Temple of Ishtar at Nineveh

The Ishtar Temple at Nineveh Author(s): Julian Reade Source: Iraq, Vol. 67, No. 1, Nineveh.
Papers of the 49th Rencontre Assyriologique Internationale, Part Two (Spring, 2005), pp. 347-390 Published by: British Institute for the Study of Iraq Stable URL: http://www.jstor.org/stable/4200586 . Accessed: 07/04/2013 17:53
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp
.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.
British Institute for the Study of Iraq is collaborating with JSTOR to digitize, preserve and extend access to Iraq.
http://www.jstor.org
This content downloaded from 146.95.253.17 on Sun, 7 Apr 2013 17:53:51 PM All use subject to JSTOR Terms and Conditions
347
THE ISHTAR TEMPLE AT NINEVEH

By JULIAN
READE
Nineveh, like modem Mosul of which it is now a suburb, lay at the heart of a prosperous agriculturalregion with many interregional connections,and the temple of Ishtar of Nineveh dominatedthe vast mound of Kuyunjik(Fig. 1). Trenchesdug on behalfof the BritishMuseum, mainlyby ChristianRassamin 1851-2, HormuzdRassamin 1852-4 and 1878-80, George Smith in 1873-4, and LeonardKing and ReginaldCampbellThompsonin 1903-5, impingedon the site. The main temple was almost completelycleared, together with an area to the north-west,by Thompson and colleagues in four seasons between 1927 and 1932 (Figs. 2-3). Many original King and Thompsonrecordsare kept in the Departmentof the AncientNear East at the British Museum;some photographic negativesare at the Royal AsiaticSocietyin London.The numerous objects from Thompson'sexcavationsare now divided between the Iraq Museum, the British Museum (where they are registeredin the 1929-10-12, 1930-5-8, 1932-12-10and 1932-12-12 to the four successiveseasons),the Birmingham collections,mostly corresponding City Museum, some were given the AshmoleanMuseumat Oxford,and the FitzwilliamMuseumat Cambridge; to other institutions,and to individualswho had contributedto the excavationcosts. Summarydescriptionsof all this work, with references,can be found in the Reallexikonder recordsincludeinformation Assyriologie (Reade 2000:407-9). King'sunpublished coveringmuch of Kuyunjik,which should with patient study be amenableto three-dimensional modelling,but thereis little about the templeitself.Thompsonin 1927-32 workedwith modestfundingbut great determination. He was acting on his own initiative,long after leavingthe BritishMuseum,which His promptand generouslyillustrated agreedto sponsorthe work but had a limitedinvolvement. reportsare the most detailedwe have of any of the Britishoperationsat Nineveh,but they are in journals which cannot have circulatedwidely abroad. They are anyhow hard to comprehend. Thompsonwas describedby Mallowan (1977: 69) as "an epigraphistby training,who had no and for Thompsonthe most important high regardfor archaeology", finds,on whichhe continued to work until his death in 1941,werethe cuneiformrecords;the restconsistedlargelyof what one of his notebooks calls "Parthian junk". He entrustedmany of the archaeological responsibilities to colleagues,and his own frustrations at the museumunderBudge'skeepership ensuredthat he gave them propercreditfor theirwork, but they had theirown lives to lead and seldompublished more about the site. There had not been the dramaticdiscoverieswhich might have justifieda long book, and we are left thereforewith the originalnotebooks and reports.The currentpaper mainly deals with the architecture and stratigraphy of the temple throughall periods,and with some of the more significant objectsfound;it amplifiesand occasionallyamendsstatementsI have made elsewhere.Thereis furtherinformation, especiallyabout later occupantsof Nineveh,which would rewardattention.I am indebtedto DominiqueCollon, Ann Searightand WilliamReade for assistancewith some of the illustrations used below.
Ishtar of Nineveh (cf. Lambert 2004/5)
The very name of Ninevehis reminiscent of the Sumerian divineappellationNin, and the status of the town as a leadingcult-centre goes back to the mid-thirdmillennium, perhapsmuchearlier. The names of the goddess of Nineveh (Beckman 1998) are written or transliterated in various ways, notably Shaush(k)a (Hurrian,attested around 2000 BC), Ishtar (best-known,used here), Mullissu (as spoken in the seventh century), Inanna and Ninlil. She was goddess of nature, passion,fertility,sex and war, and a healertoo, one of manyavatarsof the GreatGoddesswhose personality,activities,influenceand evolution are the theme of many studies (e.g. Wilcke and Seidl 1976;Groneberg1997). While widely worshipped,and especiallypopularwith Hurriansin the second millennium,as Assyrianpower grew Ishtarof Nineveh herselfbecameincreasingly a state goddess closely associatedwith the god Ashur. Almost any Assyrian representation of a largelynaked woman is liable to be seen as some manifestationof Ishtar.There are prehistoric
Iraq LXVII/ 1 ( 2005 ) = RAJI 49/2 ( 2005)
348
JULIAN READE
KUYUNJIK,|
c.640 BC: sketch
NorthGate?
/ P
~ ~~alace ~
Temple?
East Gate
50 100
200m.
about 640 BC. Author'sdrawing. Fig. 1. Kuyunjik
versions,and many more in various postures,with stylisticchanges,betweenthe third and first millennia(e.g. Gut et at 2001: 118-19, Abb. 14.154, 15.156;Reade 2002: 556-7, Figs. 4, 5, 7). AlthoughIshtarof Ninevehwas not the only goddessresidentin that city, she is surelyrepresented foundat Kuyunjikitself (Fig. 4). A pair of good candidates(BM 1905on some of the terracottas show a woman second-millennium, but presumably 47), unstratified 4-9, 456 = 98950;1932-12-10, who is naked but for a high flat hat and necklace,rests one hand on her right hip, and holds her
THE ISHTAR TEMPLEAT NINEVEH
349
Fig. 2.
"Sanctum sanctorum": Thompson's excavation office at Nebi Yunus. Print from British Museum archive.
Fig. 3.
"Excavations in progress on the site of the Ishtar Temple". Print from British Museum archive.
350
JULIAN READE
J.. 7.~~~~~
'
~~~~~~~~~~~~~~~~~~~4
middke:BM 1932-12-12, 65 = 124299 (pale brown, painted dark brown, ht 5.6; Thompson and Hamilton 1932: P1. LXVIII.13, from NN 6). Early first millennium: Lower right: BM 1904-10-9, 408 = 99375 (pale greenish, ht 5.6). Photograph by D. Collon.
figurinesfrom Kuyunjik,dated by style. Late third millennium:Upper Fig. 4. Terracotta right:BM 1905-4-9,457 = 98951 (pale greenishyellow ware, height 6.3 cm). Second millenBM 1905-4-9, 456 = 98950 middle: 47 (yellow, ht 8); Upper left: BM 1932-12-10, nium: Upper 407 = 99374(pale brown,ht 6.9); Lower (yellowishbrown,ht 11);Lowerleft: BM 1904-10-9,
351
left breast with the other; the two pieces seem to be from the same mould, and on their backs one can see the marks left by a knife that was used to cut away surplus clay before baking. Officially, however, especially in later periods, the goddess of Nineveh tends to be clothed. On the eleventh-century White Obelisk (BM 118807, Fig. 5), she is enthroned in her shrine or bedroom, which was still known by a Hurrian name, the e.nathi (Sollberger 1974: 237-8); she again has a high hat, much like the polos or mural crown worn by Hittite and Hurrian goddesses, a version of which was adopted by Assyrian queens; a long plait or necklace counter-weight hangs behind her back, and she wears a dress, at least as seen in profile. A broken ninth-century glazed tile from the Ishtar Temple shows an armed female wearing a mural crown and a plait, and followed by an attendant; a fragment showing a beard is often illustrated as joining this piece, and one text does praise the beard of Ishtar, but the join looks unlikely, and the figure may even be the queen, wearing weapons like a goddess, rather than the goddess herself (Reade 1987: 139-40, Fig. 1; Livingstone 1989: 18). The evidence for the iconography and identification of avatars of Ishtar in the eighth and seventh centuries has been presented by Collon (2001: 127, 138). Excavations The Ishtar Temple itself (Figs. 6-7) was excavated during the seasons of 1930-1 and 1931-2 (Thompson and Hamilton 1932; Gut 1995: 38-45). What had happened is clear from Thompson's reports. In 1904-5 he had located the outer courtyard of the Nabu Temple. He hoped in 1927-8 to find its library in unexcavated rooms, but they had been flattened to floor level, while to the south-west, where the inner courtyard of the Nabu Temple should have been, there were "some immensely solid and deep piers of cement and stone" of "a late period" (Thompson and Hutchinson 1929: 73). To the south-east, however, below the platform on which the Nabu Temple stood, he found some walls with palace bricks, and an important tablet including a poetic account of wars between Assyrians and Kassites (Thompson and Hutchinson 1929a: 126-33). Hoping that this was a "Palace of Ashurnasirpal", Thompson continued work there in 1929-30. He eventually realised that the walls themselves were much later, but he found a considerable amount of Assyrian material, partly reused and partly in fill. The ancient ground-surface had risen towards the south-east, and it was at a higher level in this direction that he identified the mud-brick foundations of the Ishtar Temple. He cleared its north-eastern end in 1930-1, and the remainder in 1931-2. Some of his Assyrian finds from the "Palace of Ashurnasirpal" probably derived from the Nabu Temple, and some from an early version of the North Palace nearby. Many bricks and
Fig. 5. Goddess and king in Nineveh shrine. Detail from the White Obelisk (Soliberger 1974:238, Fig. 1).
352
JULIAN READE
NINEVEH 1927-1932
North
Palace limits of excavation areaof approximate mudbrick platforms pavements, probably
all post-Assynan '_ *_outer1 court
lc lwell
post-Assynan walls (ePalace of Ashurnasirpal")
\%.TempleE
v R ^sf
B
KNi
F9 n
/
e v i
QRWX room t(Ninevite 5
F sounding
inthutuese qae Ux5(
sshar nempl
E~~~~~~~~~~~~~~~Akkadian Ashumasirpal
N
RRK?
OldBabylonian tabletsfoundN .M
SS)
_ \
Fig.6. Thompson'sexcavationsin and nearthe IshtarTemple.Author'sdrawing,afterGut (1995:II, 18,Abb.5). other terracottafittings,however,the inscriptions on which begin with the word e.gal ("palace"), may reallyhave been used in the IshtarTempleratherthan in a palace.The uncertainty is created by Ashurnasirpal's liberal but inconsistent use of e.gal, before his own name, at the start of the
inscriptions on many objects which were certainly made for temples (Grayson 1991: 359-60, 371-2, 379-85); he never does this on the better-attested stone wall-panels in his Nimrud temples, with one exception for which there is a special reason (Reade 2002a: 189). Since the Assyrians were practical people, probably few of them cared exactly what a brick said. So the Ishtar Temple is the likely source for some of the Assyrian material from the south-eastern quadrant of the area
353
J0
0 ' ;0000 a -I11 a f 11 l >1;A~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~i

.
t e
As
|
~~
I f
olUlf
bE
_____
-3?
70~~~~~~~~~~~~~~~~~~~~~~
\.S. ........
t~~-= 4 ". tr;
EE
-W
~~~~
7H
J,$
A101~
X
~
w2',,,
e
:
3'
1. A
' ' . 't
'gT1
>
354
JULIAN READE
+4
Phase 30 walls, pits and other remains
+2 0 2
4
-2
~~~~~~t
-
debris including tablets (Phase 7) and wall-pegs (Phase 20) pits, wash "Io' and // -; / I0 Cellar 3
"
fill including Phase 5 jars and figurines \ \ \'\\--'zt_\ Outer Courtyard: Laid mudbricks (Phase 20 or 27?)/// // / Phase 7 wall and
-6 -8 _8 3C
_ Phase 7 foundations
-10
-1 2 -12
-6 -18
Phase 3 (Ninevite 'ars?) 5 (Ninevite jars?) 5

Phase 2
10 -
(Phase 0?) ~~~~~~~~Pavem-ent 11 (Phase 9?) ~~~~~Pavement Ill(Phase 7) ~~~~~~~Pavement
foundations
_ scales (1 0 feet
0 5 10
in feet = c. 305 cm)
TT, west
corner
TT, 00
corner
00,
EE corner
of sectionfrom west to east throughSquaresTT-OO, Fig. 8a. Schematicreconstruction nearbyfeatures.Author'sdrawing. incorporating excavated in 1927-8, for much more of what was found during the "Palace of Ashurnasirpal" season of 1929-30, and for most of what was found in 1930-1 and 1931-2. In 1927-8 Thompson had used a radial system of planning; in 1929-32 the site was divided into 50-foot (15.24 m) squares which were assigned letters (with double letters in 1931-2) and referred to by Thompson as Sections or Sects, hereafter as Squares. The depth of architectural features and of objects was often recorded, in feet above or below datum, and objects often had their square and depth marked on them in pencil, e.g. NN 5 representing 5 feet below datum in Square NN (5 feet above datum would be NN + 5); many of the original marks are legible on objects at the British Museum. Because imperial measurements in feet and inches pervade the most important plan, Fig. 7, they are retained below, usually with a metric equivalent which is often marginally too high (1' [1 foot] = 12" [12 inches] = c. 30.5 cm). The quality of field recording was variable, as suggested by Fig. 3 and described by Mallowan (1977: 71-5), and the remains must have been bewildering, because parts of the temple had been cleared down to or below original floor-levels during ancient reconstructions, post-imperial times and early excavations. Absolute depths of structures and objects were often unrelated to age. Thompson proposed relationships between the architectural remains of the temple and the written evidence. He tended to rely on types and sizes of brick as dating criteria, but does not cite sufficient detail; the dimensions of baked bricks were found to be of little value for dating at Ashur (Haller and Andrae 1955: 8-9). A wider range of evidence needs to be considered. The data recorded by Thompson and his colleagues do make this possible. The history of the temple and its vicinity has been divided below into thirty phases, some hypothetical, and the recorded archaeological evidence for the most important developments is summarized in Fig. 8a-b. The early temples (Phases 1-5) Phase 1: Pre- and proto-history. The historical temple was located on or near the top of a mound that was already over 20 m high by the early third millennium BC. Gut (1995) has provided an authoritative analysis of the prehistoric sequence, using the records of Mallowan's Deep Sounding in 1931-2. In Square MM outside the temple, occupation began before 6000 BC. In Squares QRWX, vertically under the later temple, levels of Late or Post-Uruk occupation were reached but not widely investigated (Gut 1995: 40). The final Phase 1 building was burnt, and Phase 2 was to be founded on a layer of ash 1 foot (31 cm) thick. Clay sealings from the Early and Late Uruk periods, and early tablets, together with seals and sealings from the early and mid-third
355
+4 +4 +2
Phase wall~~~ 29 Phase 29 wall
\ ebris ol l
-wall-panel, <pavem ent
Phase 20:l InnerCourtyardlevel

(Phases 20-28)
__
Phase 28
\\
>pavenwa
= = M =
o
-2 -4 -6__ 3 A
; C
EE
Phase 30: intrusive walls, pits and other remains
-8
-10 -1 2
-14
Phase 7 foundations 10 _
Copper head and "mud platform" a spearhead (Phases 5-7) south-western wall of QRWX room (Ninevite 5 shrine?) CO3 C:3
l_ O
5_ -18 (1 0 feet = c. 305 cm)
CPhase
-20 0 X . 5 [ 10| 0, centre 0, S corner
Phase S, W corner
2 ~~~~~~~~~~~~~~~~~~Phase 1 (bevelled-rimbowls) W, centre
of sectionfrom west to east throughSquaresO-W, Fig. 8h. Schematicreconstruction incorporating nearbyfeatures.Author'sdrawing. millennium (references: Reade 2000: 395; Collon 2003), although poorly provenanced, repeatedly indicate that this area of Nineveh housed some kind of administrative centre. Phases 2-3. Early/mniddle and midcdle/late thirdl millennium.A rectangular room in Squares QRWX had walls 4.5 feet (1.4 m) thick, made of mud brick on stone foundations (Thompson and Hamilton 1932: 61-3, P1.XLVII.1-2). Thompson estimated its original internal dimensions as about 78 by 25 feet (24 by 7.7 m), which allows for a door on one side. He recorded that the room had been arched: this implies a pitched-brick vault, which would give the building a respectable place in architectural history (Fig. 9). The foundation was 20 feet (6.2 m) below datum; the walls were preserved to a maximum height of 10 or 11 feet (3.1-3.4 m), so their tops were 2.8 or 3.1 m below datum. A revetment with stone foundations on part of the exterior, 15.5 feet (4.6 m) below datum, had three layers of ash underneath it, and is the reason for treating the QRWX room as standing through two phases rather than one. In his final assessment of the site, Thompson (1934: 98) identified the QRWX room as the Ishtar Temple built by Shamshi-Adad I. He did so because of its substantial nature, and because he had dated the Phase 7 temple (attributed below to Shamshi-Adad I) to Ashur-resh-ishi I, a much later king. Gut (1995: 40) suggested that the QRWX room was probably post-Ninevite 5, but there was Late or Post-Uruk pottery at about foundation level inside and outside the room. We are not told which items were found above floor-level inside the room (of which only the north-western end was "in good condition"), but Ninevite 5 pottery was found in parts of Squares QWX at corresponding depths, albeit some of it was mixed with Greco-Parthian material (Gut 1995: 39-40, 150-1). From Square W, at respectively 8 and 18 feet below datum (2.5 and 5.5 m), there were an alabaster jar of the early or middle third millennium (Thompson and Hamilton 1932: P1. LXIV.6), and a cylinder-seal, which Dominique Collon thought might be earlier than Ninevite 5 (ibid.: P1. LXIII.10 = BM 1932-12-10, 64 = 123348; Wiseman 1962: P1.2b). All this suggests that Phase 2 was Post-Uruk or Ninevite 5, and that Phase 3 was Ninevite 5 or somewhat later, with the possibility that the building lasted into the Agade period. Gut (1995: 150-1) commented on the unusual nature of some of the Late or Post-Uruk and Ninevite 5 pottery from Square W. Miniature jars were exceptionally abundant in Squares QX; there were further alabaster jars associated with Uruk-Ninevite 5 pottery in Square H (Thompson and Hutchinson 1931: 105-6, P1.XXI.l; Thompson and Hamilton 1932: 85-6, P1. LXIV.2). These vessels would have been suitable as containers for scented oils, the use of which, for cultic and other purposes, is attested in many periods (e.g. Joannes 1993). Even if Thompson missed mud-
356
JULIAN READE
1_
-~~~~~~~~~~~~~~~~~~T
Fig. 9. Room in SquaresQRWX:Possiblythe Ninevite5 shrineof Ishtar(Thompsonand Hamilton1932:P1.XLVII.1).Printfrom BritishMuseumarchive.
brick fittings or partitions in the interior of the QRWX room, it is difficult to imagine what such a large structure was, if not a shrine. Its door could have been in the south-western wall, so that someone entering from an open space to the south-west would have turned left towards the altar; this is the "bent-axis" design, standard for third-millennium shrines in northern Iraq. The groundplans of later phases of the Ishtar Temple, while destroyed above the QRWX room itself by pitting and Greco-Parthian walls, still point towards this area as the appropriate position for the inner shrine throughout the history of the temple. A much smaller room of mud brick on stone foundations in Squares TTWW was founded 16 feet (4.9 m) below datum (Thompson and Hamilton 1932: 61-2), but "the floor was approximately - 11 feet (?)", i.e. 3.4 m (?) below datum, with a drain of baked bricks beneath it; this suggests that the floor seen by Thompson was secondary, and that this room too had two phases of occupation. Its walls were preserved up to a height of 7.5 or 8 feet (2.3 or 2.5 m), so their tops were around 2.5 m below datum. "In" this TTWW room there were at least three Ninevite 5 vessels "and ... early cuneiform clay tablets"; Gut (1995: 44) suggested that the TTWW room was later than Ninevite 5, but the tablets can be discounted because they were actually found above the room, not inside it (see below); the baked bricks in the drains, while noted by Thompson as being the same size as some of Ashur-resh-ishi I, could equally well have been earlier. The building to which the TTWW room must have belonged was partly underneath the foundations of the Phase 7 temple; the room was roughly aligned with the QRWX room, but about 80 m away, and there is no reason to link it directly with the cult in Phases 2-3. The tops of other walls, seen in Squares UV, may be Ninevite 5 too (Gut 1995: 40-1, Abb. 21). before Manishtushu.This is a hypothetical structure, presumed to Phalse4. Late third mwillenniumw, have been destroyed when Phase 5 was built. Alternatively, if the QRWX room itself remained in
357
use for many centuries,Phases 3 and 4 may be identical.Phase 4 may also be represented by some of the objectsdiscussedtogetherwith Phase 5 materialbelow.
Phase 5: Late third millennium,from Manishtushu onwards. According to fragmentary foundation
I (Grayson1987:51-5), Manishtushu, documentsof Shamshi-Adad son of Sargon,king of Agade, built or restoredthe templenamedEmenue,withinthe precinct(qaqqar) of Emashmash, together with the ziggurratof Ishtar.Clearlythe king of Agade, a city which housed its own avatarof the goddess, already acknowledgedor wished to promote the importanceof the Nineveh shrine. Manishtushu's documents(nare and temmeni)were found and then replacedby Shamshi-Adad, but not one of them has yet been identified.They were not necessarilynumerous,and perhaps some are still encased in brickwork.Their total absence, however, from such an extensively excavatedsite, would be remarkable wereit not that no reference to Manishtushu, whomShamshiAdad and probablyAshur-uballit I mentionedwith respect,has yet appearedin the IshtarTemple inscriptionsof ShalmaneserI or any later kings. Yet at least some of them surely knew of Manishtushufrom the inscriptionsof Shamshi-Adad. Perhapsthey discardedthis information, togetherwith any originalinscriptionsthey happenedto find, in order not to commemorate the unwelcomefact that Nineveh was once ruledfrom Agade (cf. J. Westenholz2004/5). Fragmentsof at least two Agade foundationdocumentshave indeed been found at Nineveh, but they were not associatedwith the fragmentsof Shamshi-Adad's foundationdocuments,and they cannotbe attributed to Manishtushu, sincethey incorporate wordingonly used in inscriptions of the later kings Naram-Sinand Sharkalisharri (A. Westenholz2000: 548-52). Five fragments, probablybelongingto a single foundationdocumentmade of a distinctivestone, slightlyrubbed down in antiquity,were found in the Nabu Templearea;the text was apparently a dedicationto Enlil, as if the documentoriginatedat Nippurand only reachedKuyunjikin the seventhcentury. A fragmentof anothersuch documentwith similarphraseologywas purchasedat Nebi Yunus. No remainsof Agade architecture have been identified,though mud-brickfoundationsof both temple and ziggurratcould have been partly incorporatedinto the Phase 7 temple. Late Agade activitiesnearbyare reflectedin a seal-impression, from SquareNN, which apparentlybore the name of an Agade court official(Gut et al. 2001: 82, catalogueNo. 120). Afterwards,in the Ur III period,Shaushaof Ninevehis mentionedin a text of Shulgi,and was the personalgoddessof a wife of Shu-Sin;anothertext refersto Tishatalof Nineveh,probablya king (references: Veenhof 2000). This, togetherwith the shrine-name e.nathi,constitutesthe primaryevidencethat Nineveh was by then ruledby Hurrians,with Shaush(k)a as principalgoddess. Phases4-5: Objects.Therewas materialof the late third millenniumunderneath at least part of the Phase 7 temple, to judge by a descriptionof pottery from a tunnel dug below the templein SquareH, and by some late third-millennium beads from below foundationlevel in SquareTT (Thompson,quoted by Gut 1995:32, 44). Many typicalpotsherds,and about 10,000beads and a few amuletsof the late thirdmillennium, includingone of lapis lazulishowinga womanin Early Dynastic III dress, were mainly found scatteredwith other materialdown a slope north-westof SquaresGH (Thompson and Hutchinson 1931:82, 107-8, P1.XXXVIII.7; Gut 1995: 27; Gut et al. 2001:76-8, 98-100). How and when the materialreachedthis position is unknown.It could have done so, for instance,during a violent episode, or duringreconstruction work. If it came from the shrine,it presumably emergedthrougha gatewaysituatedin SquareI or J at the top of the slope, not far from the entranceinto the much later Phase 20 temple. Objects of the late third millenniumin the British Museum that can be associatedwith the temple have mostly been studied in detail (Gut et al. 2001). They include pottery marked as coming from NN 7, PP 3-6, and SS 3 (ibid.: catalogueNos. 126-30, 133, 137). There were two beakersfrom underthe "3rdpavement"in Square00, one of them being inside a "largedouble pot", and this latterterm might referto an unprovenanced pair of largejars, both of which had lost theirrimsas if jammedtogether(ibid.: catalogueNos. 136, 138, 148-9). Additionalterracotta models have since been located (cf. ibid.: 86). Among them are a hollow quadrupedmarkedas comingfromNN 3 and a cartfromNN 2 (BM 1932-12-10, 689,694 = 137018,137023= Thompson and Hamilton1932:Pls. LXVII.6,LXVIII.25),and a cart from 00 5 and quadrupeds from00 3
358
JULIAN READE
1; 1932-12-10, 687-8= 137016-7= ibid.: and PP 4 (respectively BM 1932-12-12, 944 = 13841 P1.LXVII.30-31); there was anotherquadrupedfrom PP 5 (ibid.: P1.LXVII.28).The two carts waretypicalof the late thirdmillennium, are madeof pale greenish-yellow and the PP 4 quadruped and the carts are of types which belong in this period (cf. Bollweg 1999: 127-8, Abb. 119-21). or Pavement are all statedto havebeenunderthe 3rdpavement The cartsand the 00 3 quadruped III. This can be none other than PavementIII of the Phase 7 courtyard,and most of the other third-millennium ceramicsjust cited were (or in NN could have been) below the level of this pavement in the same vicinity, as suggested in Fig. 8a. Despite the lack of sure stratification (because Thompson does not seem to have actually removedpavementtiles, so that "under" the presence scopefor intrusions), merelymeans"at a deeperabsolutelevelthan",withconsequent here of all this third-millennium materialmust demonstratethat PavementIII, which is one of the pavementsdescribedas resting on "plain earth" (Thompson and Hamilton 1932: 67) or "rubbishfilling"(Mallowan,apudGut et al. 2001: 99), was laid directlyeither on levels of the The "doublepot" with a late thirdmillenniumor on debrisof this date broughtfrom elsewhere. vesselinsideit soundsas if it camefrom occupationdebrisratherthanfill. Most of thesecourtyard objectsprobablybelong in Phase 5 ratherthan Phase4. = 99372:Searight A vesselof mottledstone,probablymadein southernIraq(BM 1904-10-9,405 et al.: in preparation), may once have been dedicatedin the temple;a fragmentof it, bearinga Phase 4 or 5 inscription,was excavatedby King in his Shaft 50, probablyin this vicinity.There as coming is also a worn male head of an EarlyDynasticstatuein the BritishMuseum,registered from Kuyunjik,but the provenanceis doubtful since it belongs to a collectionwhich included much materialfrom southernIraq (BM 1856-9-3,353 = 115032: Reade 2000a:85).
Phases 5-7: The so-called Head of Sargon, and an associated spear-head
context of these objectshas been unclear(Gut 1995:38-9; Gut et al.: 2001: The archaeological 78-9). Mallowan (1936: 110) supposed that the famous head (Fig. 10; Iraq Museum 11331), might representSargon of Agade himself. Moortgat (1969: 51), without committinghimself, built the temple,he is a pointed to parallelswith Naram-Sin'siconography.Since Manishtushu defaced:the ears and beardhad been cut, seriouscandidatetoo. The head had been deliberately andone eye gougedout whilebotheyes hadlost theirinlay;the bodywas missing. the nose battered, The spear-head (Fig. 11) has attracted little attention since its original publication 55 = 123343: Thompsonand Hamilton1932:72;Thompsonand Mallowan1933: (BM 1932-12-10, with a raisedrectanguas copper.The bladeis decorated P1.LXXVIII.42).Its metalwas identified definesthe line at whichthe folded lar panelalong the midribon both sides,and a neat depression sides of the socket meet to form a hollow cylinder.A band of sheet silver,whichencirclesthe end of the socket, is now split and cracked,but seems once to have been solderedinto the shapeof a nail, which passes throughthe silver single tight ring. The band is securedby a copper/bronze into the socket and was bent inwards,evidently before or during the insertionof the shaft. were used duringthe first half of the second millenniumin Iran and Iraq Socketedspear-heads but some examplesfrom (Moorey 1971: 88; Curtis 1983: 77), as they were in Syria-Palestine, belongtowardsthe end of EarlyBronzeAge IV (Tubb 1985:194), whichprobably Syria-Palestine suggeststhat overlapswith the end of the Agade empire.So the shape of the Ninevehspear-head later than the head. it was probablybut not necessarily The silver band implies that the spear-headwas for ceremonialuse. So does an inscription which is incised on the exteriorof the socket, beginningat the shaft end. In 1932 Thompson describedit as "writtenin linear archaisticcuneiform(perhapsput on at a late date)" and he read it as "B;t-'Nin-fil", suggesting that the spear-headwas dedicated to the temple by This readingmust have been made when the metal was still coveredby corrosion. Ashurbanipal. publishedin 1933, which was It does not fit the copy, visible on the drawingof the spear-head clearlymade aftercleaning.Whereasthe silverwas protectedin the earthby the copperactingas a sacrificialanode and is in relativelygood condition, the surfaceof the copper now has the tank;aroundthe inscription itself corrosion rippledlook of metal left too long in a conservator's productsremainor have returned,the metal is distorted,and thereare modernscratcheson the although copy looks morelike e.gal ilu-su-ma, "Palaceof Ilushuma", signs.Thompson'spublished
THE ISHTARTEMPLEAT NINEVEH
359
_;r
__~~~~~~~~~~~~~~~~~~~*
I~
r headof an Agadeking(IraqMuseum 11331). Fig.10. Copper
the Old Assyrianking of this name was not, so far as we know, connectedwith Nineveh. Irving Finkel, to whom I am indebtedfor much adviceduringthe writingof this paper,considersthat the text, ratherthan being archaistic,is writtenin authenticdedicatoryscript of the late third millenniumor the first half of the second, and that, while only the first sign is entirelycertain, in that the shape and most what can be seen today is consonantwith the readinge'.galilu-su-ma, durng te survive rign 1ofManishtushuof(Andrae 193:i,n 1;ra bb Grasonm 1987:1). the traces of Ashu that seem to fit with this and the "traces"that interpretation, unambiguously the least, a plausible do not fit could be due to the disturbance. Thus it seems to at of thinae Noe first-hiand accont discvr nofthe head andsperha knownonce is wto mie:verhapsvin him, quite "House of the It is not a Sumerian such as e. name, rgasan.kalam.mal, reading. probably i temple emergtentc nheitx,rther Thopsn benor Hamchitonwitesedthen of the adremovaloysrp objetsero ther 1993: or the Ishtar at Erbil was called of the which is what Land", 90), (George Lady Temple nl e and statesie mlearth. Mlowa (1936 105imst havfoft senqurd ,oee, thairt "the hseadiel w erasiond e. Fdu6kalam.mal,restoredas the name of an unlocatedIshtarshrine(ibid.: 76). It may perhaps be relevant,however, that a poker-spearor spear-buttwas dedicatedin the Ishtar Temple at
360
JULIAN READE
.4
,~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Inscribed Fig. I 1. Top: spear-head (length26.7 cm) fromthe IshtarTemple(BM 1932-12-10, 55 = 123343).Drawnby Ann Searightand IrvingFinkel. Bottom:Detail of the inscription. Photograph by D. Collon.
lying loose in the soil on an Assyrianmud platformwithin the limits of the temple of Ishtar", and that the spear-headwas "adjacentto it". No other comparableitems were recorded,and these two weresurelyin the same stratigraphic context. Becausethe inscriptionon the spear-head was once thought to be late, we have failed to acknowledgethe significanceof its provenance, which affects that of the head too. It is given as W 7 in the caption of the 1933 drawing,i.e. SquareW at 7 feet (2.2 m) below datum. The words "copperhead" are writtenon the plan, in SquareW, over an irregular white area which should thereforebe a "mudplatform".The top of a similarwhite areajust to the north-eastis also givenas 7 feet (2.2 m) below datum.Immediately to the north there was the top of the Phase 2-3 QRWX room, the highest point of which was 2.8 m below datum, and immediatelyto the south, in SquareV, there was the top of another mud-brick wall, of Phase2 or 3, at 14 feet (4.3 m) below datum.The rest of the surrounding area is markedwith dots which seem to representthe bottom of the excavation,without identifiable features.The earliestNeo-Assyriantemplefloor (Phase 20) was about 31 cm abovedatum in the adjoiningSquare R. So this is why the head and spear-headwere protected,as their state of preservationindicates,from the fire which burnedin parts of the temple in 612 BC: they were alreadymore than 2 m deep. It is irrelevant that no tracesof the Neo-Assyrianfloors, and of the Phase 7 mud-brick foundations,survivedin SquareW itself, and that we do not know theirexact thicknessthere:they are presentfurtherwest in SquaresRS, but here the walls and floor of the
361
temple had been removedby people who mined the area for clay or valuables.Their trenches as suggestedin Fig. 8b. must have stoppedjust short of the head and spear-head, The simple explanationfor the depth, condition and associationof both objects is therefore that the head belongedto a statue that was mutilatedduringor not long after the collapseof the Agade empire,in a periodwhen it was alreadycommon practiceto vandalizethe monumentsof had been displayedand perhapseventuallyburied disgracedrulers,and that it and the spear-head in front of the shrineof the Phase 5-6 temple.They were then left unnoticedin the debriswhen the Phase 7 temple came to be built, or were consciouslydepositedat foundationlevel in that period. Originallyone or both of them could either have been made for dedicationat Nineveh, or have arrivedthere as booty, possibly even as booty from Ashur. Similarly,the collection of war booty would account for the reputedemergenceat Basetki, furthernorth, of a Naram-Sin monumentthe inscriptionon which implieserectionat Agade (Frayne 1993: 113-14). Suggestionsby Mallowan(1936: 105-6, 110; 1977:78) that the head might alreadyhave been in the temple in the Agade period, but that it was still availableto inspire one of Sargon II's sculptorsin the eighth century,were not supportedby evidence.Thompson was more logical. Suspectingthat the inscriptionon the spear-headwas added "at a late date", he proposedthat the head had belongedto one of thirty-tworoyalstatueswhichwerebroughtfrom Elamin 647 BC, some of which Ashurbanipal defaced and displayedin the South-WestPalace (Thompson and Hamilton 1932:72; Borger1996:54-5, 241). Therewould then be many possibilitiesfor the date at which the head was defaced. It might have been done when the Agade empire collapsed in Elam;or whenthe Elamitesdeportedmonuments (includingthe Naram-Sinstela)fromBabylonia; or when the thirty-twostatueswere deported,if the Assyriansresponsible thoughtthat it showed an ancientElamiteking. Ashurbanipal might then have buriedthe head and spear-head, respectfully or triumphantly, deep below the floor of the temple,with the resultthat they wereprotected when Nineveh was capturedin 612 BC. If the head was defacedin 612, as proposedby Mallowan and by Nylander(1980), it would be hard to account for its provenance,survivaland apparent The earlierdate for the inscriptionon the spear-head,however, associationwith the spear-head. eliminatesthe need for all these speculations. Another factor mentionedby Thompsonis that the head and spear-head were found "not far from the largest part of Shamshi-Adad's cylinder".This was at much the same depth, but the plan indicatesthat it was some 10 m or moredistant.Likeothermonuments visiblein the Assyrian capital cities when they were captured,it seems to have been smashedand then exposed to fire. Otherfragments of Shamshi-Adad's foundationdocumentsin the BritishMuseumappearto have sufferedin similarways, as if they had all been found by a king repairingthe templeand left on display ratherthan reburied,though they could also have been found and damagedby GrecoParthians.The copperhead and the spear-head had had a differenthistory.
The earlier second-millenniumtemples (Phases 6-7)
I. No structuralremainsof this period have been identified,but Phase 6: Before Shamshi-Adad the templeobviouslycontinuedto exist and was probablyrepaired fromtime to time.The political situation before Phase 7, as first indicated by David Oates (1968: 31), is that Nineveh was subordinateto another town, Nurrugum.The relationshipbetween Nineveh and Nurrugumis comparablewith that between Ashur, which like Nineveh was to survive and flourish, and whichlike Nurrugum Ekallatum, remainsunlocated.A possibleexplanationis that Amoriterulers with a pastoralistbackground enclosedtheir settlements with walls insidewhich both people and stock could be accommodatedin emergency; examplesare Khamira,Shibba, Rimah and Usqa near Tel'afar.If the same applies here, a possible candidatefor Nurrugummay even be Nebi Yunus, a mere 1.5 km from Kuyunjik;the Nebi Yunus mound covers some 15 hectares,there is space for a walledarea aroundit, and Neo-Assyrianstructures will have cloakedanythingearlier that may exist. In any event the Ishtar Temple, while it may have been of little interest to newcomers, maintained its statusas a cult-centre evenwhenNinevehitselfwas not a powerfulcity. At Maniin thisperiodthe nameof Nineveh, otherwise ni/ne-nu-(wa-)a, iS liableto appear as ni-ne-et (Durand 1987:224; Wu Yuhong 1994;Ziegler2004/5). The feminineform remindsone of Ishtar,but Finkel has pointed out that the readingmight be ni-ne-a~, given the variableways
362
JULIAN READE
in which place-nameswere sometimesrendered.It is not surprisingthat only one referenceto Nineveh is meanwhileknown from Kultepe,since there is no reasonto supposethat Ashur and Nineveh then had close relations,and the main caravanroutes from Ashur to Turkeywere far away, on the otherside of the Tigris;but the goddessis mentionedat Mariand Rimah(references: Dalley 2001: 156). his I. This king capturedNurrugumin the eponymy of Ashur-malik, Phase 7: Shamshi-Adad twenty-ninthyear of reign;furtherdetails about the war were providedby Nele Zieglerduring records the 2003 Rencontrein London, from fresh Mari letters(Ziegler2004/5). Shamshi-Adad rebuildingthe templeand the ziggurratof Ishtar(Grayson 1987:51-5). Unless he appropriated work begun by a rulerof Nurrugum,he acted with great speed, since he died in his thirty-third year. He visited Nineveh in person, and one of his survivinglettersordersthat wood should be for use in the temple(Dossin 1950:34-5, 41). sent there, presumably Previouskings of Ashur and Eshnunna,in various ways, had alludedto their aspirationsas family heirs of Agade (e.g. Veenhof2003: 44), and it has been suggestedthat Shamshi-Adad's had ancestralconnectionswith Agade (Charpinand Durand 1997: 372, n. 36). Shamshi-Adad temple,he emulatedthe called himself LUGAL KIS, and at Nineveh, in rebuildingManishtushu's earlieremperor.The design of the building,describedbelow, is based on southernratherthan foundationdocuments,in shape and practice.Even Shamshi-Adad's northernIraqiarchitectural besidewhichthey weredeposited(Reade 2000b). quality,may have copied those of Manishtushu also adoptedthe imperialart of Agade,examplesof whichsurvivedin centralIraq Shamshi-Adad and elsewhere(e.g. Cole and Machinist 1998: xxiii). Rainer Boehmer(Gut et al. 2001: 90-2, carvedstone pedestalfrom Kuyunjikas an Old 121-6, Taf. 2-11) has identifieda magnificently Babylonianpiece based on the style of Naram-Sin;while later reusedas a door-socket,it was of this surely made for the IshtarTemple.Boehmerhas also recognisedAssyriancylinder-seals time which are imitationsof the Agade style (Reade 1973:171-2). the Ishtar Temple is also The significancewhich Shamshi-Adadattached to reconstructing evident from his statementthat "not one of the kings who precededme had rebuiltthat temple, from the Aulum (floruitor fall) of Agade until my rule, until the captureof Nurrugum,a length of seven daru".Figuresof sixty or seventyyears have been suggestedfor ddruas "generation", multipledby seven to providea plausiblelength for the intervalbetweenthe Agade period and but sixteenor twenty years is a betterlength for a generation.We know from a Shamshi-Adad, copy of the Sumerianking-listfound at Leilan (Vincente 1995) that this documentwas present may have used reign,and his calculation not long afterShamshi-Adad's in northernMesopotamia the list with Amoritegenealogiescannot have been easy. One wonderswhether it, but correlating it is merecoincidencethat the intervalbetweenthe accessionof SargonI of Ashurand ShamshiAdad's conquestof Nurrugumseemsto be one hundredand forty years,a neat multipleof seven failed to distinguishbetweenthe two Sargons.It may be and twenty,and whetherShamshi-Adad much easierfor us to do so than it was then. own foundationdocumentshas been found in position,we Althoughnone of Shamshi-Adad's have three fragmentsfrom King and seventeenfrom Thompson (Thompson and Hutchinson by Borger1961:12;Thompsonand Hamilton1932:105-7, Pls. LXXXI1931:PI.XX.50, identified besidesone from Rassam'sexcavationsand one in Yale. Thompson's LXXXIV;see Addendum), side of the temple,and so probablydid come mostly from disturbedareas on the north-eastern King's, since Thompsonmentions in a notebook that his own 1930-1 excavationsin this area had intersectedthose of 1904. Moreover there is a join, though the break itself was ancient, between the two groups of fragments(BM 1904-10-9,365 = 99332 joining BM 1932-12-10,7a, from SquareR). A reportby King impliesa differentand much deeperfindspotfor his material field-noteonly (D'Andrea 1981:112;Reade 2000:407), but what appearsto be his corresponding inscribedvessel fragment(Phase 4-5). refersto the third-millennium remainsof Phase 7 were originallyascribedby Thompsonto Ashur-reshThe archaeological ishi I (Thompson and Hamilton 1932:63-5). He even wonderedwhetherPavementIII, in the II, one of whose brickswas courtyardof the building,mightactuallybe later than Ashurnasirpal found "under"it (ibid.: 67), but his notebook mentionsthat the brick was "two or three feet
363
from edge" of the pavement,and that Mallowan,his colleagueat the time, insistedthat its depth had no chronologicalsignificance.The ascriptionof Phase 7 to Shamshi-AdadI rests on an accumulation of circumstantial evidence,includingsome Old Babyloniantablets. The core of the Phase7 templeis whatThompsontermedthe "mainfoundation" or "platform"; it may also incorporateother phases becauseof reuse and repairs.As described(ibid.: 58, 64-5, 76, rearranged), it extended"overan area of some 300 by 150 feet" (91.4 x 45.7 m), but the plan shows that neitherlength nor width was fully excavated,and that the length was at least 106m. Its top, as preserved,was generally"at + 1 foot [31 cm] to a little below datum-level", while its base was "on approximately-5 to -6 feet [1.5 to 1.8 m]. It is usuallyabout 6 feet [ 1.8m] thick, of 18-20 courses of tawny or reddish unburnt brick". The bricks were 33-6 cm square, and 10-12 cm thick. In SquareT the platform"variedcuriously,between4 to 18 courses,doubtless of the ground".Its base may have been 3.1 m below datum in owing to the originalirregularity part of SquaresRQ, if it is "the libn [mud-brick] of Anp" (Ashurnasirpal)which "descends10 feet" accordingto one of Thompson'sfield-notes(Gut 1995:40), and the plan gives depthsof 7 to 10 feet (2.2 to 3.1 m) below datum in SquaresUV and PPUU. "The centralpart of the main foundationof unburntbrick in DD" was a "deep mass of libn, coveringan area approximately of 50 x 35 feet" (15.4 x 10.8m), which "goes down for about 19 feet [5.8 m] from its top, the lowest coursebeing on 13 feet [4 m] below datum(some 7 feet [2.2 m] below the level of the main foundation), and the top 6 feet [1.8 m] above". There was "no obvious line of demarcation" betweenthe DD mass and the remainderof the foundation,and both were "the same kind of construction"; perhapsthe formerhad been the base of the Phase 5 ziggurrat. There was no mud brick under the courtyard occupying parts of Squares NNOOPP and RRSSTT, which was paved with tiles, c. 38.52 by 5.7 cm, laid on earth (Pavement III). This pavementwas either laid after the base of the nearby temple walls had been built, or was an earlierfeature that was retainedand adapted, since a drain ran out from it to the north-west acrossthe main foundation(ibid.: 58, 67, PI.XLVII.3-4). The relationship betweenPavementIII and the drain is not clear in illustrations,but Thompson regardedthem as contemporary with one another,and the draincertainlylooks earlierthan the subsequent PavementII. So the surface of PavementIII can be regardedas a Phase 7 floor-level.An annotatedphotographconfirmsit as 2.5 feet (77 cm) below datum, while the upper end of the drain was 3.5 feet (108 cm) below datum. Sincepartsof the top of the main foundation,as preserved, werearoundone metrehigherthan the surfaceof PavementIII, they were effectivelystumpsof the originalwalls. This also applies to the brickworkprojectingstill higher in DD: what little there was of it does not look more substantialthan a wall-stub(Thompson and Hamilton 1932:P1.XLVI.2), but the floor-levelon the north-eastern side of the templemay alwayshave been higher.The base of much of the main foundation was about 1 m lower than the surface of Pavement III. Ground-leveloutside the templeto the south-westmay have been level with the base. On the south-eastthe groundseems to have been slopingaway, sincethe base is deeper,and on the north-west, in all historicalperiods, the grounddid slope from the templedownwards(sectionsin Gut 1995:43, 45). Insidethe foundationtherewere "sevencuriousbuilt 'cellars',obviouslybuilt at the same time as the foundation"(Thompson and Hamilton 1932: 66, P1.L.3). The floors of the cellars are white on the plan, as if the mud-brickfoundationextendedbeneaththem;this does not seem to have been so, however,becausea photographshows earth underCellar 7, while the interiorof the QRWX room is also white althoughthereis no mentionof its havinghad a mud-brick floor, and the excavationsin Cellar3 in SquareTT went deeperthan the base of the foundation(unless this was merelybecausethere was a pit in TT: ibid.: P1.LXIX.36). The cellarswere rectangles, like rooms withoutdoors. It is plain today that most of themwerespaces,filledwith earth,below the floors of rooms:similarlythe walls of the rooms in the Rimah templecontinuedbelow floorlevel, while the plans of the early Ashur, Anu-Adad and Sin-Shamashtemples at Ashur were largely deducedfrom spaces of this kind below missing floors (Heinrich 1982:II, Abb. 290-1, 317, 323-4). The brickwork of Cellars1-3 and 5-7 was steppedout, upwards,so that they were slightlywiderat the top, like those in the Ashur Temple(Haller and Andrae 1955:24, Taf. 34a). If the tops of Cellars 1-3 and 5-6 representroom-plans,then the externalwalls of the Phase 7
364
JULIAN READE
templewereabout 4.5 m thick,and the internalwallsabout 3 m thick,while the roomsthemselves In SquareNN the tended to be about 4.5 m wide, but there are variationsand uncertainties. position and plan of Cellar4 are anomalous,suggestingthat its date (in this form) is different it is under from that of Cellars3 and 5 on either side. Cellar7 in SquareT is also exceptional: 2 m wide, as if it underlaya corridorratherthan an ordinaryroom, and Thompsonthoughtthat device, it might possiblyhave had an entrancein its southerncorner.Perhapsit was a structural it, like the spacesinside mass surrounding intendedto accommodatedistortionin the mud-brick the ziggurratat Rimah (Oates 1967:85); this does not explainthe steppedwalls. Thompsondid not commenton what appearsto have been a comparablebut smallerfeaturewithout stepped walls in Square I, markedon the plan "sunk - 10 [feet]", i.e. 3.1 m, but he wonderedif there might be part of another cellar between SquaresT and DD, where his plan shows brickwork apparentlystepping upwardsto the south. There was more stepped brickworkin Square EE, describedas "the apparent'stairs"'(Thompsonand Hamilton 1932:70). The materialof which the mud brickfor the foundationswas made soundslike the fossil clay, abundantnear Nineveh, which would have been suitablefor a new templebuilt in the southern manner,without the typicallynorthernstone footings seen in Phases 2-3. Thompsonwondered if the central part of the foundation in DD might have been earlier than the remainder.An whetheror not some Phase5 brickwork in the structure, explanationfor some of the irregularities survives,could be that the Phase 7 templewas built as one operationbut in two stages. If so, in area of the templewas built, with the mud brick extendingnot the first stage the north-eastern only underneaththe rooms but also underneathcourtyardspaces;part of its frontagethen is courtyard.In the reflectedby the course of the drainin the northerncornerof the south-western area; but here the second stage the temple was enlargedby the addition of the south-western courtyardpavementwas laid, more quickly,not on mud brickbut on earth. Some irregularities may have resultedfrom repairsafter Phase 7, however,since therewere later kings who claimed of the temple;they mentionfindingthe recordsof earlierbuilders, to have renewedthe foundations in higherwalls. althoughthese may also have been encountered building.A courtyardat its What survivesof the Phase 7 plan suggestsa basicallyrectangular end was flankedby a single rangeof rooms, survivingas the cellarson north-west south-western and south-west,with tracesof a third range on the south-east.These definethe minimumwidth of the buildingas roughly 55 m; its length was at least 106m but possibly very little more. A UUVV, couldbe the foundation range,in Squares on the outerfaceof the south-western projection of a pair of towers flankingan outer gateway through(above) Cellar 1. This contained"in its I and Sennacherib" (ibid.: 66); a upper earth squaredstone blocks and bricks of Shalmaneser I. The brick bks". notebook specifies"full of squaredblocks, and had severalSenn. and Shalm. in later as a periods but thesefindssuit its identification gate-chamber are unidentified, inscriptions of the wall exterior temple in too, with solid flooring. The northernend of the north-western was the end of slightlywider, building SquaresNO is steppedout slightly,twice, as if this entire or as if there were towers flankinga gateway;on the other hand this is the very kind of area, The existenceof a probablyaffectedthe ground-plan. above a slope, wherelater reinforcements beadsof Phases the numerous was by suggested North-WestGate into the templenear this point In this case and evidence. the Middle favoured Neo-Assyrian Assyrian by 4-5, and is strongly as there of probably RSTUVW, been another Squares occupyingpart thereshouldhave courtyard was in the Neo-Assyriantemple,with an entranceto the shrine,whichwouldhavebeen in Squares the bottom of a tower at a QRWX. The apparenttrace of a cellarin SquareTT could represent outer and the inner courtyards. (south-western) (north-eastern) gate separating (Fig. 12) they create a ground-planmarkedly If these elements are combined symmetrically Ashur Templeat Ashur, as excavatedand restoredby Walter similarto that of Shamshi-Adad's Andrae(Haller and Andrae 1955:Taf. V) and redrawnby Heinrich(1982: Abb. 289) (Fig. 13). It looks as if the same architector school of architectswas responsibleon both occasions. At Ashur, on the centralaxis of the templeapproachedfrom the south-west,therewas an external gate, a small outer courtyard,a gate throughto a large innercourtyard,and a centralentrance to the anteroomof an inner bent-axisshrine;the innercourtyardhad additionalexternalgates, one on each side, to north-westand south-east.All these featuressuit the remainsof the Ishtar
365
50 meters
Fig. 12. IshtarTempleat Nineveh.Left: Excavatedremainsof Phase 7 platform,including walls of room of Phases 2-3 (top right) and of Phase 20 gate-chamber fragmentary and of Phase7 temple,with shrinelocatedroughlyover courtyard.Right:Suggested ground-plan QRWX room of Phases 2-3, and with north-western gate-chamber over that of Phase 20 (cf. Fig. 18). Author'sdrawing.
50m
Fig. 13. Shamshi-Adad's Ashur Temple (after Heinrich 1982: Abb. 289).
366
JULIAN READE
gate. Given the Templetoo, except that thereis no evidenceeitherfor or againsta south-eastern damagedstate of both buildings,theiroveralldimensions,c. 55 by 106+ m at Ninevehand c. 54 by 108m at Ashur, are so close that they could have been identical. Their orientationsare becauseat Ashurthereare stone footingsto differences, extremelyclose too. Thereare structural the walls, and no internalmud-brickplatform.A problemwith the parallelis the location of Cellar 7; it would have to be regardedas a structuralfeature, as outlined above, which was completelyburiedbelow the innercourtyardwhen the buildingwas complete. The Rimah temple is also relevantbecauseits foundationis dated very close to the reign of and he may have been responsiblefor it, althoughthe only known and perhaps Shamshi-Adad 1983: 65). secondarybuilding inscriptionthere was written for someone else (Howard-Carter invitedmany geometriccalculationsconDuring the course of excavation,its fine preservation and elevation;similarideas have been appliedto the Ashur cerningits proportions,ground-plan with possibilities (1988). Thereare comparable and Anu-AdadTemplesat Ashurby Stepniowski the Nineveh the IshtarTempleat Nineveh, but it is too badly damagedfor certainty.Presumably temple,like those at Ashur,Leilanand Rimah,also had externaland courtyard facadesdecorated with elaborateengagedcolumns (Haller and Andrae 1955:Taf. 39b; Weiss 1985:8; Oates 1967: Pls. XXXII-XXXIII, XXXVI). While the traditionalnorthernbent-axisapproachto the inner shrine was maintainedat Ashur, and very probablyat Nineveh too, it is plain that all these traditions.Thereare familiarwith southernarchitectural buildingswere designedby intellectuals and elites, questionsneedingfurtherexplorationsuch as what this meantfor the local priesthoods and financewere also and for cult practices;whethersouthernideas on temple administration may reflectthe evolutionof Shamshi-Adad's and to what extent such developments introduced; inscription politicalstrategy.Certainlyhis cult innovationsat Ashur,accordingto the Puzur-Sin failed to establish (Grayson 1987:77-8), promptedlocal resentment.In the end Shamshi-Adad structure plans,and the imperial any long-term effectively his empire,dyingtoo soon to implement was not resilientenough to survivehim. clay tabletsand pottery,survivefromPhase7. A podium Few objects,besidesstone inscriptions, of white stone, carvedin popular carvedin the Agade style was mentionedabove. A cylinder-seal was style with an erotic sceneand animals,Fig. 21C (Thompsonand Hamilton1932:PI.LXIII.A) debrisof 612 BC, but can be assigned found in SquareI at datumlevel, possiblyin the destruction to the Old Babylonianperiodon the basis of parallelsfrom Susa and KtlltepeLevel IB (courtesy D. Collon); given the continuity of the cult, it could well have been in the temple for over a cylinder-sealis describedas chalcedony,and the drawing thousand years. An unprovenanced publishedby Thompsonsuggeststhat it was of high quality(ibid.: P1.LXIII.3);it seemsto show introducedto a nakedgoddess. a worshipper found at Ninevehduringthe 1931-2 seasonhave tablets.Fragments Phase 7: The OldBabylonian mostly been published by Stephanie Dalley (2001); Finkel has told me more about them. in pencil as on other objects,have Provenanceswhich were markedon the tablets, presumably now been destroyedby baking to conservethe clay, but many had been safely recordedin the BritishMuseumCatalogueentries(Lambertand Millard1968);althoughsomehavebeenreplaced, in ink, the new versions tend to be shortenedor garbled.BM 137310(Lambert 1992:78), an here, Fig. 14. inventoryfrom WW6, was omittedby Dalley;it is reproduced administrative In his published report Thompson, after discussingPavementIII (Phase 7), proceeds:"In of cuneiform diggingabout this level nearthe head of the drainin Sect.00 we cameon fragments to suggest tabletsof a fairlyearlyperiod,but, until they are properlyworkedout, it is inadvisable a date closer than to say that they are probablynot earlierthan the Kassiteperiod.A few feet to the NE of the drain-head,at a higherlevel, were the greaterpart of a large syllabarytablet of Assyrianscript, 7th-centurydate [possibly BM 134526,a lexical fragmentin pre-Ashurbanipal apparentlymarked'IT.X.OO?'], and pieces of alabastervases [MiddleAssyrian,see Phase 10], of which one was carvedwith two lions" (Thompson and Hamilton 1932:68-9). A Thompson notebook, however,states more usefullythat "4 or 5 tablets of 2000 B.C. came on the line 00 (the two b[ric]kson end) s-SS-T". One of them must be BM 134535,an Old Babylonianletter, i.e. at datumlevel(followedby "18/14"whichI do not understand). whichwas marked"IT.OO.0",
367
d~~~~~~~~~~~~~1 4'~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~4
Fig. 14. Old Babylonian inventory (BM 137310). Photographs by D. Collon.
Two more are the Old Babylonian scholarly tablets, BM 134533 and 134539, which were marked "IT.TT.1I", i.e. 31 cm below datum. The remainder presumably included the unprovenanced scholarly tablets BM 134820 and 134825 (found during the same season), and 280'62 in the Birmingham City Museum (George 1979: 132-3, no. 40). Thompson's notebook further refers to a "2000 B.C. letter" in Square VV, which must have been BM 134534 or 134536; the latter was marked "IT.X.TM", but did not come from Square X which had been dug in 1930-1, because TM stands for Thompson Mallowan, i.e. 1931-2 (Thompson 1940: 86, n. 1); moreover there was also an X on BM 134526, the syllabary from 00. The broad date of 2000 BC was the one assigned in the 1930s to Hammurapi, or the Old Babylonian period, and Thompson's original assessment of these tablets was clearly correct. His published date for what must be the same group, "probably not earlier than the Kassite period", was presumably written without a clear memory of the tablets themselves; no copies or transliterations of them have been noticed in those of his notebooks which are at the British Museum. He may have been influenced by two factors. One is his eventual conclusion that the Phase 7 temple itself was to be ascribed to Ashur-resh-ishi I, which would not have suited the presence there of Old Babylonian tablets. The other is that BM 134553, which has the same provenance "00.0" as the Old Babylonian letter BM 134535, begins with two signs in pseudo-archaic script; it is actually a short votive inscription of Sargon II, as Thompson would certainly have recognised, but it may not have been clean enough to read thoroughly, and he was much more interested in the abundant Neo-Assyrian historical and library texts. Thompson also states, of the Phase 2-3 room outside the temple in Squares WWTT, that "in this chamber we found pieces of a painted pot and two of the little standless pots ... of the latest [Vth] prehistoric period, and at least two fragments of early cuneiform tablets ... of whatever date they may be" (Thompson and Hamilton 1932: 62-3). One of these must have been the inventory, BM 137310, which was marked "WW 6", i.e. 1.85 m below datum; it is dated by an eponym, Adad-bani, possibly the man who held this post just before Shamshi-Adad's death. Another fragment must be the Old Babylonian piece transliterated in one of Thompson's notebooks with the heading "WW -3 [92 cm]. Archaic Tablet"; this is the largest member of a group, BM 134537-8, probably parts of one tablet, whose provenance is given in the museum catalogue as "IT.TT-WW.2", i.e. 61 cm below datum. Thompson recorded his tablets as being "in" the Phase 2-3 room, but the top of its wall was deeper, 2.5 m below datum; so the tablets were in a later level. Individually none of these tablet provenances is satisfactory. For instance, several much later wall-pegs were also recovered from "IT.OO.0" (Lambert and Millard 1968: 91), while Square TT was pitted and produced a wall-peg at 4 feet (1.2 m) below datum (Thompson and Hamilton 1932: P1. LXIX.36) and a marble inscription "probably" of Ashurnasirpal near the bottom of Cellar 3 (ibid.: 66: text unidentified). The one suggestion of real stratigraphy is Thompson's observation that some early tablets were "near the head of the drain in Sect. 00" (which probably means they were not far above the surface of Pavement III), and that fragments of (Middle Assyrian) alabaster vessels were at a higher level. The overall picture, however, is tolerably clear, as suggested in Fig. 8C1 There was a meagre trail of Old Babylonian tablets across Squares 00,
368
JULIAN READE
SS and TT inside the temple,extendingoutside into SquaresW and WW. Those in the temple courtyardwere probablyabout the level of PavementsII and III. They must have been in one or and finallyin debrischurnedunderfoot levels,then in fill betweenpavements, more abandonment they are most when most of the pavementtiles wererecycled,but whateverthe exact stratigraphy unlikely,unless brought from elsewhere,to have been earlierthan PavementIII. Furthermore, foundationin SquareTT, Thompson'splan gives the heightof the top of the temple'smud-brick just south-eastof Cellar3, as 3 feet (92 cm) below datum,whichis 15cm lowerthan as preserved in the courtyard, abovethe pavement, the levelof PavementIII. So even tabletsthatwereoriginally could easily have moved down the slope westwardacross the walls and even out of the temple, and the pavementswere largelyremovedand the site long after 612 BC, when the superstructure was strippeddown to Old Babylonianlevels and below. This accountsfor the tabletsin Squares
vww. .
By content, the tabletsthemselvesfall into about four groups,as follows: list. BM 137310,dated by Adad-bani(Fig. 14). This documentcould be a list 1. Administrative of the temple. of booty, or evidencefor the management story, This was interpreted by Dalley as some kindof Gilgamesh 2. Scholarlytext. BM 134537-8. possibly a local version. She noted that the scriptwas "quite similar"to that of the Rimah that probablylocal clay, resembling letters.The tablet is made of a distinctive,high-quality of of many Neo-AssyrianKuyunjiktablets, and this is reflectedin the crisp preservation the signs. is about 3. Four or five scholarlytexts. BM 134533,describedby Dalley as a "tiny fragment", two-thirdsof a badly damagedtablet;Lambertand Millard(1968: 75) gave its dimensions, 7.1 by 7.7 cm, and suggestedthat it might be a letter, probablybecause of its shape, but in Sumerian; it may be a cultichymnto Ishtar Finkelhas recognisedit as writtenphonetically and d15. BM 134539has been identifiedby Finkel as as it includesthe words g]a-?a-an-m[u a cultic text also writtenin Sumerian.BM 134820and 134825also have cultic text writtenin Sumerian.The Birmingham fragment,280'62, has part of the Sumerianversion of Lugal-e. BM 134820, 134533 and 134539 have horizontalrulings betweeneach line of writing.The poor-qualityclay, especially that of BM 134533 and 134539, remindsme of tablets from who southern,not northernIraq.If they did come from the south, theneitherShamshi-Adad, who claimedto have promotedthe cult (Driverand Miles rebuiltthe temple,or Hammurapi 1955: 12-13), could have arrangedfor the dispatchof approvedcult expertsand documents. of 4. Threeletters,all in differenthands. BM 134536concernsgarlic,onions, and a distribution in form and content, those oil and lard at Nurrugum.Dalley notes that this letterresembles, Thereis no evidentconnectionwith the from Rimahand Mari from the time of Hammurapi. and the implication previouscaptureof Nurrugum, Mari lettersthat concernShamshi-Adad's centre.BM 134535is from Iluni continuedto functionas an administrative is that Nurrugum to "the son of my lord". So the likely recipientwas a king'sson who was visitingor residing in the temple,which was perhapsthe finestbuildingavailableat Nineveh;candidatesinclude as Shamshi-Adad's son, but therecould son, and Mut-Asquras Ishme-Dagan's Ishme-Dagan be others. Iluni's letter invokes Shamash and Marduk, which associates the writer with of Babylon;a man of this name was to lead Eshnunnain a rebellionagainst Samsu-iluna Babylon(Charpin1998). BM 134534mentionsa requestfor grainfrom the writer's"father" and the thereweretwo kings of Larsaof this name, one defeatedby Hammurapi Rim-Sin(?); An odd featureof BM 134534is a reference,apparentlyobsequious, other by Samsu-iluna. to "my lord LUGAL iu?"(?); no alternativeto Dalley's readingsuggestsitself. Whateverthe connotationsof the title (Hallo 1957:21-6), which was sometimesused at Eshnunnaas well it could perhapshave been appliedto otherkings too. and Samsu-iluna, as by Shamshi-Adad These tablets representa tiny proportionof the archivesto which they belonged;we cannot hope for more, since therecan only be a little soil untouchedbeneaththe remnantsof Pavements all abandonedor discardedat the same I and II of the temple.The archiveswere not necessarily to trashthe templein the warsfollowingShamshi-Adad's time. Therewereplentyof opportunities death (e.g. Charpinand Durand 1997:369, 372-3). Among the forces operatingin the area the
369
Turukkusound the least likely to have respecteda grand Mesopotamiantemple. Samsu-iluna's problemswith Iluni and Rim-Sinemergedsome 25 years after the templewas completed. It may be helpful to mention that two inscribedfragments,not fully describedin the British Museum Catalogue, are not Old Babylonian although they come from the same vicinity: BM 123444 from PP is in pre-Ashurbanipal Assyrian script, and BM 134827 from NN is an unpublished Ashurbanipalprism fragment mentioning Elam. There are several more Old Babyloniantabletsin the older "Kouyunjik Collection",and I am gratefulto JeanetteFinckefor theirnumbers(K 8755, 8765, 8860, 10468+, 10506, 13942, 18634,and 21119): it is obvious from various factors that most of them were excavatednot at Kuyunjikbut in southern Iraq, and probablythey all were. Phases6-7: Painted sherdswill seldomhave been saved,fragments pottery.Althoughundecorated of painted pottery of the early to mid-secondmillenniumare recordedfrom Kuyunjik.At the BritishMuseumthereare a few from nineteenth-century excavations,and more from Thompson; others found by him are elsewhere,and gatheringthem all togethermight be a useful exercise. Some of the painted sherds(e.g. Gut 1995:II, Taf. 96.1357-9) should probablybe dated before like examplesfrom Taya (Reade 1968:P1.LXXXVII.26-7) and Rimah (Postgate Shamshi-Adad, et al. 1997:53). In contrastthere are, from the 1931-2 season, parts of twelve very fine pottery vessels,made of a pale yellowishor pinkishbrownwarewith darkerreddishbrownpaintedstripes (BM 1932-12-12,1426-7, 1428 A-J = 139828-9, 139830A-J). The sherdsare in good condition (Fig. 15) except that two of them (139829, 139830J) have the light green stainingoften seen on materialfrom drains.Clearprovenances are writtenon only three:"NN 7" on 139830A-B, and "MM found[ation]s" on 139828.While,therefore,they cannot be treatedas a reliablegroup,they look like the remnantsof stratifiedobjectsfrom a poorly excavateddestructionlevel ratherthan randomfinds. One possibilityis that they weremostly found in or near the drainthat ran out of the templethroughSquareNN. A group of similarvesselswas found, at Rimah,abandonedin a palace which seems to have been built almost immediatelyafter the demolitionof one occupied and sacked under Hammurapi(Postgate et al. 1997:36, 55, Pls. 74-5; Oates 1972: 85-6). The historical contextsuggeststhat the replacement palaceat Rimahwas occupiedand thentemporarily abandonedin Samsu-iluna's reign. My recollectionof handlingthe Rimah vessels on site is that they were almost identicalwith those from Nineveh, as if a single factorysuppliedfine potteryto both sites. So a date in Samsu-iluna's reigncould suit both these vessels from the IshtarTemple and the lettersfrom the same place. It would not be surprising if both Nineveh and Rimah were ransackedabout this time.
The later second-millenniumtemples (Phases 8-17)
Phase8: Mid-second millennium. Thereare no recordsof repairsto the IshtarTemplefor centuries after Shamshi-Adad and Hammurapi. Ishtarof Nineveh was nonethelesshighly respectedin the long-lastingMitannianempirethat later emergedin the same region (Beckman 1998:2-7), and the buildingevidentlysurvived.Glass vesselsfromthe slope north-westof the temple(Barag 1985: 41), Nuzi Ware vases (Thompson and Hutchinson 1931: 108-9), a seal impression (ibid.: P1.XXII.7), and miscellaneous unpublished glass and faiencebeadsand other fragmentsfrom the site are Mitannianor slightlylater,dispersed duringdisturbances or renovations (cf. Tenu2004/5). Phase 9. Fragments of archaizing stone inscriptions from the templearea (Grayson 1987:115-16; possibly Reade 2000b: 87) include a referenceback to the restorationof a dilapidatedbuilding by Shamshi-AdadI, at least one of whose own inscriptionsmust thereforehave been exposed duringa new restoration.This may be ascribedto Ashur-uballit I (c. 1353-1318BC), whose work and the replacementof whose foundation documents (temmeni) are mentioned in Phase 10 inscriptions; Ashurnasirpal II (Phase 20) also mentionsAshur-uballit's work. Nineveh had been conqueredfrom Ashureither shortlybeforeor duringthe reign of Ashur-uballit. His restoration of the cult-centre was therefore in the politicaltraditionof previousconquerors,Manishtushu and Shamshi-Adad I. Physicalremainsof his work may consist of the "verysmall traces"of the tiles of PavementII, which was 1 foot (31 cm) below datum, above PavementIII, in Square00
370
JULIAN READE
| '1
\~~~~~~
0
Fig. 15. Late Khabur Ware pottery from the Jshtar Temple area (Phase 7). Top: BM 1932-12-12, 1426. Remainder: BM 1932-12-12, 1428 A-J. Drawn by Ann Searight.
371
(Thompsonand Hamilton 1932:58, 68). A stone mace-headof a Kassiteking Kadashman-Enlil, if the first of this name, could have been booty dedicatedto Ishtarby Ashur-uballit (Thompson and Hamilton 1932: 63), but it could also have belonged to Kadashman-Enlil II (Brinkman 1976: 134). were restoredfrom top to bottom by Phase 10. Areas of the templedamagedby an earthquake I (c. 1263-1234BC), according to wall-peginscriptions in whichhe mentionsShamshiShalmaneser Adad's and Ashur-uballit's work (Grayson 1987:205-7). He does not mentionShamshi-Adad's and left his own; a black stone fragmentmay be one of temmeni,but replacedAshur-uballit's, I wall-peginscriptions them(ibid.: 226). Some Shalmaneser refermoregenerallyto his restoration of the ziggurrat, wall and gate, and the depositingof his nareand temmeni, while othersdescribe him as restorer of Emashmash (ibid.: 207-9, 216-17); many of his bricksor tiles have dedications to the goddess (ibid.: 212-13). There are also "palace"bricks of Shalmaneser I from the area (Smith 1875: 140, 247), and possiblya "palace"wall-peg(Grayson 1987:217). Furtherdetail about Shalmaneser's work is providedby Phase 12 inscriptionsof Ashur-reshishi I (Grayson 1987:309-13). They referto his restorationof the facade(nameru) of the gate of the greatcourt (ki.kisal.mah); therewere lions thereby Phase 12. Thereis no recordof who first erectedlions, but theycould havebeenOld Babylonian or evenolder,like one fromthe foundations of the AshurTemplebeforeShamshi-Adad (Haller and Andrae 1955:Taf. 25b). As the favourite beasts of Ishtar they were surely in place, made of stone or glazed terracotta,by the end of I's restorations.We do not know if they were ever replacedsubsequently, Shalmaneser because texts tell us little that is specificabout the templedecoration;we have to rely on parallelssuch as the magnificent stone lions whichlaterflankedthe entranceto Ashurnasirpal II's shrineof Sharrat Niphi at Nimrud(e.g. Reade 2002a: 183, Fig. 41). Whilewall-pegshad long been known at Ashur,Shalmaneser I may have been the first king to use them in the IshtarTempleat Nineveh.Graysonsuggeststhreeversionsof theirtexts, but there could have been more. Wall-pegsof both Shalmaneser and later kings were found by Thompson. The provenancesof many are known (Lambertand Millard 1968:89-91); their depths suggest that, mixed togetherin the same debris, there were some 34 wall-pegsof Shalmaneser I, 32 of 1, 4 of Shamshi-Adad Ashur-resh-ishi IV, and 92 of Ashurnasirpal II: others are unprovenanced (ibid.: passim; Lambert 1992: 94, index, sikkatu's). Either walls remainedupright for many centuries,or later kings replacedold pegs, perhapsnot always in the right places. None of the but generallythey must have fallen from the exterior pegs was found in standingarchitecture, walls of the north-western templefrontage,the outer courtyard, and possiblythe innercourtyard; thereare more,unexcavated, presumably all roundthe outsideof the temple.Nearlyall the Ashurresh-ishiwall-pegs referringto the Lion Gate of the great court were from the north-western frontage. So the North-WestGate of the temple, the existenceof which was postulatedin the discussion of Phase 7, was probably the Lion Gate, in which case this was one area where I worked.A brickinscriptionwas reportedfrom the fill of Cellar 1 (Thompsonand Shalmaneser Hamilton 1932:66: text unidentified),which suggeststhat he repairedthat gate-chamber too. A few archaeological remainscan probablybe ascribedto Shalmaneser I. One is PavementI in SquaresNNOO in the outer courtyard,which was a scrapof tiling laid on a base of mud bricks; its surfaceseemsto have been at datumlevel (Thompsonand Hamilton1932:58, 68). Somewhere near it, "betweenSectionsOO-DD", was a lion vase (ibid.: 68-9, P1.LI.4), with other alabaster vases which must eitherbe identicalwith or closely relatedto a group of such things which was excavatedin NN, at datumlevel (Thompsonand Mallowan1933:148, P1.LXX.1-6). Therewere manymoresuchvase fragments (Searightet al., in preparation), including"two basketsof broken alabastra.MM LL and NN" accordingto the 1932 field catalogue. It sounds as if the vessels were scatteredfrom 00 north-westward, in layers of debris formed after the mud-bricktemple walls and most of PavementsI and II had been removed,but that they, like the Old Babylonian tablets,had originallyderivedfrom fill packedundera pavement.The vesselsconformto a wellknown type of Egyptianorigin,with attachedpedestalbases (lost from the lion vase itself); they held scentedoils, and bring the temple alive, as they were surelyofferedby worshippers. There are examplesof these vessels from the IshtarTemplearea at Ashur and elsewhere,and they can
372
JULIAN READE
be dated within a few decades of 1300BC (e.g. Moorey 1994: 52-3). So they are the kind of afterthe earthquake. renovations materialwhichmightwell havebeenburiedduringShalnaneser's werea leadplaquerepresenting Amongotherobjectsfromthe templearea,not usefullystratified, what seems to be a woman with partedlegs, and a lead figurineof someone in a schematically pointed hat and cloak (Thompson and Hutchinson1931:P1.XXV.7; Thompsonand Mallowan 1933:PI. LXXVIII.43);thereis also a small decoratedlead disc (BM 1932-12-12,1031= 138498, from NN 0). These recall the lead cut-outs found at Ashur, some of which show nakedwomen, pairedwithmenwearingpointedhatsandcloaks(Andrae1935a:1036,Taf.45-6); theirarchaeoloI and Tukultigical context therewas unclear,but foundationdocuments,writtenby Shalmaneser Ninurta I for a temple of the goddess of Nineveh at Ashur, were not far away (Grayson 1987: of sexualactivitieshad a specificfunction representations 196,264-5). This suggeststhat miniature in the cult of this goddess:perhapsthey were shapedfrom lead becausethis materialcould easily be reused. A substantialmud-brickstructure,5.4 m high and tombs"structure. Phases 9-15: The "vaulted of the temple about 22.5 m wide from north-eastto south-west,was built on the slope north-west in SquaresM, N and BB, with its centreroughlyoppositethe westerncornerof the temple'sinner court (Thompson and Hamilton 1932:78-80, Pls. XLVIII.2-3, LXXVIII.257).The outer walls blocking.It was firstregarded at leastfivehigharcheswith mud-brick of the structure incorporated and Nabu Templeto the north-west, as a bridgelinkingthe IshtarTemplewith the eighth-century later as "vaultedtombs". Its top was 8 feet (2.5 m) below datum, and thereforelower than the foundations of the temple, with which no connection was established.Its south-easternface, was obviouslyinserted however,is only about 8 m away from the templefrontage.This structure and a few sherdssuggest into the slope, like a terrace.It is later than the early thirdmillennium, later than the mid-second(Gut et al. 2001: 77), but how late remainsunclear.In view of the I, or another king, to evidence for earthquakes,it may representan attempt by Shalmaneser stabilizethe slope below the temple. I (c. 1233-1197Bc) referto him as builderof the house Phase 11. A few bricksof Tukulti-Ninurta implyingthat he was working of Ishtar(Grayson1987:284). Theymentionthe month tag-ri_rit(?)l, on a special festival house; this could have been an innovation,following southerntraditions. III which probablyconcernsa bit Frahm(2000) has identified,in a brokentext of Adad-nirari work.Frahmhas also offereda reconstruction back to Tukulti-Ninurta's akitu,a possiblereference the only bit akituat Nineveh),but of the whole historyof the bit akituof Ishtar(not necessarily detailsare speculative.It was probablynot in the main temple,and not excavatedby Thompson, which would help explainthe shortageof physicalevidence.We may postulatea location for the north or west of the main temple,becausethat is the generalarea wherethe bit akitu somewhere few recordsrelatingto it have tendedto emerge,but the evidenceis inconclusive. repairs,recordson I (c. 1132-1115BC), aftermentioningShalmaneser's Phase 12. Ashur-resh-ishi of Ashur-dan I the in Lion Gate the reign damaged another earthquake wall-pegs that 1987: 311-13). it who was restored himself the he (Grayson but that king (c. 1168-1133BC), as them rosettes defines the text and (ia-e-ri). of traces glaze, yellow his retain of wall-pegs Many to the roof the the from of up have raised the temple to facade, claims top Ashur-resh-ishi battlements.It had been 15 courses high, which would have been under 2 m; he removedthis, and replacedit with 50 courses,c. 6 m high.Thompsonassociatedthesestatisticswith the differing (Thompson depths of the main foundationsof the temple,which he ascribedto Ashur-resh-ishi and Hamilton 1932:63-5), but all the text reallytells us is that this king raisedthe heightof this wall-pegssome of whichmight He workedon the palace too, and there are fragmentary famade. referto eitherbuilding(Grayson 1987:314-15, 322-6). 1 ( 1114-1076BC), writtenon a tablet which was text of Tiglath-pileser Phase 13. A fragmentary as and Shalmaneser a foundationdocument,mentionsShamshi-Adad, presumably Ashur.uballit
373
kings who had worked before him on a terrace(tamla) connectedwith Ishtar (Grayson 1991: 57-9). He could havegot the earlierroyalnamesfroman IshtarTempleinscription of Shalmaneser I. This work sounds like a reinforcement of the north-western frontage of the temple, perhaps connectedwith the "vaultedtombs". Other inscriptionsof Tiglath-pileser I mention the Ishtar Templebut were mainlywrittenfor a palacenearby(Grayson 1991:35, 55). Two limestonefragments carvedwith hunt sceneswerefoundin SquaresA and P nearthe Ishtar Temple(Thompsonand Hutchinson1931:P1.XXIII.12:BM 1932-12-10, 42= 135424;Thompson and Hamilton 1932:P1.LXII.14:BM 1932-12-12, 5 = 124275).They may come from a monument such as an obelisk of the Assyriantype, with a flat or steppedtop and with carvingson the sides. The two fragments wereascribedto Ashurbanipal by Barnett(1976:49, P1.XLIV) but recognised by Poradaas MiddleAssyrian(Reade 2000c). It seemsnot unlikelythat they should be ascribed to Tiglath-pileser I, who is the firstAssyriankingknownto havelistedhis achievements in the hunt and to have mentionedthe display of narrativeart in one of his palaces;he could well have introducedthe obelisk,a type of monumentcommon subsequently but otherwisefirst attestedin the reignof his son. Assyrianobeliskshave seldombeenfoundin position,but two at Nimrudwere at eitherend of a palaceor templefaqade,wherethey had been placedat differenttimes, and one at Ashur was beside a gate. So obelisks at Nineveh were probablyno less prominent,and they seldom look as if they were damagedby the fire which burnedinside the temple in 612 BC. The most publicpositionmighthavebeenoutsidethe North-West Gate (Lion Gate), thoughthe temple had other fa9adestoo. Only two or threeearly obelisks,from the eleventhcentury,emphasizethe violence of the Assyrianking; later ones, which are mainly ninth-century, show him graciously acceptingsubmissionand tribute,whileviolencewas transferred to othermedia. Phase 14. Ashur-bel-kalaerected a limestone statue of a naked goddess (BM 1856-9-9, 60 124963), with a brief inscriptionon its back which begins e'.galand possibly impliesthat it was for publicdisplay,ina muhhi siahi (Grayson1991:108). The provenance is unsure.Rassam( 1897: 8-9; Reade 2002: 563, with more references) first describesat length a large "ditch",where he found many wall-panelsand the WhiteObelisk(Phase 16); this was betweenthe southernside of the temple precinctand the South-WestPalace. "Afterwards I found in another locality, about halfwaybetweenSennacherib's Palaceand that of Assur-bani-pal", the BrokenObelisk(BM 18569-9, 59= 118898);this must have been in or close to the main temple.Finally "we also found in the same ditch"the statue.The orderof Rassam'stext impliesthat the statuewas found with the Broken Obelisk, but his mannerof using the word "ditch"suggests that the statue may really have been found with the White Obelisk, with the paragraphabout "anotherlocality" being insertedat this point in order to link the discussionof the two obelisks. If so the statue was erectednot by the templebut somewhere near a southernentranceinto the Emashmash precinct. The limestoneBrokenObelisk,so-calledbecauseits lower half is missing,is usuallyascribedto Ashur-bel-kala (1073-1056BC). Found betweenthe two great palaces, it must have been in or near the temple, probablyin one of the wide pits that appear on contour plans, although the specific location indicatedon a plan adjustedby Rassam (Gadd 1936: 123) really refers to a monumentof Shamshi-Adad V (see Phase22). The BrokenObeliskhas a long text listingmilitary and civic achievements (Grayson 1991:99-105), sometimessimilarto those of Tiglath-pileser I whom Ashur-bel-kala had probablyaccompaniedon campaign.Grayson, following Rawlinson, supposedthat this obeliskhad originallybeenerectedat Ashur,becausethe sectionof text dealing with the king's buildingoperationsdescribeshis work at Ashur;but the start of this section is lost, and it could well have begun with work at Nineveh. Thereis a single carvedpanel showing the king in triumph; his head had apparently beendefaced,so the obeliskwas probablystill visible and vulnerable in 612 BC. Partsof both text and carvinghavedeteriorated furthersinceexcavation, a processthat may have begunwith washingin acid waterin the nineteenthcentury.Subsequent changes in its appearance,sometimesattributedto cleaning(Russell (2003: 4), are not reliable, and the older illustrations are the best (e.g. Reade 1977:P1.IV). Phase 15. A fragmentary wall-pegof Shamshi-Adad IV ( 1053-1050BC) recordsthe repairof the terrace(tamlu)of a buildingconnectedwith Ishtarof Nineveh;another,found in SquareU, refers
374
JULIAN READE
texts on severalwall-pegs, to dilapidationand to two lions (Grayson1991:118-20). Fragmentary to some from the area north-westof the temple,have been conflatedto producea text referring the repairof the faqade(nameri)of a buildingbelongingto Ishtarof Ashur, but Graysonnoted IV documentsare dated that this is unsure(ibid.: 117-18). At least two of these Shamshi-Adad in his own eponymy,c. 1053, suggestinga politicalmotive in this disturbedperiodfor what may repairs. have been insignificant Phase 16. A hymn of AshurnasirpalI (1049-1031 BC) claims that he redecoratedthe shrine in Lambert2004/5;see also Seux 1976:497-503), as many otherkingsmay have done; (reference I was also the this text was ascribedto AshumasirpalII by Menzel (1981: 117). Ashurnasirpal king probablyresponsiblefor erectingthe limestoneWhite Obelisk(BM 1856-9-9,58 = 118807), though other Middle Assyrian kings cannot be excluded (Reade 1975). An old ascriptionto II is still occasionallydefended(e.g. Grayson 1991:254-6; Russell 2003: 4), but Ashurnasirpal in the text are not compatible the iconographyis acceptedas early, and the campaignsdescribed much has been made of a II's many detailedinscriptions; with those recordedin Ashurnasirpal (the Hasanluarea of Iran), otherwiseknown only in referenceto horses from Gilzanu/Habzanu trade.Although the ninthcentury,but this need merelydocumentan earlystageof this important in the e.nathi,the WhiteObeliskhardly a captionis attachedto a carvingof the king worshipping commemorates one particularbuilding.Instead,like the BrokenObelisk,it advertisesthe king's achievementsgenerally;it does so with many narrativecarvingswhich were perhapsbased on mural decorationin a palace (Pittman 1996), and with an annalisticinscriptionon top. This is badly worn, as is much of the carvedsurface,as if the obeliskstood in the open for many years. It was found complete,lying on its side, betweenthe southernside of the templeprecinctand the South-WestPalace, and was thereforeprobably displayed at an entrance to the Emashmash precinct,hardlya place wherelooters would have lingered.Althoughthe White Obelisktoo has since excavation,it may not have been defacedin antiquity. deteriorated II (1012-972 Bc) may also have restoredthe e.nathi,accordingto two wallPhase 17. Ashur-rabi II (Grayson 1991:332-3, 337). peg fragmentsascribedto Ashurnasirpal temples(Phases18-27) Thefirst-millennium of Adad-nirari II (911-891 BC) werewrittenon four or five basalt Phase 18. "Palace"inscriptions slabs found betweenthe Ishtar and Nabu Temples(Grayson 1991: 159-60); they probablydo come from a palace. A basalt stand having the same inscriptionwas found in Square ZZ, beside SquaresXX or YY, above the IshtarTemplesite (Thompsonand Hamilton presumably 1932:P1.LXXXIII.266);anotherof slightlydifferentshape was found by Smith (Reade 2002a: II workedon the templeitself. 154, Fig. 17). So it is unclearwhetherAdad-nirari freeof a blackstone obeliskor comparable He is, however,the king to whom many fragments standing monumentfrom the Ishtar Temple area, inscribedand showing scenes of people in Grayson 1991: 161). westerndress bringingtribute, have been tentativelyascribed(references: to This is dubious.I myself(Reade 1981:151-4) have even ascribedsomewhatsimilarfragments and needsmoredetailedwork. III (810-783 Bc), but the situationis verycomplicated Adad-nirari The numberof fragmentsof this generalnature, in various kinds of stone, now split between and London (where they are mixed with some from Ashur), probably Baghdad, Birmingham most have been monuments; exceedsa hundred,and they derivefromfive or ten or moredifferent illustrated (1982: 184-6, 194-5, Abb. 139-45, 156-60). What the majorityhave by B6rker-Klahn with relativelylittle text, and that in common is that they representprocessionsof tributaries, smashed.It is not unlikelythat Ashur-dan11(934-912 BC) or his they have been enthusiastically like this in the IshtarTemple son Adad-nirarI II revivedthe practiceof placingstone monuments earlyeighthcentury.Perhaps precinct,and that their successorscontinuedto do so down to thtF completely.So the many there were later examplestoo, made of metal, which have disappeared fa9ades of the Ishtar Templewill have acquiredobelisks gradually,whenevera king wished to add one.
375
Phase 19. A few bricks and wall-pegs of Tukulti-Ninurta 11 (890-884 BC) were found north-west of the temple (Grayson 1991: 183-4, 186-7). They may have commemorated work on temple or palace; some of them, but not necessarily all, had e.gal at the start. Phase 20. Ashurnasirpal 11 (883-859 BC) was responsible for a major renovation, traces of which survived in position (Thompson and Hamilton 1932: 55, 58, 67, 69-71, 107-8, Pls. XLVI.4, XLVIII.1, XLIX.1, L.4); some are included on the schematic section, Fig. 8b. Two areas of tiling with Ashurnasirpal inscriptions in Squares RS, laid in two layers with bitumen between them, could represent the remains of the inner courtyard. Tiles further north in Square R were laid in one layer with bitumen below. Tiles had been laid "on the surface of the actual libn foundation". The pavement surface was 1 foot (31 cm) above datum. A stretch of mud-brick wall, some 2.5 m thick and 20 m long, which ran parallel with the outside wall of the temple, survived in Squares RST; its ends terminate sharply on the plan, as if both were door-jambs in good condition. The stub of a cross-wall projected from this wall northwestwards in Square I, so that the eastern corner of a room was identifiable in Squares 10. There was the lower part of an alabaster wall-panel, Fig. 16 (Thompson and Hutchinson 1931: 93-4; Thompson and Hamilton 1932: 70, PI. XLIX.3), some 4.6 m wide according to the plan, facing north-west, in position against the wall between the corner and the supposed door-jamb; its base had been set into the old mud-brick platform. It had been carved with figures whose feet were at datum level, so the floor of the room should have been slightly lower although, according to Thompson's plan, the platform itself survived to a height 2 feet (62 cm) above datum at one point in Square 0. It seems probable that the lower part of the panel was uninscribed and was left in position at Nineveh, while what could have been its upper part was found in fragments, reused in a Phase 29 wall nearby, Fig. 17 (Birmingham City Museum 118'32). This upper part had an
wall-panelcarved with tributaries (Phase 20). Print from Fig. 16. Base of Ashurnasirpal BritishMuseumarchive.
376
JULIAN READE
Fig. 17. Phase 29 wall built from carvedand inscribedfragmentsof Phase 20 panels, the base of one of which is visible in the backgroundto the left. Print from BritishMuseum archive. Ashurnasirpal II inscription in triplicate on its reverse, which is now on display, and is a little over 4.55 m wide (courtesy Phil Watson); Thompson recognised its other face, now built (if preserved at all) into the museum wall, as sculpture destroyed by fire. A cut in the platform at the base of the cross-wall suggested that there had once been at least one more wall-panel in this room, facing south-west beyond the corner. The inscription (Grayson 1991: 306-10) states that the king had conquered the land of Mehru, and cut timber there for the roofs of Emashmash; the same conquest, somewhere near Mount Amanus, appears in the account of the king's Lebanon campaign between 875 and 867 BC (ibid.: 219). The inscription adds that the temple built in Emashmash by Shamshi-Adad was dilapidated, so Ashurnasirpal planned and rebuilt it from top to bottom, made it bigger than before, and provided Ishtar with a new throne in her sanctuary (e.ku). There were several variant inscriptions on stone describing Ashurnasirpal's reconstruction of the temple (ibid.: 310-15), but the fragmentary and scattered texts have not been classified satisfactorily. Surviving inscriptions may normally have been from the reverses of wall-panels, against walls and therefore unaffected by fire when the temple was destroyed, but some panels had inscriptions on both faces (e.g. Birmingham 227'78). Some inscriptions may be from the foundation documents (na.ru.a)mentioned in one of the many wall-peg texts (Grayson 1991: 328-37, 361-5). The latter too are hard to classify: some of the longer ones refer to Shamshi-Adad, and others mention the e~.nathi in connection with Ashur-uballit and possibly Ashur-rabi II; Shalmaneser I is not mentioned. It looks as if
377
50 meters Fig. 18. Ishtar Temple at Nineveh. LeJi: Suggested reconstruction of Phase 20 North-West Gate and inner courtyard, superimposed on remains of Phase 7 platform (cf. Fig. 12). with Suggested ground,lostshrinestill locatedroughlyabove QRWX room of Phases 2-3. Rig/ht:
Author'sdrawing. plan of Phase7 temple,with similararrangement. Ashurnasirpal's work was done in several stages, and his inscriptions specify those kings whose own inscriptions had been recovered in the course of each operation. There are also some Ashurnasirpal terracotta corbels in the shape of clenched fists which tend to appear in English excavation field-notes as "hands of Ishtar" (ibid.: 375-6). There are bricks with various inscriptions, one of which mentions the c.nathi (ibid.: 381-5). The positioning of Ashurnasirpal's pavements and wall-panels directly on the mud-brick foundation of the temple confirms his claim to have completely renewed at least part of the building: he must have removed earlier pavements and rebuilt some walls, and he will also then have been free to adjust the ground-plan. The surviving wall-panel in the 10 room helps define the plan, since it showed four large-scale figures, "the king (Ashurnasirpal) receiving three tributary kings, the first wearing sandals, the second barefooted, and the third with upturned shoes", who were moving right from the eastern corner of the room. There were very similar figures in Ashurnasirpal's slightly later palace at Nimrud, shown approaching the doors of the throneroom (Budge 1914: P1.XXVIII). The theme and arrangement suggest that the North-West Gate of the temple led into the 10 room, and that Thompson's plan is correct in implying the existence of a door-jamb behind the king; here there would have been a door into the inner courtyard. Perhaps the composition was extended with more tributaries to occupy the entire north-eastern side of the gate-chamber, with a similar balancing composition on the south-western side. The gate itself would have been Ashurnasirpal's version of the earlier Lion Gate. If this reconstruction is correct, the ancient Cellar 7 was certainly buried below the inner courtyard. Evidence for Ashurnasirpal's work in the outer (south-western) courtyard is restricted to wallpegs found in the debris there, so he repaired the faces of walls, but part of the Phase 7 wall in Square DD, or a replacement of similar appearance, was still standing to a height of 1.8 m above datum. If Ashurnasirpal kept the existing Phase 10 floor in this area, which was at datum level, both it and the floor of the north-western gate-chamber would have been, appropriately, slightly below the level of the inner courtyard.
378
JULIAN READE
IshtarTemplecan be obtainedfrom architecA little more informationabout Ashurnasirpal's tural fragmentswhich were found out of position. Among fragmentsof alabasterwall-panels, found north-westof the temple itself, are some on which the 14-line temple inscriptionis in of carving,Fig. 19 (Thompsonand Hutchinson1929:80, Pls. VI-VII; 1929a: betweentwo registers 118, P1.XLI.4-10; possibly 1931: P1.XXI.10=BM1930-5-5, 232=136997; Thompson and 45; Weidner1939:107,Abb. 86; Russell 1998:18, Hamilton 1932:P1.LXIII.22= BM 1932-12-10, and, in a lowerregister, Fig. 9, 204-5, 240, Nos. 214-20). Thereis a lion-huntin an upperregister, a libation over an animal originallyidentifiedas a lion, but perhapsmore like a bull; another fragment(Fig. 19, top right) may show the head of a bull in a hunt. Thompsonsaw all these as parts of one panel, but they could belong to a pair of panels,over 3 m high, with hunts of lions and bulls above and libations below. A smallerversion of the pair, but with the directionof at Nimrud(Budge 1914:Pls. XII, XIX). was latercarvedin the throneroom movementreversed, the The themeof the carvingswould obviouslyhave suiteda room or buildingwhichemphasised role of Ishtar in the hunt; the bit akitu seems improbable,since there is no referenceto havingworkedon it. Ashurnasirpal An inscriptionon the reverseof the hunt panels had been written in unusuallylarge script upsidedown (Thompsonand Hutchinson1929:80; 1929a:118, 121,P1.XLIII.49-53). Thompson Adadpositivelystatedthat it was "a dedicationto Nabu by an unknownking",and he suggested nirari III or Sargon, both of whom worked on the Nabu Temple.It is difficultto see how he could have learnt this from the seven scrapsof text of which he publishedcopies, Fig. 20, and which even Borger (1967: 536) declined to identify. Another possibility is that these were renovations;this Ashurnasirpal panels from the Ishtar Templerecycledduring seventh-century do not seem to have been burnt, as in would explain why the carvings,although fragmentary, is that the text on theirsecondarypositionsthey would have faced a wall. An unlikelyalternative scripton panelsreusedduring the reverseconstitutesuniqueevidencefor the use of monumental repairsafter 612 BC. The other main group of materialwhich may be ascribedto Ashumasirpal,apart from freeglazedtiles and bricksreusedin later of ninth-century consistsof fragments standingmonuments, buildingsnorth-westof the temple.Assyrianglazed decorationswere usuallythough not always displayedout of doors, and in other templesthey weresometimeson the platformswhichflanked the main entrances.Some of those which Sargonwas to place in the AshurTemplehad scenesof warfare(Andrae 1925:P1.6), and a likelyposition for some of the Ninevehtiles (Thompsonand Hutchinson 1931: Pls. XXVIII, XXIX.1-5, 10(?), XXX-XXXI, XXXII.1-5; Nunn 1988: Abb. 125-30) would have been in the inner courtyard,in front of the shrine of Ishtar. They thereare scenesof warfare, includeone scene of warfarewith a guillocheframebelow. Otherwise chevronframesabove and below, and one tile tribute,and triumph,with typicallyninth-century representingthe goddess or the queen. There are ninth-centuryglazed bricks also (ibid.: P1.XXIX.7-9, 11-12), which could have been placed at a higher level on doors or below but thereare far too few to permita reconstruction. crenellations, III (858-824 BC) indicatesthat he did some work in the Phase21. A single brickof Shalmaneser temple (Grayson 1996: 170), and Thompson'snotebook mentions that two of his bricks were of Gate. Two joining fragments found near SquareVV, so perhapshe workedon the South-West with a dedicationto Ishtarby Shalmaneser III, possiblypartof a statue, fine whitestone inscribed were found (Grayson 1996: 114-15; Thompson and Hamilton 1932: P1.LXXXIX.302: BM 1932-12-10, 9-10); one of themwas I foot (31 cm) abovedatumin SquareI, so thismonument a suitableposition for royal statues. may have been placedin the north-westgate-chamber, Phase 22. A fragmentof a black stone monument,probablya stela, which was inscribedwith side of the temple V (823-811 BC), was found on the northern partof the annalsof Shamshi-Adad area by Loftus (Grayson 1996: 180-8: BM 1856-9-3, 1510- 115020). As explainedelsewhere originalexcavationplan (Reade and Walker1982:115), its positionis markedon the unpublished in the BritishLibrary,at a point whereanotherversionstates "portionof obelisk discoveredby Mr Rassam".It is unclearwhetherthis monumentwas in the templeor the palace.
379
Fig. 19. Ashurnasirpal lion-hunt, possible bull-hunt, and libation scene fragments (Thompson and Hutchinson 1929: Pls. VI-VII).
380
JULIAN READE of Ashr spttScLdtuytw(Pt.) L6r9edtaractzvs, L.ns o. Wok Inscriptiam.

44" ap t.
4-9
k f50AO1c:tt (/s
51 O Onk 50
hea
.3V4. 5Z A
/: d
hunt and libation scenes (Thompson and Fig. 20. Inscriptionon reverseof Ashurnasirpal Hutchinson1929a:P1.XLJII.49-53). Phase 23. Bricks of Adad-nirari III (810-783 BC) state that he worked on a palace at Nineveh, completing his father's work, and that he built the Nabu Temple (Grayson 1996: 218-20); work on the latter occupied the years 788-787 BC (Millard 1994: 58), an indication of the speed with which such things could be done. A fragmentary text suggests that he repaired the bit akitu (see Phase 11). Although Adad-nirari may have been responsible for one of the obelisks discussed above (Phase 18), the best-preserved monument of his reign surviving from the Jshtar Temple is a stela of yellow limestone, now partly reddened by heat, which describes a major royal grant to the official Nergal-eresh in 797 BC (Grayson 1996: 2 13-16). Like Babylonian kudurru monuments, it was presumably placed in the temple as a permanent record under divine protection; this was effective because, although the name of Nergal-eresh was deleted on monuments found elsewhere, there is no indication that this stela was damaged before 612 BC. Phase 24. Sargon II1(721-705 BC) rebuilt the Nabu Temple, possibly in 713 BC. There is an inscription, written towards the end of his reign, which mentions Jshtar of Nineveh in a list of pious works (Thompson 1940: 86-8). According to Ashurbanipal (Borger 1996: 254-5), Sargon built (or restored) the bit akitu for Ishtar. A substantial block of limestone, tentatively ascribed to Sargon or Sennacherib by Borger (1967: 536), and to Sargon by Frahm (2000: 77), was found between the Ishtar and Nabu Temples (Thompson and Hutchinson 1929a: 120, P1. XLII.43); it refers to a bit akitu. Phase 25. Sennacherib (704-681 BC), on a damaged stone tablet dated c. 702 BC from the slope below the temple, records that he restored the temples of gods including Sin, Ningal, Shamash, Aya (four in one building, as confirmed by Ashurbanipal, apud Fuchs 1996: 291), and the lady of Nineveh, and he refers back to Ashurnasirpal as a previous builder; he probably claims to have used stone and to have rebuilt the place from top to bottom (Frahm 1998: 109-10). Fragments of Sennacherib foundation cylinders found on the temple site belong, according to Frahm (1997: 45-6, 137), to editions dated to 702 and 700 BC, the building sections on which described the South-West Palace and general improvements to Nineveh. Sennacherib's use of such unfocused texts as foundation documents of temples would suit his disdainful attitude to religion as a mere political tool. Physical evidence for Sennacherib's work on the temple is modest. A brick inscription was reported from the fill of Cellar 1 (Thompson and Hamilton 1932: 66: text unidentified), which suggests that he repaired the south-western gate-chamber. There was a stone slab fragment, with a text beginning e.gal, from Square I, 1 foot (31 cm) below datum, which was about level with the floor in the north-western gate-chamber (ibid.: 114, No. 262 =271). The inscriptions on two bricks from the temple site simply give Sennacherib's name and titles (ibid.: 116). These do not reveal what if anything he rebuilt from top to bottom, but the reference to limestone may mean that he reinforced the foundations on the exterior. The survival of the Ashurnasirpal wall-panel in the north-western gate-chamber suggests that at least part of the ninth-century building was still structurally sound. A broken Sennacherib text on a "small white limestone tablet", dated 690 BC, mentions a bit akitu, Eshahullezenzagmukam, in connection with the Nergal Gate in the outer wall of Nineveh (Ali and Grayson 1999); the excavator mentioned, during the 2003 Rencontre, that the stone was part of the paving of the courtyard of the gate. So it was reused, e.g. during repairs about 613 BC.
381
the version Frahm(2000: 75-7) suggestedthat this bit akitu might have temporarily superseded on Kuyunjik. to Emashmash Phase26. A list of the piousworksof Esarhaddon (680-669 BC) includesa reference on an Esarhaddonprismfragmentfrom (Borger 1956:94). No buildinginscriptionis preserved the templearea (ibid.: 38, Nin. D). Phase 27. Texts of Ashurbanipal(668-631 BC) provide furtherevidence on the bit akitu. The buildinginscriptionon fragmentsof PrismT, includingone dated 646, which were found northwest of the Ishtar Temple,describeshow he restoredand decoratedwith glazed bricks the bit akitu of the goddess which had been built by Sargon (Borger 1996: 132, 254-5, 257: his military BM 121006+ 127889); a later account mentions that the glazed bricks represented triumphs(Fuchs 1996:268-70, 291). Other Ashurbanipal prism fragmentsfrom the area have been assignedby Borgerto the PrismA edition (c. 644 BC), the buildingsection on which refers to the nearbyNorth Palace, but no prism fragmentcan really be assignedto a single "edition" unless it includesboth annalisticand buildingsections.We also know that the goddess travelled to the bit akitu by chariot (Livingstone 1989: 19), and that ceremoniesof this kind involved triumphal public processions (e.g. Weissert 1997: 346-50), but the location of the building remainsunclear. There are indeed glazed bricks of the eighth or seventhcenturythat were found betweenthe Ishtar and Nabu Temples (Thompson and Hutchinson 1929: 81; 1929a: P1.LVII.358; 1931: P1.XXVI.2,4; Nunn 1988:171-2, Abb. 131-8), but none of themis recorded as showingnarrative scenes. Instead they are bricks in high relief, a new techniquein Assyria. The subject-matter and ... vineswithgrapes"; includes"a wingedbull,a figureholdingthe usualpannikin, illustrations also show a probablesphinx,a bush or tree, and an enigmaticman whose head is bare (unlesshe wore a diadem,now lost), whose hair is arrangedin a late eighth- or seventh-century bun, and and earlierkings. whose right hand is raisedin the finger-snapping gestureused by Sennacherib While magic figuresare unexceptional, the scale and subject-matter of these fragmentsrecallthe glazed-brick panels from Khorsabadon which Sargon'sname was represented by symbols (e.g. Finkel and Reade 1996). The Nineveh brickslook like anotherroyal name, from anothertemple fa9ade.Other such high-reliefbricksshowing "mysticfigures"were used in one of the drainsof North Palace, built c. 646 (Rassam 1897: 222); these were presumablysurplus Ashurbanipal's stock from a temple;one brickwhich may come from this group shows a sphinxand is unglazed (BM 138719). A later inscription of Ashurbanipal(Fuchs 1996: 291) also describes how he decorated Emashmash,the IshtarTemple,with gold and silver, and how he worked on the ziggurratand outer doors. This work is listed with that in many other cities, and must have precededthe operationdescribedin the main buildinginscription,at the end of the text (ibid.: 258-9, 295). This latteris again concernedwith Emashmash, a formerking [Ashurnasir]pal, dilapidation,and restoration with stone (ibid.: 258-9, 295). The text was writtenon four largestone slabs, possibly door-sills,found in fragments "chieflytowardsthe S.E. side of the greatfoundation".It has many similaritiesto the text on a fragmentary prism found north-westof the temple (Borger 1996: 199-200); the buildinginscriptionactuallywrittenon what appearsto be a detachedfragmentof the selfsameprism,dated by the eponymof c. 634 BC (BM 122613:Millard1968:111, P1.XXVII; Reade 1998:257; cf. Borger1996:356), awaitsan editor. At least nine paving-stones, possiblyof the same date, have a much shortertext which refersto enlargingthe court (kisallu)of the Ishtar Templeand paving it with stone (Streck 1916:274-7; exampleslisted by Borger 1996:354). It so happens that, in the outer courtyard,Thompson identified"the remainsof the latest Assyrian restoration of the Temple ..., two adjacent libn pavements between 00 and DD on the top of the older courtyardof burnt brick ... the level of their top being 1 foot (31 cm) above datum"(Thompsonand Hamilton1932:58, 71, P1.XLVI.3);theirpositionis suggestedon Fig. 8a. An outdoor mud-brick"pavement" is impossible.Maybe therewere two layers of mud brickon top of which a more resistantsurfacewas once laid (as happenedwith PavementI), and this was Ashurbanipal's stone pavement.He could well have describedhis work in the area as a renewal
382
JULIAN READE
the walls, to judge by the numberof since the latterhad at least refurbished of Ashurnasirpal's, could also have recycled This is the timewhenAshurbanipal his wall-pegsfound in the courtyard. possiblyfor use in the Nabu Templein the outercourtyard wall-panels, hunt-scene Ashurnasirpal's (Streck 1916:272-5; Borger1996:353-4). of which he was layingvery similarpaving-stones must have finely carved with magical figures,probably seventh-century, An incense-burner, stood at a temple doorway (BM 1930-5-8,218: Thompson and Hutchinson 1931: P1.XXVII; Reade 2002a: 151, Fig. 14). There are two stone heads of women of similardate: one belonged to Layard (BM 1848-11-4,317 = 118897), the other was found by Smith in the temple area in the shrine,likea comparable (S 2496= BM 135106);theyprobablybelongedto statuesdedicated 1970:Taf. 19,20a-c). Two fragmentary statuefromnearthe IshtarTempleat Ashur(Strommenger apotropaicfigurineswere publishedfrom the 1930-1 season (Thompson and Hamilton 1932: P1.LXVII.I and 25 = BM 1932-12-10,48); one is part of a dog from "surfaceof mound", so like the North Palace,but the other is from SquareN, 2 feet might have derivedfrom somewhere (62 cm) below datum,and probablycame from the IshtarTemple.
The Late Assyrian temple archives
While the evidencefor the various ancient librariesof Kuyunjikis very confused and needs mixed" (e.g. Lambertand Millard 1968: ix; careful treatment,they were never "inextricably Parpola 1986:223; cf. Reade 2000: 421-6). It would be easy to collect togetherthe remainsof tablets which were kept in the IshtarTemplein the seventhcentury(Reade 1986:217-8). The and might help in our grasp of Assyrian bulk of the materialmay be earlierthan Ashurbanipal, historyif the projectwere of interestto a philologist.Whetherany prismsand tablets intellectual in the templelibraryis unclear. were also preserved with buildinginscriptions Specifically,to judge by the provenancesrecordedby Thompsonand preservedby Lambert and Millard(1968: 89-93), the libraryshould have containedmost of the cuneiformtexts, apart from prisms,wall-pegs,wall-panelsand other buildinginscriptions,found on Kuyunjikduring BM 121033-121131 found and 128017-128152, numbered 1928-32. Some of the tablet fragments in 1927-8, are likely to be from the Ishtar Temple;others will be from the Nabu Templeand House"). Many more of North Palace,with a few possiblyfrom the outer town ("Sennacherib's they are most of the the tabletsfrom the 1929-30 season are likelyto be from the IshtarTemple: and 139344-9.The 1930-31excavation with numbersin the rangesBM 122624-122657 fragments was largelyinside the IshtarTemplesite; tabletswhich should be from this season have numbers The 1931-2 excavain the rangesBM 123356-123409,123442-5, 123482-8, and 134774-134810. tions were largelyinside or near the IshtarTemple,but there were also soundingsin the SouthWest Palace and outer town: tablet fragmentshave numbers in the ranges 134496-134604, 134811-134833,137310-11 and 138659.Once the nature of the libraryis better understood,it should be possible to identify more tablets from it among those excavatedby Rassam, Smith and King. recordsfrom the Neo-Assyriantemple,althoughtwo docuThere are very few administrative ments relate to its possible function as a bank (Mattila 2002: 102-3). There is also a group of lions, which were used on very large and obviouslyofficialstamp-sealimpressions,representing 1995: 189); they are unprovenanced, storagejars (Herbordt 1992:142,Taf. 19;Curtisand Reade of from various excavationsat Nineveh. What looks like another them, describedas a "lion modelledin relief", was found in SquareC north-westof the temple,Fig. 21A (Thompsonand Hutchinson 1931: 83, PI.XXII.3: Iraq Museum);it would have been easy, for the excavators to supposethat it had been modelledratherthan handlinga singleone of these seal impressions, were it Ishtar's the lion was Since animal, seems quite likely that the lion stamp-seals stamped. There is also a sherd paintedwith a stridinglion, Fig. 21B used by her temple administrators. (ibid. PI.XXII.2: 1930-5-8, 194), and one royal sealing (Thompson and Hamilton 1932: P1.LXIII.5).
Other structuresin the Emashmashprecinct
The precinctextendedbeyondthe main temple.At Nineveh,as at Ashur,the templemust have edge of the precinct;this is because it overlooked a been on or close to the north-western
383
CJ~
All. ~~~~~~
Ci/8
iwtn moJdId in \4I Aeie/ (\)
\/
D
C
c
ttO.
Wht
00
6. Cho&lco...
seaL)
sLcrc etL7ik
toio
swt.
(e)
(Cq4.tie
0 WLit_
a::nJ Seat.
.%
#9
_ .9:
*9
.*
* . .
, S9
* * ,,
SowsZ(cbo7i3C>);
Ckam6er/L
C.25,6/o&r
Floor . (?$).
by Thompsonand his staff,probablyall completedin the evenings Fig. 21. Typicaldrawings at Nineveh. A. Lion seal-impression? (Thompson and Hutchinson 1931: P1.XX11.3); B. Paintedsherd(ibid.: P1.XXII.2); C-D. Seals(Thompsonand Hamilton1932:P1.LXIII.1-2); E. Carvedbox (Thompsonand Hutchinson1931:P1.XXII.9). Drawingslike these, not ideal but far betterthan nothing,madepromptpublication possible:each usuallyhas a provenance and scale, and items kept in Iraq were annotatedwith (B) for Baghdad. depression, beyond which the Nabu Temple was eventually built, and tribute presumably entered the temple through the Lion Gate on this side. The other prominent building in the precinct must have been the ziggurrat of Ishtar. ShamshiAdad I's text, though incomplete, seems to say that Manishtushu had built a ziggurrat in Emashmash (Grayson 1987: 51-5), which would make it the earliest "ziggurrat" actually known as such. Shamshi-Adad's version was bigger and better, and he called it Ekitushkuga. In the later periods the ziggurrat was separate from the main temple, like the main ziggurrat at Ashur. While some texts of Shalmaneser I do refer in one breath to the repair of both temple and ziggurrat at
384
JULIAN READE
Nineveh after an earthquake (Grayson 1987: 208), no ziggurrat is mentioned in the very numerous texts of Ashurnasirpal II describing his own repairs to the temple. Sennacherib seems to refer to it once simply as the ziggurrat and once as the ziggurrat of the house of Ishtar (Luckenbill 1924: 99, 102). It is the pride of Nineveh in a Neo-Assyrian prayer to the goddess (Livingstone 1989: 18). Ashurbanipal too refers to the ziggurrat of Nineveh, which he probably redecorated; he calls it Ekibikuga rather than Ekitushkuga (Fuchs 1996: 269, 291). The Sennacherib reference gives the position of the ziggurrat in relation to the South-West Palace, suggesting that as at Ashur the ziggurrat was roughly opposite the South-West Gate of the main temple; modern contours suggest the presence of a high structure in this part of Kuyunjik. In attempting to locate the Ishtar ziggurrat itself more precisely, I have sometimes mistakenly quoted (Reade 2000: 407) King's 1904 statement that he found, about 10-13 m down "in the centre of the mound", a wall "solidly built of rough hewn stones", and that "in a layer of debris near the foot of the wall were found fragments of stone and black basalt bearing inscriptions in very archaic characters" (D'Andrea 1981: 112). There are such pieces from King's excavation, the fragments of the Shamshi-Adad inscription, which mention the ziggurrat, but as noted above (Phase 7), these probably really came from the temple site. This Ishtar ziggurrat must be distinguished from the ziggurrat belonging to the Adad Temple, which was not far away: many Ashurnasirpal wall-pegs referring to his restoration of the Adad Temple were found just west of the Ishtar Temple (Grayson 1991: 337-9), and a Sargon II prism recording his restoration of the Adad ziggurrat was found in the region of the South-West Palace (Fuchs 1998: 4-5). Sennacherib (Luckenbill 1924: 99), by relating the position of the Kidmuri Temple to the SouthWest Palace, showed that it was east of the Ishtar Temple. Menzel (1981: 1, 121-2; Reade 2000: 409-10) suggested that the Kidmuri Temple and various other shrines were also included in Emashmash. One may have been the shrine of Gula: part of ajar, impressed with a seal representing this goddess, emerged in the 1927-8 season, Fig. 22 (Thompson and Hutchinson 1929a: P1. LVII.340; Collon 2001: 122). The White Obelisk, and possibly Ashur-bel-kala's naked goddess, were found about 100 m south-south-west of the temple; they would presumably have been suitable decoration for a gate into the precinct. This would have looked down towards the old South-West Palace on the right and the main East Gate of Kuyunjik on the left. The difference in floor-level between the exterior of the Ishtar Temple itself and the South-West Palace in the seventh century was about 6 m; this is estimated from the elevation of the palace floor in domestic quarters south-west of the ziggurrat, which was c. 23.5-25 feet (7.2-7.7 m) below datum (Thompson and Mallowan 1933: 72-4, P1.CVI). Some Neo-Assyrian wall-panels, which showed a royal procession with acolytes and musicians moving up and down a slope (Barnett et al. 1998: Pls. 473-96), were found near the
BRO,NZE LAMP.
Fig. 22.
Seal-impression on sherd, showing Gula (Thompson and Hutchinson 1929a:
P1.LVII.340). Fig. 23. Lampfrom one of the Phase 30 buildings(Smith 1875:140).
385
with the platform WhiteObeliskand may have lined the sides of a bridgeconnectingEmashmash South-West Palace. of Sennacherib's
The destructionof the temple (Phases 28-30)
of the Templeby fire,presumably at the Destruction Phase28: The "Thedestruction fall of Nineveh. of 612 BC, was obvious on the face of the trenchin Sects. O-N-M as one looked S.W. Here the ash, 1 foot thick (31 cm), was inclinedat an angle of 450 from the libn to a furtherdepth of 4 feet (123 cm), and ran along for about 20 feet (6.2 m) away from the libn" (Thompson and Hamilton 1932:73; section in Gut 1995:43, Abb. 24b). This quantityof ash suggestsan origin inside the temple.Therewas more ash elsewhere,but none necessarily from the destruction.The face of Ashurnasirpal's wall-panel in Squares 10 had been affected by fire (Thompson and Hamilton 1932:70), and must have remainedvisible till 612 BC. The floor surfaceson eitherside wall in SquaresRST seem the only ones above or beside which, though not of Ashurnasirpal's mentionedby Thompson,destructiondebrismight have been found in position, as suggestedin Fig. 8b. Besidesthe possiblestatue of Shalmaneser III, thereare three smallerobjectswhich may have been close to the positionsthey had in 612 BC. One is the Old Babyloniancylinder-seal with an eroticscene, Fig. 21C. A stone handlein the shapeof a ram'shead, with tracesof yellow paint, has the same provenance(ibid.: P1.LXII.12 = BM 1932-12-10, 46). Thereis also the greaterpart of a fine limestoneoil-jar, from Square S at 1 foot (31 cm) below datum (ibid.: P1.LXIV.I = BM 1932-12-10,93), but its date is unsure. Naturally many of the other items which were distributed pellmellacross the site, includingscrapswhich the reportsdo not illustrate,are likely to have been left behind in 612 BC, and some may have been in the temple for centuries.Items from abroadincludea carvedstone box of the ninth or eighthcentury,perhapsfrom Carchemish, Fig. 21E (ibid.: P1.XXII.9, cf. Woolley 1969:P1.28.3-4), and a seventh-century stone cosmetic palettefrom the Levantcoast (Thompsonand Hutchinson1931:PI.XXII.4 = BM 1930-5-8,207, cf. H. Thompson: 1972). A broken chalcedony cylinder-sealhas an inscriptionsuggestingit belongedto an Elamiteor other Iranian,Fig. 21D (Thompsonand Hamilton 1932:P1.LXIII.2BM 1932-10-12, 670 = 136999).
Phase29: Seventh-sixthcentury. The temple,and possiblythe ziggurrat,stood on the high spine of Kuyunjik.There is no evidenceof deliberatedemolitionof the walls in 612 BC, but they were exposed to erosion. Their deteriorationis recordedin the scatteredand broken wall-pegs of Assyriankings.Thompson'snotes describea wall, visibleon his plan in Square0, whichabutted against the only standing wall-panel;it was made of "4 courses of chunks of marble, often inscriptionand sculpture,all brokenup", as shown in Figs 8b and 17. This recallsrepairsto the nearbyNabu Templesoon after 612 BC (Thompson and Hutchinson1929a:106-7), and similar elsewhereon Kuyunjik(references: reoccupation Reade 2000:428). So perhapsthe wall in Square O was built not long after 612 BC, while some templewalls were still upright,but it was thought to be Parthian(Thompsonand Hamilton 1932:108). Phase 30: Sixth centuryand later. Wishful thinking tended to assume that walls incorporating Assyrian building materialswere Assyrian in date (e.g. Thompson and Hutchinson 1931: 86; Thompson and Hamilton 1932:69), but many were really Greco-Parthian or later. Even in his final report Thompsondeprecatedthe possibilityof "wholesaledepredations" of tiles from the temple site (which might have seemed to detractfrom the value of his excavations),while also concludingthat "later occupantsof the mound ... found in (the temple) a plentifulsupply of well-puddled clay, and theirdepredations and the action of the weatherhad removedthe greater part of the libnwallingwhich had once formedchambersof the Temple-building on its surface" (ibid.: 57, 68); most of the facingmaterialssuch as bricks,tiles and stone panelshad indeedbeen removed.Some narrowwalls above the ancientmud brick around Square00 were ascribedto the seventhcenturyBC (ibid.: 72); these, however,like some others, and like the small buildings alwaysrecognised as Greco-Parthian in a streetcrossingthe QRWX area (ibid.: Pls. XCI-XCII), can only have been builtafterthe main superstructure of the templehad beenflattened.Xenophon did not comment on any settlementat Nineveh in 401 BC (Anabasis3: 4, 10), and Thompson
386
JULIAN READE
continuitybetweenthe temple of Phase 27 and these later found no evidencefor architectural structures. It does not necessarilyfollow that the temple precinctentirelylost its sanctity.In the GrecoParthianperiodtherewas clearlya majorshrinea little to the south (Reade 1998a:69-76). There were also tombs in the vicinity (Curtis 1976); they must have belongedto people who did not bury their dead outside city-walls,but this does not tell us the context. Smith (1875: 139-42), east of the centreof Kuyunjik,found "a smallchamberlike a shrine,solidly workingsomewhere built of stones and cement,the walls plasteredover and coveredwith a patternof lines disposed in lozenges. The shape of the chamberwas square, two cornersbeing ornamentedwith square pilasters,and at one end was a largecircularrecess.In the chamberI found a bronzelamp with two spouts for wicks",Fig. 23 (S 2495 = SOC79 = BM 131164)."I was not satisfiedthat this was reallyan Assyrianbuilding.... Verynear the chamber,I found the capitalof a largecolumnbut traced no buildingto which it could have belonged".St John Simpsonkindly tells me that the lamp is early Islamic;the room does sound like a shrine or chapel. Later Thompson was to on the mound(Reade 1998a:75, Fig. 12). capital somewhere photographa largeGreco-Parthian Smith also found "a squarechamber... built up of stones from the Assyrianbuildingsnear it. dedicatedto of Ashurbanipal All along the walls were placedsmall squareslabs with inscriptions the goddesssof Nineveh, none of them in their originalposition".Smith'stwo rooms must have been dismantledafter he left the site. In the templearea Thompsonfound many Greco-Parthian terracottas,which includeerotic scenes suitablefor the goddess, and stone statuettesof varying quality (Thompson and Hutchinson 1931: P1.XXIII.10; Thompson and Hamilton 1932: material.No Pls. LXII.8, LVII.23,LXVIII.1), but many sites of the periodproducecomparable shrineto Mullissuor Aphroditeon Kuyunjik,but we do not doubt there was a Greco-Parthian know its location. Conclusion Anyone travellingtoday from southernto northernIraq is immediatelyaware of enteringa are most obviously,the land is no longerflat, and palm-trees very differenttype of environment: less scarce. Five thousand years ago, when the wildlife and vegetation of both regions were degraded,the contrastmay have been greaterstill. The domainof the naturegoddessof Nineveh will have been distinctive,but she had enoughin commonwith SumerianInannato be identified whichgrew with her. Ishtarof Ninevehmust have begunlife as a local goddess,with a reputation she became earlier, perhaps becauseNinevehwas a regionalcentre.Duringthe thirdmillennium, famous outside this region. She was respectedby rulers of Agade and Babylon as well as by temple. built her a magnificent Hurriansand Hittites.Shamshi-Adad When Nineveh fell underAssyrianrule and becamethe second city of an expandingkingdom, restoredthe temple,and later kings took greatcare to maintainit properly.Just as Ashur-uballit Ashur, once the god of an obscure mountain, became a supremeimperialdivinity, Ishtar of Nineveh also evolved. In Late Assyrianofficialsourcesshe seems less the motherand more the amazon. While her temple was still the place where women made dedications,the exerciseof Assyrian power was displayed on stone panels, obelisks and glazed tiles of which numerous monumentsmeantto the enemieswho capturedNineveh fragmentssurvive.What these narrative out of doors and accessible,they werebrokenup in condition: is current in 612 BC reflected their similarthemesin royalpalaces.It is doubtful the than showing wall-panels muchmorethoroughly of imperial the goddessand hercult, stripped Still of worship. a remained place whetherthe temple name of the of commemorates town Karamleis the survived: must nearby have today baggage, Mullissu. Thompson. in this paperhas beenderivedfromthe recordsof Campbell Most of the information They have turnedout, like the recordsof Mallowan'sDeep Soundingstudiedby Gut (1995), to includeevidencefor They apparently detailedto allow for extensivereassessment. be sufficiently sequenceof the later IshtarTemple, a Ninevite 5 shrine;for the ground-planand architectural contextsof the so-calledHead of to the AshurTemple;for the archaeological and its relationship texts from the temple;for some of the contentsand decoration Sargonand of the Old Babylonian of the Late Assyriantemple;and for what happenedto the buildingafter 612 BC. This was a
387
considerableachievementfor the philologistdescribedby Mallowan as having "no high regard for archaeology".Uncertainties remainbut, of all the buildingson Kuyunjik,the IshtarTemple is the one most completelyexcavatedand most fully published.
Addendum
BM 1932-12-10, 424 = 123481,a whitishlimestoneflake (convex, not from a cylinder),found in 0(?) 3, has the Shamshi-Adad text (iii.26-iv.3) in slightlylargerscript,identifiedin consultation with Dr FaroukAl-Rawi.
References Ali Yaseen Ahmad and A. K. Grayson 1999 Sennacherib in the Akitu House. Iraq 61: 187-9. Andrae, W.
1925 Coloured ceramics from Ashur.London.

1935 Die jiungerenIschtar-Tempel in Assur. Mitteilungen der Deutschen Orient-Gesellschaft73: 1-12. 1935a Die jungeren Ischtar-Tempel in Assur (Wissenschaftliche Ver6ffentlichung der Deutschen OrientGesellschaft 58). Leipzig. Barag, D. P.
1985 Catalogue Asiaticglass in the BritishMuseum of Western I. London.

Barnett, R. D.
1976Sculptures from the NorthPalaceof Ashurbanipal at Nineveh.London.

Barnett, R. D., E. Bleibtreu and G. Turner
1998Sculptures from the Southwest Palaceof Sennacherib at Nineveh.London.

Beckman, G. 1998 Istar of Nineveh reconsidered. Journal of CuneiformStudies 50: 1-10. Bollweg, J.
1999 Vorderasiatische im Spiegel der Terracottaplastik Wagentypen bis zur altbabylonischen Zeit (Orbis
Biblicus et Orientalis 167). Freiburg/Gottingen. Borger, R. 1956 Die Inschriften Asarhaddons, Konigs von Assyrien (Archiv fur Orientforschung, Beiheft 9). Graz.
1961Einleitung in die Assyrischen Konigsinschriften. Leiden/Koin. 1967Handbuch derKeilschriftliteratur, BandI: Repertorium dersumerischen undakkadischen Texte.Berlin. 1996Beitrdge zumInschriftenwerk Wiesbaden. Assurbanipals.
B6rker-Kldhn, J.
1982Altvorderasiatische Bildstelen undvergleichbare Felsreliefs (BaghdaderForschungen 4). 2 vols. Mainz

am Rhein. Brinkman, J. A.
1976 Materialsand studiesfor KassitehistoryI: A catalogueof cuneiform sourcespertainingto specific monarchs of the Kassitedynasty.Chicago.
Budge, E. A. W.
1914Assyrian in the BritishMuseum: sculptures Reignof Ashur-nasir-pal. London.

Charpin, D. 1998 Iluni, roi d'Esnunna. NABU 1998(1), No. 29: 33-4. Charpin, D. and J.-M. Durand 1997 A??ur avant l'Assyrie. MARI 8: 367-91. Cole, S. W. and P. Machinist
1998Letters fromprieststo thekingsEsarhaddon andAssurbanipal (State Archives of Assyria13). Helsinki.

Collon, D.
2001 Catalogue Asiaticseals in the BritishMuseum: of the Western seals V: Neo-Assyrian Cylinder andNeoBabylonian periods.London.
2003 The Ninevite 5 seal impressions. In E. Rova and H. Weiss (eds.), The origins of North Mesopotamian
civilization: Ninevite5 chronology, economy, society(Subartu9): 249-69. Turnhout.
Curtis, J. E. 1976 Parthian gold from Nineveh. In British Museum YearbookI: The classical tradition:47-66. London. 1983 Some axe-heads from Chagar Bazar and Nimrud. Iraq 45: 73-81. Curtis, J. E. and J. E. Reade (eds.)
1995Art and empire:Treasures from Assyriain the BritishMuseum. London.

Dailey, S. 2001 Old Babylonian tablets from Nineveh; and possible pieces of early Gilgamesh epic. Iraq 63: 155-67.
D'Andrea,M. M. 1981 Lettersof LeonardWilliam King, 1902-1904; introduced, editedand annotated withspecialreference
to the excavations of Nineveh (MA thesis, University of Wisconsin). River Falls.
388
JULIAN READE
Dossin, G. 1950 Correspondancede Samsi-Addu (Archives Royales de Mari I). Paris. Driver, G. R. and J. C. Miles 1955 The Babylonian laws II: Transliteratedtext, translation,philological notes, glossary. Oxford. Durand, J.-M. 1987 Villes fantomes de Syrie et autres lieux. MARI 5: 199-234. Finkel, 1. L. and J. E. Reade 1996 Assyrian hieroglyphs. Zeitschrift fur Assyriologie 86/2: 244-68. Frahm, E. 1997Einleitung in die Sanherib-Inschriften(Archiv filr Orientforschung, Beiheft 26). Wien. 1998 Sanherib und die Tempel von Kuyunjik. In S. M. Maul (ed.), Festschriftfur Rykle Borger zu seinem 65. Geburtstag am 24. Mai 1994: Tikip santakki mala balmu (CuneiformMonographs, 10): 107-21. Groningen. 2000 Die Akitu-Hdiuservon Ninive. NABU 2000/4, No. 66: 75-9. Frayne, D.R. 1993Sargonic and Gutianperiods (2334-2113 BC) (The Royal Inscriptions of Mesopotamia: Early periods 2). Toronto/Buffalo/London. Fuchs, A. 1996 Die Inschrift vom IUtar-Tempel.In Borger 1996: 258-96. 1998 Die Annalen des Jahres 711 v. Chr. (State Archives of Assyria Studies 8). Helsinki. Gadd, C. J. 1936 The stones of Assyria. London. George, A. R. 1979 Cuneiform texts in the Birmingham City Museum. Iraq 41: 121-40. 1993 House Most High: The temples of ancient Mesopotamia. Winona Lake. Grayson, A.K. 1987 Assyrian rulers of the third and second millennia BC (to 1115 BC) (The Royal Inscriptions of Mesopotamia, Assyrian periods 1). Toronto/Buffalo/London. 1991 Assyrian rulers of the early first millennium BC, Part 1 (1114-859 BC) (The Royal Inscriptions of Mesopotamia, Assyrian periods 2). Toronto/Buffalo/London. 1996 Assyrian rulers of the early first millennium BC, Part II (858-745 BC) (The Royal Inscriptions of Mesopotamia, Assyrian periods 3). Toronto/Buffalo/London. Groneberg, B. R. M. 1997 Lob der I.tar: Gebet und Ritual an die altbabylonische Venusgottin: Tanatti Iftar (Cuneiform Monographs 8). Groningen. Gut, R. V. 1995 Das prahistorische Ninive: Zur relativen Chronologie der fruhen Perioden Nordmesopotamiens (Baghdader Forschungen 19). 2 vols. Mainz am Rhein. Gut, R. V., J. E. Reade and R. M. Boehmer 2001 Ninive-Das spate 3. Jahrtausend v. Chr. In J.-W. Meyer, M. Novak and A. Pruss (eds.), Beitrdge Archdologie WinfriedOrthmanngewidmet: 74-129. Frankfurt am Main. zur Vorderasiatischen Haller, A., and W. Andrae 1955 Die Heiligtumer des Gottes Assur und der Sin-Sama1-Tempel in Assur (Wissenschaftliche Veroffentlichung der Deutschen Orient Gesellschaft 67). Berlin. Hallo, W. W. 1957 Early Mesopotamian royal titles: A philologic and historical analysis (American Oriental Series 43). New Haven. Heinrich, E. 1982 Die Tempel und Heiligtumer im alten Mesopotamien: Typologie, Morphologie und Geschichte (Denkmaler Antiker Architektur 14). 2 vols. Berlin. Herbordt, S. 1992Neuassyrische Glyptik des 8.-7. Jh. v. Chr. (State Archives of Assyria Studies 1). Helsinki. Howard-Carter, T. 1983 An interpretation of the sculptural decoration of the second millennium temple at Tell al-Rimah. Iraq 45: 64-72. Joannes, F. 1993 La culture materielle a Mari (V): Les parfums. MARI 7: 251-70. Paris. Lambert, W. G. 1992 Catalogue of the cuneiform tablets in the Kouyunjik Collection of the British Museum: Third supplement. London. 2004/5 Igtar of Nineveh. Iraq 66 (2004): 35-9 = RAI 49/1 (2005): 35-9. Lambert, W.G. and A. R. Millard 1968 Catalogue of the cuneiform tablets in the KouyunjikCollection of the British Museum: Second supplement. London.
389
Livingstone, A. 1989 Court poetry and literary miscellanea (State Archives of Assyria 3). Helsinki. Luckenbill, D. D. 1924 Sennacherib, king of Assyria: Annals (Oriental Institute Publications 2). Chicago. Mallowan, M. E. L. 1936 The bronze head of the Akkadian period from Nineveh. Iraq 3: 104-10. 1977 Mallowan's Memoirs. London. Mattila, R. 2002 Legal transactions of the royal court of Nineveh II (State Archives of Assyria 14). Helsinki. Menzel, B. 1981 Assyrische Tempel (Studia Pohl: Series Maior 10). 2 vols. Rome. Millard, A. 1968 Fragnents of historical texts from Nineveh: Ashurbanipal. Iraq 30: 98-111. 1994 The eponyms of the Assyrian empire 910-612 BC (State Archives of Assyria Studies 2). Helsinki. Moorey, P. R. S. 1971 Catalogue of the ancient Persian bronzes in the Ashmolean Museum. Oxford. 1994 Ancient Mesopotamian materials and industries: The archaeological evidence. Oxford. Moortgat, A. 1969 The art of ancient Mesopotamia. London/New York. Nunn, A. 1988 Die Wandmalerei und der glasierte Wandschmuck im alten Orient (Handbuch der Orientalistik 7/l/2/B6). Leiden. Nylander, C. 1980 Earless in Nineveh: Who mutilated "Sargon's" head? American Journal of Archaeology 84: 329-33. Oates, D. 1967 The excavations at Tell al Rimah, 1966. Iraq 29: 70-96. 1968 Studies in the ancient history of northern Iraq. London. 1972 The excavations at Tell al-Rimah, 1971. Iraq 34: 77-86. Parpola, S. 1986 The royal archives of Nineveh. In K. R. Veenhof (ed.), Cuneiformarchivesand libraries:223-36. Leiden. Pittman, H. 1996 The White Obelisk and the problem of historical narrative in the art of Assyria. Art Bulletin 78: 334-55. Postgate, C., D. Oates and J. Oates 1997 The excavations at Tell al Rimah: The pottery (Iraq Archaeological Reports 4). London. Rassam, H. 1897 Asshur and the land of Nimrod. New York/Cincinnati. Reade, J. E. 1968 Tell Taya (1967): Summary report. Iraq 30: 234-64. 1973 Tell Taya (1972-3): Summary report. Iraq 35: 155-87. 1975 Ashurnasirpal I and the White Obelisk. Iraq 37: 129-50. 1977 Shikaft-i Gulgul: Its date and symbolism. Iranica Antiqua 12: 33-44. 1981 Fragments of Assyrian monuments. Iraq 43: 145-56. 1986 Archaeology and the Kuyunjik archives. In K. R. Veenhof (ed.), Cuneiform archives and libraries: 213-22. Leiden. 1987 Was Sennacherib a feminist? In J.-M. Durand (ed.), La femme dans le Proche-Orient antique: 139-45. Paris. 1998 Assyrian eponyms, kings and pretenders, 648-605 BC. Orientalia 67: 255-65. 1998a Greco-Parthian Nineveh. Iraq 60: 65-83. 2000 Ninive (Nineveh). Reallexikon der Assyriologie 9 (5-6): 388-433. Berlin/New York. 2000a Early Dynastic statues in the British Museum. NABU 2000(4), No. 73: 82-6. 2000b Early foundation records from the Istar Temple at Nineveh. NABU 2000(4), No. 75: 86-7. 2000c Assyrian sculptures in the British Museum: Technical notes. NABU 2000(4), No. 78: 88. 2002 Sexism and homotheism in ancient Iraq. In S. Parpola and R. M. Whiting (eds.) Sex and gender in the ancient Near East (Comptes rendus, Rencontre Assyriologique Internationale 47): 551-67. Helsinki. 2002a The ziggurrat and temples of Nimrud. Iraq 64: 135-216. Reade, J. E. and C. B. F. Walker 1982 Some Neo-Assyrian royal inscriptions. Archivfur Orientforschung28: 113-22. Russell, J. M. 1998 Thefinal sack of Nineveh. New Haven/London. 2003 Obelisk. Reallexikon der Assyriologie 10 (1/2): 4-6. Berlin/New York. Searight, A., I. L. Finkel and J. E. Reade In preparation Assyrian stone vessels and related material in the British Museum. Seux, M.-J. 1976 Hymnes et prieres aux dieux de Babylonie et d'Assyrie. Paris.
390
JULIAN READE
Smith, G. 1875 Assyrian discoveries. London. Soliberger, E. 1974 The White Obelisk. Iraq 36: 231-8. Stepniowski, F. M. 1988 Metrologische und geometrische Interpretationen der Grundrisse sakrale Bauwerke in Assur. Mitteilungen der Deutschen Orient-Gesellschaftzu Berlin 120: 173-88. Streck, M. 1916 Assurbanipal und die letzten assyrischen Konige bis zum Untergange Nineveh's (Vorderasiatische Bibliothek 7). 3 vols. Leipzig. Strommenger, E. 1970 Die neuassyrische Rundskulptur(Abhandlungen der Deutschen Orient-Gesellschaft 15). Berlin. Tenu, A. 2004/5 Ninive et A?gur a l'epoque medio-assyrienne. Iraq 66 (2004): 27-33 = RAI 49/1 (2005), 27-33. Thompson, H. 0. 1972 Cosmetic palettes. Levant 4: 148-50. Thompson, R. Campbell 1934 The buildings on Quyunjiq, the larger mound of Nineveh. Iraq 1: 95-104. 1940 A selection from the cuneiform historical texts from Nineveh (1927-1932). Iraq 7: 85-131. Thompson, R. Campbell, and R. W. Hamilton 1932 The British Museum excavations on the Temple of Ishtar at Nineveh, 1930-31. Annals of Archaeology and Anthropology 19: 55-116. Thompson, R. Campbell, and R. W. Hutchinson 1929 A century of excavation at Nineveh. London. 1929a The excavations on the Temple of Nabu at Nineveh. Archaeologia 79: 103-48. 1931 The site of the Palace of Ashurnasirpal at Nineveh, excavated in 1929-30 on behalf of the British Museum. Annals of Archaeology and Anthropology 18: 79-112. Thompson, R. Campbell, and M. E. L. Mallowan 1933 The British Museum excavations at Nineveh, 1931-32. Annals of Archaeology and Anthropology 20: 71-186. Tubb, J. N. 1985 Some observations on spearheads in Palestine in the Middle and Late Bronze Ages. In J. N. Tubb (ed.), Palestine in the Bronze and Iron Ages: Papers in honour of Olga Tufnell: 189-96. London. Veenhof, K. R. 2000 Ninive 5. Akkadische und altassyrische Periode. Reallexikon der Assyriologie 9 (5-6): 433-4. Berlin/New York. 2003 The Old Assyrian list of year eponyms from Karum Kanish and its chronological implications (Publications of the Turkish Historical Society VI.64). Ankara. Vincente, C.-A. 1995 The Tall Leilan recension of the Sumerian king list. Zeitschriftfir Assyriologie 85: 234-70. Weidner, E. F. 1939 Die Reliefs der assyrischen Konige (Archiv filr Orientforschung, Beiheft 4). Berlin. Weiss, H. 1985 Tell Leilan on the Habur plains of Syria. Biblical Archaeologist 48: 5-34. Weissert, E. 1997 Royal hunt and royal triumph in a prism fragment of Ashurbanipal. In S. Parpola and R. M. Whiting (eds.), Assyria 1995: 339-58. Helsinki. Westenholz, A. 2000 Assyriologists, ancient and modern, on Naramsin and Sharkalisharri. In J. Marzahn and H. Neumann (eds.), Assyriologica et Semitica: FestschriftfiurJoachim Oelsneranlasslich seines 65. Geburtstages am 18. Februar 1997 (Alter Orient und Altes Testament 252): 545-55. MOnster. Westenholz, J. G. 2004/5 The Old Akkadian presence in Nineveh: Fact or fiction. Iraq 66 (2004): 7-18 = RAI 49/1 (2005): 7-18. Wilcke, C., and U. Seidl 1976 Inanna/Iftar. Reallexikon der Assyriologie 5(1/2): 74-89. Berlin/New York. Wiseman, D. J. 1962 Catalogue of the WesternAsiatic seals in the British Museum: Cylinderseals I: Uruk-Early Dynastic periods. London. Woolley, C. L. 1969 CarchemishII: The town defenses. London. Wu Yuhong 1994 The localisation of Nurrugum and Ninet = Ninuwa. NABU 1994(2), No. 38: 35-7. Ziegler, N. 2004/5 The conquest of the holy city of Nineveh and the kingdom of NurrugOmby Samsi-Addu. Iraq 66: 19-26=RA149/1 (2005): 19-26.

Temple of Ishtar at Nineveh

Uploaded by

Copyright:

Available Formats

Temple of Ishtar at Nineveh

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Temple of Ishtar at Nineveh

Uploaded by

Copyright:

Available Formats

The Ishtar Temple at Nineveh Author(s): Julian Reade Source: Iraq, Vol. 67, No. 1, Nineveh.

THE ISHTAR TEMPLE AT NINEVEH

c.640 BC: sketch

about 640 BC. Author'sdrawing. Fig. 1. Kuyunjik

THE ISHTAR TEMPLEAT NINEVEH

THE ISHTAR TEMPLE AT NINEVEH

post-Assynan walls (ePalace of Ashurnasirpal")

QRWX room t(Ninevite 5

THE ISHTAR TEMPLE AT NINEVEH

0 ' ;0000 a -I11 a f 11 l >1;A~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~i

t~~-= 4 ". tr;

Phase 30 walls, pits and other remains

Phase 3 (Ninevite 'ars?) 5 (Ninevite jars?) 5

(Phase 0?) ~~~~~~~~Pavem-ent 11 (Phase 9?) ~~~~~Pavement Ill(Phase 7) ~~~~~~~Pavement

in feet = c. 305 cm)

THE ISHTAR TEMPLE AT NINEVEH

Phase wall~~~ 29 Phase 29 wall

-wall-panel, <pavem ent

Phase 20:l InnerCourtyardlevel

Phase 30: intrusive walls, pits and other remains

5_ -18 (1 0 feet = c. 305 cm)

-20 0 X . 5 [ 10| 0, centre 0, S corner

2 ~~~~~~~~~~~~~~~~~~Phase 1 (bevelled-rimbowls) W, centre

Fig. 9. Room in SquaresQRWX:Possiblythe Ninevite5 shrineof Ishtar(Thompsonand Hamilton1932:P1.XLVII.1).Printfrom BritishMuseumarchive.

THE ISHTAR TEMPLEAT NINEVEH

THE ISHTARTEMPLEAT NINEVEH

r headof an Agadeking(IraqMuseum 11331). Fig.10. Copper

THE ISHTAR TEMPLEAT NINEVEH

THE ISHTAR TEMPLEAT NINEVEH

THE ISHTAR TEMPLEAT NINEVEH

THE ISHTAR TEMPLE AT NINEVEH

Fig. 14. Old Babylonian inventory (BM 137310). Photographs by D. Collon.

THE ISHTAR TEMPLEAT NINEVEH

THE ISHTAR TEMPLEAT NINEVEH

THE ISHTAR TEMPLE AT NINEVEH

THE ISHTAR TEMPLEAT NINEVEH

THE ISHTAR TEMPLE AT NINEVEH

THE ISHTAR TEMPLE AT NINEVEH

JULIAN READE of Ashr spttScLdtuytw(Pt.) L6r9edtaractzvs, L.ns o. Wok Inscriptiam.

THE ISHTAR TEMPLEAT NINEVEH

THE ISHTAR TEMPLE AT NINEVEH

iwtn moJdId in \4I Aeie/ (\)

Seal-impression on sherd, showing Gula (Thompson and Hutchinson 1929a:

P1.LVII.340). Fig. 23. Lampfrom one of the Phase 30 buildings(Smith 1875:140).

THE ISHTAR TEMPLEAT NINEVEH

THE ISHTAR TEMPLEAT NINEVEH

1925 Coloured ceramics from Ashur.London.

1985 Catalogue Asiaticglass in the BritishMuseum of Western I. London.

1976Sculptures from the NorthPalaceof Ashurbanipal at Nineveh.London.

1998Sculptures from the Southwest Palaceof Sennacherib at Nineveh.London.

1982Altvorderasiatische Bildstelen undvergleichbare Felsreliefs (BaghdaderForschungen 4). 2 vols. Mainz

1914Assyrian in the BritishMuseum: sculptures Reignof Ashur-nasir-pal. London.

1998Letters fromprieststo thekingsEsarhaddon andAssurbanipal (State Archives of Assyria13). Helsinki.

civilization: Ninevite5 chronology, economy, society(Subartu9): 249-69. Turnhout.

1995Art and empire:Treasures from Assyriain the BritishMuseum. London.

THE ISHTAR TEMPLEAT NINEVEH

You might also like

(Phase 0?) ~~~~~~Pavem-ent 11 (Phase 9?) ~Pavement Ill(Phase 7) ~~~~~Pavement