PART 1

PALACES OF THE ANCIENT

NEAR EAST AND

THE LATE-ANTIQUE

MEDITERRANEAN WORLD

"SEAT OF KINGSHIP"/"A WONDER TO BEHOLD":THE PALACE AS CONSTRUCT

BY FOLK DEFINITION, THE PALACE IS WHERE THE RULER

resides. In the successive kingdoms of ancientMesopotamia, however-Sumerian, Akkadian,Babylonian, Assyrian-the palace was the seat ofmany activities: administrative, bureaucratic, in-dustrial, and ceremonial as well as residential. Inbrief, it was an "institution," not just a "resi-dence"; part of the state apparatus, not merely acontainer of state apartments.

The word for palace in Sumerian and Akkadi-an, the languages of ancient Mesopotamia, iscomposed of the Sumerian sign for "house" fol-lowed by the adjective "large, great" (.gal).1 TheAkkadian borrowing is not a literal translation(where Akkadian "house" = bitu and "large" =rabi), but rather is formed from the Sumerian(ekallu), emphasizing the composite term as itsown cognitive category. At base, the underlyingadjective denotes scale, but may also be seen toreflect elevated (enlarged) status and function,such that a more accurate translation might be"the Great House," as opposed to merely "the bighouse."2

In a Mesopotamian gloss, preserved in the year-name of an Old Babylonian ruler of the eigh-teenth century B.C. and written in Sumerian, apalace the king has just constructed for his high-est administrative official is referred to as worthyof being "the seat of his own kingship" (ki-tuSnam-lugal-la-na), while a slightly earlier hymn inSumerian speaks in the voice of a ruler, whodescribes how the foundations of his rule weremade strong "in the palace of kingship, in mypure, good seat" (&gal-nam-lugal-laki-tu-ki-dul 0 -ga-ga).3 These two references imply far moreextended functions for a royal palace than mere-ly the royal residence; the use of the abstractnoun, "kingship," suggests that the palace is thecenter from which rule is exercised and in whichthe state is run.

I hope to demonstrate here that issues of mor-phology and decorative program are tied to thisextended function. Were the collection of essaysin this volume and the conference it preservesdevoted to the palace in the ancient Near East,contributors would each be taking a particularregion, period, function, or form-as has been

IN THE ANCIENT NEAR EASTBY IRENE J. WINTER

done in the Islamic contributions that follow.Instead, I will attempt to cover a comparable sub-field within a single paper. The service an over-view of the ancient Near East can provide forIslamicists, whose areas of interest coincide inlarge measure with the major geographical unitsof the ancient Near East, is to offer a relativelybroad survey of trends and a relatively detailedbibliography (see Appendix). This will permit aperspective on continuity and change across his-torical, cultural, and religious divides, and alsothe possibility of pursuing particular aspectsthrough further reading. The highlighting ofselected examples and aspects of ancient NearEastern palaces will lay the foundations for inqui-ry into whether or not there are areas of signifi-cant overlap in the pre-Islamic and Islamic NearEast, and I hope stimulate others to seek furthercontinuities and observe meaningful changes asparticular interests arise.

Regional divisions of the ancient Near Eastwere no less distinctive in ethnic, linguistic, andcultural character than those in later, Islamicperiods. The multiple functions of a Mesopota-mian palace are thus not necessarily characteris-tic of royal residences in the neighboring city-states of northern Syria and southeast Anatolia,even though they were also referred to as ekallu.4

These "palaces" were considerably smaller in scalethan theirAssyrian and Babylonian counterparts,had their own characteristic forms, and do seemto have been simply residences. In an interestingturn-around, when the Assyrian rulers of the firstmillennium B.C. adopted and adapted that West-ern form, they did not also call it a palace-presumably on the principle that you cannot calltwo different things by the same name-and socoined a special term, bit-hilani (on which, seebelow), in order to distinguish it from their ownlarger complexes. Wherever possible, I shall referto these regional differences, particularly withrespect to Anatolia and Syria-Palestine; but forpurposes of time and space will leave out signifi-cant aspects of ancient Egypt, on the one hand,and ancient Iran, on the other. Throughout, Ishall concentrate mainly on the Mesopotamiansequence-the most complete and perhaps also

IRENEJ. WINTER

the most developed we have for the ancient NearEast.

In selected examples of palaces from the Urukperiod to the Neo-Babylonian (from ca. 4000 toca. 500 B.c.), focus will be on the three aspectsreferred to above: form and space, includingtechnology; decorative program; and function.At the same time, the reader is referred to severalrecent specialized studies that present far moredetailed descriptions, analyses, and illustrationsthan are possible here.5

As is the case for much of the pre-modernIslamic world, the direct information we haveabout the ancient Near Eastern palace comeslargely from often incomplete archaeologicalremains, amplified by textual reference. UnlikeIslam, however, there is no living tradition-hence no contemporary practitioners or con-structions to aid in interpretation or recon-struction. Our popular impressions of thepre-Islamic palace come from biblical referenc-es, on the one hand (as, for example, the descrip-tion of Solomon's palace in Jerusalem that fol-lows the account of his temple, I Kings 7:1-12),and from British watercolor reconstructionsof Assyrian palaces at Nimrud and Ninevehfollowing excavations in the mid-nineteenthcentury, on the other.

In the latter, we see splendid, multi-storiedstructures, elaborately decorated. Because theground plans and ground-floor sculptures werebased upon excavated remains, they are general-ly reliable; however the rest of the elevation islargely invention: oftentimes a hodge-podge cop-ied from other known ruins, such as Persepolis inIran.6 It is also striking when one compares therestoration drawings of Nimrud and Nineveh,how very much they resemble the drawings donefor the restoration of the facade of BuckinghamPalace in the 1820s! One is forced to concludethat the draftsmen responsible for the Assyrianwatercolors were themselves "(re-) constructing"palaces according to their own contemporarydesires and imagination-in particular an imag-ination that saw the Assyrian "empire" in themirror of the then contemporary British empire.

The account of Solomon's palace does containa good deal of useful information: from his lavishuse of cedar in construction (7:3), to the pres-ence of three rows of windows, suggesting multi-story facades (7:4), the relationship betweenportico, where the king sits in judgment, andinner courtyard (7:7), and the separate resi-dence for his wife, pharaoh's daughter (7:8).

This information has not received much atten-tion, largely because, as is also the case withIslamic research, scholars have concentratedmainly on religious areas and buildings. Yet, notonly for the Solomonic period but throughoutthe ancient Near Eastern sequence, to establish anew state, or capital, both a temple to the primarydeity and a palace had to be constructed.

The Uruk Period (ca. 4000-3000 B.C.), in whichthe early stages of a complex social hierarchy andlarge-scale urbanization have been observed, is alogical starting point for examining the Mesopot-amian palace.7 However, at the type-site of Uruk/Warka on the lower Euphrates, although archae-ologists have recovered a large complex of build-ings identified as temples, with characteristictripartite plan, bent axis approach in the cella,altar, and podium,8 nothing clearly recognizableas a palace has yet been discovered. One anoma-lous structure has been excavated in the sacred(Eanna) precinct, levels V-IVa; known as build-ing 11 or "Palace E," it is square in plan, with alarge central courtyard surrounded by banks ofrooms (resembling more than anything the laterIslamic four-iwan building type with very shallowiwans; see fig. 1) .9 The plan is clearly distinct fromthat of a temple, so the building has been suggest-ed as a possible palace. The problem is that wehave no corroborating textual evidence for thebuilding, nor even for the existence in the earlytexts of a title that clearly designates a ruler; so thebuilding could well house "administrative" activ-ities and still be related to the religious com-plex. 1° How the Mesopotamian state and a desig-nated ruler emerged is far too complex to discusshere, although most scholars agree that somesort of hierarchical organization in governancemust have been operative in the Uruk period. Atthe same time, since archaeological work at thesite of Warka has concentrated on the sacredquarter, it is certainly possible that a palace existsin unexcavated areas. Only further fieldwork canhelp to determine whether we are faced with anaccidental absence in the archaeological record,ora meaningful absence in the historical record."

The first buildings to be clearly identified aspalaces date from the third phase of the EarlyDynastic Period (ca. 2600-2430 B.c.), and coin-cide with the earliest textual evidence for titlesdenoting rule: Sumerian lugal, "king," and ensi,"steward" or "governor," a regional title equiva-lent to king in the hierarchy of governance. Thebest (and earliest) archaeological evidence comes

28

"SEAT OF KINGSHIP"/"A WONDER TO BEHOLD"

from Kish, the legendary city where kingship asan institution is said to have "descended fromheaven."' 2 Although there seems to be a greatdeal of variation in overall plan, the two buildingsdesignated as palaces contain a large number ofrooms of differing size and shape, suggestingmany functions, and all seem to share what is latera defining characteristic of palaces in Mesopota-mia-a large central courtyard.'3 Since this is aperiod in which autonomous city-states were dis-tributed across the Mesopotamian alluvium, it isto be anticipated that each central city wouldhave had its own palace, however incomplete thepresent evidence.

In the succeeding, Akkadian period (ca. 2334-2154 B.C.), political development within the peri-od marks a significant change toward a central-ized nation-state, incorporating formerlyautonomous city-states within a single polity un-der the hegemony of Agade. Unfortunately, thecapital of Agade has not been definitively identi-fied or excavated. A number of other sites haveproduced large buildings dated to the Akkadianperiod and identified as palaces (e.g., Khafaje,Tell Asmar, Tell al-Wilayah, possibly Assur), all ofwhich have features in common: at least onecentral court, perimeter walls with a primaryentrance, evidence of residential use togetherwith other activities.' 4 The most complete palaceplan preserved is that at Tell Brak, a site in theHabur region of northern Mesopotamia.' 5 Thebuilding is identified through bricks stampedwith the name ofNaram Sin, king ofAgade (2254-2218 B.C.) .As reconstructed, itis essentially square,ca. 80 meters a side, but incomplete to the southand southwest, and known only from founda-tions. Nevertheless, features preserved are com-mon to other palaces: massive perimeter walls;single, monumental entrance on axis with largecourtyard, surrounded by rooms; and at leastthree additional, smaller courtyards also flankedby banks of rooms (fig. 2).16 However, since Brakis a site at the very northern periphery of Akkadi-an political influence, this building is probablymore a fortress-cum-governor's-palace/provin-cial administrative center than a royal seat. If weexpand our definition of palace to include notonly residences associated with the exercise ofpower by the highest absolute authority, but alsoby the highest localauthority in any given politicalstructure, then provincial governors and depen-dent local princes can certainly also occupy "pal-aces." Indeed, the Old Babylonian year-namecited above, which referred to the residence of a

high-ranking official as a palace [.gal], suggeststhat the extended administrative functions per-formed in such a building may be the mostoperative variable in defining the term, whichshould then be applicable to more than the royalseat, even within the capital. 7 In any case, whilethe Brak building may well be called a palace, it isconceivable that the Akkadian kings' palace(s)back in Agade would have been larger, perhapsless regular, the exterior walls perhaps less mas-sive, and might well have contained a greatervariety of room types, correlated with a widerrange of activities. In short, it is not clear that onecan generalize from this plan to the capital, or forthe period as a whole.

The recently excavated site of Tell Mardikh(ancient Ebla) in North Syria provides us withfurther evidence that palaces of this period(roughly the late Early Dynastic/early Akkadianperiod) were not merely residences, great housesof local hierarchical rulers, but were also centersof political and administrative activity. In theroyal palace of level IIB1 were found hundreds ofcuneiform tablets the contents of which rangefrom treaties with foreign rulers to daily econom-ic records, all carefully stored on shelves and inbaskets within specially designated archives. 8

While no extant southern Mesopotamian palacehas produced such archives, this is likely to be aresult of the palaces having been cleaned out andoften razed to their foundations to permit subse-quent building. The demonstrated epigraphicalrelationship between the Ebla tablets and textsfound in non-palatial contexts in Mesopotamia,in conjunction with known political and militaryevents that link the two regions, allows us to positthe existence of similar palace archives in con-temporary and even earlier Sumerian and Akka-dian palaces as well.'9 That the practice continuesinto the early second millennium is evident fromthe accumulated tablets and sealings found in thepalace at Tell Leilan in the Habur region ofNorth Syria, the contents of which attest to abroad network of communications between re-lated polities of the Old Babylonian/Old Assyri-an period.2 0

Happily, a relatively well-preserved palace hasbeen excavated at Mari on the middle Euphrates,which was apparently in use over a number ofreigns from the late Ur III/Isin-Larsa to the earlyOld Babylonian period (ca. 2000-1758 B.C.). Itis in this palace, consisting of some three hun-dred rooms and courts, that all of the spatialconfigurations plus decorative schemes and

29

IRENEJ. WINTER

administrative functions that characterize laterAssyrian and Babylonian palaces can be observedas part of a coherent complex of features. 21 Clear-ly recognizable is the primary entrance into alarge paved courtyard, with subsidiary roomsranged around that space and an additional,smaller inner court (see plan, fig. 3). While notall scholars agree on the functions attributed tospecific rooms or areas of the palace,22 there is nodoubt that ovens and food-storage features indi-cate the residential nature of the building. In-deed, a small group of cuneiform tablets found inthe northwest wing of the palace attest to thedelivery of delicacies for visiting dignitaries.2

When that evidence is seen in conjunction withthe later Assyrian administrative texts known asthe Nimrud Wine Lists, it is clear that at any givenmoment the palace household included largenumbers of individuals-members of the royalfamily, high court officials, eunuchs, guards, work-ers, and visitors-all of whom were being fed andprovisioned from palace stores.2 4 In addition, avast collection of administrative texts has beenpreserved in rooms around the outer court. Therange of subjects covered by these texts makes itclear that the palace was engaged in administer-ing the king's own estates and production indus-tries, as well as affairs of state.2 5

The Mari palace also preserves for us the firstappearance of a particular constellation of for-mal reception suites well known from palaces inlater periods. At Mari, the reception suite is setparallel to the northern end of the inner court(see fig. 3, court 106 and rooms 64 and 65). Acentral doorway connects the first room to thecourtyard. There is evidence of a podium on thesouth wall opposite that central door. The podi-um was plastered and whitewashed, giving it spe-cial prominence, and could have been used ei-ther as a base for a statue, or, more likely, as aplatform for the throne of the ruler himself, forthose occasions that called for him to be in fullview, and with a full view of the courtyard. Theinner room included a second podium on theshort, westwall, which then faced down the lengthof the room toward an elevated niche that, it hasbeen suggested, may have contained an image ofthe local goddess, Ishtar. Identical suites, with theinnermost room being the formal throne roomand a throne base preserved along a short wall,are also to be found in Assyrian palaces of thefirst millennium (see below). What is more, evi-dence of a developed program of decoration inwall paintings is preserved at Mari that also

echoes the decorative programs of laterAssyrianpalaces.

In the Mari palace, two sets of wall paintingswere found fallen from the northern wall of court106 that gives access to the throne-room suite.2 6

One set preserves what is likely to be a royal figureattendant upon a sacrifice; the other what seemsto be a scene of investiture of a ruler, identifiableby his headgear and garment, by the goddessIshtar, within an elaborate setting of trees andplants (see detail, fig. 4). This last scene wasplaced on the facade just to the right of thedoorway entering the throne-room suite, and al-Khalesi has suggested that the space depicted inthe painting in fact replicates the physical spaceof the inner throne room. Especially if this is so,but even if the iconography merely asserts thespecial selection of the ruler by the goddess ingeneral terms, the presence of the "investiture"painting suggests that the facade of the throne-room complex serves as an important convey-ance for statements of royal rhetoric and stateideology-a pattern we will see in both Assyrianand later Babylonian palaces.

A third set of paintings comes from room 132,a small chamber opposite the main entrance offthe large court 131. The floor level of this cham-ber is raised slightly and the entrance empha-sized by concentric semicircular steps thatjut outinto the courtyard. This special focus, in combi-nation with the fact that the imagery includes thefigure of a ruler pouring libations before a seateddeity, has led to suggestions that the chamber isa small chapel. I find such a suggestion persua-sive, particularly as both textual and other ar-chaeological evidence attest to the presence ofritual spaces in other palaces. 27 In Hittite Anato-lia as well, the king and queen were expected toperform certain ritual acts daily, and Guiterbockhas adduced the likelihood of a sanctuary as aregular feature of the palace.28 A ritual functionhas also been attributed to certain suites withinthe Assyrian palace of Assurnasirpal at Nimrud; 29

and in historical times, the association of a royalchapel with the palace is certainly well known.The Mari palace thus adds evidence for an impor-tant religious component to supplement the var-ious aspects of the Mesopotamian complex.

Most recently, Margueron et al. and Pierre-Muller have published an additional series ofpaintings, unfortunately fragmentary, that ap-parently decorated a reception suite in a sec-ond story in the southeastwing of the same palace(above areas E, F, and room 120 on the plan, fig.

30

"SEAT OF KINGSHIP'/"A WONDER TO BEHOLD"

3).30 Its fragmentary nature and the difficulty ofreconstruction notwithstanding, this group ofpaintings is extraordinary for containing in itsrepertory a number of motifs-ruler in combatwith a lion, ruler victorious over a fallen enemy,and ruler receiving some delegation of approach-ing individuals-that find direct counterparts inthe more complete decorative programs of laterAssyrian palace reliefs. This implies an iconogra-phy of rule associated with palace decoration inuse over at least a thousand-year period. Further-more, the presence of a second story that in-cludes a reception space fits well with both theSolomonic description cited above and later tex-tual and pictorial evidence. In particular, in theeleventh-century B.c. Egyptian text ofWenamun,the Egyptian envoy who visits Byblos on the Med-iterranean coast is received by the local prince inthe upper chambers of his palace, from the win-dows of which one could see the sea.31 Both thetexts and this new evidence from Mari suggestthat we may be missing quite essential parts ofNear Eastern palaces known only from groundfloors or foundations, and we should thereforenot try to distribute all palatial functions acrossthe ground-floor plan as if that were the entiretyof the building.

Recent analyses of wall construction at Marisuggest several rebuildings and a long period ofuse before Mari was conquered by Hammurabi ofBabylon in his thirty-fourth year (ca. 1758 B.c.);

Moortgat had earlier argued that the wall paint-ings in the palace reflected different phases aswell.3 2 This raises an additional caution for thescholar, for, while it is important to emphasizethe degree of continuity that exists from thesecond into the early first millennium B.c. withrespect to some aspects of spatial configuration,decoration, and function, it is also the case thatwe cannot assume each palace represents a sin-gle, coherent program. It is only in the Neo-Assyrian period that we have a sufficient numberof examples and degree of preservation to at-tempt fuller readings of attitudes toward rule andto assess experiences of authority-what I havecalled elsewhere "royal rhetoric"-as articulatedin palace construction and decoration.

During the early second millennium B.c., theregion around the upper Tigris, near modernMosul, had established its political independencefrom the south. In the early first millennium, thisarea constituted the heartland of Assyria, fromwhich, in a series of military maneuvers over aperiod of some three hundred years, the state

expanded its territory until it reached from theZagros in the east to the Mediterranean in thewest, and from the Taurus in the north to Baby-lon and Egypt in the south and southwest. Overthis period, virtually every successive ruler initiat-ed the construction of a new palace, as the capitalshifted from Assur to Nimrud, to Khorsabad, andfinally to Nineveh. Although there are no expla-nations for these shifts in the several preservedAssyrian royal inscriptions, they have been un-derstood as a function of statecraft. As withSolomon in Jerusalem, a new ruler establishedthe authority of his reign in part through palaceand other building campaigns.

The complete circuit of city walls has beentraced at Nimrud and at Khorsabad, with enoughpreserved at Nineveh to suggest that a similarpattern prevailed (fig. 5). Essentially, rectilinearenclosure walls pierced by gates in all directionssurrounded large areas. Set into and sometimesbreaking the line of the exterior wall were twotypes of construction, often at different ends ofthe city: a raised citadel containing royal palace (s)and temples, and a building known in Akkadian/Assyrian as an ekal-maarti, a fortified palace, orarsenal, based on the general plan of a residentialpalace, but often larger in scale and with a sim-pler distribution of rooms around each court-yard.3 4

The consistency of the Assyrian pattern high-lights the difference from neighboring first-mil-lennium citadels. In the Assyrian case, the royalpalace and citadel are set into the rectangularperimeter wall, often overlooking a distinctivenatural feature in the landscape, like a river; inthe capital cities of the principalities of NorthSyria and southeast Anatolia to the west, as in thekingdom of Sam'al at Zincirli, the pattern israther to contain the citadel and palace at thecenter, a round perimeter wall more or less equi-distant at all points from the citadel enclosing thelower town.-5 In all cases, the citadels are raised,and access is limited via controlled routes andgates, in a way similar to later Islamic practice inthe Near East. The royal palaces of the ancientNear East are also themselves frequently set onraised platforms, so that a continuous sequenceof physical elevations may be read as progressiveelevations in status.3

In the case of building techniques and materi-als, a combination of environmental factors pluswealth, labor force, and extension of trade net-work conspires to dictate materials and methods,which present significant regional variation.

31

IRENEJ. WINTER

Throughout Mesopotamian history, the primarybuilding material was mud brick-making use ofthe most abundant natural resource in a regionof virtually no stone or construction-size wood. AtZincirli, set in the foothills of the amply woodedAmanus mountains, stone foundations are over-laid with walls that combine wooden beams andbrick. On the Anatolian plateau to the northwest,stone construction was common. In Assyria, brickwas used for bearing structural loads; however,proximity to sources of stone and wood allowedrulers under the influence of the west and north-west to introduce stone revetments and orthostatslabs as decorative skins on the walls of theirpublic buildings, and to employ a variety of pre-cious woods as well. From the Middle AssyrianPeriod (second half of the second millenniumB.C.) through the Neo-Assyrian Period (first halfof the first millennium B.C.), kings describe indisplay texts the lavish construction materialsassembled for their palaces and, in the laterphase, actually depict on palace reliefs the cut-ting of wood and quarrying of stone blocks forsculpture. 3 7

Limited by the preservation of buildings, weare reduced to reconstructing facades, lighting,and roofing methods from the occasional an-cient representations of architecture, in monu-mental scale on palace reliefs or in miniature oncylinder seals.38 From these images it would seemthat massive exterior walls with niched facadesand crenellated tops were prevalent at least fromthe mid second millennium. And, although post-and-lintel construction was likely to have beenthe principal way of spanning space and bearingweight, there is again evidence for vaulting inpalaces from the same period.3 9 At Khorsabad,both complete barrel vaults and intact archesover major entries were well preserved, as wasarched wall construction in the so-called Gover-nor's Palace at Nimrud.40 All of this suggests thatwhile the antecedents of Islamic constructionmay be found most immediately in the greatarches of the Sasanian period at sites like Ctesi-phon, the beginnings of that tradition may havereached considerably farther back into antiquitythan is generally acknowledged. Indeed, I shouldnot be at all surprised if one day we find evidenceof simple dome construction in the ancient NearEast as well!

Great attention was paid in the Assyrian palacesto the scale and decoration of major gateways andentrances, including threshold inscriptions andthe colossal human-headed bulls and lions that

flanked principal doors (fig. 6).41 It is not clearwhether these great stone colossi actually carriedthe weight of doorway arches, or, like the ortho-stats, simply lined the walls; but their iconographyis one of menace and protection. The placementof monumental stone sculpture at doorways seemsto have been borrowed by the Assyrians from theWest-where gateway lions and sphinxes areknown from second-millennium Hittite sites onthe Anatolian plateau, and then later, from first-millennium Neo-Hittite citadels, like Zincirli, withwhich the Assyrians came into contact during theninth-century military campaigns in the area.4

The larger lesson to be learned from interac-tions of this sort, which must surely be relevantfor subsequent periods as well, is that in someaspects of architectural practice, like the shapesofperimeterwalls and placement of elite citadels,neighboring states may, despite contact, remaindistinct; however, in other aspects, like buildingtechniques, materials, or decorative schemes, theymay change once contact is established. In thecase of Assyria and the West, Assyrian palaceconstruction owes a good deal to foreign contact.In the case of Assyria and the East, by contrast-as seen at the site of Hasanlu in northwest Iran,exposed to Assyrian contact around the sametime-we see the converse: Assyrian elementswere adopted, as illustrated by the addition ofglazed plaques and new porticoes to embellishlocal building facades.43 In this latter case, itwould seem that there was a desire to emulatepractices associated with the major political forcein the region. In the case of the Assyrian adapta-tion of Western elements, there may also havebeen some positive charge associated with theincorporation (appropriation?) of a highly devel-oped tradition just as the Assyrian polity wasexpanding.

In later historical periods, it is possible to doc-ument the spread of new techniques and modesof construction as part of the general dissemina-tion of architectural knowledge and practice.While this is not possible for the ancient NearEast, I do wish to underscore the importance ofseeking to distinguish between transmission ofknowledge as part of practice and culturallycharged borrowings that carry with them coordi-nates of reference and meaning. Furthermore, Iwould stress the fact that regionally distinct tradi-tions in morphology and decoration are not fixed,but rather, within the constraints of resourcesand cultural practice, can respond to historical/political exigencies.

32

"SEAT OF KINGSHIP"/"A WONDER TO BEHOLD"

The earliest of the relatively well-preserved Neo-Assyrian palaces, the Northwest Palace of Assur-nasirpal II (883-859 B.C.) at Nimrud, sits on thewestern edge of the citadel, overlooking the Ti-gris. Although many rooms of the western sectorhave been eroded away, the basic configurationof the ground plan can be read (see fig. 7).Typical of most Neo-Assyrian palaces, it conformsto the basic type established by the palace at Mari,in which space is divided into two main sectorsorganized around an outer and an inner court-yard.44

Dividing the two courts on an east-west axis isthe throne-room suite of two long rooms, onewith a throne base preserved in situ (cf. room B onplan). Primary access to the throne room is viathe large outer courtyard, where monumentalpylons with flanking door guardians mark theentrance. This entrance is on the long, northwall, necessitating a 90-degree turn to face thethrone on the short, east wall. The pattern ofaccess and layout is one seen already in the palaceat Mari, and is repeated in the reception/throne-room suites of all of the majorAssyrian palaces."Oppenheim suggested that this represents a con-scious modeling of the royal audience chamberupon the bent-axis plan of the early Sumeriansanctuary, in effect sanctifying the ruler withoutformally deifying him.4 In any case, the ruler isnot on axis with, or visible from, the outer court,as he would have been in the anteroom at Mari,and as was traditionally the case in the straight-axis throne rooms of the later Neo-Babylonianand Persian periods, where the king was seatedopposite the main door (for example, in theSouthern Palace of Nebuchadnezzar II in Baby-lon, fig. 8).47 These traditions of visibility andsight-lines, both of and by the ruler, can becomesignificant indicators of cultural and nationalattitudes toward authority and the person of theruler, and analysis of patterns thereby goes be-yond description toward the reconstruction ofexperience within the built environment, as Is-lamicist Eric Schroeder called for nearly fortyyears ago,"4 and as Guilru Necipoglu pursues inher paper here.

Subsequent Assyrian rulers also built palaceson the citadel at Nimrud, but the remains are toofragmentary to read the complete plans. WhenSargon II (721-705 B.C.) decided to shift thecapital to Khorsabad, he constructed the citadelin such away that his palace and attached templeswere the only buildings at the highest level, withsubsidiary palaces and administrative buildings

in a separate, lower enclosure (see reconstruc-tion of the citadel, fig. 9, and plan of the palace,fig. 10). Sargon preserved the organizational prin-ciple of two main courtyards; however, his throneroom (room VII on plan) is no longer betweenthe two courts, but rather is set longitudinallyalong the southwestern wall of the inner court.Nevertheless, the configuration of the throne-room reception suite remains constant, as it didthroughout the Neo-Assyrian period.4 9

As noted at the beginning, a very different sortof royal palace is found in contemporary Neo-Hittite and Aramaean sites to the west of Assyria.Here, small self-contained structures are markedby columned-portico entrances into banks oflateral rooms, often with service rooms at one orboth ends (for example, Hilani III at Zincirli, fig.11). They either stand independently or aregrouped around enclosed courts (as in the "Up-per Hilani" complex at Zincirli) .50 The columnedportico and limited size are characteristic of royalbuildings in Syria and Palestine from at least themid second millennium onwards. While the pol-ities they represent are considerably smaller thanthe large urban states of Babylonia and Assyria,the reduced size of the Syro-Palestinian palaces isnot merely proportional to their population orterritory. Even considering that they may wellhave stood several stories high, on the model ofMari and the Wenamun text cited above, 51 thelimited number and type of rooms and spacessuggest that they could not have served as manyand diverse functions or constituents as an Assyr-ian palace.

It is presumably this smaller type of royal palaceto which the Assyrian ruler, Tiglath Pileser III(744-727 B.c.) referred when he declared that hehad constructed a palace in the western manner,which he called a lit-hiliini. Sargon II also claimsto have constructed such a building at Khorsa-bad. The literature on this building type in Assyr-ia is long,5 2 and I shall refer to it only briefly here;but I believe it may have ramifications for thelater development of the iwan in the Islamicarchitectural tradition, as well as serving as animportant historical case of cultural borrowing.

Scholars have debated just where Sargon's bit-hilini might have been located and how its prin-cipal features might be recognized. The mostsalient feature of the plan of known westernpalaces, the columned portico, has been themarker sought by most scholars, largely becauseit is observable on the ground! It is this featurethat connects the building type to the iwan: and

33

IRENEJ. WINTER

when, as at Zincirli, several buildings are groupedaround one court, we may in fact see antecedentsfor the three- or four-iwan building.5 3 However, itis possible that the term hilani could be related tomodern Hebrew and ancient Ugaritic hln, "win-dow," and may therefore actually be identifiedless by its columns than by a multistoried facadewith windows, such as is described in I Kings 7 forthe palace of Solomon and as has been recon-structed at Mari, with perhaps the second-storyoverhang supported by a columned portico. Insuch a case, the bt-hilani may not always reflect aseparate building, but rather a suite or complexincorporated in the main palace.

One possibility for the Assyrian 1bt-hilaniis thatthe building was not in the city or on the citadelat all, but rather was located in some landscapedarea outside the walls, as is depicted in a huntingpark on one of Sargon's reliefs (fig. 12) and on arelief of Assurbanipal from room H of his NorthPalace at Nineveh, where small pavilions withcolumned porticoes stand amid trees and water-courses.5 4 A second candidate for the bit-hilani atKhorsabad is a small, free-standing structure (of-ten labeled a temple, but on no solid evidence)thatis set on the western corner of the citadel (seeplan, fig. 10). Yet another possibility is that theattached suite of rooms at the northwestern endof the royal palace, which extends out beyond theline of the city wall, constituted a specially desig-nated wing (= rooms 1-8 on plan, fig. 10).

A clue to the character of the structure may becontained in Tiglath Pileser's description of thebuilding as built "for his pleasure," that is, despitethe formulaic nature of this phrase, which is usedby several kings, one is led to think of the struc-ture, free-standing or attached, as distinct fromthe official apartments and reception areas. Thiswould apply to all of the three possibilities notedabove. Along with the separate building on thecitadel and the park pavilion, the attached suiteof rooms at Khorsabad would lend itself well torepose. The northwest edge of the citadel looksout over the course of the river Khosr, therebyproviding both view and fresh air. Sargon tells usthat he laid out a landscaped park at Khorsabad,the siting of which would most appropriately bebeyond the city to the northwest. In addition, theorientation of room 7 is such that its doorway isaligned with that of room 4, to look out to thenorthwest; and it is precisely in the reliefs of room7 that we see banquet scenes and an elaboratefrieze of the king's hunters in a park. Moreover,the trees and river that are represented in the

hunting park conform to Sargon's description ofthe park he created, which he tells us explicitlywas modeled on a western landscape. How betterto enjoy the park than in a western-style struc-ture?

However the iit-hilani in its original or bor-rowed form may be identified in future, its asso-ciation with leisure and park land introduces theconnection of gardens and purposeful landscap-ing to Assyrian palaces. This is attested by theMiddle Assyrian period, when Tiglath Pileser Irecords taking both hardwood and fruit trees"from the lands over which I had gained domin-ion," and filling the orchards of Assyria withthem, while in a second text he records plantinga royal garden for his "lordly pleasure, in themidst of which he built a palace."5 5 The traditionis perpetuated throughout the Neo-Assyrian peri-od, as seen from the Banquet Stele text of Assur-nasirpal II, in which the king, describing thefounding of the new capital at Nimrud, enumer-ates the various trees and plants gathered in histravels and incorporated with abundant watercanals into luxuriant gardens at home.5 6 Scholarsof the ancient Near East are just beginning tolook for archaeological evidence of such land-scaping, particularly associated with palaces;5 7

but the line from Assyrian to Babylonian to Achae-menid to Islamic palace gardens and orchardscan atleastbe affirmed. Terms utilized in describ-ing these early gardens all denote pleasure andjoy. It should come as no surprise that in theseearly periods no less than in later times, wealthand power would be associated with manage-ment of the landscape for purposes of delecta-tion, notjust mere sustenance-especially in anenvironment where gardens were expensive anddifficult to establish and maintain; but one canalso go a step further in suggesting that such adisplay could be part of a public statement ofwealth, power, and even territorial appropriationthrough reference to the lands of origin of thevarious trees and plants collected.

As noted above, Assurnasirpal II also refers tothe varied types of wood he employed in hispalace at Nimrud, and like the gardens, the build-ing materials would have conveyed wealth andpower independent ofnarrative content. Far moreexplicit statements of wealth and power, not tomention political ideology, are possible and at-tested in the decoration of the actual palacebuildings, through the addition of applied verbaland visual messages.

It is in the incorporation of inscribed texts into

34

"SEAT OF KINGSHIP"/"A WONDER TO BEHOLD"

the scheme of palace "iconography" thatAssyrianpractice may come closest to later Islamic prac-tice. Although calligraphic scriptwas never devel-oped in the ancient Near East to the extent thatit was in later Islamic periods, one does see adistinctly "lapidary style" employed for the palacetexts, which needs further study as part of theoverall visual effect in the decorative program asa whole (visible, for example, on the block sur-rounding the doorway colossus, fig. 6). Russellhas done the most complete study to date of therole played by various sorts of inscriptions in asingle palace, where the ruler makes use of eachtype of text for different rhetorical purposes.5For regions brought into the Assyrian polity thatretained local rulers, we find bilingual inscrip-tions on palaces thatjuxtapose Assyrian Akkadi-an to the local language. This is not unlike thesituation described by Catherine Asher (see arti-cle in this volume) for Mughal India, where alocal Hindu ruler could include inscriptions inboth Persian and Sanskrit. The ancient NearEastern and South Asian cases show intriguingsimilarities in that both evince significant differ-ences in nomenclature in the local versus theofficial court language. One ruler of Guzana tothe west of Assyria, for example, is referred to as"king" in the local Aramaean, but only as "gover-nor" in the Akkadian, appropriate to his subordi-nate status vis-A-vis Assyria.5 9

The extent to which the Assyrians developedthe application of orthostat stone relief carvingsto exterior and interior palace walls was unprec-edented and, as noted above, seems to be derivedfrom contact with North Syria and Anatolia. TheAssyrians employed limestone and alabaster intheir carvings; the Syrian and Anatolian sitesoften used basalt as well as limestone, and it isinteresting that the alternation of black and whitestone for decorative purposes, as on the LongWall and Herald's Wall at Carchemish,6 ° as well asthe use of lions as door or gateway figures, can stillbe attested in the same general region well intothe Islamic period (on which, see the Soucek,Redford, and Tabbaa articles in this volume).

The Assyrian orthostats stand some two metershigh, and are carved in relatively high relief,often incorporating inscriptions over or as part ofvisual representations. Traces of color suggestthat they were originally painted; and Layard'saccount of the throne room ofAssurnasirpal II inthe Northwest Palace includes references to frag-ments of plastered wall paintings along with thereliefs-presumably from the upper parts of the

wall surfaces and the ceiling (see reconstruction,fig. 13; although note that the human figures aresmall in relation to reliefs). In addition, Assurna-sirpal's Banquet Stele mentions the decorationof his palace with glazed brick and with bronzedoor bands, examples of which have been foundelsewhere,6 and Postgate, following the specula-tions of Reade, argues for the probability oftextiles, no longer extant, as having been anotherimportant medium of palace decoration.6

Numerous studies in recent years have investi-gated the sorts of political and cultural messagesarticulated in Assyrian decorative programs. Thesemessages are conveyed by placement, as well as bycontent. Composite, protective creatures stand atdoorways and at corners. Within rooms, the king'sfigure is often given prominence opposite door-ways, or in the center of a wall, regardless of thesubject of the scene.6 3 On one such example, ascene of the Assyrian king Sennacherib receivingprisoners in the field after the siege of Lachish,we see the enthroned ruler positioned exactly inthe middle of the northwest wall (fig. 14). In thefield above the king's face is a rectangle contain-ing four lines of explanatory text. The introduc-tion of textual labels into visual narratives, firstattested on reliefs in the eighth century, served toemphasize the image or the narrative momentwith which they are associated, thereby comple-menting or augmenting the visual program. 64

When we look at the sorts of motifs commonlyrepresented in palace decoration, many com-monalities occur across the entire range ofpreserved evidence. Foremost among these is thepresence of the palm tree, either as an indepen-dent element or in association with the image ofthe king himself. At Mari, as in the Babylonianpalace of Nebuchadnezzar II (604-562 B.C.), thepalm occurs in a repeating frieze on the outerfacade of the throne room, flanking the centraldoor through which the ruler could be visible atselected times.5 In the Northwest Palace ofAssur-nasirpal, the same tree is depicted throughoutthe palace, most frequently flanked by symmetri-cal genii.66 In the throne room, which constitutesa special case, the king himself is shown duplicat-ed on either side of the tree, presumably partici-pating in its ritual care (fig. 15). If Castriota iscorrect in suggesting that on some occasions thetree stands metonymically for the ruler, preciselybecause maintenance of the fertility of the landthrough proper ritual performance is a majorfunction of kingship,6 7 then the repetition of themotif in Babylon and throughout the Northwest

35

IRENEJ. WINTER

Palace is not merely for purposes of decoration,but also conveys the powerful message that rule isgrounded in nature, i.e., in cosmology.

The motif of the king and tree is accordedpride of place in the throne room of Assurnasir-pal; it appears both directly behind the king onhis throne on the eastern wall and directly oppo-site the major doorway of the north wall. Thronerooms, as relatively public ceremonial and politi-cal loci, are especially likely to be highlyinvested with charged imagery-in the best pre-served cases, incorporating a number of motifsthat in total reflect the full panoply of royalactivities and attributes.? In the throne roomof Assurnasirpal II at Nimrud, scenes of huntand battle are distributed along the long, southwall and parts of the north wall, with the kinghimself depicted at the far west end. I have ar-gued elsewhere that the assemblage of imagescan be read as a unified program, recapitulatingin both content and structure the king's "Stan-dard Inscription" that is written over every slab(see, for example, on throne-room slab B.23, fig.15), and signifying all of the major attributes ofthe ruler appropriate to his stewardship of thestate: ritual performance, virile strength, militaryvictory, and statecraft. 69 Russell has recently dem-onstrated that in the later reign of Sennacherib,this lexicon was expanded to include new themesrelated to civic construction, which convey moreexplicit messages pertaining to maintenance ofthe "center, " i.e., the capital, in contrast to earlierformulations that emphasized the maintenanceof the state through territorial acquisition andthe establishment of boundaries.7 0

These Sennacherib reliefs are situated in thelarge court giving onto his principal throne room.As noted for Mari, the courtyard wall that doublesas throne-room facade is especially adapted toproclamations of rule, and in the Neo-Assyrianpalaces of Assurnasirpal II and Sargon, at least(e.g., fig. 16), we find particular attention givenon that wall to processions of foreign delegationsbearing tribute to the ruler-a topos conveyingthe ruler's ability to command both wealth andstately attention.7

Through verisimilitude in landscape elementsand dress, military narratives are made to bemore than generic victories; they refer to actualcampaigns of the king's reign. The representa-tion of at least a half-dozen separate campaigns inthe reliefs ofAssurnasirpal II's throne room con-firms the king's account in his Banquet Stele ofhow he depicted on his palace walls the "glory of

my heroism across highlands, plains and seas." Inlater Assyrian reigns (as, for example, the Lach-ish siege of Sennacherib, fig. 14), these militaryscenes proliferated throughout the entire pal-ace. By concentrating them in the throne roomduring the reign of Assurnasirpal and by placingthe throne room itself at the center of the palace,the ninth-century king conveyed the fundamen-tal message that, as the throne room is the heartof the palace, so the palace is the heart of thestate.

The use of extended decorative programs asvehicles for the articulation of ideology is notunusual in the history of royal palaces, and manyart historical studies have attempted to recon-struct those programs, along with their ideologi-cal underpinnings. 2 In the palaces of the ancientNear East, the "official" public statements aboutthe ruler and the state as they appear in thedecorative program serve to underscore the insti-tutional nature of the palace as part of the largerstate apparatus.7 3 To the extent that "the palace"can serve as metonym for the ruler (as "the WhiteHouse" does for the American president), andthereby for the state, the palace is the source ofideology; and to the extent that the palace is thephysical manifestation of a program of royal rhet-oric, it is also the vehicle for that ideology. Thus,we return to issues of function, and the role of thepalace-qua-institution, with which we began.

Clearly, the ruler and a large extended house-hold resided in the palace and had to be accom-modated. On the practical level, this requiredprivate apartments, cooking facilities, and stores.Evidence exists that the king's mess includedlarge parties of his sons and officials and thatallocations were made for the rest of the palaceon a regular basis.7 4 In addition, periodic festivi-ties must have been organized, which would havenecessitated the banqueting of very large num-bers of individuals, as on the occasion of theinauguration of Assurnasirpal II's new capital atNimrud, when the king records he fed some70,000 people for ten days.

The size of the Assyrian throne room (ca.10 x 45 meters for both Assurnasirpal II andSargon II) and its decoration argue for its natureas a public reception suite in which the rulerwould give audiences, although whether on aregular basis or occasionally is uncertain. Visualevidence that the ruler at least received selectedmembers of the court and highly placed officialsis preserved in scenes carved on cylinder seals,which show individuals presented before the

36

"SEAT OF KINGSHIP"/"A WONDER TO BEHOLD"

seated ruler-a tradition that must have beennot unlike the Mughal darbarillustrated in sever-al miniatures of the seventeenth-century court ofJahangir and Shah Jahan (see Necipoglu, figs.26-28) .75 Scenes on reliefs showing AssurnasirpalII with cup or bowl in hand may attest to the king'sjudicial function, in keeping with a long-standingiconography of the ruler renderingjust decisionsin Mesopotamia. 7 6 Whether the ruler exercisedlegal office inside the palace (perhaps in thethrone room) or outside is not certain for thisperiod, however. On the basis of the Solomonicreference (see above, with regard to I Kings 7:7)and a Sumerian literary text regarding the leg-endary king Lugalbanda, who "takes [or exercis-es] office in the outer courtyard, in 'the Gate thatBrings in Myriads',"7 7 one is encouraged to seekpossible evidence for similar use of the palacecourtyard and gate in the Assyrian period. It mustalso be considered that the throne room couldhave functioned as a venue for legal hearings.

Tribute scenes on courtyard facades leadinginto the throne rooms at Nimrud and Khorsabad(e.g., fig. 16) both illustrate and assert state re-ception of foreign delegations. That the palacesserved as the repositories of such gifts, along withthe booty seized in foreign campaigns, we knowfrom Assurnasirpal's repeated assertion that hebrought precious metals and other rich booty tothe palace, and also from Sargon's statement thathe restored the Northwest Palace in order toplace in it booty from his victory over Carchemishin 717.78 Not unlike later palaces, from FatehpurSikri to Versailles, the display of valuable goodsand elaborate appointments served as signifiersof the success of the ruler, and hence of thestate.

Tribute scenes, overall decorative program,and display all attest not only to specific events,butalso to the very fact that the palace was the sitewhere statecraft was conducted. Texts from Nim-rud further document that the extended house-hold supported in the palaces included minis-ters, administrative officials, and scribes, whosejob itwas to run not only the palace but the state.79

Assurbanipal's famous "library" at Nineveh mayreflect the special case of an unusually literateruler; however, the archives of other palaces,such as Mari, strongly argue for the palace asrepository of central state records as well.

The iconography of room 132 at Mari and thesuite around room G at Nimrud, as was men-tioned earlier, may suggest that some roomsin the palace were devoted to ritual activities,

and I suspect that if any new palaces were tobe investigated with modem excavation meth-ods, we would find considerably more evidenceto support such a contention. The importance ofritual performance and court ceremonial in oth-er times and places argues strongly for the pres-ence of such spaces within the palace.8 If we wereto include procession as part of ceremonial dis-play, then the fact that the processional routefrom the Ishtar Gate to the temple of Marduk atBabylon passes along the east wall of Nebuchad-nezzar's palace (see fig. 8) could imply an activerole for the palace and/or the king in the proces-sion.81

Finally, I would argue that a significant compo-nent of function is "affect": the impact of thebuilding upon subjects of the state and uponforeigners. Lackenbacher has studied Assyriannarratives of royal building activities, with partic-ular focus on palace construction. 8 2 In a numberof instances, rulers take credit for innovations intechnique (such as Sennacherib's reference tobronze casting for column bases) or include state-ments about how skillfully the palaces have beenconstructed.8 3 In addition, rulers express person-al gratification concerning their palaces (e.g.,"palace of myjoy" and "my royal residence that Ilove") .84

Perhaps most important of all, we find refer-ences to intended impact. Assurnasirpal refers tohis new palace as "fitting and splendid," "palaceof all the wisdom of Kalhu" (Nimrud) .85 The kingis clearly celebratory; but at the same time heproclaims the palace as concentrating within it-self all that is of value in the capital. Nearly twohundred years later, Sennacherib calls his newresidence "Palace without a Rival."86 He says ofthe limestone reliefs, "I made them objects ofastonishment"; of his colossi, "I made them awonder to behold"; and of the palace as a whole,"To the astonishment of all peoples I raised aloftits head. "87

The importance of this phrase, "to [or for] theastonishment of all peoples" (ana tabrat kizSatnze) cannot be too strongly emphasized. It is anexact translation of a Sumerian formula of refer-ence to impressive building, largely applied totemples in the earlier periods (u6-di un ar). InNeo-Assyrian usage, both temples and palaces areso described, but it is especially characteristic oftexts referring to new palace constructions Itwould be interesting to survey extant attestationsto see whether it is possible to determine a timewhen "astonishment" was accorded to palaces as

37

IRENEJ. WINTER

well as temples, and whether this correlates withany significant developments in the Mesopotami-an state and in the institution of kingship. Garellihas discussed some of the attributes of royalpalaces intended to astonish. He noted that stan-dardwords for "beautiful" or "well-built" were notused for royal buildings; rather, one finds a vo-cabulary focusing on qualities also applied to theperson of the king. By Garelli's account, becausethe palace was the work of the king, it was pos-sessed of the same "splendor" and "majesty" asthe king himself." However, it should be remem-bered that it is the ruler's own voice in the textsthat articulates the qualities of the palace, so thatthe king actually imbues the building-or assertsthat his buildings are imbued-with those veryqualities which he would also have ascribed tohimself. The palace is thus set up as a mirror ofthe king. It is a physical manifestation of theruler's power and ability to build; and at the sametime, by having built so impressively, the ruler hasfurther demonstrated his power and ability tocommand resources, induce astonishment, andcreate a fitting seat of government-in short, torule. The rhetorical function of the palace, asexemplified through its affect, is, I would argue,as essential as its residential, administrative, pro-ductive, and ceremonial functions.

Throughout the preceding survey, I have tried todemonstrate that morphology, decorative pro-gram, and function are not independent vari-ables. Rather, room type, organization of space,individual decorative motives, and overall deco-rative scheme are fundamentally linked to func-tion. Given the nature of the archaeological andtextual record, in any scholarly study of the pal-ace we are limited to the expression of the royalvoice, which privileges rhetoric and intentionover actual practice. Obviously, as is all too wellknown in modem times, buildings can be poorlydesigned for anticipated functions. Equally, whenbuildings are secondarily occupied, or when his-torical events precipitate change, then their orig-inal form and decoration can either constrainfunction or have little relationship to new usage.Nevertheless, recovery of the ideal schema andthe associated originating rhetoric is a necessaryfirst step toward any critique of the fit betweenintention and actual practice.

In the Mesopotamian schema, and apparentlyin the Hittite one as well,"° the palace was con-ceived as incorporating a bundle of activitiesand functions: residential, political, administra-

tive, industrial, ritual, ceremonial, and affective.Storage and display of surplus and luxury goodsserved as extensions of elaborate decorative pro-grams that articulated state ideology, and spaceswere designed to meet the functional needs ofthe palace as an institution.9' Limited compari-sons with other palace types have suggested theimportance of regional diversity; but at the sametime, it can be demonstrated that regionally spe-cific building forms and decorative practices couldbe transmitted across regional boundaries undercertain political or cultural conditions. There aremany ancient Near Eastern palaces not men-tioned in this brief survey, and many palaces thathave been mentioned have received less thanadequate description or analysis. Because I havebeen sketching with a broad brush, there hasbeen a tendency to emphasize similarities in theMesopotamian sequence across some two thou-sand years. In many ways this is not unjustified.From early royal hymns in Sumerian to later royalinscriptions in Assyrian Akkadian, indicationsare that the palace was construed as the seat ofkingship, not merely as the residence of the king.Nevertheless, micro and macro shifts in form anddecoration need to be studied more closely withrespect to the many political changes in statedevelopment over this long period. What I havetried to stress throughout is the role of the palacewithin the context of the state and the rhetoricalfunction of the palace as embodiment ofthe state.

If there turn out to be significant continuitiesin building materials and techniques, decorativeprograms (especially for non-figural motifs ofsymbolic value), and ceremonial/administrativefunctions from the pre-Islamic to Islamic periodsin the Near East, beyond the few I have notedhere, it will not surprise me at all; nor will I besurprised if distinct regional traditions withinIslam actually reflect recognizable subdivisionsin earlier periods as well. At the same time, onewill want to take care to distinguish featuresapparently similar over time that are merely theconsequence of relatively limited ways of repre-senting/organizing authority (what evolutionarybiologists call spurious homologies) from fea-tures that truly represent continuity in underly-ing concepts and traditions. When continuitycannot be demonstrated, it then becomes neces-sary to account for the differences as artifacts ofdiffering historical practice. I confess that a sig-nificant part of my mission in the foregoingsurvey has been to convince historians of Islamicarchitecture (and culture) that the pre-Sasanian

38

"SEAT OF KINGSHIP"/"A WONDER TO BEHOLD"

pre-Islamic world should be included in theirscholarly purview. Yet it has also become appar-ent that, however much we acknowledge thesignificant divide between before and after theintroduction of Islam, students of the ancientNear East have much to learn from consideringthe more complete historical record of Islamicpractice-by which I mean not only buildingpractice, but also cultural and political practice.

In the end, what is so clear as to be obvious, butstill needs to be stated, is that any study of thepalace, whatever the historical period, is funda-mentally linked to the study of concepts of au-thority and rule. To understand the palace, onemust see it as the locus of a particular practice ofgovernance. Furthermore, when continuities of

morphology and/or decoration occur acrossspatio-temporal boundaries, one cannot imme-diately assume continuity in meaning; wheneverpossible, it is necessary to establish associativesignificance independently. Conversely, it is pos-sible that differences in morphology and/or dec-oration nevertheless represent quite similar so-cial and political systems.

For the ancient Near East, the play in thesubtitle of this survey was a conscious one: thepalace is both a physical and a mental construct,both builtand construed. Itis at once the concen-trated center of rule, "the seat of kingship," andalso the concrete expression of rule, "worthy ofbeing" the seat of kingship, "for the astonishmentof all peoples," "a wonder to behold."

39

IRENEJ. WINTER

Notes

My sincere thanks to Jfilide Aker, Jak Cheng, MarianFeldman, and Ann Shafer for permission to cite unpub-lished works produced as seminar papers at HarvardUniversity, and toJuilide Aker and Margaret Sevenkofor tough readings of an earlier draft of this paper. Mythanks also to Barbara N. Porter for her generosity inproviding a copy of her paper on the date palm for useby me and my students prior to its publication. Conver-sations with many colleagues will have found their wayinto a number of aspects of this broad survey. I amgrateful to all.

1. For the non-specialist, it may be useful to note thatSumerian-speaking peoples precede Akkadianspeakers in the lower Tigris-Euphrates valley. Thetwo languages and peoples mixed in the thirdmillennium B.C., after which time we find Sumeri-an loan words in Akkadian, and evidence for Akka-dian influences and words in Sumerian as well.Akkadian briefly (ca. 2300 B.C.) and then ultimate-ly (ca. 1900 B.C.) replaced Sumerian and serves asthe language of Babylonia and Assyria. Akkadianscribes and scholars continued to be taught Sum-erian, copying old inscriptions aspartof the schoolcurriculum, and preserving literary and ritual textsthrough translation and bilingual word lists. Bilin-guals continued into the early first millenniumB.C., by which time Sumerian was no longer a livinglanguage. The convention in a modern scholarlytext is to indicate Sumerian words in boldface typeand Akkadian words in italics.

2. Note that the Sumerian word for temple, by con-trast, consists merely of the logogram for "house"followed in construct by the name of the deity, asin E.dInanna, "the house of the god(dess) Inanna."My translation of "Great House" for the Mesopot-amian palace should not be confused with a struc-ture so indicated and distinct from the palace/ekallu in Hittite tradition (see H. G. Giiterbock,'The Hittite Palace," in Le palais et la royauti: XIXCrencontre assyriologique internationale 1971, ed. P.Garelli [Paris, 1974], 305-6).

3. J. V. Kinnier-Wilson, TheNimrud Wine Lists:A Studyof Men and Administration at the Assyrian Capital inthe 8th Century B.c. (London: British School ofArchaeology in Iraq, 1972), 45: year 34 of Samsui-luna of Babylon; G. Pettinato, review of Romer,Sumerische 'Konigshymnen, in Archiv fir Orientfor-schung (1992), 208-10, line 99, re Lipit-Ishtar ofIsin.

4. And when multiple functions are characteristic ofHittite palaces of the late second millennium B.C.

on the Anatolian plateau (cf. Guiterbock, "HittitePalace"), this is not to say that they are the same

functions. Each historical and cultural case needsto be examined first in terms of internal evidencefor associated activities.

5. E.g., Ernst Heinrich, Die Paldste im alten Mesopota-mien, Denkmaler antiker Architektur 15, Deut-sches Archaologisches Institut (Berlin, 1984);Seton Lloyd and H. W. Muller, AncientArchitec-ture (New York: Rizzoli, 1986); Jean Mar-gueron, Recherches sur les palais misopotamiens del'ige du bronze, 2 vols. (Paris, 1982); Rudolph Nau-mann, Architektur Kleinasiens: von ihren A nfangen biszum Ende der hethitischen Zeit (Tiibingen, 1955).

6. James Fergusson, The Palaces of Nineveh and Persep-olis Restored (London:John Murray, 1851), fig. 31and frontispiece.

7. HansJ. Nissen, The Early History of the Ancient NearEast: 9000-2000B.c. (Chicago, 1988), ch. 3, 39-64.

8. 0. Tunca, L 'architecture religieuse protodynastique enMisopotamie, Akkadica, Supplementum II (Louvain,1984); Ernst Heinrich, Die Tempel und Heiligtiimerim alten Mesopotamien (Berlin, 1982).

9. See plan in M. S. B. Damerji, TheDevelopment of theArchitecture of Doors and Gates in Ancient Mesopota-mia, trans. T. Takaso and Y. Okada (Tokyo: Insti-tute for Cultural Studies of Ancient Iraq, 1987),fig. 147; also published in Heinrich, Die Tempel, fig.118a, and Nicholas Postgate, Early Mesopotamia:Society and Economy at the Dawn of History (London:Routledge, 1992), fig. 4.2.

10. See discussion in Postgate, Early Mesopotamia, 137ff., and also reference to the issue of the connec-tions between the origins of royalty and the palacein early Mesopotamia in O. Tunca, review of Mar-gueron, Recherches sur les palais, in Zeitschrift firAssyriologie 74 (1984): 318.

11. See on this period, Heinrich, Die Paldste, 9-13,where some possible candidates for palaces arenoted, based on morphological comparisons withlater palaces, but without definitive evidence.

12. See on Kish, E. Mackay, A Sumerian Palace and the"A "Cemetery at Kish, pt. 2 (Chicago, 1929); P. R. S.Moorey, 'The 'Plano-Convex Building' at Kishand Early Mesopotamian Palaces," Iraq 28 (1964):83-98; Heinrich, DiePaliiste, 14-28; Postgate, EarlyMesopotamia, 137.

13. For a discussion of the spatial properties of thatcentral court in Palace A at Kish, seeJ. Margueron,"Remarques sur l'organisations de l'espace archi-tectural en Mesopotamie," in L'archiologie de l'Iraqdu dibut de l'ie'poque niolithique d 333 avant notre ire,

40

'SEAT OF KINGSHIP"/ "A WONDER TO BEHOLD"

Colloques internationaux du Centre National dela Recherche Scientifique, no. 580 (Paris: Editionsdu CNRS, 1980), 165-68.

14. See Heinrich, Die Paliste, 43.

15. See David Oates, "Excavations at Tell Brak, 1978-81," Iraq 44 (1982): 187-204, which also includesbibliography for the early excavations of the build-ing.

16. Heinrich, Die Paliiste, fig. 22.

17. This would certainly seem to be true at the Assyriansite of Khorsabad, where the royal palace situatedon the uppercitadelwas replicated in smaller scalein several buildings on the lower terrace. One ofthe lower buildings has been identified as theresidence of the king's brother and chief vizier;the other possibly belonged to his son as crownprince (Gordon Loud, "An Architectural Formulafor Assyrian Planning, Based on the Results ofExcavations at Khorsabad," Revue d'Assyriologie 33[1936]: 153).

18. See on Ebla the many publications of P. Matthaie,for example, Itesori diEbla (Rome and Bari: EditoriLaterza, 1985), 25-54 and figs. 14-36, 41-47.

19. It is indeed unfortunate, given the number ofadministrative texts and literary texts preservedfrom the Ur III period (ca. 2112-2004 B.c.) insouthern Mesopotamia, that we have not recov-ered any of the major palaces of Ur III rulers (cf.Postgate, Early Mesopotamia, 141). It is thereforedifficult to estimate how characteristic of the peri-od was the provincial palace built by an Ur IIIgovernor of Eshnunna in the Diyala River regionto the northeast (Henri Frankfort, Seton Lloyd,and Thorkild Jacobsen, The Gimil-Sin Temple andthe Palace of the Rulers at Tell-Asmar, Oriental Insti-tute Publications 43 (Chicago, 1940).

20. Dominique Parayre and Harvey Weiss, "Cinq cam-pagnes de fouilles a Tell Leilan dans la hauteJezireh (1979-1987): Bilan et perspectives," Jour-nal des Savants (Jan.-June 1991), 3-26. Unfortu-nately, however, not enough of the plan of thepalace has been excavated to permit comparisonwith other buildings.

21. See on Mari the initial excavation reports ofAndreParrot, MissionArchiologiquedeMari,vol. 2, pt. 1: LePalais: architecture, pt. 2: LePalais: peintures murales,and pt. 3: Le Palais: documents et monuments (Paris,1958-59); and more recently, Heinrich, DiePalste,68-81; Margueron, Recherches sur les palais, 209-380 (the most extensive discussion to date); Marie-Henriette Gates, 'The Palace of Zimri Lim at

Mari," Biblical Archaeologist (June 1984), 70-87.Interpretive studies include Y. al-Khalesi, The Courtof the Palms: A Functional Interpretation of the MariPalace, Bibliotheca Mesopotamica 8 (Malibu, 1978);Jean-Marie Durand, "L'organisation de l' spacedans le palais de Mari: le tmoinage des textes," inLe systime palatial en Orient, en Grice et d Rome,Universite des Sciences Humaines de Strasbourg,Travaux du Centre de Recherche sur le Proche-Orientet la Grce antiques 9 (Leiden, 1987) ;JeanMargueron et al., "Les appartements royaux dupremier &tage dans le palais de Zimri-Lim," M.A.RI.6 (1990): 433-62; Beatrice Pierre, "D&cor peint t

Mari et au Proche-Orient," M.A.RI. 3 (1984): 223-54, and Beatrice Pierre-Muller, "Une grande pein-ture des appartements royaux du palais de Mari(Salles 219-220)," M.A.R.I. 6 (1990): 463-558.

22. See chart in Gates, BiblicalArchaeologist (June 1984)for divergent opinions.

23. M. Burke, "Une reception royale au palais deMari," Revue d'Assyriologie 53 (1959): 139-46.

24. Kinnier-Wilson, Nimrud Wine Lists, 32 ff.

25. Postgate, Early Mesopotamia, 149.

26. See on this Anton Moortgat, "Wandgemalde imPalaste zu Mari und ihre historische Einordnung,"BaghdaderMitteilungen 3 (1964): 68-74, al-Khalesi,The Court of the Palms, and Pierre, "Decor peint aMari."

27. See especially, Frankfort et al., The Gimil-Sin Templeand the Palace of the Rulers, where the palace of theprovincial governor of Eshnunna when the citywas under the hegemony of Ur included twoshrines, one presumably to the local deity, theother to the deified ruler of Ur.

28. Guiterbock, "Hittite Palace," 310.

29. Marc Brandes, "La Salle dite 'G' du palaisd'Assurnasirpal II a Kalakh, lieu de ceremonierituelle," Actes de la 17eime Rencontre AssyriologiqueInternationale (Gembloux, 1970), 147-54. Note,however, that Julian Reade is not persuaded byBrandes's argument that the suite around room Gwas used for ritual lustrations and suggests ban-quets as an alternative ("Assyrian ArchitecturalDecoration: Techniques and Subject Matter," Bagh-dader Mitteilungen 10 [1979]: 85). The argumentfor both is based upon the imagery of the seatedking holding a phiale-like bowl on the reliefs, andwhether he is lustrating or drinking is not certain.However, on scenes where individuals are clearlybanqueting, as in the reliefs of Sargon II at Khors-abad, their drinking vessels are obviously being

41

IRENEJ. WINTER

brought to the lips, which is not the case here, soI tend to be more persuaded by the symbolic thanthe literal in the present scene.

30. Cited above, n. 21.

31. Cited in I.J. Winter, "Art as Evidence for Interac-tion: Relations between the Assyrian Empire andNorth Syria," in Mesopotamien und seine Nachbarn:politische und kulturelle Wechselbeziehungen im AltenVorderasien, ed. Hartmut Kuihne, Hans J. Nissen,andJohannes Renger, Berliner Beitrge zum Vor-deren Orient, vol. 1, pt. 1 (Berlin, 1982), 363.

32. Moortgat, "Wandgemalde im Palaste zu Mari";also discussed in Moortgat, The Art ofAncient Meso-potamia: The Classical Art of the NearEast (London,1969), 82-84.

33. Khorsabad is a particularly good example since,like Samarra and Fatehpur Sikri, it was built andoccupied by a single ruler, Sargon II of Assyria.The phenomenon of new palace construction bysuccessive rulerswas noted by Oppenheim, AncientMesopotamia, and has been discussed in part bySylvie Lackenbacher, Le roi btisseur: Les ricits deconstruction assyriens des origines d Teglatphalasar 3(Paris: Editions Recherche sur les civilisations,1982): 76, as away of demonstrating that the rulersurpassed all of his predecessors.

34. Fort Shalmaneser, excavated at Nimrud between1949 and 1963, is a good example of the type--onwhich, see M. E. L. Mallowan, Nimrud and ItsRemains (Aberdeen, 1966) ,vol. 3, plan VII. For thecity walls, with citadel and ekal-miiarti set into theperimeter, see plan in Moortgat, Art of AncientMesopotamia, fig. 102.

35. Naumann, ArchitekturKleinasiens, fig. 259.

36. Note also the point made by Postgate, Early Meso-potamia, 137-40, that in the earlier periods inMesopotamia, the location of the palace variedsignificantly from one city-state to another andfrom one period to another: sometimes locatedclose to the traditional temple/sacred precinct,sometimes in newer sectors of town, away from theolder traditional areas, and this was closely tied tothe relationship of a particular ruler to the state.

37. Cf., for example,A. Kirk Grayson, RoyalInscriptionsof Mesopotamia 2 (Toronto, 1991): 38-45, for thepalace of Tiglath Pileser I atAssur and 227-28 forthe Northwest Palace of Assurnasirpal at Nimrud;John M. Russell, "Bulls for the Palace and Order inthe Empire: The Sculptural Program of Sennach-erib's CourtVI at Nineveh," Art Bulletin 69 (1987):520-39, and Sennacherib's Palace without a Rival at

Nineveh (Chicago: University of Chicago Press,1991). Although the Neo-Assyrian palaces are farbetter preserved, it would seem from the MiddleAssyrian texts that kings employed many of thesame building techniques as are archaeologicallyattested later-for example, the lining of the wallswith basalt and alabaster slabs, and the installationof large gateway figures, noted in the Tiglat PileserI inscription cited above.

38. See especially Edith Porada, "Battlements in theMilitary Architecture and in the Symbolism of theAncient Near East," in Studies in the History ofArchitecture Presented to Rudolf Wittkower (New York:Phaidon, 1967), 1-12; and Margueron, Recherchessurlespalais, pt. 2, ch. 2, on open and closed spacesin the Mesopotamian palace.

39. Heinrich, DiePalste, 89-91 and fig. 46; Margueron,Recherches sur les palais, 521-22; H. Gasche andW. Birschmeier, "Contribution l'6tude de lavofite en briques crues," Akkhadica 24 (1981): 1-16;and Damerji, Doors and Gates, 116 ff., for doorswitharched construction.

40. Heinrich, Die Paldiste, fig. 76.

41. On attention to doors and gates in general, theirconstruction and plan, see Damerji, Doors andGates.

42. On this, see Winter, in Mesopotamien und seineNachbarn, esp. 356. Note, however, that Readerightly points out Assyrian colossi have Mesopota-mian prototypes as well in smaller-scale free-stand-ing figures of glazed terra cotta or stone ("AssyrianArchitectural Decoration," 18).

43. It should be noted, however, that interior spatialconfigurations did not change. See on this I. J.Winter, "Perspective on the 'Local Style' of Hasan-lu IVB: A Study in Receptivity," in Mountains andLowlands:Essays in theArchaeology of GreaterMesopot-amia, ed. L. D. Levine and T. C. Young,Jr. (Malibu:Undena Press, 1977), 371-86.

44. The literature on this palace is vast, as it is one ofthe best preserved in the ancient sequence. For arecent study of the architectural basis for recon-struction, see Richard Sobolewski, "Beitrag zurtheoretischen Rekonstruktion derArchitektur desNordwest-Palistes in Nimrud (Kalhu)," in Palastund Hiitte: Beitrdge zum Bauen und Wohnen im Alter-tum (Mainz: von Zabern, 1982), 237-50. For thisparticular two-court plan, see Moortgat, Art ofAncient Mesopotamia, 127, following Loud, "AnArchitectural Formula," 156, in which the out-er court is referred to as the "gate-court" or biibnuin Akkadian, the inner court as the "house" or

42

"SEAT OF KINGSHIP"/"A WONDER TO BEHOLD"

"residential court," Akk. bitanu. My objections tothis terminology would take too long to arguehere, but in summary can be related to the likeli-hood that the rooms flanking the innercourtwerealso devoted to ceremonial or public functions,while the thickness of the walls (some 5 meters)argues again for a second story, with the residen-tial quarters more likely to have been located onthe upper floor(s).

45. Geoffrey Turner, "The State Apartments of LaterAssyrian Palaces," Iraq 32 (1970): 177-213, anddiscussion in I. J. Winter, "Reading Concepts ofSpace from Ancient Mesopotamian Monuments,"in Concepts of Space Ancient and Modern, ed. KapilaVatsyayan (New Delhi: Abhinav, 1991), 57-73.

46. AncientMesopotamia, 328, as cited in Winter, "Read-ing Concepts of Space," 63.

47. See on this Marian Feldman,'The Presentation ofKingship: The Neo-Babylonian ThroneroomFacade of the Southern Palace at Babylon," un-published paper, 1992, as well as S. Polony and G.Winkler, "Statische Untersuchung des Thronsaalsder Siidburg in Babylon," Mitteilungen der Deut-schen Orient-Gesellschaft99 (1968): 55-58. The Bookof Esther of the Hebrew Bible, 5:1-2, records thisstraight axis in use on into the Achaemenid periodof the second half of the first millennium as well,in its description of the Persian king seated on "hisroyal throne in the royal house opposite the gate,"and from there seeing Esther standing in thecourtyard.

48. "Scientific Description of Art," a review of DonaldWilber, Architecture oflslamic Iran, inJournal of NearEastern Studies 15 (1956): 93-102.

49. Turner, "State Apartments of LaterAssyrian Palac-es," cited above, n. 45.

50. See the lengthy discussion in Naumann, Architek-tur Kleinasiens, 354-78, for the sites of Zincirli,Carchemish, and Tell Halaf, among others. Morerecently on Tell Halaf, seeJeanny V. Canby, "Gu-zana (Tell Halaf)," in Ebla to Damascus: Art andArchaeology of Ancient Syria, ed. Harvey Weiss (Wash-ington, D.C., 1985), 332-38. Indeed, in the earlier,Hittite palace of the second millennium as well,although itwas apparently a large complex, Gfiter-bock has suggested that residence and administra-tive quarters may have been in separate structures(cf. Gfiterbock, "Hittite Palace," esp. 308).

51. Cf. reconstruction in Naumann, Architektur Klein-asiens, fig. 455.

52. See Winter, "Art as Evidence for Interaction."

Note also thatJ. Borker-Klahn ("Der bit-hilani imbit-ahiiri des Assur-Tempels," Zeitschrift fiirAssyri-ologie70 [1980]:258-72) has recentlydocumentedan instance in which a bi t-hilaniwas associated witha temple of the Neo-Assyrian period in Assur. Shereconstructs the building (Borker-Klfhn, "Der bit-hilani," fig. 4) with portal sculpture flanking alarge entry, and an upper colonnaded balconyabove the entrance.

53. I have even wondered whether there might not besome etymological relationship between the Ara-maean/Akkadian hilani and Arabic ()iwan, butsuch speculation goes well beyond my own compe-tence.

54. Richard D. Barnett, Sculptures from the North Palaceof Ashurbanipal at Nineveh (London: British Muse-um, 1976), pl. XXIII (room H).

55. See Grayson, Royal Inscriptions, 27, 55, and, for asimilar account by Assur-bel-kala, 105.

56. Cf. Donald J. Wiseman, "A New Stele of Assur-nasir-pal II," Iraq 14 (1952): 24-39, and morerecent translation in A. K. Grayson, Assyrian RoyalInscriptions, pt. 2 (Wiesbaden: Harrassowitz, 1976),para. 678, as well as discussion in Ann Shafer, "AMonument to the Center of Empire: Ashurnasir-pal II's 'Banquet Stela' in Context," unpublishedpaper, 1992.

57. See the early study by A. Leo Oppenheim, "OnRoyal Gardens in Mesopotamia,"Journal of Cunei-form Studies 24 (1965): 328-33; also Donald J.Wiseman, "Palace and Temple Gardens in theAncient Near East," in Monarchies and Socio-Reli-gious Traditions in the Ancient NearEast, ed. H. I. H.Prince Takahito Mikasa (Wiesbaden: Harrasso-witz, 1984), 37-43; and discussion in Lackenbach-er, Le roi batisseur, 124-29.

58. For the palace of Sennacherib at Nineveh, seeRussell, Sennacherib'sPalacewithoutaRival ch. 2,7-33; see also Pamela Gerardi, "Epigraphs and Assyr-ian Palace Reliefs: The Development of the Epi-graphic Texts," Journal of Cuneiform Studies 40(1988): 1-35, for a special case of epigraphs ap-plied to the narrative reliefs; andJulian E. Reade,"Sargon's Campaigns of 720, 716 and 715 B.c.:Evidence from the Sculptures," Journal of NearEastern Studies 35 (1976): 95-104, for the couplingof distinct types of texts with equally distinct typesof imagery in the palace of Sargon II at Khorsabad.In I. J. Winter, "Royal Rhetoric and the Develop-ment of Historical Narrative in Neo-Assyrian Re-liefs," Studies in Visual Communication 7, 2 (1981):2-38, I have attempted to demonstrate that in thepalace of Assurnasirpal II at Nimrud, even the

43

IRENEJ. WINTER

structure of the text finds parallels in representa-tional practice. In addition, shifts in grammaticalstructure when referring to the ruler (from thirdperson singular, as part of narrative accounts, tofirst person singular as self-presentation) are in-dicative, and not without parallels to the accountby Sheila Blair (see article in this volume) ofIlkhanid texts concerning the ruler (that shiftfrom third-person narratives to second-personexhortatives).

59. Ali Abou-Assafet al., La statue de TellFekherye et soninscription bilingue assyro-aramienne (Paris: EditionsRecherche sur les Civilisations, 1982), 64, re line19 in the Akkadian text and line 13 in the Aramaic.

60. Winfried Orthmann, Untersuchungen zurspdthetitis-chen Kunst (Bonn: Habelt, 1971), 501-2, 503-5,and pls. 24, 26-28.

61. Glazed bricks as part of palace decoration areattested in the Middle Assyrian period by TiglathPileser I, who continues a palace begun by hisfather, adding a facade of glazed brick (the color)of obsidian (black), lapis lazuli (blue), pappardilu-stone and parutu-alabaster (which must be yellowand white, respectively, for that is the color palettefound on later exterior glazed facades and walls).For the Assurnasirpal text, see Grayson, AssyrianRoyalInscriptions, 2:para. 677; indeed, the BanquetStele on which the reference occurs could itself beconsidered part of the palace program, as it stoodin niche EA of the great courtyard D of the palace,just adjacent to the throne room (see Shafer,"Monument to the Center of Empire"). The onlyproblem with this is uncertainty, given the lateroccupations of the palace, that this was the stele'soriginal location.

62. For discussion of the throne-room textiles, seePostgate, Early Mesopotamia, 143.

63. Julian E. Reade, 'The Architectural Context ofAssyrian Sculpture," Baghdader Mitteilungen 11(1980): 86.

64. Gerardi, "Epigraphs and Assyrian Palace Reliefs."

65. Feldman, "Presentation of Kingship."

66. See on this motif, now, Barbara N. Porter, "SacredTrees, Date Palms and the Royal Persona of Ashur-nasirpal II,"JournalofNearEastern Studies52 (1993):129-39.

67. See David Castriota, "Divinity, Kingship and Me-tonymy: The Imagery of the Sacred Tree in the Artof Syria and Mesopotamia in the Second Millenni-um B.C.," unpublished paper.

68. Unfortunately, however, not all throne rooms inAssyrian palaces have been completely recovered.The reliefs of the throne room of Sargon at Khors-abad (room VII on plan) were lost in an accidentafter their removal, those of Assurbanipal at Nin-eveh are poorly preserved, those of Tiglath Pileserwere dismantled in antiquity to be reused in a laterpalace. It is also likely that no matter how completethe archaeological remains of a given buildingmay be, we will never fully recover all of theelements that contributed to the decorative pro-gram. The loss of textiles as a contributing factorhas been noted. In addition, the actual thrones onwhich the ruler sat could well have been decoratedwith motifs appropriate to the ideology of rule.This is certainly the case with the two decoratedstone throne bases that have been preserved: oneof Shalmaneser III from Fort Shalmaneser (see P.Hulin, 'The Inscriptions on the Carved Throne-Base of Shalmaneser III," Iraq 25 [1963]: 48-69,and the recent analysis by Michelle I. Marcus,"Geography as an Organizing Principle in theImperial Art of Shalmaneser III," Iraq 49 [1987]:77-90), the other of Sargon II from Khorsabad(discussed in Winter, "Royal Rhetoric," 19 and fig.17). Neither of the throne bases adds a rhetoricalelement not also preserved in wall decoration,although the emphasis on state diplomacy on theShalmaneser III base offers a different nuance byvirtue of its primacy.

69. Winter, "Royal Rhetoric" (text in Grayson,AssyrianRoyal Inscriptions, 2:paras. 650-53). (For a differ-ent reading of this throne room, see Luc Bachelot,"La fonction politique des reliefs neo-assyriens,"in Marchands, diplomates et empereurs: Etudes en hon-neurdePaul Garelli, ed. D. Charpin and F. Joannes[Paris 1991], 109-28.)

70. Russell, "Bulls for the Palace," 520-39.

71. In the palace of Assurnasirpal atNimrud, there isclear indication by dress that the tribute bearersare foreigners from the West. While just one na-tional contingent is represented, at least in what ispreserved of the court D facade, Assurnasirpalmentions that envoys from twelve countries wereinvited to his inauguration festivities in the text ofhis banquet stele, which was set up in that samecourtyard. It is to be expected that all would havebrought gifts, and I see the throne-room facade asa kind of presaging of the delegations anticipatedat the inauguration. Sargon II, who restored theNorthwestPalace at Nimrud, similarly represents atribute-bearing foreign delegation on the facadeof his throne room (cf. Heinrich, DiePaldste, fig. 93= the west wall of court VIII, leading into thethrone room, room VII, our fig. 16). An indicationof the sort of high-level gifts between rulers

44

"SEAT OF KINGSHIP"/ "A WONDER TO BEHOLD"

appropriate to the opening of a new palace ispreserved in an exchange of letters between theking of Babylon and the king of Egypt in the latersecond millennium, in which the Babylonian kingdeclares that he has just built a new "house," isplanning a "house-opening," and invites the Egyp-tian king to attend. The Egyptian king responds bysending luxurious furnishings for the new palace,including ebony furniture overlaid with ivory andgold (cf. William L. Moran, The Amarna Letters[Baltimore and London: TheJohns Hopkins Uni-versity Press, 1992], EA 3 and EA5), very much ona parwith the chandelier sent by Queen Victoria tothe Ottoman court on the completion of the Dol-mabahce Palace in the nineteenth century (forwhich reference I am indebted to Juilide Aker).That these tribute scenes have a broader valencethan merely diplomatic gift exchange has beensuggested in a most interesting paper byJak Cheng,in which he argues that the representation oftribute is a means of proclaiming the stability ofthe economic base of the state through the effec-tive accumulation of wealth, and as such consti-tutes an important trope in the iconography of thesuccessful ruler (cf. "Tribute Scenes in the Pro-gram of Legitimation by Sargon II of Assyria atKhorsabad," unpublished paper, 1992).

72. For example, Jonathan Brown and W. H. Elliott,on the palace of Philip IV of Spain, A Palace for aKing: The 'Buen Retiro'and the Court ofPhilip IV(NewHaven and London: Yale University Press, 1986).

73. As defined and discussed by Louis Althusser, "Ide-ology and Ideological State Apparatuses," in Leninand Philosophy and Other Essays, trans. B. Brewster(London: New Left Books, 1971), 131-49 (origi-nally published in French, 1966).

74. Kinnier-Wilson, Nimrud Wine Lists, 32 ff.; Postgate,Early Mesopotamia, 145.

75. See discussion in Irene J. Winter, 'The King andthe Cup: Iconography of the Royal PresentationScene on Ur III Seals," in From Image to Insight:Studies in Honor ofEdith Porada, ed. M. Kelly-Buccel-lati (Malibu: Undena Press, 1986), 263, and Visha-kha N. Desai, Life at Court: Art for India's Rulers,16th-1 9th Centuries (Boston: Museum of Fine Arts,1985), figs. 12 and 13, for the Mughal paintings.

76. Winter, "The King and the Cup," 253-68; Post-gate, Early Mesopotamia, 150.

77. ThorkildJacobsen, "Lugalbanda and Ninsun,"Jour-nalof Cunieform Studies 41 (1989): 69-86, esp. lines58-59, pp. 71 and 73.

78. Grayson, Assyrian Royal Inscriptions, 2:para. 653

(=Assurnasirpal II, Standard Inscription); D. D.Luckenbill, Ancient Records of Assyria and Babylonia2 (Chicago, 1927): para. 138 (=Sargon II).

79. Kinnier-Wilson, Nimrud Wine Lists, 95 ff.; Postgate,Early Mesopotamia, 141.

80. Cf. Giilru Necipoglu, Architecture, Ceremonial andPower: The Topkapi Palace in the Fifteenth and Six-teenth Centuries (Cambridge, Mass.: MIT Press,1991), 120-22, 140-41. See also in this volume,Necipoglu, figs. 2[17], 10a [1], 21 [16], planswhich show small mosques in the Ottoman, Safa-vid, and Mughal palaces.

81. On the processional route and recent archaeolog-ical work in Babylon, see H. Trenkwalder-Piesl,"The Procession-Street of Marduk in Babylon,"Sumer 41 (1985): 36-40.

82. Le roi bdtisseur, esp. 73-81; see also Lackenbacher,LePalais sans Rival: Les ricits de construction en Assyrie(Paris: Editions la DEcouverte, 1990), passim.

83. See, for example, the Chicago Assyrian Dictionary,vol. 'N', p. 187, entry under naklis Sennacheribrefers to an earlier palace, whose construction hadnot been artistic/skillful enough, Sa ekalli . . .epista1 la naklatma; his son Esarhaddon declaresthat he "had [a palace] built skillfully as his royalseat and for the pleasure of his lordship," anamfisab arriitiya u multa'fiti belutija naklis usepima.This is clearly a continuation of Middle Assyrianrhetorical practice, as when Tiglath Pileser I statesthat he built his palace at Assur with understand-ing and skill, decorating it in a splendid fashion,and making it fitting as a royal residence (cf.Grayson, Royal Inscriptions of Mesopotamia, 2:TP1.4.65-66, 77-89). In some of the royal correspon-dence from the later reigns of the Neo-Assyrianperiod, there is also evidence that rulers were notonly kept informed but engaged through corre-spondence in decisions regarding constructionand decoration. See, for example, Simo Parpola,State Archives of Assyria 1 (Helsinki: Helsinki Uni-versity Press, 1987): nos. 60, 61,110,133, and StateArchives of Assyria 5 (Helsinki: Helsinki UniversityPress, 1990): nos. 15, 56, 282, 293, concerningSargon II and the work on Khorsabad.

84. Cited in Lackenbacher, Le roi btitisseur, 74.

85. Grayson, Assyrian Royal Inscriptions, 2:paras. 680and 653, translates it as "in splendid fashion," andpara. 682, the "palace full of wisdom." I find morepersuasive the translations of Ann Shafer in "Mon-uments to the Center of Empire"): in the firstinstance because it is more grammatically accurateand in the second because it implies an "epithet"

45

IRENEJ. WINTER

applied to the palace, as opposed to mere descrip-tion.

86. Literally, "Palace for which there is no second [or,no equal] .

87. Luckenbill, Ancient Records ofAssyria and Babylonia,2:paras. 367, 389, 394.

88. See, for example, Grayson, Royal Inscriptions ofMesopotamia, 2:296, where Assurnasirpal uses acomparable phrase for his Sharrat-Niphi temple atNimrud.

89. Paul Garelli, "La conception de la beautk en Assy-rie," in Lingering over Words: Studies in Ancient NearEastern Literature in Honor of William L. Moran, ed.Tzvi Abusch, John Huehnergard, and PiotrSteinkeller (Atlanta, Ga.: The Scholar's Press,1990), 176 and passim, with respect to terms liketasribtu, "splendor, majesty, sumptuousness," from

the verb Sarahu, "to render majestic, sumptuous,imposing."

90. Giiterbock, "Hittite Palace."

91. Clearly, more work needs to be done in systematicstudy of the shifting role of the palace as institu-tion over time in the ancient Near East, as seenthrough changes in decorative program, room,and spatial distribution, sight lines, and avenuesofapproach. What is interesting is that the evidenceprovided by Mari and the Neo-Assyrian palacessuggests the palace decorative scheme consistednot merely of a single, generalized message appro-priate to the "seat of kingship," but of a series ofaccumulated messages, communicated in individ-ual rooms and areas, that were specifically tailoredto the function of associated spaces. One musttherefore "read" the sum of those messages to getcloser to the overall rhetorical program of thepalace.

46

"SEAT OF KINGSHIP"/"A WONDER TO BEHOLD"

APPENDIX

Bibliography for the Study of Ancient NearEastern Palaces

GENERAL TEXTSCassin, Elena,Jean Bottero, andJean Vercoutter, eds.

Die altorientalischen Reiche, vols. 2 and 3. Frankfurt amMain, 1966.

Ellis, Richard S., ed. A Bibliography of MesopotamianArchaeological Sites. Wiesbaden, 1972.

Frankfort, Henri. The Art and Architecture of the AncientOrient, 4th ed. New York, 1970 (originally published1954).

Lloyd, Seton and H. W. Muller. AncientArchitecture. NewYork: Rizzoli, 1986.

Moortgat, Anton. The Art of Ancient Mesopotamia: The

Classical Art of the Near East. London, 1969.Naumann, Rudolph. Architektur Kleinasiens: von ihren

Anfangen bis zum Ende der hethitischen Zeit. Tfibingen,1955.

Nissen, HansJ. TheEarly History of the Ancient NearEast:

9000-2000 B.c. Chicago, 1988.Oppenheim, A. Leo. Ancient Mesopotamia: Portrait of a

Dead Civilization, rev. ed. Chicago: University ofChicago Press, 1977.

Postgate, Nicholas. Early Mesopotamia: Society andEconomyat the Dawn of History. London and New York:

Routledge, 1992.Strommenger, Eva. 5000 Years of the Art of Mesopotamia.

New York, 1964.

PALACES (+ TEMPLES AND TEXTS)

Abou-Assaf, Ali, Pierre Bordreuil, and A. R. Millard. Lastatue de Tell Fekherye et son inscription bilingue assyro-aramenne. Paris: Editions Recherche sur lesCivilisations, 1982.

Abu-es-Soof, Behnam. "Further Investigations in Ashur-nasir-pal's Palace," Sumer 19 (1963): 66-67.

Albenda, Pauline. The Palace of Sargon, King of Assyria.

Paris, 1986.Al-Khalesi, Yasin. The Court of the Palms: A Functional

Interpretation of the Mari Palace. BibliothecaMesopotamica 8. Malibu, 1978.

Andrae, Walther. Das wiedererstandene Assur. Leipzig:J. C. Hinrichs' Verlag, 1938.

Bachelot, Luc. "La fonction des reliefs n6o-assyriens,"In Marchands, diploma tes et empereurs. Etudes en honneur

de Paul Garelli, ed. D. Charpin and F. Joannes, 109-28. Paris, 1991.

Barnett, Richard D. Sculptures from the North Palace ofAshurbanipalatNineveh London: The British Museum,1976.

Borker-Klahn, Jutta. "Der bit-hilani im bit-Sahuri desAssur-Tempels," Zeitschriftfiir Assyriologie 70 (1980):258-72.

Botta, P. E. and M. E. Flandin. Monument de Ninive.Paris, 1849 [=Khorsabad!].

Brandes, Marc. "La Salle dite 'G' du palais d'Assur-nasirpal II a Kalakh, lieu de c6r6monie rituelle," Actesde la 17ime Rencontre Assyriologique Internationale.Gembloux, 1970, 147-54.

Burke, M. "Une reception royale au Palais de Mari,"Revue d'Assyriologie 53 (1959): 139-46.

Canby, Jeanny V. "Guzana (Tell Halaf)." In Ebla toDamascus: Art and Archaeology of Ancient Syria, ed.H. Weiss, 332-38. Washington, D.C., 1985.

Cassin, Elena. "Le palais de Nuzi et la royaut6d'Arrapha." In Le palais et la royauti: XIX RencontreAssyriologique Internationale 1971, ed. P. Garelli, 373-92. Paris, 1974.

Castriota, David. "Divinity, Kingship and Metonymy:The Imagery of the Sacred Tree in the Art of Syriaand Mesopotamia in the Second Millennium B.c."Unpublished ms.

Cheng, Jak. "Tribute Scenes in the Program ofLegitimation by Sargon II of Assyria at Khorsabad."Unpublished paper, 1992.

Damerji, M. S. B. The Development of the Architecture ofDoors and Gates inAncientMesopotamia, trans. T. Takasoand Y. Okada. Tokyo: Institute for Cultural Studies ofAncient Iraq, 1987.

Durand,Jean-Marie. "L'organisation de l'espace dansle palais de Mari: le tmoinage des textes," in Lesystime palatial en Orient, en Grce et i Rome. Universitedes Sciences Humaines de Strasbourg: Travaux duCentre de Recherche sur le Proche-Orient et laGrce antiques, 9. Leiden 1987.

Eickhoff, T. Kar Tukulti Ninurta. Abhandlung derDeutsche Orient-Gesellschaft, Nr. 21. Berlin: MannVerlag, 1985.

Feldman, Marian. "The Presentation of Kingship: TheNeo-Babylonian Throneroom Facade of the SouthernPalace at Babylon." Unpublished paper, 1992.

Fergusson, James. The Palaces of Nineveh and PersepolisRestored. London:John Murray, 1851.

Frankfort, Henri, Seton Lloyd, and ThorkildJacobsen.The GimilSin Temple and the Palace of the Rulers at Tell,Asmar. Oriental Institute Publications 43. Chicago,1940.

Fritz, Volkmar. "Paliste wahrend der Bronze- undEisenzeit in Palastina," ZeitschriftdesDeutschesPaldstinaVereins 99 (1983): 1-42.

Garelli, Paul. "La conception de la beaut6 en Assyrie."In Lingering over Words: Studies in Ancient NearEasternLiterature in Honor of William L. Moran, ed. T. Abusch,J. Huehnergard, and P. Steinkeller, 173-77. Atlanta:The Scholar's Press, 1990.

Gasche, Hermann, and W. Birschmeier. "Contributiona l'6tude de la vofite en briques crues," Akkadica 24(1981): 1-16.

Gates, Marie-Henriette. "The Palace of Zimri Lim atMari," BiblicalArchaeologist (June 1984): 70-87.

47

IRENEJ. WINTER

Gerardi, Pamela. "Epigraphs andAssyrian Palace Reliefs:The Development of the Epigraphic Texts," Journalof Cuneiform Studies 40 (1988): 1-35.

Grayson, A. Kirk. Assyrian Royal Inscriptions, Pt. 2.Wiesbaden: Harrassowitz, 1976.

Gullini, G. "Struttura e spazio nell'architetturamesopotamica arcaica: da Eridu alle soglie delProtodinastico," Mesopotamia 5/6 (1970-71): 181-279.

Gufiterbock, Hans-Gustav. 'The Hittite Palace." In Lepalais et la royauti, ed. Paul Garelli, 305-13.

Heinrich, Ernst. Die Tempel und Heiligtiimer im altenMesopotamien. Berlin, 1982.

· Die Paldste im alten Mesopotamien. DenkmalerAntiker Architektur 15, Deutsches ArchiologischesInstitut. Berlin, 1984.

. "Mass und Ubermass in der Dimensionierungvon Bauwerken im alten Zweistromland," Mitteilungender Deutschen Orient-Gesellschaft 99 (1968): 5-54.

· Schilfund Lehm. Studien zur Bauforschung 6.Berlin, 1934.

Hulin, Peter. "The Inscriptions on the Carved Throne-Base of Shalmaneser III," Iraq 25 (1963): 48-69.

Kinnier-Wilson, J. V. The Nimrud Wine Lists: A Study ofMen andAdministration at theAssyrian Capital in the 8thCentury B.c. London: British School of Archaeologyin Iraq, 1972.

Klengel, Horst. "K6nigtum und Palast nach den Alalah-Texte." In Le palais et la royaute, ed. Paul Garelli, 273-82.

Koldewey, Robert. Das wiedererstehende Babylon. Leipzig1914 (Eng. trans. The Excavations at Babylon, trans.A. S. Johns, London 1914).

Lachemann, Ernest. "Le palais et la royaute de la villede Nuzi: les rapports entre les donnes archeologiqueset les donnes epigraphiques." In Lepalais et la royaute,ed. Paul Garelli, 359-72.

Lackenbacker, Sylvie. Le roi bitisseur: Les rcits deconstruction assyriens des origines d Teglatphalasar III.Paris: Editions Recherche sur les civilisations, 1982.

. Le Palais sans Rival: Les rcits de construction enAssyrie. Paris: Editions la Decouverte, 1990.

Lampl, Paul. Cities and Planning in the Ancient NearEast.New York, 1963.

Levine, Louis D. "Cities as Ideology: The Neo-AssyrianCentres of Ashur, Nimrud, and Nineveh," The Societyfor Mesopotamian Studies Bulletin 12 (1986): 1-7.

Liverani, Mario. "The Ideology of the Assyrian Empire."In PowerandPropaganda (Mesopotamia 7), ed. MogensTrolle Larsen, 297-317. Copenhagen: AkademiskForlag, 1979.

Loud, Gordon. Khorsabad I:Excavations in thePalace andat a City Gate. Oriental Institute Publications 28.Chicago, 1936.

. "An Architectural Formula for AssyrianPlanning, Based on the Results of Excavations atKhorsabad," Revue d'Assyriologie33 (1936): 153-60.

-- and C. B. Altman. Khorsabad II: The Citadel andthe Town. Oriental Institute Publications 40. Chicago,1938.

Luckenbill, D. D. AncientRecords ofAssyria andBabylonia,2 vols. Chicago, 1926-27.

Mackay, Ernest. A SumerianPalace and the "A " Cemetery atKish, Pt. 2. Chicago, 1929.

Mallowan, M. E. L. Nimrud and Its Remains, 3 vols.Aberdeen, 1966.

Marcus, Michelle I. "Geography as an OrganizingPrinciple in the Imperial Art of Shalmaneser III,"Iraq 49 (1987): 77-90.

Margueron, Jean. "Les palais de l'ige de Bronze enMisopotamie." In Le palais et la royauti, ed. PaulGarelli, 11-26.

. "Remarques sur l'organisations de l'espacearchitectural en Mesopotamie." In L'archiologie del'Iraq du debut de l'epoque neolithique d 333 avant notreire, 157-69. Colloques internationaux du CentreNational de la Recherche Scientifique, No. 580.Paris: Editions du CNRS, 1980.

· Recherches sur les palais misopotamiens de l'ige dubronze, 2 vols. Paris, 1982.

. "Enceinte sacre et palais," Histoire et archiologie80 (February 1984): 26-29.

, Beatrice Pierre-Muller, and M. Renisio. "Lesappartements royaux du premier etage dans le palaisde Zimri-Lim," M.A.RI. 6 (1990): 433-62.

Matthiae, Paolo. I tesori diEbla. Rome and Bari: EditoriLaterza, 1985.

. "The Reception Suites of the Old SyrianPalaces." In De la Babylonie d la Syrie, en passant parMari: Milanges offerts dJ-R. Kupper, ed. O. Tunqa, 210-28. Liege, 1990.

Meissner, Bruno and D. Optiz. Studien zum BitHilani imNordpalast Assurbanipals zu Ninive. Abh. Berl. 1939,18. Berlin, 1940.

Meuszynski, Janos. Die Rekonstruktion der Reliefdar-stellungen und ihrer Anordnung im Nordwestpalast vonKalhu (Nimrud). Mainz am Rhein, 1981.

~-- . "La facade de la salle du tr6ne au Palais Nord-Ouest a Nimrud," Etudes et Travaux 5 (1979): 5-13.

--- , and H. Abdul-Hamid. "Ekal Assur-nasir-palz II:First Report on Reliefs of Rooms 'B' and 'L'," Sumer30 (1974): 111-19.

Moorey, P. R. S. 'The 'Plano-Convex Building' at Kishand Early Mesopotamian Palaces," Iraq 28 (1964):83-98.

Moortgat, Anton. "Wandgemalde im Paliste zu Mariund ihre historische Einordnung," BaghdaderMitteilungen 3 (1964): 68-74.

Moran, William L. The Amarna Letters. Baltimore andLondon: TheJohns Hopkins University Press, 1992.

Oates, David. "Excavations at Tell Brak, 1978-81," Iraq44 (1982): 187-204.

Oppenheim, A. Leo. "On Royal Gardens in Mesopo-tamia,"JournalofCuneiform Studies 24 (1965): 328-33.

48

"SEAT OF KINGSHIP"/"A WONDER TO BEHOLD"

Orthmann, Winfried. Untersuchungen zur spdthetitischenKunst. Bonn: Habelt, 1971.

Paley, Samuel M. and Richard P. Sobolewski. TheReconstruction of the Relief Representations and TheirPositions in the Northwest Palace at Kalhu (Nimrud) II.Mainz am Rhein, 1987.

Parayre, Dominique and Harvey Weiss. "Cinqcampagnes de fouilles i Tell Leilan dans la hauteJezireh (1979-1987): Bilan et perspectives," Journaldes Savants (Jan.-June 1991), 3-26.

Parrot, Andre. Mission Archiologique de Mari II(1): LePalais: architecture. Paris, 1958.

-- -. Mission Archiologique de Mari II (2): Le Palais:peintures murales. Paris, 1958.

· Mission Archiologique de Mari 11(3): Le Palais:documents et monuments. Paris, 1959.

Pedersen, Olof. Archives and Libraries in the City ofAssur,vols. 1-2. Uppsala, 1985-86.

Pierre, Beatrice. "D&cor peint Mari et au Proche-Orient," M.A.R.I. 3 (1984): 223-54.

Pierre-Muller, Batrice. "Une grande peinture desappartements royaux du palais de Mari (Salles 219-220)," M.A.R.I. 6 (1990): 463-558.

Polony, Stefan, and G. Winkler. "Statische Unter-suchung des Thronsaals der Suidburg in Babylon,"Mitteilungen derDeutschen Orient-Gesellschaft99 (1968):55-58.

Porada, Edith. "Battlements in the MilitaryArchitectureand in the Symbolism of the Ancient Near East." InStudies in the History of Architecture Presented to RudolfWitthower. New York: Phaidon, 1967, 1-12.

Porter, Barbara N. "Sacred Trees, Date Palms and theRoyal Persona of Ashurnasirpal II," Journal of NearEastern Studies 52 (1993): 129-39.

Reade,Julian E. "Sargon's Campaigns of 720, 716 & 715B.C.: Evidence from the Sculptures,"Journal of NearEastern Studies 35 (1976): 95-104.. "Assyrian Architectural Decoration: Techniques

and Subject Matter," BaghdaderMitteilungen 10 (1979):17-49.

- -. "The Architectural Context of AssyrianSculpture," BaghdaderMitteilungen 11 (1980): 75-87.

Russell,John M. "Bulls for the Palace and Order in theEmpire: The Sculptural Program of Sennacherib'sCourt VI at Nineveh," Art Bulletin 69 (1987): 520-39.

. Sennacherib's Palace without a Rival at Nineveh.Chicago: University of Chicago Press, 1991.

Schroeder, Eric. "Scientific Description ofArt," a reviewof Don ald Wilber, Architecture ofIslamicIran, in Journalof Near Eastern Studies 15 (1956): 93-102.

Shafer, Ann. "A Monument to the Center of Empire:Ashurnasirpal II's 'Banquet Stela' in Context."Unpublished paper, 1992.

Sobolewski, Richard P. "Beitrag zur theoretischenRekonstruktion der Architektur des Nordwest-Palastes in Nimrud (Kalhu)." In Palast und Hiitte:Beitrige zum Bauen und Wohnen im Altertum, 237-50.Mainz: von Zabern, 1982.

Trenkwalder-Piesl, H. "The Procession-Streetof Mardukin Babylon," Sumer 41 (1985): 36-40.

Tunca, O. L'architecture religieuse protodynastique enMisopotamie. Akkadica, Supplementum 2. Louvain,1984.

.Review of Margueron, Recherches surles palais... , in ZeitschriftfiirAssyriologie 74 (1984): 306-12.

Turner, Geoffrey. "The State Apartments of LaterAssyrian Palaces," Iraq 32 (1970): 177-213.

Unger, E. Babylon: Die heilige Stadt nach der Beschreibungder Babylonier. Berlin, 1970.

Winter, Irene J. "Perspective on the 'Local Style' ofHasanlu IVB: A Study in Receptivity." In Mountainsand Lowlands: Essays in the Archaeology of GreaterMesopotamia, ed. L. D. Levine and T. C. Young, Jr.,371-86. Malibu: Undena Press, 1977.

. "Royal Rhetoric and the Development ofHistorical Narrative in Neo-Assyrian Reliefs," Studiesin Visual Communication 7, 2 (1981): 2-38.

"Art as Evidence for Interaction: Relationsbetween the Assyrian Empire and North Syria." InMesopotamien und seineNachbarn:politischeund kulturelleWechsel-beziehungen imAlten Vorderasien, ed. H. Kuihne,H.-J. Nissen, andJ. Renger, 355-82. Berliner Beitrigezum Vorderen Orient, Band I, Teil I. Berlin, 1982.

"The Program of the Throneroom ofAssurnasirpal II." In Essays on Near Eastern Art andArchaeology in Honor of Charles Kyrle Wilkinson, ed.P. O. Harper and H. Pittman, 15-31. New York, 1983.

."The King and the Cup: Iconograpy of theRoyal Presentation Scene on Ur III Seals." In Insightthrough Images: Studies in Honor of Edith Porada, ed.Marilyn Kelly-Buccellati, 253-68. Malibu: UndenaPress, 1986.

. "Reading Concepts of Space from AncientMesopotamian Monuments." In Concepts of SpaceAncient and Modern, ed. Kapila Vatsyayan, 57-73.New Delhi: Abhinav, 1991.

Wiseman, D.J. "A New Stele ofAssur-nasir-pal II," Iraq14 (1952): 24-39.

. "Palace and Temple Gardens in the AncientNear East." In Monarchies and Socio-Religious Traditionsin the Ancient Near East, ed. H. I. H. Prince TakahitoMikasa, 37-43. Wiesbaden: Harrassowitz, 1984.

49