Carved Ivory Furniture Panels from Nimrud:

A Coherent Subgroup of the North Syrian Style

IRENE J. WINTER
of ArtHistory,The Universityof Pennsylvania
Department

IN THE COLLECTION of The Metropolitan Museum Annals from Assurnasirpal II (883-859 B.C.) to Assur-
of Art are a number of carved ivory plaques excavated banipal (668-627 B.C.)and is also frequently depicted
at the Assyrian capital of Nimrud, and dating to the on reliefs, such as that of Sennacherib from Nineveh
early first millennium B.C. The pieces have been ac- (704-68I B.C.), in which the king's soldiers carry off
quired from the British School of Archaeology in Iraq beds, chairs, footstools, and tables from a conquered
in return for the Museum's support of renewed excava- citadel (Figure 5).2
tion at the site from 1949-63, under the direction of Sir The pieces from Room SW7 were originally called
Max E. L. Mallowan. Included in the group are ten bed-heads by Mallowan; later, Mallowan and Geor-
plaques (for example, Figures I, 2, 3) and two complete gina Herrmann convincingly argued that, based upon
panels (for example, Figure 4) that belonged to a group the dimensionsof the complete panels and the distances
of approximately nineteen decorated pieces of furniture between them as found in the room, they may rather
stacked in Room SW7 of Fort Shalmaneser and appar- have been the backs of chairs.3 While it is clear that
ently abandoned when the building was destroyed in these more or less rectangular panels were not part of
612 B.C.' the typical couch with a C-shaped headboard such as
That these pieces were discovered stored in a major the one Assurbanipalreclines upon in his garden scene
building in the Assyriancapital is not surprising.From (Figure 6) or theone representedin a camp scene on a
the ninth through the seventh centuries B.C. there was relief ofTiglath Pileser III (744-727 B.C.),4 we cannot
a steady stream of ivory, both finished products and definitely say that there were no beds in this period with
tusks,arriving in Assyria as booty and tribute from sur- rectangularhead- or foot-boardssimilar to those known
rounding nations. This is documented in the Royal from Egypt and Syria in the Late Bronze Age or as

i. The most characteristic pieces from SW7 were published by 2. For Assurnasirpal II, see Luckenbill, AR I, ?? 459, 475-477,
M. Mallowan in JV&RII, pp. 485-515. The entire collection was 479; Shalmaneser III: AR I, ?? 585, 593, 625; Adad Nirari III:
then presented in Mallowan and G. Herrmann, SW7. The works in AR I, ? 740; Tiglath Pileser III: AR I, ?? 769, 804; Sargon II:
the Metropolitan Museum are: 59.I07.3 (SW7 no. 3, ND7917); AR II, ?? I7, 45, 172; Sennacherib: AR II, ?? 240, 366; Esarhad-
58.31.I (SW7 no. 7, ND6376); 59.107.6 (SW7 no. 8, ND7963); don: AR II, ? 527.
59.107.7 (SW7no. 39, ND7925); 59.107.10 (SW7no. 51, ND79o8, 3. SW7, pp. 3-9.
plaque 3); 59. I7.4,5 (SW7 no. 63, plaques i, 2 ND7951); 58.31.2 4. R. D. Barnett and M. Falkner, The Sculptureof Assurnasirpal
(SW7 no. 67, ND6368); 59.107.15 (SW7 no. 87, ND7579); from the Centraland Southwest
II, Tiglath Pileser III and Esarhaddon
59. 07.8 (SW7 no. 89, ND7930); 59.107. I (SW7 no. 95, ND79Io); Palacesat Nimrud(London, 1962) PI. LX.
59.107.2 (SW7 no. 105, ND79o0).

25

The Metropolitan Museum of Art

is collaborating with JSTOR to digitize, preserve, and extend access to
Metropolitan Museum Journal ®
www.jstor.org
found in an eighth-century tomb at Salamis.5However, STYLISTIC AFFINITIES
the form that the SW7 chairs would have taken does
seem to be more consistentwith the high-backed chairs An examination of the figured plaques makes it clear
seen on reliefs ofTiglath Pileser, Sennacherib, and As- that this collection belongs to the North Syrian group
surbanipal, as well as on reliefs from Carchemish and of ivory carvings, as originally defined by F. Poulsen
Zinjirli in North Syria. In many cases these chairs, de- and R. D. Barnett in referenceto the ivories discovered
picted in profile, clearly include decorative carved pan- by F. Loftusin the South East, or Burnt, Palace at Nim-
els in the lower portions and below the armrests, so rud in I853. Poulsen distinguished the North Syrian
that decorated backs would also be expected.6 from the Phoenician ivories discoveredby A. H. Layard
Apparently the decorated panels of SW7 were origi- in 1848-49 in the North West Palace on the basis of
nally backed on wood, which has not been preserved. details and motifs related to those on reliefs at Car-
The panels average about 85 x 55 cm. in size and chemish, Zinjirli, Marash, and Tell Halaf-for exam-
consist of several plaques mounted together: a center ple, hairstyles, floral elements and trees, female
section of from four to six contiguous plaques, usually sphinxes, and musicians in procession toward a seated
framed at top and bottom by narrow strips. The whole figure-and a general absence of Egyptian elements.7
is then generally bound by two or three vertically ar- Poulsen's distinction was developed further by Bar-
ranged plaques at right and left (Figures 4, 7, 8, 9). nett, who noted not only the absence of Egyptian fea-
The notion ofjoining several plaques to form a panel tures, but also the presence of Hurrian and Hittite ele-
is a logical outcome of the size of the original tusk. The ments inherited from the second millennium. Barnett
practice can be observed earlier, in the Ras Shamra also described the physiognomical features that char-
bed of about 1400 B.C., and continued into later times, acterize the style: oval face, high receding forehead,
where, for example, a throne made for the archbishop large eyes and nose, small pinched mouth, and little or
of Ravenna in the Carolingian period is comprised of a no chin.8
seriesof panels made up of five ivory plaques framed by These features are closely paralleled on first-millen-
decorated horizontal and vertical strips, all set in wood nium objects from North Syria (Figures I , 2).9 Sig-
(Figure io, from the throne base). nificantly, the same conventions occur on the SW7
The subjects of the SW7 plaques are remarkably plaques. The SW7 ivories also include several motifs
consistent: variations on seated or standing figures, that can be found in Barnett'sSyrian group: for exam-
generally grasping the tendrils of a plant. In many ple, the chariot scene (Figures 9, 13) and the human
cases, a small winged sundisk appears at the top. On figures holding a blossom in each hand (Figures 2, I2,
several panels, plaques of symmetrical volute trees I4), in addition to the women's characteristic long
frame the figured plaques at either side; when there are garment, vertically striated, with beaded borders,
vertical frames at the sides, they are generally com- clearly longer at the back than at the front (Figures 3,
prised of two or three superimposed figured plaques, 4, 12). The group is, according to Poulsen'soriginal cri-
as in Figure 9. The upper strip is occasionally decorated terion, quite free of the Egyptian elements that are
with a long winged sundisk, the bottom strip more found in a typical Phoenician-style plaque (Figure 15).
rarely with a narrative scene. Nevertheless, there are minor differences between

5. H. Baker, Furniturein the AncientWorld(London, 1966) figs. 7. F. Poulsen,DerOrient Bildkunst

unddiefriihgriechische (Leipzig,
83-85, 132-135 (XVIII Dynasty); C. F. A. Schaeffer, "Les fouilles 1912) pp. 38-53.
de Ras-Shamra/Ugarit, 15e, I6e, 17e campagnes," Syria XXXI 8. R. D. Barnett,CNI,pp. 40-44.
(1954) pls. viii-x; V. Karageorghis, Excavationsin theNecropolisof 9. Compare the reliefs from Tell Halaf, Carchemish,and
SalamisIII (Nicosia, 1974) pl. LXVII. Zinjirli in Orthmann, USK, pls. 8-i I, 15-I9, 56-59, and the
6. See B. Hrouda, Die Kulturgeschichte des assyrischen
Flachbildes Zinjirlisilverplaquesin von Luschan,AiS V, p s. 46k, 47d. The
(Bonn, 1965) pls. 14, 15. The suggestion that the SW7 panels were relationshipbetweenthese plaquesfrom Zinjirliand the Syrian-
parts of chairs rather than beds is further supported by the slightly style ivorieswas noted by O. W. Muscarella,"Hasanlu, I964,"
concave curve of the panels across their width (noted in SW7, p. 3), TheMetropolitan Museum of ArtBulletin25 (1966)figs.5, 6.
more appropriate for a backrest than a bedboard.

26
FIGURE I FIGURE 2 FIGURE 3

Ivory plaque. Room SW7, Fort Ivory plaque. Room SW7, Fort Ivory plaque. Room SW7, Fort
Shalmaneser, Nimrud. The Shalmaneser. The Metropoli- Shalmaneser. The Metropoli-
Metropolitan Museum of Art, tan Museum of Art, Rogers tan Museum of Art, Rogers
Rogers Fund, 59.107.4 Fund, 59.107.7 Fund, 58.31.2

27
 /~ -i p

**-"*'(
gF "YE
r
'.
.~~~~-y i,
.at" ?

? ^-^ ^ .; * , -e -.

__ _____
___
 FIGURE 4 plant forms, particularly trees, tend to be extremely
Ivory panel. Room SW7, Fort Shalmaneser. The curvilinear: human limbs are often awkwardly articu-
Metropolitan Museum of Art, Rogers Fund, lated; and the figures are generally larger in scale,
59. 07.3 crowding the plaques (Figures 4, 7, 8). All of these
qualities are consistently exhibited within the SW7
group, and suggest that the SW7 ivories forma coherent
the SW7 and the Loftus ivories that cannot be explained subgroup of the North Syrian style.
merely on the basis of different intended use, the Loftus It is therefore significant that of all the similarities
collection consisting of small objects (mirror handles, that can be cited between the ivories of SW7 and the
pyxides) and the SW7 panels belonging to furniture. fixed monuments of North Syria, the parallels cluster
These differences include variations in male and female around two sites within twenty kilometers of each
hairstyles and in men's garments, all of which, although other: Zinjirli and Sakce Gozu.o1
paralleled in Assyrian and North Syrian reliefs, are not Closely comparable are the plaques representing
to be found among the Loftus ivories. In addition, SW7 seated women (Figures 8, I6) and a stela carved with
the same subject of the time of Bar Rakib (about 740-
FIGURE 5 725 B.C.), found outside Hilani I at Zinjirli (Figure
Drawing of a relief: Palace of Sennacherib, Nine- 17).1 The stela bears a winged sundisk in the field, as
veh (original lost)
Io. The complete range of parallels in theme and details be-
6 tween the SW7 ivories and material from other North Syrian and
FIGURE
Assyrian sites is presented in SW7, pp. 19-35, 39-6I. I shall dis-
Relief. Palace of Assurbanipal, Nineveh. The cuss only those parallels with works from Zinjirli and Sakce Gozii.
British Museum, no. I24920 (courtesy the Trus- ii. See also SW'7,p. 33 and nos. 46:2, 3; 47-50.
tees of the British Museum)

- --:.,.. ,
Nff :. :.--

• " `f- " ''': - " -, j

?? . *

'~ ' ~ ~'. ~

:~,,

- ;;,u,: '~"~?-..:
";;" ".'
,,5 ' . r?

'.~
~ ~~~~ -'-. ....
?
~ ~
1*~
"
- ~~~
...
..~~~~~~~~~~~~~~
~~~~~~~~~
?"r ,.. -pZr' 1
, 2' .

'* '
F3 :~~~~~~~~.4" 't?-? "C:-~?' ..;_2'*; .
? ~ ~ ~ .'-- ,
,
~?-~i:S~t~?
"":'
"."::' '
':~('=r'i, '~" ?'~
."Y~'"':4? ;~I?.
'~..~ ,fflo-ol
'-'........-
....---il L;.':' .' "~----
~ -~ ~-"~:
....~ ......
'
:-. ,.It lc ~?'-: ~:,- ~~~~~~~~~~~~~~~~~
_..... "-"'" ,'.
??: 1.' -,~ . ,.......~;..
i, ~ .....
I;ci~~~~~:..,,,
~~I~~~6~~~C~~~
~.I ~ ~t
~~~~ ..:..
~~~~~~~~~~~-..%
.~~~~f ~ -a r. ....
"':~~

!trr ? ,'"i u ? Le .; . . .?,

-. r_; - .,,,,.,
~. . . -...
......,,.....~ ,,.. ,,~-..",&.~"..,
~,? ,..~

29
 FIGURE 7
Ivory panel. Room SW7, Fort
Shalmaneser. Iraq Museum,
Baghdad, 61898 (photo: Wm.
Collins Sons and Co., London)

" . -:
???; iT1 ?r r
!- dl
r?l -.-?-
:* ;?- -?? ''cSk';; )?-lb-C?-L*C
-i: ?*'*
c. t-.
i ?. P. v i-?=
?5. i.C u? :??7

FIGURE 8

Ivory panel. Room SW7, Fort Shalmaneser. Iraq Museum, 62721 (photo: Wm. Collins Sons and Co.)
FIGURE 9
Ivory panel. Room SW7, Fort
Shalmaneser. Iraq Museum, 62722
(photo: Wm. Collins Sons and Co.)

FIGURE 10

Ivory panel. Throne of the Arch-

bishop Maximian. Cathedral of
Ravenna
 ?_nk _i

' ~ ' a number of plaques from SW7. The SW7 women

. '"?~?do
7 BfkL
C ^?B -all sit on square stools, feet resting on footstools. On
some plaques they grasp the tendrils of plants before
i| , -
~ ^ 4 '? j2"
? ;A them with one hand, holding an object in the other;
generally, however, they reach toward a small table set
^e '
*fi.tlBxatg Hl !~ min the branches of a flowering tree. The tables are cross-
i .'
4 !'
,
3
i legged, terminating in bull's feet; they are invariably
.^:t,
' '.ifi ^
"jflM^ ': Ipiled't] high with dishesand layers of bread. The arrange-
ment of food on the tables, even to the inclusion of a
??;i XK ,,_. . v low, footed bowl, is also seen on the table before the
?
:*1..,Ti~ : seated woman of the Zinjirli stela; the tables' shape,
F
; .,
? - crossed legs, and vertical support piece are likewise
J~i l1' ^i
*
IH1 [-.^ r similar. The fleecy tasseled cloth that covers the Zin-
; chair and the shaped footstool are also depicted
~~~jirli
:^g^p^^^^ ^4
on the ivories, and the single lotus blossom with short

FIGURE II
Silver plaque. Zinjirli. Staatliche Museen, Berlin - - .
(photo: courtesy Vorderasiatische Abteilung, .i
Staatliche Museen) ,i ^

FIGURE I22

Ivory pyxis fragment. Burnt Palace, Nimrud. .

The British Museum, no. 131141 (courtesy The
Trustees of the British Museum)

FIGURE I3 . ?

Ivory pyxis. Burnt Palace, Nimrud. The British^ t

Museum, no. 18173 (courtesy The Trustees of . _ 1
the British Museum)

? *. . y curvilinear stem, held by the Zinjirli woman in her

. .1.^
. Mh~
t. .
left hand, is likewise paralleled on one of the plaques.'2
A scene similar to that of the ivories and the Zinjirli
stela is repeated on a silver plaque from Zinjirli (Figure
I I). Another closely related representationis found on
a badly weathered relief found by J. Garstang near the
base of the mound of SakqeGozuii,where a similar table
in a banquet scene is outfitted with food and a low,
footed bowl.'3 Unfortunately, it is not possible to deter-
mine what sort of garment the seated figure is wear-

.r 12. For the trellis-fringe tassels, see SW7, no. 46:2 (my Figure
~
')? ....'" -8) and p. 33; for the footstool, SW7, no. 47 (my Figure i6); for
the lotus blossom, SW7, no. 48.
13. J. Garstang, "Excavations at Sakge Geuzi in North Syria:
Preliminary Report for 19o8," LAAA I (I90o8)pl. xxxv: I.

32
FIGURE 14 FIGURE 15
Ivory plaque. Room SW7, Fort Shalmaneser. Ivory plaque. Room SWI2, Fort Shalmaneser.
Iraq Museum, 62705 (photo: Wm. Collins Sons Iraq Museum, 65508 (photo: Wm. Collins Sons
and Co.) and Co.)

33
 FIGURE I6 FIGURE 17
Ivory plaque. Room SW7, Fort Shalmaneser. Stela. Zinjirli, about 735-720 B.C. Staatliche
Iraq Museum, 60553 (photo: Wm. Collins Sons Museen, VA2995 (photo: courtesy Vorderasia-
and Co.) tische Abteilung, Staatliche Museen)

ing, although the chair seems identical in shape to that reliefs and sculpture from Zinjirli and Sakce Gzii of
of the stela, and to the sort of chair the SW7 panels the period of Bar Rakib-for example, on the female
probably once adorned. The figure, seated to the right, sphinx orthostat from Hilani II at Zinjirli and on the
reaches out with the right arm toward the footed bowl, male sphinx relief from Sakce Gozi (Figure i8).14
as does one of the women on the SW7 plaques (Fig- Seated male figures are rare in the SW7 assemblage.
ure I6). When they do occur, they are without the small table
All of the women on the SW7 ivories have the same and generally reach into the branches of a tree. How-
hairstyle: three or four long corkscrew curls down the ever, on two plaques, the cushioned chair has a high,
back of the head behind the ear, with one long curl fall-
ing in front of the ear, and short curls on the brow. It is 14. USK, pls. 64b, 5 1b. Note especially the identical single curl
a style of men's and women's coiffure that is typical of in front of the ear and the four behind in 51 b.

34
FIGURE 18
Relief. Palace portico, Sakce
Gozu. Second half of the eighth
century B.C. Archaeological
Museum, Ankara, 1811

FIGURE 19
Reliefs. Palace facade, Sakce
Gozii. Second half of the eighth
century B.C. Archaeological ~i..r - e r- A
;~,C~~u=r --.-
Museum, Ankara I807,
i8io

35
 stool. As I point out later, the use of bulls in this context
may be a reference to the identity of the seated figure.
This is the only example of a seated figure holding a
raised bowl or cup, as on the Zinjirli stela. It can be
compared as well with a fragment of a relief of Bar
Rakib himself, also from Zinjirli, on which the king
also holds an open bowl in his outstretched hand.17
The seated figure of the ivory panel is flanked by two
plaques on each side depicting winged griffin-demons
holding lustral cones and buckets, and "heros" with
small animals over their shoulders, all facing the cen-
ter. The griffins' open beaks with extended tongues,
hair and topcurls, and downward curving short wings
are most closely paralleled by the griffin-demons on the
palace orthostats at Sakce Goziu (Figure I9).18 The
tradition of carrying an animal over the shoulder oc-
curs on ninth-century reliefs at Carchemish, in the pro-
cession of male figures on the King's Gate, and on one
of the orthostats of the Citadel Gate at Zinjirli.'9 An
even closer parallel is to be found on the orthostat from
Bar Rakib's Nordhallenbau at Zinjirli, where despite
differences in the proportions of the figures, a virtually
identical gazelle with curved horn is carried on the

15. SW7, no. 51:1, 51:4.

16. SW7, no. 52.
17. USK, pl. 67d.
18. Similarities in stance, position of the upper hand holding
the cone, and proportion of the figures outweigh differences in
dress or the fact that the Sakce Gzii griffin-demons have four
wings. It is significant that in both cases the wings are shown both
raised and lowered. There are examples within the SW7 group of
four-winged creatures, for example, SW7, nos. 3, 67, and 68, which
depicts a male genius wearing a short kilt. Similar griffin-genii are
also depicted on a pair of a-jour plaques from the temple of Haldi
at Altintepe and from Toprak Kale, in Urartu (see T. Ozgiiu,
FIGURE 20 AltintepeII [Ankara, 1969] pl. xxxII and CNI, W.I 3). They share
with the Sakce Gzii reliefs and the SW7 ivories the same hair and
Relief. Nordhallenbau, Zinjirli. About 735-725 forehead curls, distinctively curving wings, and open beak; the
B.C. Staatliche Museen, VA3007 (photo: cour- Toprak Kale griffins wear a tunic beneath a long belted skirt iden-
tical to that worn in SW7, no. 25. Considering the closeness of the
tesy Vorderasiatische Abteilung, Staatliche Mu-
Urartean works to those from SW7 and Sakce Gzii, and the close
seen) political ties between North Syria and Urartu through the first
half of the eighth century B.C., I would suggest that the stimulus
angled back like that of the Zinjirli and Sakce Gozii for the Urartean ivories, or possibly the ivories themselves, came
from North Syria. Evidence for this from various contemporary
examples,I5 and on a third plaque we see again the sources is gathered in my unpublished doctoral dissertation,
same footstool. 6 "North Syria in the Early First Millennium B.C., with Special
On the central plaque of a complete panel (Figure 7), Reference to Ivory Carving" (Columbia University, I973) pp.
a male figure is seated on a cross-legged stool that ends I25-I30.
19. C. L. Woolley and T. E. Lawrence, CarhemishII (London,
in bull's feet, like the table in Figure I6. The male figure
1921) pls. B.22b-B.24; AiS III, pl. xxxvII: upper right; USK, pls.
rests his feet on the backs of two bulls that form his foot- 3oe-h, 57b.

36
 FIGURE 21
Relief. Enclosure gate, Sakce Gozii. About mid-eighth century B.C. Staatliche Museen, VA97I (photo:
courtesy Vorderasiatische Abteilung, Staatliche Museen)

shoulders of one of the king's attendants (Figure 20.)20 on the Assyrian, worn by the second lion-slayer of the
The hero's distinctive hairstyle-three rows of large Sakce Gozii hunt relief. However, G. Herrmann has
spiral curls-can best be compared to that of one of the convincingly argued that the animal bearer's over-
lion-slayers on the hunt relief of the Sakce Gozii palace garment is rather made of a woolly, looped fabric.23
enclosure (Figure 2I).21 The overgarment worn by the Indeed, the general form of every garment repre-
animal bearer from SW7 superficially resembles both sented on the Sakce Gozii hunt relief is duplicated on
the long coat of Assyrian armor that is represented as the SW7 plaques: the belted, open skirt of the spearman
horizontal rows of rounded lappets to indicate metal comparable to those on a number of plaques that show
scales,22 and the cut-away garment obviously modeled men with buckets grasping branches of the tree;24 the

20. AiS IV, pl. 63; USK, pl. 64c. It is not significant that the 22. Seen on reliefs: Barnett and Falkner, Sculpture,pls. LXVII,
gazelle's head on the Zinjirli relief is not turned while that of the cxxII; actual pieces of iron scale armor were uncovered by Mallo-
SW7 plaque is. On the reliefs of the King's Gate at Carchemish, wan in Fort Shalmaneser (N&R II, p. 490, fig. 336). Such scale
the heads of animals carried by male figures in procession all vary armor was also used for horses (W. Lambert, "Sultantepe Tablets
in position. The turn of the head is therefore not necessarily a VIII: Shalmaneser in Ararat," Anat. Stud. I I [1961] pp. 150-I5);
criterion of style or date. however, as both wool and armor are appropriate materials for
21. A similar hairstyle, although with tighter curls and some- horse coverings, it is not possible to conclusively identify this looped
times long hair at the nape of the neck, is worn by the children of or lappet pattern on the SW7 ivories.
Araras on reliefs of the King's Buttress at Carchemish (D. G. 23. SW7, pp. 23-24. This seems especially clear in the similar
Hogarth, CarchemishI (London, I914] pi. B.7; USK, pl. 3if). In garment worn by men in the flowerpot helmet series of ivories, in
all cases, the individuals who wear their hair in this manner are which the lappets and loops are very much like the border of the
beardless, and at Carchemish they are definitely children. This garment (Figures 2, 14, 24). These loops can also be observed at
may therefore be a convention for representing youth-in this case, the border of the cloak of Bar Rakib, which was surely of fabric,
the youthful hunter/hero. I believe it is a phenomenon distinct not metal, as depicted on a relief from Zinjirli (AiS IV, pl. LXVII;
from the "Gilgamesh" hero often shown on Assyrian reliefs-for USK, pi. 66c). G. Herrmann in SW7 does reserve the possibility
example at Khorsabad-where the figure is represented frontally, that one of the representations (no. 21) may actually indicate
with hair arranged in large spiral curls to the shoulders, and whose armor, however, as the lappets are so clearly visible.
antecedents go back to Akka.dian cylinder seals (T. A. W. Madh- 24. SW7, nos. 23, 24.
loom, The Chronology of Neo-AssyrianArt [London, 1970] p. 86).

37
 -00. ---m

FIGURE 22
Hunt scene, detail of Figure 9 (photo: Wm. Collins Sons and Co.)

short kilt with a diagonal flap from which tassels fall bull, collapsing on one knee, is as large as the entire
between the legs, which is also worn by four male fig- chariot complex before him. The other two bulls are
ures on a complete panel ;25and the belted wraparound also enormous and press against the upper and lower
coats of the men in the chariot worn also by two of the borders of the plaque.
figures in the chariot of the SW7 hunt scene (Figure 22). The similarity between this hunt plaque and the hunt
This hunt scene occurs on a horizontal plaque that relief from Sak;e Gzii has been pointed out by a num-
makes up the base of one of the complete panels (Fig- ber of scholars.27 The winged sundisk above the chariot
ure 9), and shows four men in a chariot chasing two of the relief is identical to the sundisks in the upper
bulls before them and attacking a bull to the rear of the plaques of the hunt panel: in both the disk has a beaded
chariot. The prancing horses wear blankets and border, pendant volutes on either side, and a fan of tail
medallionlike tasseled ornaments at the shoulder. The feathers between. The volute curls spring directly from
chariot is hitched by a yoke and an elliptical draft-pole the curving pinions of the wings, while two registers of
decorated with panels of rosettes. The six-spoked wheel wing feathers extend to either side.28 The horses of both
of the chariot is set toward the rear of the chariot box. the ivory plaque and the relief are in the same posture
The driver and a bowman about to shoot wear the and wear the same paraphernalia; the chariots are sim-
wraparound woolly coat; they and a third, partially ilarly shaped and appointed, with the same draft-pole;
hidden, figure face the two bulls ahead. Details of the
archer's gear are carefully delineated. He wears a 25. SW7, no. 65.
26. Also noted in SW7, p. 69.
leather wrist guard on the left hand and a finger guard 27. For example, N&R II, p. 490; D. Ussishkin, "On the Date
on the right hand, which pulls back the bowstring. The of a Group of Ivories from Nimrud," BASOR 203 (1971) pp. 22-27.
28. This is in distinction to the winged disk as represented in
only other representation I know of on which these fea-
Assyrian glyptic (E. Porada, Corpusof Near EasternSeals in North
tures have been so scrupulously recorded is one of the AmericanCollectionsI: The Collectionof the PierpontMorgan Library
reliefs of Bar Rakib from the Nordhallenbau at Zinjirli [Washington, D.C., 1948] nos. 640-646, 648-650), in which the
(Figure 23). There, the walking bowman has his bow Hathor curls are replaced by a short curlicue that appears just
above the sun while the tail feathers sometimes turn at either end
slung over the left shoulder, while he carries two arrows into volutes with tendrils extending down from the disk. It is also
in his right hand and the finger and wrist guards in his distinct from the winged disk on Phoenician-style works (CJVI,
left.26 S.146), where the disk is generally flanked by uraei. Thus the
A fourth figure in the chariot, wearing a long gar- "Hathor" curls and clearly defined pinions serve to mark the North
Syrian style in the early first millennium B.C. (see I. Winter,
ment with a fringed shawl over one shoulder, leans out "Phoenician and North Syrian Ivory Carving in Historical Con-
over the back of the chariot to spear a bull behind. The text: Questions of Style and Distribution," Iraq38 [1976] pp. 4-6).

38
and the four rosettes on the pole of the plaque may be The other plaques that accompany the SW7 hunt
comparable to the four rosettes in the field of the relief. scene in Figure 9 include a winged female genius in the
The only major difference is that the chariot wheel has center and eight male figures in short tunics with open
eight spokes on the relief, and only six on the ivory. coats who grasp the curving branches of a flowering
On the relief, the hunted lion is attacked by two fig- tree. This last theme is by far the most common of the
ures on foot. However, the diagonal thrust of the spear representationsin the SW7 group. On the vertical side
held by both hands of the hunter on the right of the plaques, the men stand on several rows of a scale pat-
relief is very similar to that of the spear of the hunter on tern-a common convention for rendering earth or
the ivory. The most extraordinarysimilarity in the two ground since Early Dynastic and Akkadian times in
representationsis in the disproportionatelylarge size of Mesopotamia.29The particular pose of reaching to-
the hunted animals. ward the plants with the arm extended to eye level or
above finds its closest parallel on a bone tube carved in
FIGURE 23 Syrian style from the Loftus collection.30It may also be
Relief. Nordhallenbau, Zinjirli. About 735- compared to the orthostat from the palace facade at
725 B.C. Staatliche Museen, VA 3000 SakCeGozu, where two men, shown with both a winged
disk and a plant, reach upward to grasp the bud issuing
from a curving tendril (Figure I9).
The standard tree in the SW7 collection consistsof a
long, undulating stem from the shoots of which issue
different types of flowers. The figures on the plaques
either grasp one or two of the plant's tendrils or reach
toward the flowers. The motif of a man with a tree
occurs from earliest times in the art of the Near East,
and need not be documented here. The rounding of
forms and the exaggerated curves are characteristic of
the North Syrian style.31However, the particular kind
of coiling "tendril" tree is unique. The only close paral-
lels to our trees are the papyrus plants with twined
stalks that flank a frontally posed man in a long dress
on an ivory from Arslan Tash, and the even more
tightly twined tendrils twisted around a central stalk
on an ivory in the Loftus collection.32On neither of
these, however, do the tendrils curve out beyond the
main trunk as they do on the SW7 plaques.
One of the flowers that grows on the trees is a leafy
clusterthat resemblesthe palmette plants of Phoenician
ivories, yet omits the symmetric volutes at the stem
juncture that are always indicated on Phoenician ex-
amples (compare Figures i and I5). This simple pal-
mette has an antecedent in the treesfromthe Investiture

29. H. Frankfort, CylinderSeals (London, 1939) pls. xvII: c, h;

xvIII: a, g, j, k; xix: a, c.
30. CNI, S.69.
31. D. Opitz and A. Moortgat, Tell HalafIII (Berlin, 1955) pl.
94a; CNI, S.I45, S.47, S.69; AiS V, pl. 46k.
32. F. Thureau-Dangin et al., Arslan Tash Atlas (Paris, 1931)
pi. xxxIII: 43; CNI, S.3.

39
 FIGURE 24
Ivory plaque. Room SW7, Fort Shalmaneser.
The British Museum, no. 132692 (courtesy the
Trustees of the British Museum)

mural at Mari, dated to the early second millennium,33

and is found also on a basalt relief from the Bit Hilani
at Carchemish, showing a bullman grasping the long
stem of a palmette plant very much like ours.34On this
basis, I have suggested it as a criterion for the North
Syrian style. These leafy palmette clusters are often
combined on the same panel with other flowersthat are
representedeither as two outcurving petals with a cen-
tral round bulb, or as several outcurving pointed petals
(Figure 9). On a few occasions, all three types issue
from the same tree (Figure 8), while on one complete
panel, the three vertical side plaques each bear a tree
with a different one of the three types of flower.35
Within the SW7 group, a distinctive series of plaques
depicts men wearing odd flowerpot helmets and cork-
screw curl hairstyles. Of these, several show men grasp-
ing the tendrils of the typical tree (for example, Figures
2, I4), while others represent two identical figures
standing side by side, holding hands (for example, Fig-
ure 24).36 These distinctive pieces include a register
above the human figuresin which a winged siren, hold-
ing a blossom in each outstretched hand, appears in the
place of the sundisk. The plaques also include lower
registersin which a couchant female sphinx or lion and
sometimes a stylized floral element are represented.
The garments of these male figureshave already been
discussed in regard to the woolly coat that resembles
armor. The only concrete parallel for the composite
form of the flowerpot helmet is to be found on a ninth-
century orthostatfrom the Citadel Gate at Zinjirli (Fig-
ure 25), although it is possible that one of the female
heads in the round and a female sphinx carved in relief
from the Nimrud ivories wear similar headdresses.37

33. A. Parrot, Sumer(New York, 1959) fig. 346.

34. C. L. Woolley and R. D. Barnett, Carchemish III (London,
1952) pl. B.49a; USK, pl. 35d.
35. SW7, no. 46, plaques I, 4 (my Figure 8); no. 2, plaques 1-3.
36. SW7, nos. 38, 39, 40, 42, 43, 44.
37. N&R II, fig. 499, from room SEi, Fort Shalmaneser; fig.
525, from room SW37.

40
 FIGURE 25 FIGURE 26
Relief. Citadel gate, Zinjirli. Ninth century B.C. Relief. Hilani III, Zinjirli. Second half of the
Oriental Museum, Istanbul, 771I (photo: Ori- eighth century B.C. Oriental Museum, 7730
ental Museum) (photo: W. Schieli, Deutsches Archaologisches
Institut, Istanbul)

The hair of these figures is rendered differently from own beard curls and in the ringlets of his attendant
that of the rest of the SW7 group: three long corkscrew musicians, with their tightly coiled spiral corkscrews
curls down the back of the neck and similar curls in the ending in open curls-are closest to the treatment of
beard. It is a hairstyle known from several North Syrian curls on the ivories, while the attendant figure from
reliefs of the eighth century B.C., associated by Akurgal Hilani III provides a good parallel to the way that the
with the "Aramaean" style.38 It can be seen on the curls curve up rather than simply hang straight down
orthostats of the two men opposite a tree (Figure 19) or at the nape of the neck. The same sort of long curls bil-
the two male sphinxes (Figure I8) from the palace at lowing out in the back and long beard curls are care-
Sakce Gozu, as well as on reliefs from both the Nord- fully represented on Syrian captives being subdued by
hallenbau of Bar Rakib39 and Hilani III (Figure 26) at soldiers of Tiglath Pileser III in his wall paintings from
Zinjirli. The Nordhallenbau reliefs-both in the king's Til Barsib,40 and on foreign mercenaries in one of the

38. E. Akurgal, The Art of Greece:Its Origins(New York, 1968) 39. AiS V, pls. 60, 62.
P- 53. 40. F. Thureau-Dangin, Til BarsibAtlas (Paris, 1936) pl. xxiv.

41
 FIGURE 28

Winged female figure, detail of Figure 2

FIGURE 27
Relief. Palace of Tiglath Pileser III, Nimrud. WORKSHOPS
Archaologisches Institut der Universitat, Zurich,
1916 Clearly, all of the SW7 plaques contain related scenes
or elements. At the same time, in composition as well as
king's battle reliefs from Nimrud (Figure 27); it is a in execution, the flowerpot helmet series seems to con-
style distinct from the Assyrians' own coiffure. stitute a separate unit within the collection, suggesting
The winged female figures that appear in the upper that it was conceived and executed apart from the rest.
registersare most unusual (Figure 28). Their faces are Criteria for individual workshops have been put forth
broad and round, typical of Syrian female representa- in the recent publication of SW7.42 Distinctions drawn
tions, as on the female sphinx column bases of Zinjirli indicate that within a single workshop, different plaques
and Sakce Gozii (Figure 29). The proportions of out- of the same panel may have been done by different
spread wings and tail feathers to head bear striking re- hands, and therefore that a workshop contained several
semblance to the bronze siren cauldron attachments artisans.43 One principal workshop is distinguished,
that have been attributed to North Syrian manufacture whose work consists of the majority of pieces, including
(Figure 30).4' In this case, the human attributes are the hunt panel (Figure 9). The seated man and griffins
placed directly above the standard winged disk, so that panel (Figure 7), the flowerpot helmet group (Figures
the disk becomes the creature's breast, and she is seen 2, 14, 24), a panel that shows men actually standing in
as if flying head-on. Placement of these sirens in the the branches of the tree,44 and another panel on which
same position on the ivory plaques as that occupied by kilted men and long-robed women stand upon a com-
the simple sundisksuggests a connection in meaning as posite floral winged disk with rosette center,45 are con-
well. Similarly, the couchant sphinxes and lions in the sidered the products of different workshops. Evidence
lower registers can be compared to the sphinxes that for this is very convincing, as each of the subgroups in-
occupy the space below the chairs of seated figures on cludes features that differ in rendering and in concep-
other SW7 plaques. tion from those of the main group.

41. 0. W. Muscarella, "The Oriental Origin of Siren Cauldron 42. SW7, pp. 35-39.
Attachments," HesperiaXXXI (1962) pp. 317-329; H.-V. Herr- 43. SW7, no. 3 and p. 73.
mann, Die Kesselderorientalisierende
Zeit (Berlin, 1966) pp. 64-67. 44. SW7, no. 21.
For a discussion of the blossoms held by the sirens on the ivory 45. SW7, no. 65.
plaques, see below, p. 45.

42
 Nevertheless, as each of these subdivisions still main-
tains its closest parallels with the eighth-century reliefs
of Zinjirli and Sakce Gozii, it does not seem necessary,
as argued in the publication, to ascribe to these work-
shops differences either in time or place. Once the raw
material was available in an area, it is possible that a
single center could support a number of coexisting
workshops,46 all of which could demonstrate a range,
not only in the quality of workmanship but also in style,
from archaistic to "modern" and including the rare,
innovative, or unique piece.47
A similar case can be observed in the reliefs of the
Nordhallenbau at Zinjirli. There is a clear difference in
proportion, attention to detail, and overall quality in
the relief of the seated Bar Rakib and his scribe, the
iW 2V musicians, and the attendants with jug and archer's
I gear (see Figure 23)48 from the other attendant figures
characterized by the short and squat gazelle bearer
FIGURE 29 depicted in Figure 20. Apparently, at least two sculp-
Column base. Hilani III, Zinjirli. Second half of tors worked on the orthostats for the building, the more
the eighth century B.C. Oriental Museum, 7731 competent working on the more prestigious royal slab
(photo: Oriental Museum) and its adjacent blocks. A similar situation pertained in
the carving of the Parthenon metopes in fifth-century
Athens, where different sculptors executed individual
slabs, which then reflected varying degrees of expertise
and ranged greatly in style from archaic to classical.49

46. As, for example, in Old Babylonian Sippar, where there

were apparently rows of contiguous goldsmiths' shops (A. L. Op-
penheim, "Trade in the Ancient Near East," VlnternationalCongress
of EconomicHistory,Leningrad,1o-14 August 197o [Moscow, 1970]
p. 18, note 46).
47. As far as quality is concerned, despite the same motif of men
grasping trees, note the sharp contrast between the awkward,
columnar figures of SW7, nos. 64:1-4, with their sticklike arms,
and the well-proportioned and modeled figures of nos. 65-66. More
subtle distinctions in quality can be made among closely related
plaques such as SW7, nos. 2, 4, 5, 19, 21, 26, 38, 40, 46, in all of
which a figure grasps the tendril or branch of a tree. In the more
FIGURE 30
successful examples, the stalk bendsjust below the hand, suggesting
Bronze cauldron attachment. Olympia. Eighth the weight and pull of the figure's arms, whereas in less successful
century B.C. National Museum, Athens, B3040 ones, the curves are arbitrary.
48. AiS IV, pls. 60-62, figs. 257,259; USK, pls. 63b-d, f-h; 64a.
(photo: National Museum)
49. F. Brommer (Die Metopendes Parthenon[Mainz, 1967] p.
174) speaks of a period at the most of eight years, possibly only five,
and quotes C. Picard to the effect that if all of the metopes had been
found in isolated instances, the range would have been extended
over thirty years, from about 460 to 430 B.C., rather than from
447 to 439 at most.

43
ANALYSIS AND INTERPRETATION
OF DECORATIVE SCHEME

That the SW7 panels were all conceived as part of a

single "program" is equally clear when one examines
their content. For, whatever the stylistic variation in
individual works, the theme of a seated or standing
figure with a tree and some form of the winged disk
remains constant.
When one looks at the collection as a whole, it is pos-
sible to come a bit closer to defining the underlying
program. None of the human figures wears the head-
gear that would immediately identify a deity.50 Yet
traditionally, the seated position is reserved for divine
or highly prestigious personages.5'Significantly, when
there is a single seated figure or a pair in a panel, they KJ
occupy the central plaque or plaques. Furthermore,
the pair of bulls that compose the footstool of the seated
male on the griffin panel (Figure 7) are precisely the
animals associated with the Syrian storm god from the
Hittite Empire through the Roman period.52Finally,
at least one of the seated women (Figure 8) holds in her
outstretched hand a large ring similar to that held by a
seated female goddess on a relief of Tiglath Pileser III FIGURE 3I
from the Central Palace at Nimrud (Figure 31) .53 The Drawing of a relief. Palace of Tiglath Pileser III,
same ring is held by a seated figure of Ishtar in the pro- Nimrud (original lost)
cession of gods from Sennacherib'srock relief at Maltai,
as well as in the upper portion of the stela of Esarhad-
also associatedwith Astarte; on equestrian bridle orna-
don (680-669 B.C.) found at Zinjirli.54
ments, the goddess often stands on lions or lions' heads,
At Maltai, all of the male gods hold both a ring and
as does the Ishtar figure from the processionat Maltai.58
a staff; female deities hold the ring alone. On both the
Maltai and Esarhaddonreliefs, Ishtar is the only seated
50. As, for example, the headdresses worn by divine figures on
figure. Her chair is straight-backed,without arms, and ninth-century reliefs from Zinjirli and Carchemish, and by male
its lower portions are decorated with mythological fig- sphinxes of the Sakce Gozu palace (USK, pls. 23a, b; 50c; 53c;
ures. The shape of the chair is not unlike some of the 58d, f ).
51. Frankfort, CylinderSeals, pls. xiv: e, f; xx: b, j; xxiv: f
chairs represented on the SW7 plaques55and is also
(Akkadian); xxv: d, e, i (Ur III); xxvi: k; xxvii: a, b, g (Old
similar to the sort of chair we would expect the SW7 Babylonian); xxx: k, 1 (Kassite).
panels to have decorated. In addition, as was previously 52. 0. R. Gurney, The Hittites (London, I952) p. 141 ;J. Gar-
stang, The HittiteEmpire(London, 1929) p. 304, fig. 42.
mentioned, a number of chairs and stools depicted on
53. Barnett and Falkner, Sculpture,pls. xcii, xcIII.
the SW7 ivories include winged sphinxes between the 54. Hrouda, Kulturgeschichte, pl. 40:I; AiS I, pl. I, fig. 4.
legs of the chairs, while a single example contains a 55 SW7, no. 51.
complex floral arrangement of lotus blossomssimilar to 56. SW7, nos. 41, 48, 50, 49.
57. CNI, p. 85, note i; E. Porada, "Review of Barnett, Cata-
those in the tree before the seated figure.56The associa- logueof NimrudIvories,"AJA 64 (1959) p. 93.
tion of Mesopotamian Ishtar and the West Semitic god- 58. SW7, no. 5 I:1. The association of the goddess with lions is
dess Astarte with the female sphinx has been clearly discussed in H. J. Kantor, "A Bronze Plaque with Relief Design
from Tell Tainat," JNES 2 I (1962) pp. 93-117, esp. pp. oo100-101.
demonstrated.57The lion, which replaces the sphinx on Again, this association continues in Syria into Roman times (see
some of the plaques in the flowerpot helmet group, is Garstang, Hittite Empire,pp. 302-305).

44
The spindles held by a seated female on one of the tween the fertility of the land and care for the "sacred
plaques is often associatedwith Astarte'sclassical coun- tree."63 On those SW7 plaques that contain a tree
terpart, Aphrodite/Venus, although the spindle is also alone, the basis for the complex voluted plant is the
an attribute of Anath in second-millennium texts from palm tree, provider of important staples of life in arid
Ugarit.59The walking genius figures are clearly raised climates.64 The two most frequent blossoms on the
to a "mythological" place by the addition of wings, plaques are the abbreviation of the tree-the palmette
while the wingless figureson one of the complete panels -and the lily or lotus. The lotus is most often depicted
hold short, curved batons often held by the gods, as on as a single blossom. It appears held, on a short stem, in
a relief from Assur.60Thus, although it is not certain addition to growing on trees, and is occasionally shown
that any of the figures on the SW7 plaques represent as part of an alternating chain of buds and flowers.
gods or goddesses, their attributes seem to associate One may wonder why the lotus plant is singled out
them with the divine. and why it appears in the bud and blossom chain. The
Some sort of plant form occurs in each of the figured pattern is often taken simply as an ornament, or else as
plaques. The importance of the plant elements is indi- a representation of two stages in the life cycle: birth
cated on the plaques where there are two men alone (bud) and maturity (flower). However, I would sub-
(for example, Figure 24), as a separate register is pro- mit that the choice of this particular plant is an explicit
vided at the bottom in which an individual floralvolute reference to the fact that the most commonly repre-
is depicted. Several plaques also have lower registers sented type, the Caerulea, or blue lotus, opens and closes
that contain bud and lotus or plant elements (for ex- daily, flowering from sunrise to midday,65and hence is
ample, Figures i, 8). In general, the figures reach to- a constant reminderof regeneration. Thus, the bud and
ward or grasp the curving branches of a tree. On one lotus do not representspecific stages in the plant's life,
panel, the figures actually stand in the branches, while so much as its daily renewal. Its connection with the
on several plaques, the trees spring from a schematic sun is clear as well, as the blue lotus bud opens precisely
representationof earth, indicated by a pattern of over- at dawn and closes at midday, when the sun is most
lapping scales.6' destructive.66On the ivories, these symbols of regenera-
The role of the "sacred tree" in the tradition of the tion and life cycles are quite consistentwith the presence
ancient Near East is well established.62Neo-Assyrian of sacred trees and deities, as well as with the sundisk,
texts from our period refer to the close association be- discussed below.67

59. See SW7, no. 51:2, and CML, Baal II ii and 3 and p. 15. 6I. SW7, nos. 38, 39.
In this context. Mallowan has discussed at length two possible 62. H. Danthine, Le palmier-dattier et les arbressacrisdansl'icono-
interpretations of the "banquet scene," which he associated with ancienne(Paris, 1938); G. Widengren, The
graphiede l'asie occidentale
the laden tables depicted on several SW7 plaques with seated King and the Tree of Life in AncientNear EasternReligion (Uppsala,
women. I do not feel we are provided with enough evidence here I951).
to associate this particular scene with either funerary repasts or 63. CNI, p. 89.
victory celebrations (SW7, pp. i -16), but we must recognize the 64. SW7, nos. 2:4, 9; 21: i, 6; 22:I, 6; 89-94. See Oppenheim,
many situations in which such "meals" play a significant role. In AncientMesopotamia,p. 312, with regard to the date palm's ability
addition to the situations cited by Mallowan, there is the "sacred to withstand the brackish water and salinated soil characteristic of
marriage" (S. N. Kramer, The SacredMarriageRite [Bloomington, southern Mesopotamia, and the date's value as a staple of the
I969] p. 79), seduction scenes in general, such as when Ishtar ancient diet.
wrests a number of attributes from Enlil at Eridu (S. N. Kramer, 65. G. A. D. Tait, "The Egyptian Relief Chalice," JEA 49
SumerianMythology[Philadelphia, I961] pp. 65-68), and offering (1963) p. 96.
scenes, such as food provided for the gods (A. L. Oppenheim, Let- 66. The bud and lotus therefore is an appropriate symbol for
tersfrom Mesopotamia[Chicago, 1967] no. I44). In all cases, the funerary monuments. Tait ("The Egyptian Relief Chalice," p. 99)
laden table had a positive significance. The unifying principle notes that the Egyptian Caeruleachalice is frequently shown in the
might be described as reflecting the abundance that results from ritual of the dead, and not in ordinary banquet scenes, although
a good relationship with the land, and hence with the gods; how- he does not explicitly connect the Egyptian concern for the after-
ever in isolated instances it is not always possible to reconstruct to life to that particular blossom.
which specific narrative an individual representation refers. 67. E. R. Goodenough (JewishSymbolsin theGreco-Roman Period,
60. SW7, no. 65, as compared with Madhloom, Chronology, pl. IV [Princeton, 1954] pp. 25-43) distinguishes between "live sym-
LX:4. The men on the panel also hold curved batons that may be bols," charged with meaning, as opposed to decoration. In the
paralleled by objects held by divine figures elsewhere. ancient Near East, the repetition of symbols, even of a decorative

45
 Although the rosette does not occur frequently on ace.7' The hunt relief from Sakce Gozu was also highly
the SW7 ivories, there is one notable instance where visible, reconstructed by Ussishkin as one of a pair set
the rosette fills the center of a floral sundisk placed at on either side of the gateway to the palace enclosure.
the bottom rather than at the top of the plaques in a Thus, even if we cannot identify the figures in the char-
complete panel.68 In addition, four rosettes appear on iot or the precise significance of the scene,72 we can
the draft-pole of the chariot on the hunt panel (Figure perceive that it is nonetheless culturally charged and
22), which, as noted above, may be comparable to the that the implied power, prowess, and capacity to per-
presence of four rosettes in the field of the Sakce Gozii form are related to the same cultural concerns as pro-
hunt relief (Figure 21). The association of the rosette viding fertility for the land.
with the Mesopotamian goddess Ishtar and with the We come then to the role of the winged sundisk and
Syrian mother-goddess Kubaba is clear. It appears atop siren figures within this context. On the relief of two
the headdress of the seated Ishtar on the relief of Tiglath men opposite a tree from the palace facade at Sakce
Pileser III from Nimrud (Figure 31) and on Kubaba's Gozii (Figure I9), tendrils ending in lotus flowers actu-
polos and robe in reliefs from Zinjirli and Carchemish.69 ally extend down from the sundisk to be grasped by the
As with the lotus, the significance of this particular figures below. This is also common on Neo-Assyrian
plant becomes clear once one observes the rosette in cylinder seals.73 On the SW7 ivories, the disks do not
nature. Plants that grow with radiating leaves lying have pendant blossoms, but consistently appear in the
close to the ground (called basal leaves, or rosettes) are field above figures who grasp tendrillike branches or
among the hardiest of the plant kingdom, living in con- flowers on a tree. On one panel, every plaque contains
ditions unsuitable for most plants, resisting weather a winged disk in addition to the large winged disk por-
changes, and reproducing rapidly.70 Thus the rosette, trayed on a horizontal plaque that extends across the
like the bud and lotus, represents that which endures entire panel (Figure 9). In the flowerpot helmet group,
and generates, and, if the botanical analogy is well the disk takes the form of a female siren holding blos-
taken, is an appropriate symbol for goddesses associated soms (Figure 28); on another panel, the disk appears in
with fertility. an unusual floral configuration.74 Since the association
The presence of rosettes on the Sakce Gzii relief and of the griffin and the sun is attested in classical sources
on the SW7 chariot supports the suggestion that the and has been suggested also in the ancient Near East,75
disproportionately large size of the hunted animals re- it is possible that the griffin-demon guardians of the tree
flects a significance beyond that of a secular hunt. Con- on the seated man panel (Figure 7) may be standing in
temporary kings of Assyria are shown participating in for the sun's presence. If this is so, then this panel, which
royal hunts as part of their iconography of power and is attributed to a different workshop from most of the
success; the reliefs prominently displayed in the pal- SW7 pieces, might well represent a rare, not necessarily

nature, served to intensify a situation and even to heighten the reliefs in H. Frankfort, Art andArchitecture of theAncientOrient(Bal-
drama, as can be seen so clearly in literary compositions. Repetition timore, 1954) pp. 99-o01. Two reliefs with divine symbols on the
further implies continuity, precisely by not showing the unique, draft-pole strap (Hrouda, Kulturgeschichte, pl. 26:3, 4) further sup-
but rather by exemplifying principles which repeat and endure. In port the significance of this occurrence on the ivory (my Figure 22).
this case, then, the bud and lotus, and particularly the bud and 72. The annals of Sennacherib record that on the doors of the
lotus chain, would be a "live" symbol, the pattern synonymous king's palace there were depicted scenes of Assur going to battle in
with the recurrent principle it embodies. a chariot, with the divine Amurru as charioteer and the "victorious
68. SW7, no. 65. Note that this occurs also on the sundisk of prince" (presumably Sennacherib) also in the chariot. Rakib-el,
the Zinjirli seated woman stele (Figure 17). the charioteer of El, was a major divinity in the Aramaic pantheon
69. M. Riemschneider. Die Welt der Hethiter(Stuttgart, 1954) (H. Frankfort, Kingshipand the Gods[Chicago, 1948] p. 327). For
pl. 48; USK, pi. 58c; Woolley and Barnett, CarchemishIII, pls. additional references to gods, particularly weather-gods, in chari-
B.39a, 64. For the rosette's association with Inanna/Ishtar, see ots, see M. Weinfeld, " 'Rider of the Clouds' and 'Gatherer of the
M.-T. Barrellet, "Les deesses ai6les et armees," Syria XXXII Clouds.' " Journalof theAncientNear EasternSocietyof ColumbiaUni-
(1955) P. 244. versity5 (1973) pp. 421-427.
70. E. L. Palmer, "Basal Rosettes," NaturalHistory(November, 73. Porada, Corpus,nos. 640-645.
I960) pp. 36-45. 74. SW7 no. 65.
71. See Assyrian kings hunting in AR, passim, discussed for 75. CNI, pp. 73-77.

46
innovative, but independent solution to a problem, Hittite king were installed in Aleppo and in Carchem-
where the ancient ivory carver was faced with a general ish. Illustrative of this process and particularly apt for
theme and selected an individual way of dealing with it. the present discussionis a text from Ras Shamra of that
The sundisk as represented on SW7 plaques is distin- time, in which the Hittite sun-goddess of Arinna is re-
guished by two large spiral curls framing the disk. These ferred to as the "Shepesh" of Arinna-the name of the
curls are generally associated with the characteristic Canaanite sun-goddess.80
hairdo of the Egyptian goddess Hathor. The allusion, Thus, on those occasions when a female figure hold-
as we shall see, is clearly conscious and most appropri- ing blossoms appears above the winged disk on SW7
ate within the context of the SW7 panels, for in her own ivories (Figures 2, I4, 24), Mallowan has identified her
culture Hathor represents both "the creative force of with Shepesh.81In drawing upon the West Semitic tra-
the sun and at the same time embodies the fertile as- ditions of the Ugaritic texts, it is assumed that the first-
pects of vegetation."76 millennium tradition in Syria did not radically change
This particular form of the sundisk first developed in from the second millennium, although no comparable
Anatolia during the period of the Hittite Empire (I400- body of literature exists from the later period.
1200 B.C.).77 While the symbol is generally seen in asso- Who is this goddess Shepesh and what were her func-
ciation with representations of the king, the presence of tions? In the poem of Ba'al, Shepesh is not a major fig-
Hathor curls reflects Anatolian assimilation of Egyp- ure. She is called "luminary of the gods," and as such
tian elements at this time,78 especially appropriate serves as a messenger for E1.82 She is sometimes de-
since the Hittite sun-deity was also female. It is pre- scribed in her destructiveaspect, "burning hot," so that
cisely at those first-millennium North Syrian sites that "the furrowsin the field are cracked with drought."83
had strong Hittite and Neo-Hittite traditions that one But she is also enlisted as a helper of Anath, accom-
finds examples of the winged and voluted disk.79 The panying the goddess down into the earth to retrieve
blend of West Semitic with Hittite traditions clearly Ba'al and thus return fertility (and life) to the earth.84
had begun during the syncretistic period of the Hittite In the "Pantheon List" from Ras Shamra, she is cited
Empire in the late second millennium, when sons of the immediately after Anath.85

76. E. L. B. Terrace and H. G. Fischer, Treasuresof EgyptianArt wereHittitesin residencein the Egyptiancourtin the earlyfour-
from theCairoMuseum(London, 1970) p. 48. Barnett (CNI, p. 82) teenthcentury.
suggested that these curls as worn by female figures on ivories from 79. In one example(Woolleyand Lawrence,Carchemish II, pl.
the Loftus Collection may indicate the Canaanite goddess Qedesh A. 6:i), the centerdisk containsthe four-pointedstar relatedto
attested on second-millennium monuments, who is often shown the Hittite"signeroyale."The continuationof second-millennium
with Hathor locks and sundisk on her head, holding lotus flowers. motifsintoNorthSyrianartof thefirstmillenniumis mostapparent
Although W. Helck (Betrachtungen zur grossenGottinundden ihr ver- in the banquet scenes and representationsof seated figures:for
bundenenGottheiten[Munich, 197I] p. 217) has since shown that example,the reliefsof Yagriand Alaca Huyuk (H. Bossert,Alta-
Qedesh was not a deity but rather an epithet, "qedesh," or holy, natolien[Berlin,1942]figs.571, 516), that continuedin the ninth-
that was applied to the goddess Astarte, the association with princi- centuryreliefsof Zinjirli,Carchemish,and Marash,as well as in
ples of fertility is clear in any case. the eighth-centuryexamplescited earlierfrom Zinjirli,Marash,
77. See the representation of Tudhaliya IV (I250-1220 B.C.)at and Tell Rifa'at (USK, pls. 57c, 2ic, 43i, 45b, 66d, 45d, g, 46a,
Yazilikaya, in E. Akurgal, Art of theHittites (New York, 1962) pl. 47d, 48i).
xxix and fig. 78. 80. R. Dussaud, Les ReligionsdesHittiteset des Hourrites,p. 335,
78. K. Bittel, in an address to the Columbia University Seminar cited in N&R II, p. 496, note 59.
on the Archaeology of the Eastern Mediterranean, Eastern Europe 8I. SW7, pp. i6-18; N&R II, pp. 496-498.
and the Near East (December 13, 1973), speculated on the possi- 82. CML, Baal III*C i 15.
bility that the Temple of the Storm-God at Bogazkoi owed its form 83. CML, Baal Vv 17-18, II viii 21-23, III ii 24-25, III iii
as a freestanding sanctuary surrounded by a walled complex of 24-iv 3.
storerooms to Egyptian prototypes, resulting from contacts be- 84. CML,Baal Ii 7-15, III iv I7-20. For a study of the sun-
tween the two regions, especially during the late fourteenth and goddess,see A. Caquot, "La divinite solaireougaritique,"Syria
early thirteenth centuries B.C. In a more recent paper by K. K. XXXVI (1959) pp. 90-10I, in which the goddess is described as a
Riemschneider, "Who Taught Babylonian to the Egyptians?," minordeityservingas an "elementof scientificexplanation"in the
presented before the meetings of the American Oriental Society, fertilitymyths.
Philadelphia, March I6, 1976, it was further suggested that there 85. J. Nougayrol et al., UgariticaV (Paris, I968) pp. 45, Ii,
20, 21.

47
 It is tempting to rely upon these Ugaritic texts, the are generally associated with Astarte, goddess of fer-
Ba'al myth in particular, and to see in the ivories analo- tility, on both second- and first-millennium monu-
gies to Ba'al the storm-god and his sister, Anath, pro- ments.88 I have already referred to attributes of Astarte
viding fertility to the earth, aided by the sun-goddess in conjunction with representations of the seated
Shepesh. It is equally tempting to see in the seated male women and female sphinxes on several plaques from
figure of the griffin panel (Figure 7) Ba'al himself, his the SW7 group. However, it must be noted that at pres-
feet resting on the bulls of the storm-god, cup in hand, ent it is not possible to determine if blossoms per se
and flanked by animal-bearers, as if this were an illus- were restricted in association to the single goddess, or if
tration of the occasion described in the myth in which, they could serve as attributes of several female divini-
after Ba'al's defeat of Yam, "[they] held a feast for him ties.89
and gave him to drink; [they] gave a cup into his Finally, the assimilation of the two chief goddesses of
hand."86 the Ugaritic pantheon, Anath and Athirat (Astarte), is
Yet before one applies a specific story to the repre- attested during the late second millennium90 and is
sentations on the SW7 panels, it must be pointed out reflected in the myth of Ba'al where both Anath and
that there are similarly enticing elements in other myths Athirat carry Ba'al's cup.91 Given this assimilation, it is
of the mid-second millennium B.c. For example, Shep- significant that in the Ugaritic literature, Anath is
esh also plays a beneficent role in the story of "Shahar sometimes described as a bird of prey with outstretched
and Shalim"-twin gods identified as dawn and dusk wings.92 A particularly suggestive reference occurs in
and associated with viticulture, at whose birth Shepesh the myth of Aqhat, in which Anath joins a flock of
makes "tendrils abound with... and with grapes."87 eagles hovering above the hero during a meal, in order
The image of the twin figures in flowerpot helmets to steal his divine bow.93 Aqhat, who is associated with
springs to mind (Figure 24), particularly in regard to fertility, is warned in the use of his weapon that he must
the siren figures in the sundisk, and the tendrillike offer the first fruits of the chase to "his" (presumably
quality of the branches. Ba'al's) temple.94 His father, Danel, is a chthonic deity;
It is also necessary to explore the possibility that the like Aqhat, Danel is associated with the produce of the
winged creatures do not represent the sun-goddess at earth.95 One could relate this story as well to the SW7
all. The blossoms held by siren figures on our plaques group: winged female sirens with outstretched eagle's

86. CML, Baal Vi 4-I9. head and arms but no body are the sirens with blossoms incised on
87. CML, p. 23; Shahar & Shalim i 25-26. tridachna shells that have been found throughout the Near East
88. Kantor, "A Bronze Plaque with Relief Design from Tell and the Mediterranean. However, the figure on the shells has never
Tainat," pp. 93-117; R. D. Barnett, "North Syrian and Related been identified (see the bibliography in S. Stucchi, "Un nuovo
Harness Decorations," in K. Bittel, ed., Vorderasiatische Archdologie frammento di Tridacna Incisa," Bolletino d'Arte 44 [I959] pp.
(Berlin, 1964) pp. 21-26. There is a small silver pendant from I58- 66).
Zinjirli, on which a nude frontal female, her arms bent at the 90. W. F. Albright, The Archaeology and Religionof Israel (Balti-
elbow, holds a blossom in each hand (AiS V, pl. 47f ). Although the more, 1942) pp. 74-75. For monuments, see I. E. S. Edwards,
figure is not identified by inscription, she is iconographically the "A Relief of Qudshu-Astarte-Anath in the Winchester College
same as those represented on the equestrian plaques discussed by Collection," JNES 14 (1955) pp. 49-51, pl. iI, where a nude god-
Kantor and Barnett. It should be noted also that Astarte is the dess holding a blossom in her right hand and a serpent in her left
only second-millennium goddess mentioned in a first-millennium is shown standing on the back of a lion. The accompanying inscrip-
text-as one of the six Phoenician deities called to witness the tion identifies her as all three goddesses.
treaty of Esarhaddon with Ba'alus of Tyre, about 671 B.C. (R. 91. CML, Baal Vv 33-34. In another Ras Shamra text describ-
Borger, Die InschriftenAssarhaddons, KonigsvonAssyrien[Graz, 1956] ing a banquet of the gods (R. S. 24.258, translated by C. Viroul-
p. 109, line 18). leaud, in Nougayrol et al., UgariticaV, pp. 545-551), Anath and
89. A clothed frontal female holding blossoms is represented on Astarte prepare a meal together for (presumably) El.
the Mahrada stele, found outside of Hamath (CNI, fig. I4; USK, 92. See Virolleaud in Ugaritica V, pp. 553-555, regarding
pl. 38g). The inscription identifies her as "Great Queen of HA- R.S. 24.252.
country." According to J. D. Hawkins, "This could be a writing 93. CML, Aqhat III i 20-21; III i 30-32.
of 'Hamath,' as Egypt is elsewhere written MI-country-i.e. 94. CML, Aqhat V 36-38, and p. 53 note 13.
Mizri. The Great Queen in that case must be a goddess, and the 95. Implied in CML, Aqhat III i 15, I i 30-3 1. Driver notes on
chief goddess of Hamath was called Pahalatis in other contempo- p. 8 that Anath clearly intends to revive Aqhat after he is inadvert-
rary inscriptions, i.e., Ba'alat" (personal communication). Unfor- ently killed (III i 16); he must have been resurrected at the end of
tunately, it is not clear to what extent Ba'alat was considered a the poem, which is now missing.
separate deity or an epithet. Closest to our winged females with

48
 ing of the same symbol.96 This equation of sirens and
winged sundisks would seem to be supported by the
- --.4 recent discovery of a pair of bronze equestrian orna-
..i
ments in a late eighth-century tomb at Salamis, Cyprus,
on which a female siren figure is seen over the head of a
nude "mistress of animals" (Figure 32).97 On other
equestrian ornaments of similar date, this same nude
female is frequently represented with a simple winged
disk in the field above.98 As on the SW7 plaques, then,
the two elements seem to be interchangeable in other-
wise standard compositions. Nevertheless, there does
remain the enigmatic representation on the Bomford
plaque-an equestrian frontlet of North Syrian style
although of unknown provenance-where we see
. I -. ?w
.
.-P - clearly a harpylike creature with the head of a woman
...5 I
and the body of an eagle, including talons, in the field
above a nude female.99
It is indeed unfortunate that no relevant first-mil-
lennium literary texts have been preserved with which
to compare the representations on the SW7 ivories. The
Ugaritic texts cited above appear related to the scenes
represented on the ivories, although no definite corre-
lations can be made. This may be because the repre-
sentations are not specifically parallel to a text; or, if
they are, because we do not have the text. In this re-
gard, it is important to keep in mind that the ancient
ivory carver would have taken a common cultural
vocabulary for granted and would not necessarily have
been explicit in his literary allusions.'1?
)
--

96. Mallowan (SW7, p. I7) has suggested that if the winged

siren figures represent the female sun-goddess, then perhaps the
FIGURE 32 winged disk alone represents the sun in its male aspect, a reference
to the male solar deities of neighboring Babylonia and Assyria.
Drawing of a bronze equestrian ornament, Sala- With this I must disagree, first because there is no reason to assume
mis. Eighth century B.C. Cyprus Museum, Nico- such a reference in ivories produced in North Syria, and second
sia (photo: courtesy Director of Antiquities, Cy- because all of the winged disks retain the feminine association of
the Hathor curls, which are absent only when the siren figures with
prus Museum) their own hair curls are placed above.
97. The pieces are badly corroded (see Karageorghis, Excava-
tionsin theNecropolisof SalamisIII, pl. LXXXIX). Despite the solution
wings as Anath; animal bearers as Aqhat; men with the I wonder whether the siren's
presented by published drawings,
plants to emphasize fertility; and two like figures in appendages are spread talons or blossoms held in outstretched
the flowerpot helmet series, not as twins but rather as hands.
Danel and Aqhat. 98. Examples occur from Nimrud (J. J. Orchard, Ivoriesfrom
Nimrud (i949-1963), FasciculeI: : EquestrianBridle-HarnessOrna-
At the present time, the evidence seems to be
ments[Aberdeen, 1967] pl. xxvIIm)and from Gordion (R. S. Young,
weighted toward an identification of the SW7 siren fig- "The 1961 Campaign at Gordion," AJA 66 [1962] pl. 46).
ures with the sun-goddess only because of the appar- 99. Barnett, "North Syrian and Related Harness Decorations,"
pl. 11:3.
ently interchangeable places occupied by the sirens and Ioo. On Babylonian narrative art being allusive rather than
the simple winged sundisks with Hathor curls, suggest- depictive, see A. Perkins, "Narrative in Babylonian Art," AJA 6I
ing that the simple disks are but an abbreviated render- (1957) P. 55.

49
 FIGURE 33
Relief. Palace of Tiglath Pileser III. Detroit Institute of Arts, 50.32

Thus, while we cannot for the present match pre- elements, figuresassociatedwith trees,and laden tables.
cisely the representations on the ivories with known All of these elements reflect a common meaning: the
texts, what is clear is the general consistency in the- daily rising of the sun, like the daily opening of the lotus
matic material from panel to panel: winged disk, plant in its floweringseason, reflectsthat continuity necessary

FIGURE 34 FIGURE 35
Relief. Palace of Tiglath Pileser III. The British Relief. Palace of Tiglath Pileser III. The British
Museum, no. 118907 (courtesy the Trustees of Museum, no. II8905 (courtesy the Trustees of
the British Museum) the British Museum)
for life, which then requires sustenance from the earth
whose abundance is celebrated by the laden tables.1I0
Nor is the hunt out of place in this context. The success-
ful hunt, like war, ends in victory; and, like the motif of
the bud and lotus, implies the cycle of death which
must preceed rebirth. It is striking that both the god-
desses Anath and Astarte and the storm-god as well
combine in themselves the dual aspects of fertility and
war; as if fertility too was won through struggle and
thus the achievement thereof was to be celebrated as a
victory after battle.I02
Despite individual variations, then, the SW7 ivories
were clearly produced according to a single icono-
graphic program. This unity in conception strongly
suggests that the scenes not only had a meaning, but ^
also a purpose. Although the links to fertility might
FIGURE 36
imply that the pieces were more appropriately to be Column base. Palace, Tell Tainat. Second half
associated with beds than with chairs, it is equally pos-
of the eighth century B.C.Antioch Museum, 6020
sible that the panels decorated chairs that served some
(photo: Oriental Institute, University of Chicago)
special function, such as specific repasts or ceremonies.
The chairs may even have been part of temple furni-
ture; they are significantly like those upon which the from Building II at Tell Tainat, which are dated after
goddesses sit on reliefs from Maltai and elsewhere, and the Assyrian annexation of Pattina by Tiglath Pileser
seem to exemplify the "seat for a god [with] a rest at in 743 B.C. (Figures 24, 36).107
[its] back," presented to Athirat in the poem ofBa'al,103
while in the inventory texts of Assurbanipal from Nine- ioi. It is during this period in Egypt and Phoenicia that the
veh, mention is made of ceremonial furniture used in corollary motif of the infant Horus, called the "infant sun," seated
on a lotus, became popular in art. Tait ("The Egyptian Relief
connection with the cults of specific deities.104
Chalice," p. I35) specifically associates the motif with rejuvena-
tion.
DATING AND HISTORICAL I02. This may explain why both warriors and caretakers of the

PERSPECTIVE sacred tree wear the same garment on the SW7 plaques. The
observation was made, but not pursued, by G. Herrmann in SW7,
p. 26; Mallowan, SW7, p. o0,does refer to the men grasping trees
As far as the date of the ivories is concerned, the as "powerful warriors," without indicating, however, why he calls
them this.
chronology of North Syria in this period is essentially
103. CML, Baal II ii 31-32.
linked to that of Assyria. The best parallels for Assyrian
104. A. R. Millard, "Fragments of Historical Texts from Nine-
elements on the SW7 plaques come from the reliefs of veh: Ashurbanipal," Iraq XXX (1968) p. io8: (B.M. 83-1-18,
Tiglath Pileser III-such as the hairstyle of foreign 6oo + B.M. 123425): bed of... wood with precious stones for
mercenaries and prisoners (Figure 27); the common Bel and Beltiya "for performing marriage"; bed of ebony overlaid
with gold for Marduk.
court garment of long-skirted robe with fringed shawl 105. Madhloom, Chronology, p. 69.
wrapped around the waist and over one shoulder (Fig- Io6. Thureau-Dangin, Til Barsib Atlas, pi. XLVII. Furniture
ures I, 33); sandal type;l05 the armor worn by horse- styles did not change substantially from the ninth through the
seventh centuries B.C. and so cannot be used in chronological
men (Figure 34); a possible variant on the flowerpot
arguments.
helmet worn by another horseman (Figure 35); and a I07. Stylistic similarities, as pointed out by Herrmann, SW7,
tasseled cloth covering the king's throne, similar to that p. 88, may be seen in the open mouth, crouching posture, and clear
distinction of the ruff from the rest of the mane. For the date of
used on the seats of the ivory plaques.106 In addition,
Building II at Tell Tainat, see R. C. Haines, Excavationsin thePlain
the best comparison for the couchant lions that appear of Antioch:The StructuralRemainsof theLaterPhases(Chicago, 197 I)
on several plaques is with the lions of the column bases p. 66.

51
 Clearly, the parallels from Zinjirli are from the and Bar Rakib. This would accord roughly with the
period of Bar Rakib, king of ancient Sam'al from about date of about 740-730 proposed by Mallowan and
730 to 720 B.C. Striking similarities have also been ob- Herrmann as the most reasonable period for the ivor-
served with the hunt relief of the palace enclosure of ies' production. "
Sakce Gozu, demonstrated by Ussishkin to date to the At this time, Bar Rakib of Sam'al was a vassal of
first half of the eighth century.108 In a paper presented Tiglath Pileser of Assyria. An inscription written by
before the American Oriental Society in I972, I sug- Bar Rakib for his father, Panamu II, describes how the
gested that the relationship apparent in reliefs from earlier king had been killed "running at the side" of the
Zinjirli and from the palace at Sakce Gozu was due to Assyrian king's chariot; in other words, while fighting
close political ties between the two sites, and that both with the Assyrian army, presumably against Damascus
had probably been part of Sam'al. In any event, the in 733.112The representation of foreign bowmen on the
reliefs from the two sites were products of the same reliefs of Tiglath Pileser III (Figure 27) thus takes on
workshop and are alike in overall style and specific greater significance as an illustration of the type of
details of rendering as well as in conception.109 Thus, event recorded in the Aramaean inscription. And as
the palace of Sakce Gozu, whether the site was politi- further support for the suggestion of the ivories' manu-
cally allied with Zinjirli or not, evidently belonged to facture in Sam'al, in a study of foreigners represented
the same cultural tradition and looked, presumably, to on Assyrian reliefs, M. Wafler has shown that the par-
the larger site for its craftsmen, much as Solomon, in ticular hairdo of corkscrew curls (as seen on several of
building his temple and palace at Jerusalem, called in the ivories as well as on the "foreign" bowmen) is pe-
artisans from Tyre. 10 culiar to men of Sam'al.113
I would suggest, therefore, that the SW7 ivories, as a While we cannot definitely conclude that the ivories
coherent subgroup within the North Syrian style of the could not have been begun before Tiglath Pileser's
early first millennium B.C., were produced in the same reign, or subsequently, in the reign of Sargon II (722-
cultural context as the Zinjirli and Sakce Gozu reliefs- 705 B.C.), it is likely that they were executed and deliv-
in Sam'al, during the reigns of both Tiglath Pileser III ered to Nimrud during Tiglath Pileser's time. Mallo-

io8. D. Ussishkin, "The Dates of the Neo-Hittite Enclosure at Hilani II with the gateway lions of the Hallenbau P, also at
Sakcagozii," BASOR pp. 18I (1966) 15-23. In his article cited in Zinjirli, thus establishing the stylistic unity of sculpture from a
note 26, Ussishkin tried unconvincingly to show that the ivories building dated to the reign of Bar Rakib with work from Sakce
should have been carved prior to the Sakce Gozui huht relief, be- Gozu. To this may be added close similarities in the various repre-
cause of the six-spoked chariot wheels and the form of the chariot sentations of Bar Rakib, as well as attendant-figures from Hilani
(BASOR 203, pp. 26-27). His arguments have been refuted by M. III, with the Sakce Gozu king relief, and in the female sphinx
A. Littauer and J. Crouwel ("The Dating of a Chariot Ivory from column bases from both sites (see my unpublished dissertation,
Nimrun Considered Once Again," BASOR 209 [1973] pp. 27-33), pp. 207-2 I0).
who demonstrate that the profile of the chariot box on the ivory is I Io. I Kings 5.
one known from the reign of Tiglath Pileser III on, and that al- I I. SW7, p. 62. The reference on p. 64 ofSW7 to the presence
though the six-spoked chariot wheel is generally associated with the of "Phoenician or Aramaean" fitters' marks on the backs of several
ninth century, there is at least one chariot of the period of Tiglath plaques supports my argument. These markings were not recorded
Pileser that has only six spokes. Ultimately, Littauer and Crouwel at the time of their excavation before they were masked in protec-
opt for a date in the reign of Sargon II of Assyria (722-705 B.C.)for tive bandages. However, since the plaques are in the North Syrian
chariotfirstappearsin
the hunt ivory,becausethe four-passenger style, the markings are probably not Phoenician. The presence of
reliefs at that time. However, it must be emphasized that the ivories Aramaean signs would seem to rule out a provenance such as Car-
are not Assyrian, but rather North Syrian. Just as the chariot with chemish, where the Hittite hieroglyphic script was still used, and
an eight-spoked wheel appeared earlier in North Syria than in would be consistent with the hypothesis that the ivories were pro-
Assyria, the four-passenger vehicle may well have been developed duced in Aramaean Sam'al.
first in Syria and subsequently adopted in Assyria. I believe that 112. For the inscription of Panamu, see H. Donner and W.
the evidence amassed by Littauer and Crouwel is consistent with Rollig, Kanaandische undAramaischeInschriftenI (Wiesbaden, I962)
the strong arguments for dating the rest of the SW7 group within pp. 223-224, no. 215.
the reign of Tiglath Pileser III. I Darstellungen,Alter
I3. M. Wafler, Nicht-Assyrerneuassyrischer
109. Akurgal, in Greece,p. 60, noted the relationship between Orient und Altes Testament 26 (Neukirchen-Vluyn, 1975) p. 186
the female sphinx orthostat of Hilani II at Zinjirli and the male and figs. 104, io6, 107.
sphinx orthostats of Sakce Gozu, as well as that of the lions from

52
wan makes it clear that Fort Shalmaneser was used by by the presence at Al Mina on the Syrian coast of ivory
Tiglath Pileser."4 What is more, in 738, the king of tusks partially sawed for carving. It is supported in the
Sam'al had been included in Tiglath Pileser's tribute Assyrian records where receipt of ivory tusks is recorded
lists, in which gifts of ivory were noted.15 from both Carchemish and Pattina, suggesting that the
The SW7 ivories may well have been tribute for the raw materials necessary for local production were at
Assyrian king from his vassal at Sam'al, either Panamu hand.120 An implication that the existence of various
or Bar Rakib, although, as it is assumed that Bar Rakib centers of ivory production was recognized in antiquity
lived into the reign of Shalmaneser V (726-722 B.c.), itself is contained in the annals of Shalmaneser III of
it is possible that the furniture was a gift presented to Assyria, in which the chronicler carefully distinguishes
Shalmaneser on his accession to the throne. That such among the types of ivory furniture taken as booty from
a gift would have been considered appropriately lavish different cities: ebony furniture set with ivory from
is implied in the annals of the various Assyrian kings Carchemish, ivory furniture overlaid in silver and gold
who prized ivory furniture as booty or tribute. It is even from Bit Adini, and inlaid ("tamlu") ivory furniture
more eloquently suggested in the poem of Ba'al, in the from Damascus. I2 It is therefore highly likely that upon
joy with which Athirat received her gifts of a chair, further stylistic analysis, additional subgroups within
footstool, and table, fit "for a god.""6 the Syrian style of ivory carving can be attributed to
other centers of production, and that other media, for
example metalwork, would also yield to subdivision.
SUMMARY AND CONCLUSION Second, in addition to implying the existence of many
ivory-working centers, the SW7 assemblage provides
Now that all of the threads of stylistic parallels for the us with information about the organization of craft pro-
SW7 ivories, their iconographic content, and chrono- duction during this period, unfortunately so absent in
logical placement have been drawn together, the sig- contemporary documents. For example, we can see that
nificance of the group becomes clear. For in this single a single center contained multiple workshops, and that
collection we are provided with not one, but four con- an individual workshop could include several crafts-
clusions regarding fine ivory work in the early first mil- men.
lennium B.C. Third, the closely related panels in the SW7 collec-
First, as regards centers of production, Barnett, in his tion exhibit a range in quality and treatment of the
publication of the Nimrud ivories excavated by Loftus single iconographic theme that allows us to glimpse
and Layard, suggested that the entire group was carved solutions chosen by individual artists within the pro-
at Hamath, which he proposed as the center of ivory gram's specifications and that suggests the limits of any
working in the early first millennium."17 This was later rigidly linear notion of stylistic development through
supported by Riis, who cited waste flakes of ivory, as if time.
from a workshop, found in the excavation of the palace
I14. N&R II, p. 469.
at that site.II8 Whether or not there was ivory carving
I I5. AR I, ?? 772, 8o0.
at Hamath is not within the scope of this paper to deter- I I6. CML, Baal II ii 29-3I. For the value placed on fine ivory
mine. By distinguishing a significant subgroup within work, see R. S. 25.421 in Nougayrol et al., UgariticaV, pp. 315,
the Syrian style and attributing it to Sam'al, however, 3 7, in which a goddess's beauty is likened to the perfection of an
ivory panel.
I do suggest the existence of at least one local center of I 17. CNI, p. 46, and reaffirmed in R. D. Barnett, "Hamath and
manufacture. Since Sam'al was not one of the larger or Nimrud," Iraq 25 (1963) pp. 81-84.
wealthier states of North Syria in the ninth and eighth I I8. P. J. Riis, SukasI (Copenhagen, 1970) p. 169.
centuries B.C., such a conclusion implies a model of 119. See P. K. Hitti, Historyof Syria (London, 1951) p. 49, and
A. H. M. Jones, "Asian Trade in Antiquity," in D. S. Richards,
multiple centers of production of fine ivory work at this ed., Islamandthe Tradeof Asia (London, I970) pp. I-Io, especially
time-on the pattern of Syrian cities in more recent p. 8.
I20. Al Mina: CVI, p. 165, note I. Balawat: L. W. King, The
times, where most luxury production is carried out in
BronzeReliefsfrom the Gatesof Shalmaneser, King of Assyria(London,
independent operations in each major center.119 The 1915) pls. xxvIII, xxxIII.
validity of this suggestion for antiquity is strengthened I21. AR I, ?? 475, 476, 740.

53
 And finally, fourth, the very recognition of a "pro- has been speculated that the front panel illustrated here
gram"-appropriate iconographic themes related to is the product of a Syrian workshop, or of Syrian crafts-
classes of objects-makes clear the close ties that must men in Constantinople during the sixth century A.D.'22
have existed between other classes of objects and their This opens up the specific problem discussed here to
decoration at this time. This observation unites the the broader context in which the throne of Maximian
SW7 ivories with a corpus of monuments as far-reach- is in a very real sense the heir to the Syrian ivory-carv-
ing as the chair of the archbishop of Ravenna, with its ing tradition of the first millennium B.C. It suggests,
panels of the evangelists and John the Baptist (Figure despite gaps in the archaeological record, a continuity
Io), as well as the sculptural programs of classical and in tradition of considerable importance in the later his-
medieval architecture. In fact, it is perhaps particu- tory of the art of fine ivory work, of which the SW7
larly apt to close with the throne of the archbishop, as it panels represent a significant part of the earlier phases.

ACKNOWLEDGMENTS

An earlier version of this study formed a portion of my doc- Anat. Stud.

toral dissertation,"North Syria in the Early First Millennium Anatolian Studies
B.C., With Special Reference to Ivory Carving" (Columbia AR
University, 1973). I wish to acknowledge the constant support D. D. Luckenbill, Ancient Recordsof Assyria and Babylonia
and help of the Ancient Near East Department of The Metro- I, II (Chicago, 1926, 1927)
politan Museum of Art, and in particular the encouragement BASOR
of Prudence O. Harper, who through ProfessorEdith Porada Bulletin of the AmericanSchools of Oriental Research
of Columbia University originally provided the opportunity CML
for me to begin work on this material. I am also grateful to G. R. Driver, Canaanite Myths and Legends (Edinburgh,
Georgina Herrmann and Oscar White Muscarella, for their I956)
valuable comments after reading the manuscript, and for CNI
stimulating conversations over the course of several years. I R. D. Barnett, Catalogueof the Nimrud Ivories in the British
would further like to thank Sir Max Mallowan and Barbara Museum (London, 1957)
Parker, with whose kind permission I was able to examine JEA
then unpublished ivories at the Institute of Archaeology, Journal of Egyptian Archaeology
London, in the spring of I971 when, as a Fellow of the JNES
Samuel H. Kress Foundation, I had the opportunity to pur- Journal of Near Eastern Studies
sue research on several of the problems discussed here. LAAA
LiverpoolAnnals of Archaeologyand Anthropology
N&R
M. Mallowan, Nimrud and its Remains (New York, I966)
SOURCES ABBREVIATED SW; 7
M. Mallowan and G. Herrmann, Ivories from Nimrud
AiS (1949-9631), Fascicule III: Furniturefrom SW7, Fort Shal-
F. von Luschan, Ausgrabungenin Sendschirli(Berlin, 1893- maneser(Aberdeen, 1974)
I943) USA
AJA W. Orthmann, Untersuchungenzur spdthetischen Kunst
AmericanJournal of Archaeology (Bonn, I971)

I22. R. Hinks, CarolingianArt (London, I935) pp. 43, 44.

54