, “found in Egypt;” Amandry 1958a, 16, n. 55; 1959, 42, n. 39); a limestone
plaque depicting a death scene, in the Hague (v. Bissing 1930, 226 ff., fig. 1a,
b; Svoboda 1956, 64); two stone lions said to have been found at Leontopolis
(fig. 17) (Cooney 1953b, 17ff., figs. 1–4; Amandry 1958, 16, n. 55), and a stone
25 The Toukh-el Garmous silver rhyton is obviously of Greek, Hellenistic, style and there-
fore cannot be included as an Achaemenian find in Egypt: Luschey 1938a, 770ff.; 1939, 74,
n. 422, 111, n. 632; Adriani 1939, 352, 359; Svoboda 1956, 64. Amandry 1958a, 16, n. 55; 1959, 51,
n. 101 refers to it as “de type acheménide,” which is true only in the sense that the rhyton is
based on Achaemenian types.
26 Both Botti (1956, 147 ff. and Amandry 1958a, 16, n. 55) believe that an Egyptian bust
amulet (Cooney 1953b, fig. 7; Hoffmann 1958, fig. 6), in Brooklyn; a faience
rhyton fragment also in Brooklyn (Hoffmann 1958, fig. 3); another in the Lou-
vre (Roes, 1952, fig. 4; Hoffmann 1958, fig. 3); and finally a limestone lion head,
also in the Louvre (Roes 1952, fig. 7).27
What is the evidence that these objects came from Egypt? The Oriental
Institute plaque was said by its former owner to have been acquired in Egypt;
the Hague plaque “nach unverdächtigen Zeugnis des Händlers Cassira um
1909 in Mithrahine, der Statte der alten Memphis, zu tage gekommen;” the
Louvre rhyton was once in the Clot Bey collection; as for the lion head, I
have not been able to get any data; unless there is solid information to the
contrary, it should be treated as an acquired, not an excavated, object. The
Brooklyn amulet, although listed by Cooney (1953b, fig. 7) in the photograph
caption as “Egyptian,” is presented in footnote 14 as “Provenience unknown;”
it was purchased from a dealer in New York City. The faience rhyton in
the same museum was originally given to the Peabody Museum in Salem,
Massachusetts, sometime after 1798 and subsequently, in 1948, was given as
a gift to the Brooklyn Museum (Hoffmann 1958, 10); in other words, it has
no known provenience. We are left with the lions in Brooklyn. According
to Cooney they were bought by Wilbour from a Dr. Fouquet of Cairo, who
had bought them in 1885, a year after they were said to have been found
by peasants and sold to a dealer. Collectively, the information we have
that the lions derived from Egypt is circumstantial. Nevertheless, certain
points are worth noting as they may help in determining provenience. The
lions were acquired in Egypt; they are large enough almost to preclude
the need to transport them from, say, Syria or points further east; and
they were said to have come from a sanctuary of the god Mahes, where it
would be appropriate to have representations of lions. It is therefore highly
probable that Brooklyn’s Leontopolis lions indeed came from that site and
they deserve at least a parenthetical listing among the Achaemenian objects
from Egypt.
At the same time, I still believe it is justified to have omitted the many
fine silver vessels said to have come from Pithom/Tell el Maskhuta (fig. 18);
I do not find it possible to connect firmly the silver with the alleged find
in 1947 (Cooney 1956, 43ff.). Here an Egyptologist might eventually be able
to clarify the nature of these objects and produce information that might
justify an Egyptian provenience.28
27 I would like to acknowledge the help of Ch. Desroches-Noblecourt and Robert Bianchi
Shrine in Egypt,” JNES XV, I (1956), 1–9, accepts the objects in Brooklyn as definitely from
Tell el Maskhuta and reaches historical conclusions concerning “the earliest known records
of Arab life in Egypt” (p. 9). W.J. Dumbrell, “The Tell el-Maskhuta Bowls and the ‘Kingdom’
of Qedar in the Persian Period.” BASOR 203 (October 1971), 33–44 comes to the same conclu-
sions. I wish to thank Richard Fazzini for these references.
29 The helmet and horsebit were not included in Appendix B because I was narrowly
provenience of the silver statuette from R.-B. Wartke of the Staatliche Museen zu Berlin,
DDR, whose cooperation I gratefully acknowledge. Wartke informs me that “Die persische
Silberfigur, VA 4852 ist bei uns ohne Herkunftsangabe verzeichnet (Ankauf): ‘nach Angabe
von Herrn Prof. Zahns aus Soloi/Cilicien.’ ” The italics are in the original; whatever the origin,
a western origin for the piece is excluded. Casson (or the editor) inadvertently wrote Sicily
rather than Cilicia.
Wartke also supplied me with the following information requested concerning a gold
bracelet “aus Sardes” published first by Luschey 1938a, fig. 5, and also by Otto 1944, notes 2
and 3; Pudelko 1933–1934, 8B; and Amandry 1958a, 14, n. 46: “Das von Ihnen zitierte Stück
aus Sardes gehört zu einem Paar Goldarmreifen und befindet sich jetzt in Antiken-Museum
1066 chapter forty
lion heads, which “nach Angabe eines Vorbesitzers aus Thessalien stammt,”
probably was actually made in Asia Minor, although it reached Thessaly
in antiquity. The heads look very much like pussy cats rather than lions
and to my eyes are not Achaemenian. Nevertheless, the use of unexcavated
objects to discuss ancient trade is unjustified on its own merits. I have
elsewhere discussed the gold bracelet in Karlsruhe as a purchased object
without provenience, although many have claimed it as an object excavated
at Corinth (Muscarella 1977a, 195).
Conclusions
14, reports on an Old Persian Inscription on a clay tablet said to have been found in a garden
in 1937 in Gherla, Roumania. On pp. 13 f. Harmatta discusses—but rejects—the possibility
that the tablet is a forgery (see below, Section III, no. 5); a thermoluminescence test of the
tablet would help resolve the issue. See also R. Frye, The Heritage of Persia (New York, 1963),
117, n. 96, who seems to accept the authenticity of the inscription.
excavated and unexcavated achaemenian art 1067
If the aims of this paper have been achieved, it has been demonstrated how
facile it has been for uninvestigated and undocumented statements to rise
above the level justified by the evidence, thereby diverting scholars from
1068 chapter forty
Abbreviations
AA Archaeologische Anzeiger.
AfO Archiv für Orientforschung. Graz.
AMI Archaeologische Mitteilungen aus Iran.
BASOR Bulletin of the American Schools of Oriental Research.
BICS Bulletin of the Institute of Classical Studies of the University of London.
BMQ British Museum Quarterly.
DLZ Deutsche Literaturzeitung.
ESA Eurasia Septentrionalis Antiqua.
ILN Illustrated London News.
JANES Journal of the Ancient Near East Society of Columbia University.
JAOS Journal of the American Oriental Society.
JFA Journal of Field Archaeology.
JNES Journal of Near Eastern Studies.
MMA The Metropolitan Museum of Art, New York.
RLA Reallexikon der Assyriologie.
ZFA Zeitschrift für Assyriologie.
ZDMG Zeitschrift der Deutschen Morgenländischen Gesellschaft. Leipzig.
excavated and unexcavated achaemenian art 1069
Bibliography
Adriani, A.
1939 “Rhyta,” Bull. Sociéte Royale d’Archéologie Alexandrie X, 350–362.
Amandry, Pierre
1958a “Orfèvrerie achéménide,” Antike Kunst 1, 1, 9–23.
1958b “Royal Persian Gold …,” ILN, December 17, 1139–1141.
1959 “Toreutique achéménide,” Antike Kunst 2, 2, 38–56.
1963 Collection Hélène Stathatos III (Strasbourg), 260–272.
1965 “Orfévres grecs à la Cour du Grand Roi?” 8e Congrès Inter. d’Archéologie
Classique (Paris), 581–587.
Barnett, R.D.
1960 “Ancient Oriental Goldwork,” BMQ XXII, 29–30.
1968 “The Art of the Bactrians and the Treasure of the Oxus,” Iran. Antiqua VIII,
34–53.
v. Bissing, F.W.
1930 “Totenstele eines persischen Grossen aus Memphis,” ZMDG 84, 226–238.
v. Bothmer, Bernard
1960 Egyptian Sculpture of the Late Period (Brooklyn).
Botti, G.
1956 “Buste di un Dignitarie della XXVII Dynastia nel Museo Egizio de Firenze,”
Boll. d’Arte 2, 147–149.
Calmeyer, Peter
1969 Datierbare Bronzen aus Luristan und Kirmanshah (Berlin).
1972 “Hamadan,” RLA (Berlin), 64–67.
1974 “Zu einiger Vernachlässigten Aspeckten Medischer Kunst,” Proc. of the IInd
Annual Symposium on Archaeological Research in Iran (Teheran), 112–127.
Casson, Stanley
1938a “Achaemenid Sculpture,” Survey of Persian Art I, 346–366.
1938b “Achaemenid Metalwork,” Survey of Persian Art I, 367–376.
Musée Cernuschi
1948 Iran, Piéces du Musée de Teheran (Paris).
1070 chapter forty
Cooney, John
1953a “The Portrait of an Egyptian Collaborator,” Bull. Brooklyn Museum XV, 2, 1–
16.
1953b “The Lions of Leontopolis,” Bull. Brooklyn Museum XV, 2, 1, 17–30.
1956 Five Years of Collecting Egyptian Art (New York).
Culican, William
1965 The Medes and Persians (New York).
Dalton, O.M.
1964 The Treasure of the Oxus (London).
Filow, B.D.
1934 Die Grabhügelnekropole bei Duvanli in Sudbulgarien (Sofia).
Frankfort, Henri
1950 “A Persian Goldsmith’s Trialpiece,” JNES 9, 111–112.
1955 The Art and Architecture of the Ancient Orient (Penguin).
Ghirshman, Roman
1962 “Le Rhyton en Iran,” Artibus Asiae XXV, 57–80.
Goldman, B.
1951 “Achaemenian Chapes,” Ars Orientalis II, 43–54.
Herzfeld, Ernest
1926 “Eine neue Darius-Inschrift aus Hamadan,” DLZ 42, 2105–2108.
1928 “A New Inscription of Darius from Hamadan,” Mem. of The Arch. Survey of
India 34, 1–7.
1930 “Ariyarana König der Könige,” AMI 2, 113–127.
1935 “Eine Silberschüssel Artaxerxes I,” AMI 7, 1–8.
1937 “Die Silberschüssel Artaxerxes I und die goldene Fundementurkunde des
Ariaramnes,” AMI 8, 5–51.
1938 “Bronzener Freibrief eines Königs von Abdadana,” AMI 9, 159–177.
1941 Iran in the Ancient East (Oxford).
1968 The Persian Empire (Wiesbaden).
Hoffmann, Herbert
1958 “Fragment of a Faience Rhyton,” Bull. Brooklyn Museum XIX, 3, 10–12.
Kantor, Helene
1957 “Achaemenid Jewelry in the Oriental Institute,” JNES XVI, 1–23.
excavated and unexcavated achaemenian art 1071
Kent, R.G.
1950 Old Persian (New Haven).
Lefebvre, M.G.
1923 Le Tombeau de Petosiris III (Paris).
Luschey, H.
1938a “Achämenidisch-persische Toreutik,” AA, 760–772.
1938b “Griechisch-persische Metallarbeiten in Antiquarium,” Berliner Museen
LIX, 76–80.
1939 Die Phiale (Bleicherode am Hertz).
1968 “Lowe von Ekbatana,” AMI (n.f.) 1, 115–122.
Mellick, Machteld
1973 “Excavations at Karataş-Semayük and Elmali, Lycia, 1972,” AJA 73, 3, 293–
303.
Moorey, P.R.S.
1971a A Catalogue of the Persian Bronzes in the Ashmolean Museum (Oxford).
1971b “Towards a Chronology of the Luristan Bronzes,” Iran, IX, 113–129.
1974 Ancient Persian Bronzes (London).
De Morgan, J.
1896 Mission scientifique en Perse (Paris).
Mostafavi, M.T.
1953 Historical Monuments of Hamadan (in Persian) (Teheran).
Olmstead, A.T.
1948 History of the Persian Empire (Chicago).
Otto, H.
1944 “Ein achämedischer Goldwidder,” ZfA 14 (48), 9–22.
1072 chapter forty
Paper, H.H.
1952 “An Old Persian Text of Darius II (D2Ha),” JAOS LXXII, 169–170.
Pope, A.U.
1935 “Recent Persian Art,” ILN, March 2, p. 1.
1945 Masterpieces of Persian Art (New York).
1948 “Recently Found Treasure …,” ILN, July 17, 58–59.
Porada, Edith
1965 The Art of Ancient Iran (New York).
Pudelko, G.
1933–1934 “Altpersische Armbänder,” AfO 9, 85–88.
Ridgway, D.
1977 “Manios Faked?” BICS 24, 17–30.
Roes, Anna
1952 “Achaemenid Influence upon Egyptian and Nomadic Art,” Artibus Asiae, XV,
17–30.
Rostovzeff, M.
1922 Iranians and Greeks in South Russia (Oxford).
1931 Skythien und der Bosporus (Berlin).
Schefold, K.
1938 “Der Skythische Tierstil in Südrussland,” ESA XII, 2–78.
Schmidt, E.
1953 Persepolis I (Chicago).
1957 Persepolis II (Chicago).
1970 Persepolis III (Chicago).
Smirnov, Y.I.
1934 Der Schatz von Achalgori (Tiflis).
Stucky, R.
1976 “Achämenidische Ortbänder,” AA, 13–23.
Svoboda, B.
1956 in B. Svoboda, D. Cončev, Neue Denkmaler antiker Toreutik (Prague), 9–89.
Terrace, E.L.B.
1962 The Art of the Ancient Near East in Boston (Boston).
1966 “Ancient Near Eastern Art,” The Pomerance Collection of Ancient Art (Brook-
lyn Museum), 13–57.
excavated and unexcavated achaemenian art 1073
Tuchelt, K.
1962 Tiergefässe in Kopf-und Protomengestalt (Berlin).
Vanden Berghe, L.
1959 Archéologie de l’Iran ancien (Leiden).
Vernier, E.
1907 La Bijouterie et la Joaillerie égyptiennes (Cairo).
1927 Catalogues Général des Antiquités Egyptiennes (Cairo).
Wakeling, T.G.
1912 Forged Egyptian Antiquities (London).
Walser, Gerold
1966 Die Völkerschaften auf den Reliefs von Persepolis (Berlin).
Wilkinson, C.K.
1949 “The Art of the Ancient Near East,” Bull. MMA 7, 7, 186–196.
1955 “Assyrian and Persian Art,” Bull. MMA 13, 8, 213–224.
Zahn, Robert
1930 “Neuerwerbungen im Antiquarium,” Berliner Museen LI, 147–149.
1074 chapter forty