/
U       R       K       E       S H
The   First   Hurrian Capital
By Giorgio Buccellati and Marilyn Kelly-Buccellati
A n early Hurrian myth, preserved in a Hittite version, / \     tells the story of a young god, Silver, who lives JUJ   with his mother somewhere in the countryside /       \ away from the cities. He has a quarrel with his friends, who taunt him because he has no father. Prompted, his mother tells him that his father lives in the big city:
Oh Silver! The city you inquire about, I will describe to you. Your father is Kumarbi, the Father of tire city Urkesh. He resides in Urkesh, where he rightfully resolves the lawsuits of all the lands. Your brother is Teshup: he is king in heaven and is king in the land. Your sister is Sauska, and she is queen in Nineveh. You must not fear any of them. Only one deity you must fear, Kumarbi, who stirs up the enemy land and the wild animals (adapted from Hoffner 1990:46-47).
Let us consider a possible etiology. We may assume that the story's original setting is in the mountains, where silver, the metal, is actually mined. There are contrasts among the small groups of mountain people. A group that controls silver establishes contacts with the cities of the plain, claiming ethnic affiliation (the chief god of the city, Kumarbi,
A view of the Tur Abdin (the southern edge of the Taurus range) from Tell Mozan (the surface of the tell is visible in the foreground). The large saddle in the mountain range marks the location of the modern city of Mardin, where a strategic pass leads to rich copper mines (Ergani) to the North. (All illustrations courtesy of the International Institute for Mesopotamlan Area Studies.)
is the fattier of Silver); accepting the urban rule of law (Kumarbi administers justice); and paying allegiance in return for defense (Kumarbi is in control of enemies, both human and animal).
Urkesh
All told, this myth would seem to provide a rather transparent idealization of the relationship between mineral resources and their commercial exploitation! At any rate, se non e vero, e ben inventato: this scenario may not be true, but it describes well the broader geo-political situation in northern Syro-Mesopotamia as we can reconstruct it in the early historic periods. Metals (especially copper) were being mined in the Taurus, where urbanization proper had not yet developed, and were shipped to the urban markets in the plain. Some of these cities, in the piedmont area, were more likely than others to serve as gateways for this trade. Urkesh holds a privileged place in the myth. Its location at the site
Biblical Archaeologist 60:2 (1997) 77
il trade routes along the Euphrates and in the Khabur : Urkesh was at the crossroads of two major arteries— nnecting the Mesopotamian plains with the land, and East-West along the southern edge of the
High Mound indicating the major excavation areas, difference in elevation (some 15 m) between temple nit, in the center of the tell, and building AK, on the . The major phase of the Temple dates to about 2450 ing AK dates to about 2200 bce. This would ighly to the ancient skyline, with the temple ! urban landscape.
l underscores its strategic position on the trade orth to south. Whatever the merits of the pro-5y, the fact remains that Urkesh was the city of ial god of the Human pantheon, and that his /ith the urban world is explicit and detailed. It such mythical prestige of Kumarbi's city was siderable cultural and political significance from eriods.
rust have been important politically as well. Sev-e known from this city, and the Hurrian title ;ted only in connection with this city name. In is the only Hurrian city for which one could line of kings at the end of the third millennium id Kelly-BucceUati [henceforth B&K-B]1988:31-ipect, the title of our article needs a qualification.
Supporting walls
Perimetral walls
A Plan of Temple BA showing stone ramp, in front of the doorway which leads to what is known as a bent-axis approach to the sanctuary. The section at the top shows a projection of the pitched roof as reconstructed on the basis of calculations derived from the nature and size of the walls.
V One side of a double sided stele with a plowman behind a draft animal and his dog. It was found near Temple BA, and it probably dates to about the same time period (2450 bce). Note the effective rendering of movement in the gesture of the plowman pushing the plow into the furrow.
It could be taken in its double meaning that Urkesh was the earliest Hurrian capital and that it was the first one to have been positively identified. The latter is true enough; the former makes an exaggerated claim. It is in fact quite likely
that other major urban centers in the piedmont area of the Taurus, from Tell Chuera to Nineveh, which were contemporary with Urkesh, should have been independent from it politically and also ruled by Hurrian dynasts. But even if, and when, evidence for such new Hurrian capitals may be found, in the current state of our knowledge Urkesh certainly holds a position of predominance.
Tell Mozart
In 1984, when we set out to choose a new site to excavate in Northern Syria, after some ten years of work at Terqa, we picked Tell Mozan, proposing that it might be, precisely, Urkesh (B&K-B 1988:25, 38). In 1995, after eight seasons of excavation, we were able to confirm our hypothesis: the evidence came from small and fragile seal impressions, discarded on the floor of a large room in a well preserved building, which we labeled AK (B&K-B 1995b; 1996a; see also below). From the nature of the deposit and the layout of the building, we first assumed this to be a storehouse. While this initial assumption is still valid, what appears now is that this storage area is attached to a much larger architectural complex, which we take to be the royal palace. But more about this later. First, a bird's eye view of the site and the major finds to date.
During the first seasons, work was concentrated on the area of the city wall encircling the high mound and the temple of the lion, Temple BA. The city wall measured at least six meters high and eight meters thick. Directly in front of this wall, builders constructed a smooth glacis. Both the function of the wall and the glacis were negated when, at a later date, inhabitants dumped material from a burned building in the moat; it is dated by the door sealings and pottery found in the burnt deposit to Early Dynastic (ED) HI (2600-2350 bce; B&K-B 1988:65-82). We think that this wall was replaced by a structure surrounding the Outer City.
Ceramics and a few seal impressions date the latest well preserved level in Temple BA, with its stone offering table, to just after the abandonment of the city wall. A stone lion emerged from the destruction level of this phase of the temple (B&K-B 1997a:61-62). A certain amount of realism in the rendering of its hair pattern characterizes its style.
Just outside the temple, in a disturbed level under the sod layer, excavators unearthed a double-sided, round-topped stele (Kelly-Buccellati 1989). One side shows an animal herd very realistically as being in motion. The opposite side depicts a plowman behind his draft animal with a dog above him. What is unique here is the fact that the plowman is using a diagonal line to push off from, as it were, in his motion forward. The dynamism of each scene is striking as is its subject centered on a simple act of daily life. Daily life scenes are a common feature of the seal impressions recently excavated by us and discussed below.
In 1990 we started the excavation of a step trench on the western side of the mound. At the base of this trench we discovered a large building containing a number of seal impressions. This stratified corpus of inscribed seal
Biblical Archaeologist 60:2 (1997)
Biblical Archaeologist 60:2 (1997)
71
tabled us to identify the site as ancient Urkesh. iged to the reigning Human dynasty: the king, reen, Uqnitu; and several royal courtiers. The yal nurse, Zamena, have Human names while one courtier have Akkadian names, lumber of inscribed seal impressions from ; palace AK reached about 170, out of a total ie 1000 impressions. These 170 impressions m a total of only 17 seals. This does not mean, we have a mathematical proportion of 10 seal >r each seal. Far from it: the statistical disper-Teater. What emerges from these statistics is eal impressions belonged to the king, namely i as many as 5 seals were used. In other words, 2re stored on behalf of the king, and almost i sealed with a different seal. Many more objects se) were sealed on behalf of the queen; how-of her seals was frequent (with 34 rollings), f containers, namely 81, were sealed on behalf i, and specifically 28 with two distinct seals of rse (totalling 17 rollings for one seal and 11 for with the seal of the queen's chief cook; and idual whose title was not given. From what then, it appears that the storehouse was used he benefit of the queen and her household.
pressions
s and the Urkesh Dynastic Program eals from Urkesh offer a new phenomenon in im art: that is, the use and repetition of visual expression to dynastic concerns. This dynas-carried out through the seal designs of the nd two of the royal courtiers. Illustration of id the succession to the throne are paramount jphy of these Urkesh dynasts (B&K-B 1996a; als convey the more abstract of the two con-epicting power, through a juxtaposition of the i powerful wild animal, the lion, which is also an important god, Nergal.1 A scene of pre-and a warrior are also motifs indicating 3 1996b:77). The queen's seal (ql)—with her Tied on the backs of two servants—conveys }pt with different modalities. The other major iround the royal succession. In one of the king's instructed from three fragments, he is seated A'ith the crown prince, shown as a child, p in a gesture of filiation and dependence. The ed a star behind him in the field. The prince lead of a lion reclining at the foot of the throne is lion is rendered in realistic detail, especially :he way his mane is shown. A vessel stands t of the lion; it may be connected with the over-leld by a standing figure facing the scene. This finely articulated profile, may be human, but of the pose and the rigidity of the stance ed in the context of the overall scene. The
A Seal of King Tupkish with a lion reclining at the foot of his throne. The crown prince stands on the lion's head and touches the lap of the king in a dynastic gesture of dependence and filiation.
V Seal of the king (whose name and Hurrian title endan appear in the inscription) showing two attendants one of whom is holding a ball of thread (wool?) on the outstretched palm of the hand.
bull under the inscription box faces the figure rather than the king.2 The figure then may be interpreted as that of a deity. If this is true, the lion shown being fed by a deity emerges as a dynastic emblem, linking figuratively and symbolically the ruling king and the crown prince.
Another of the king's seals (kl), reconstructed from four rollings, depicts two attendants, one of whom is carrying what probably is a ball of wool or thread on the palm of an outstretched hand. On the left of the seal there may be a table and a seated figure, probably the king. The hat of one of the figures and the shape of the table leg can be paralleled with an uninscribed Urkesh sealing from the same floor. While a ball of thread may appear to be a strange icono-graphic element on the seal of the king, the role of textiles in the wealth accumulation of the city of Ebla is well known. On the much later Apadana at Persepolis, the final two
Inscribed Seals of the Royal Court King Queen Queen's Court
(k1-k6) (q1-q8) (h1-h4)
Corpus of inscribed seals belonging to the king and to the queen of Urkesh, and to members of the queen's court, showing the number of seals belonging to each. The seals depicted in this chart are identfied in the text by their owner (e.g., k = king) and numerical position in the columns (e.g., k2 = the second seal in the king's column).
Biblical Archaeologist 60:2 (1997)
Biblical Archaeologist 60:2 (1997)
81
elfth delegation also carry thread (Walser these seals fit into the Urkesh dynastic pro-3y stress filiation and power. The theme of le other hand, is explicitly depicted in the queen identified thus far. f the queen (q4), she is shown on the left ! braided hair3 style decorated with a braid i the end. A small girl, with the same bair irnament, touches her lap. Another seal of iffers a very similar scene with the addi-the corpus of queen's seals from this single ieven variations of this scene exhibiting iynastic continuity. A different seal of the i number of members of the royal family ne crown prince depicted on the seal of the me lap touching gesture) as well as a small r lap. With this scene we now have three long the figures: the prince, a princess, and children are important symbols of the con-nasty both through the succession to the i and the consolidation of power through ■r example, the daughter of Sargon, Enhed-ie en-priestess at Ur.-> A smaller child appears the nurse of Uqnitum, who has the Hur-nena. In this case Zamena is holding onto on the lap of the queen. In the field is an r, shown also on the royal family seal, which ■ royal male children.
intimacy shown through the queen's hold-Id in her seal and those of her nurse is one ing aspects of these royal seals. Touching < the older children to the king and queen, the tie through the royal line or in the case mgh the royal hierarchy. While the Urkesh al children, and courtiers are themselves a seals, as clearly indicated by the setting in A'e have are "portrait-like" representations als (B&K-B 1997b:196-99). Their depictions conographic settings, with some variation acteristics, were sufficient for their imme-ion even by those who could not read the il owners in the inscriptions. In the seal of ng the royal family, the head of the king is :ad of a figure seated in a similar fashion )wn similar to the feathered crown worn by . In the corpus of over 1000 seal impressions, ;ad which is a possible portrait of king Tup-
Seals
ribed seals, a number of scenes render com-ither being performed singly or in pairs. In es, two standing figures engage in an activ-rith a tall container while a third makes owl. Notably, the upturned elbow holds the ion in the design.5 One inscribed docket— ininscribed seal showing a figure with an
ilical Archaeologist 60:2 (1997)
(top) The Urkesh royal family (q2)
with queen Uqnitum and a small child on the left and Tupkish and the crown prince on the right, (center) Drawing of a portion of an impression made by the same seal as above. During the initial phase of our research we had not been able to identify any seal impression giving the face of king Tupkish. A fragment discovered at a later date seems to fit this scene, and is now tentatively identified as providing just such a portrait, (bottom) Urkesh impressions show people engaged in a variety of simple, every day activities. Here two individuals are making something in a tall vat while a third stirs a bowl.
Seal of the cook of the queen (h3): the name of the female cook was probably written in the inscription box above the young animal at the far right of the impression. Depiction of the professions of the seal owners is one of the hallmarks of Urkesh glyptic. The churning woman represents the cook; the butcher on the right holds a knife and a hornless victim.
arm extended over a table facing a seated figure holding a cup—may indeed refer in some way to the iconography of the queen's seals. The emphasis on the details of the gestures and also in some cases the iconography of the hat and dress links these types of uninscribed seals with those of the monarchs. It is also possible, given the example of the royal seals, that some themes on the uninscribed seals reflect the professions of the seal owners.
While a number of the uninscribed seals do render themes common from the south, as in the Shamash and the Etana seals, most of them are new, either of the dynastic and related types just discussed or a more schematic style. In this category belong scenes that emphasize the geometric frame around the figural scene or contain a number of discreet elements which can be the disarticulated heads of animals (D. M. Matthews, 1997:136-37, his Brak Style). These motifs have links with the same types of motifs found on the later Nuzi style seals and painting from Nuzi itself. The continuity in the visual arts extended to the architectural traditions, discussed below. It is now clear that there was a distinct type of northern art and architectural tradition existing at least from the middle of the third millennium. Urkesh Glyptic Styles
From the iconography and style, it appears that the seals of the queen, the nurse Zamena, and the cook were carved in the same workshop, if not by the same seal carver. In them, there is an emphasis on fitting the inscription box within the overall design. This is clearest in a queen's seal where her title is simply N1N, Queen (ql). Here two servants literally carry the inscription and therefore her name and title, on their backs. On Zamena's seals (hi and h2), two cases of the inscription box are shortened to accommodate the
head of the human-headed bull. The cook's seal (h3) boasts the two servants, the woman churning and the butcher, facing the inscription box where she is described as the female cook of Uqnitum. While the integration in the seal design of text and iconography is one of the characteristics of Akkadian art, the Urkesh emphasis on the integration of the cultural meaning, as well as form, is rare.6
In the seals of the queen and the two courtiers connected with her, the figures are more expressionistic than those of the king, with the proportionately large eyes, heads and hands, narrow faces with long chins. They possess an emphasis on gestures, as in the outstretched and elongated arm of the attendant above the table, the nurse holding the child by the wrist while touching the lap of the child, or the hands of the woman churning shown in the up-and-down motion by depicting the hands at different heights. This workshop produced new variations of body positions best exemplified by the bending figures in ql and h3. All the seals of the queen and her nurse Zamena, except ql, have motifs which are more specifically oriented toward the concept of succession. In fact, q3-8 are all variations of the scene whereby the queen is shown receiving homage from her daughter. This scene includes an attendant holding a bag or jar with an elongated arm extended toward the queen, also possibly an exaggerated gesture of homage.
The seals of the king are in general of a higher quality as they are more finely carved and, from the meager evidence for his seals from our corpus, appear to have a more varied content. The dramatically rendered prince standing on the head of the lion in the best preserved of the king's seals reflects the same intense interest in gestures to convey the deeper symbolic meaning as the queen's seals. However, the whole scene goes beyond the seals of the queen in showing a very dramatic royal setting. The seal carvers working for the king certainly were aware of the style of the queen's seals; the two figures in kl have similar characteristics in the rendering of the heads.
In terms of style, all the royal seals emphasize realistic details as shown in the carving of the horns and wool of the goat in the royal family seal or the boar under the inscription of one seal of the queen. The rendering of the lion on two seals of the king emphasizes the details of the lion's mane. Baskets are depicted in the seal of the queen's cook and in one seal of the queen herself. The rendering of realistic details also marks Old Akkadian seals from the south, but in the Urkesh royal seals the imagery is very different.
There is another striking aspect of the seal iconography of Urkesh: the motifs of the inscribed seals reflect the professions of the seal owners. In the cook's seal, a woman bends over a basket in which are placed two jars; she is churning.7 The profession of the cook is rendered through the action of this bending woman. The butcher with a kid and a knife stands on either side of the inscription naming the cook's profession. The nurse, too, had seals that visually indicated her profession as named in her inscriptions.
Biblical Archaeologist 60:2 (1997) 83
lis the theme includes the birdman, holding a staff, and a nude woman, both standing above a supine human figure. Two ain the same scene, except that one is a mirror image of the other. Such reversal of identical iconographic motifs was our first rkesh iconography could include such variations.
Seals
ice for all these seals comes only from their lay which are in most cases small and Thus far we have excavated over 1000 jm a single floor of the building; a rough esti-number of identifiable original seals represented s around eighty. The seals were rolled on clay «es, jars, bags, and baskets. The seal impres-glimpse of part of the Urkesh adrninistration, of the queen since the greatest proportion of ieals were hers. Officials of the queen (and • directly connected to her) must have had the al goods in her name. These officials probably r duties elsewhere in the city or in surrounding ey sealed the containers. In this building, how-ners were opened by breaking the clay sealings, :ause the goods in the containers were redis-e within the building itself. Only one sealing :wo rollings: one of the cook and one of the mber of seals belonging to the king are lim-)es not appear to have been a principal area h bureaucrats linked to him directly. iat seal carvers working for the dynasts carved ->d royal seals in Urkesh. Only one inscribed been carved in the south; it resembles ED III idian southern models. The seal owner has an te, Innin Shadu. He certainly is an important Jrkesh court given the large number of his seal ;cavated from the floor of this building. Other ably arrived on containers from outside the t of the city; this is most certain in the case of ig showing the typically Akkadian motif of g-
s working in Urkesh must have been famil-adian art. Akkadian art influenced some the dynastic art of Urkesh such as the empha-f we take into consideration the round-topped nstratified, near Temple BA, then we can say
that the artists of Urkesh also considered the rendering of movement an important aspect of their art (Kelly-Buccellati 1990). The stele renders an everyday scene with an extraordinary movement of the plowman impelled forward by his own action of pushing from a diagonal line in the composition (see above, p. 79). There is no comparable scene in Akkadian art. Thus, while there are some superficial similarities between Akkadian art and that of Urkesh,8 the differences are so profound as to indicate that the spirit which enlivened the art of Urkesh was generated by local concepts and ideas and was not an imitation of southern models. The reappearance of these concepts in the first millennium art of the Neo-Hittite states in northern Syria offers a further indication of its original autonomy and continued dynamism.
What can the Urkesh dynastic program tell us about Akkadian art in the south? This type of visual communication could certainly have a place in the strong imperial concepts which were central to late Akkadian art. However no hints of this type of art can be found in Akkadian cylinder seals.9 In all probability, the Hurrian court at Urkesh itself developed and employed this dynastic pictorial vocabulary. Hurrian Artistic Style
In summing up the characteristics of this art, it is clear that artists from Urkesh developed a new visual language. Their art stresses two aspects of the internal vision of the Hurrian dynasty. Foremost is the concern for the royal succession as shown in the seals of the queen, but also in those of the courtier who was most interested in the succession, the royal nurse. In addition, on at least one of the king's seals there is a very clear statement of this same concern.
The other aspect of the dynastic program indicated on these seals is the power of the dynasts as shown through images of internal harmony and strength. In our corpus, this is not as widely indicated as the first emphasis. On one of the queen's seals, her servants are literally carrying her name and title on their backs. The king, by portraying himself seated on his throne with a reclining lion, connects himself with the
All the Urkesh glyptic evidence has been pieced together from over 1000 tiny sealings the size of which is indicated in this photo. The seal shown is one of the fragments from which the composite scene was derived that is shown here on page 82. Only the most meticulous excavation and the individual examination of thousands of clay lumps could hope to salvage such ephemeral material culture.
power of the lion, but more importantly with the deity symbolized by the lion. A presumably later king, Tishatal, also associated himself with a lion.
Other characteristics of this art include its naturalism, especially in the rendering of specific details, the emphasis on gestures to communicate the important message in the seal designs, the expressionism of the queen's workshop seals, and the connection of the profession with the iconography of the official's seal design.
Even when artists of Urkesh employ subjects prevalent in the south, they can construct the designs in a different manner, and, notably, sometimes carve them both in the positive and the negative as is the case with the birdman and the nude woman.
These new data help us place the question of "Hurrian art" on a new footing. Whether or not this art can be described as "Hurrian art" is too long of a discussion for this forum. Much of what is new in the dynastic program is art developed for a Hurrian dynast, his queen, and the royal courtiers connected with her. A number of irninscribed seals, too, exhibit similar characteristics, so that evidence for a body of artistic works connected with this dynasty is now being built up. How widespread this art is in the north in the third millennium is unclear. However, the idea of a visual vocabulary used to express dynastic concerns still existed in the first millennium in the north Syrian city states of the Neo-Hittite period, a weighty indication of the power of this idea in the north.
Schematic style seals represent a different style, one to
this point unconnected with the royal seal designs. These seals were more widespread in third millennium northern art. They had a resonance which was more continuous in the north as it existed still in the art of the Nuzi period which in turn influenced later art. We find examples of the schematic style prominently utilized at Urkesh and Brak, but also to some extent outside the northern area at Mari. A sealing from Ebla contains a border of disarticulated heads shown full-face, while geometric frames also turn up there (Kelly-Buccellati 1996). In other words, the northern schematic style emphasizing stock elements and geometric frames was at home in the north but is reflected in third millennium sites elsewhere.
While the Hurrian presence at Urkesh is strong and politically powerful, there is as yet little third millennium Hurrian evidence from Brak. Therefore we cannot connect, on the basis of our present evidence, the schematic style with Hurrian artists or Hurrian patrons in the same way that we can connect the dynastic art of Urkesh with its Hurrian royal patrons. Based on present evidence, the emerging picture of art in late third millennium northern Syro-Mesopotamia presents at least two styles first clearly indicated in the stratified context of the AK building at Urkesh. The first is the dynastic art linked to the Hurrian dynasty and at this point not found outside Urkesh. The second is a more widespread schematic style which had its roots in ED III glyptic but was more important in the Akkadian period. This style continued to dominate in the north into the Nuzi period. It cannot be linked specifically to Hurrian artists or patrons, although at least in the Nuzi period, it was probably also employed by artists who were Hurrian.
The Identification of Urkesh as Mozan
The most important epigraphic data consist of the inscriptions found on the impressions of cylinder seals: from these we have been able to identify ancient Urkesh with the site of Tell Mozan. This is all the more remarkable as the fragments that give us these inscriptions are very tiny, in a poor state of preservation, and so fragile that their very recovery was extremely difficult. One thing is certain: given the nature of the evidence, the identification of Urkesh could only have happened as the result of methodical and controlled excavations. Such an identification would in fact have been possible if the bronze lions of Tish-Atal had been traced immediately to Mozan; but no chance discovery would ever have brought to light the minute fragments of seal impressions with which we have been dealing in our excavations.
Even upon recovery, the reading of the pertinent inscription boxes proved very difficult. There are only three very
Biblical Archaeologist 60:2 (1997)
Biblical Archaeologist 60:2 (1997)
83
an
m
>f the same h, and a ing of the >ased on all ts.
	
	MW=Uti
	
3 on which the name Urkesh is found. Two ; to the inscription box of the king's seal ]f a globular object. The sign keS presents the y of being written with a sequence of strokes il, though not unattested. Finally, the fact that as to be read as a mirror image to make sense le the initial reading even more difficult. The >elongs to the inscription box of the seal with are four other inscribed seals of the king, em does the name Urkesh appear. The recon-ing of the legend with the royal name and part of the name of the city is still missing, ght be rendered in English transcription by as follows: U[r]kesh. However, the cuneiform 'eographical name is so distinctive that there >ut the restoration and therefore about the r the site.
llyptics: The Positioning of the Inscription
anexpected features of Urkesh glyptics is rrelation between scene and inscription.
the nuances of the dynastic program as por-eals of the king and the queen and in the }f pertinent figurative elements in the seals
of the nurse. But there is more. For just as he way in which the cuneiform inscription n many of these seals. We will illustrate here
the two most interesting examples.
In its early stages, cuneiform was written from top to bottom. We can verify this mode of writing on large monuments, which have an unequivocal right side up (e.g., the so called Obelisk of Manishtushu), and/or whenever the inscribed portion is combined with a figurative scene, as with the Code of Hammurapi or with seals. Assyriologists visiting the Louvre and looking at the Code are more likely to turn their head than to flip the stele sideways*. But with small items like tablets or seal impressions we generally hold the object in such a way that we can read from left to right. Accordingly, an epigraphist's view of a seal is regularly sideways.
The practice of top-to-bottom writing on seals is so universal in the third millennium that our seal of the queen reference (ql) came as a real surprise, since it was obviously read from left to right already in antiquity. It is the seal that spells out the title queen after her name, Uqnitum. Given the special interest of Urkesh glyptic artists in incorporating the legend with the scene design, it seems tempting to suggest that such an innovation in legend carving, unique within the third millennium and rare later, was intended for a special effect, which is underscored by the marked horizontality of the box. Let us contrast the seal of the queen with the others, where the inscription box serves as a vertical element which terminates on either side the sequential order of figures in the main composition. In most cases, the inscription box is linked with a minor compositional motif, which completes the vertical effect of the box, especially when the two are made to dovetail with each other, as in the case of the nurse's seals. In the seal impression of the queen, however, the frame is placed horizontally as if it were to be read at the same time that one looks at the scene. The horizontality of the inscription box is emphasized by the fact that there is no secondary motif beneath the legend. This absence of a secondary motif may itself have a specific connotative function, namely, to stress the burden that the rank of the queen places on her female servants, who are made to carry-on their back the frame with her name and title fully legible.
The seal of the cook (h3) presents yet another variation in the way in which the inscription boxes are integrated in the scene. We have, in fact, two distinct halves of the box. One is tall, and it occupies the entire field. In this case, however, the vertically of the box may serve a different connotative function than that of framing the scene on either side. It may in fact convey the exact opposite meaning: it gives the impression of a stele, centrally positioned so as to serve as the focal point faced directly by the two servants of the cook, the butcher and the churning woman. The two vertical framing elements, on either side of the composition, are provided by the smaller half of the box, which was abraded on the seal but must have contained the name of the chief cook. The box, placed vertically, sits on the back of the bent woman, without a minor compositional element, as if imitating the position of the inscription box on the seal of the queen!
Queen's seal (q1) where her official title is given with the Sumerogram NIN, "queen." The imagery seems to emphasize her official status, in that two female servants carry her inscription on their backs. The unity of idea and artistic design, the correlation between inscription and scene, represents a notable feature of glyptic style at Urkesh.
The Figurines of Urkesh
More than 300 small terra cotta sculptures have been found iri the Royal Storehouse and in adjacent areas of building AK at Urkesh. Third millennium floors have yielded both animals and human-like ' figures. ..
The anthropomorphic figurines can be grouped tvpologicallyaccording to the way they are terminated. A good number share the characterise . tics of the "gaming pieces" found in early strata at ; ivra, Tobier's "hitherto unknown amorphic type." A different type of torso, outside this typology,.but from the Royal Storehouse; has parallels at Tell Chuera. These pieces—and most of them are broken—are modelled inconsistent manner. Most are made from the same local clay and baked. Surface finishing techruques"include scraping with a sharp instrument and pinching and smoothing With fingers. A variety of incisions in the clay represent ; pelt, orifices, and the separation of body-parts. Dots provide decoration or indicate bodyrparts . and are applied with blunt or pointed sticks and reeds.
!• The animal figurines are remarkable for their realism, indicating that the figures were sculpted bv artisans who were familiar with the way these animals looked and behaved. Herbivores hold their heads high above the grasses of the steppe; a bridled stallion turns his head to the right, refusing to pose for an official portrait. There are many animals represented in the Urkesh storerooms. The relationship of body-parts within each animal . genus is consistent and permits identification. As might be expected, the animal figurines at Urkesh . include equids, dogs, sheep, and goats. Given that panthers roamed the steppe in the third millennium and that seals represent Tupkish, king of Urkesh, enthroned with a lion by his chair, we may hot be so surprised to find many felines in the ' figurine corpus. What is startling, however, is the realism of these small creatures. We can distinguish among the various species of eqiiid represented; ■ and we can trace the signs of domestication. ' The Urkesh figurines' were all found in the' .'same'third millennium strata in the Queen's Storeroom or adjoining rooms. We have, then, an "ideal single context to speculate responsibly about how. the Urkesh figurines were used. We are ■working on this puzzle—using the available textual clues—as of this writing. ■ Rick HaOser
iblical Archaeologist 60:2 (1997)
Biblical Archaeologist 60:2 (1997)
87
5 and Mirror Writing.
econd feature which is unique to Urkesh glyp-n the third millennium, and quite rare later: g. In our corpus, there are three different types :e.
/pe consists of paired matches. We have, in ,vo different sets of seal impressions which are age of each other. Clearly, we have the same als of the queen with her daughter (q4 and q5) . man (p. 84), except that in each pair one scene Tiage of the other.
i type consists of a single set of seal impres-in be identified as being mirror images because i only be understood when it is reversed. Such the king's seal with the bearer of a globular le can appreciate now the point made earlier ing in a seal inscription being normally from . rather than from left to right. When reading lirror image of the original impression, what lot the sequence of signs on a line, but rather of lines in a box. On the seal impression as le cuneiform signs are upside down (for n and the da); and the sequence does not make le title endan is followed by a personal name, ? king o/Tupkish. But in the mirror image of mpression the signs are right side up, and the re proper sequence: Tupkish, king of Urkesh. i explanation for such a doubling up of seal je the desire to identify different but related e retaining the same basic iconographic motifs, also consider another, quite different expla-s suggested by the third type of mirror imaging: doublet. Exemplified by the pair of seals belong-n's nurse, Zamena, this kind of doublet presents rly in the inscription box and not in the scene he two seals from which these composite seal ;rive are almost indistinguishable. The iconog-cal, and so is the sequence of lines in the legend size of one is slightly larger than the other, •iations appear in the writing. The signs on middle box) are right side up on the original m, whereas the signs on the other seal are for upside down. This clearly suggests that only 3 would have to be read as a mirror image. Lt the sequence of words? This is precisely what clue for the origin of this very curious Urkesh r-image writing. It may be (and we must stress :al nature of our suggestion) that the seal at top a an inverted word order, of a type which is ian noun phrases.10 If so, the seal at top would n version, so to speak, and the one below the idian version of one and the same prototype: rse-Zamena's seal versus the seal of Zamena, qnitum. Alternatively: the version at top would irrian on the seal impression and in Sumerian in the seal itself; and conversely for the version
"Uqnitum's nurse Zamena's seal"
Hurrian (?)
Akkacfian (?)
"seal of Zamena nurse of Uqnitum"
		
		
m		
Another case of two seals with a special type of reversal: while the scene is the same in both seals, in the seal at top the cuneiform signs are reversed in the impression (hence they would have been in the positive on the seal). The tentative suggestion is advanced that the same inscription might have been read with differing word order depending on whether it was understood as Hurrian or as Akkadian.
Building AK during excavation. A light canvas cover, draped over a light structure of poles and ropes, protects from the heat and the blinding light of the Summer sun. Excavations are regularly held in the Summer, because of the greater availability of workmen, who are for the most part either students or farmers. As excavation proceeded northward (to the left in the photo), the incorporation of the storehouse into a much larger complex became increasingly likely. Floors continued on the other side of the storehouse wall for nearly two dozen meters. At that point, excavators exposed traces of steps leading up to the floors, and the floors were thus able to be perceived as the threshold or porch of a monumental entryway.
below. This is a mere hypothesis, but, if correct, we would have evidence of a rather sophisticated use of multilinguism in the royal court of Urkesh and in its scribal circles.
The Architectural Setting of Building AK: Storehouse or Palace?
The building which we have labeled AK is located at the lowest end of the stepped trench on the east side of the mound. We had assumed that we were digging in a storehouse, to judge from the nature of the finds (especially the discarded sealings) as well as from some aspects of the architecture (for example, the fact that the walls were not plastered). It must have been, however, an important storehouse, given its large dimensions and the organic arrangement of its plan. In fact, the large number of inscribed seals referring explicitly to the king, his wife, and her retainers left no doubt that it served the needs of the royal court.
Because the southwestern portion of the building had been badly eroded, with rainwater settling in the south-
western corner, little if anything was found there, except for the foundations of the walls and the subfloors. However, this situation afforded us a good view of tire stratigraphic sequence before we even started excavating within the room. In anticipation of what these floor deposits might contain, we planned for a particularly careful excavation. It was only in this manner that we could retrieve a rich glvptic harvest, after collecting and reviewing individually thousands upon thousands of clay lumps, all potentially qualifying as seal impressions.
During our last season, in the summer of 1996, we had planned to extend our excavations to the north, assuming that we might find a few service rooms clustered around a courtyard. But it was to be quite different. There was no trace, in Sector F, of a wall to the west. In the place where we had expected a perimeter wall, the floors continued as if over a threshold. Some 22 meters to the north, we found clear remnants of steps leading up to this threshold. A round brick platform, and another possible platform, seemed to suggest the presence of column bases, though this remains quite
Biblical Archaeologist 60:2 (1997)
Biblical Archaeologist 60:2 (1997)
tly, our conception id changed. And so /e: instead of orient-X we now look at it dew of the entrance, mumental entrance eeping with a mere i end, we were only not really to test, a /hole building may 1 palace. It is for test-s that we are gearing ir the excavations in V. msis
some of the compar-that assists us in lace hypothesis. There }onents of the build->art in which we are \e entrance; (b) the accessed from the ; also as a transition sectors); (c) a service ctly from a and/or b; 'ice area, found only
urselves from the two-Tint mentality against ie Margueron (1982) ued, we can ask: What rnction of the entrance g? One type found in resented especially in I hilani model" gives 1 relief to the entrance: focal point in the exte-as a decompression nside (both psycho-latically). In the south, ce is like a panel which tween buttresses within and the decompression takes place through ridors which lead to a deep and recessed s a terminal point of arrival. Excellent bit hilani type of building existed at Zinjirli ctions recently published in Parker 1996:215): Drticoed entrances of the two buildings are, 1 examples of bit hilani structures—but some Che earliest known example of such a struc-his possible Urkesh example, is the one from m and Sigrist 1997).
lot the interpretation of building AK as a alid we will find out during our upcoming ' case excavations there are extremely promis-comitant factors lend special significance
Kite photograph of AK with the traces of the monumental entrance on the lower right (1996 excavations). The sand bags replace "negative walls," i.e., ancient walls which were robbed of their stones subsequent to the abandonment of the building. (The sand bags at bottom serve to protect the excavation area from heavy water runoff during the winter.)
to the stratigraphy of building AK:
(1) The storehouse was built de novo, according to a unified building plan, and it is likely that the same obtains for the remainder of the building.
(2) The earliest floor deposit of the building is very well preserved.
Plan of the AK building as excavated and projected by us before the 1996 season.
(3) Later floor deposits contained within the walls of the same building rise to a height of almost two meters: this indicates a long and continuous period of occupation.
(4) The building extends well into the mound, so that it is a reasonable expectation that we may recover it in its entirety, and preservation promises to be excellent as one proceeds to the east and north of the area exposed so far. It is true that the mound is sloping up in the same direction. This means that we should brace for many a season of excavations if we want to uncover the entire building, and if we want to proceed with the care to which we are accustomed.
From all indications, the process will be well worth it. The strata that overlay the building belong in the late third and early second millennium, and such stratigraphic continuity is of particular interest in view of recent theories about a possible environmental catastrophe at the end of the third millennium (Weiss, Courty et al. 1993) The earliest floors
of our building date to about 2200 bce: to be more precise, Carbon-14 determinations favor a slightly later date (2175 bce), while stylistic and epigraphic considerations favor a slightly earlier date; 2200 is a preliminary and arbitrary average between the two measures. Obviously, the lower the date, the greater is the amount of deposition to be compressed in the centuries marking the turn of the millennium.
In the coming season (summer 1997), we will literally cross the threshold of the palace, entering thereby into the nerve center of one of the great seats of power of ancient Syro-Mesopotamia. We expect to encounter, on the basis of ample, if later, comparative material from northern Syria, a monumental stairway and porticoed entrance leading into a ceremonial reception area, where the throne room was presumably located. To the right of the ceremonial reception area, the two service quarters functioned, as we originally assumed, as storage areas, but for goods that were destined to immediate use by the royal court. In this sense, our "storehouse" is not a long-term warehousing depot, but rather the provisioning center for the immediate needs of the court. Because of the presence of many sealings belonging to the female cook of the queen (i.e., the chief supervisor of the royal kitchen), and because of the large number of sealings that had been placed on jars, the commodities stored might have been predominantly foodstuffs destined for the royal kitchen.
Urkesh and the Hurrians
Even before our excavations and the consequent identification of the site, Urkesh was well known as the earliest documented Hurrian urban center; as the only city for which we have a known sequence of Hurrian kings during the third millennium; as a mythical city, in the proper and specific sense that it was identified in Hurrian mythology as the residence of the primordial god of the their pantheon, Kumarbi. We can add now that it is also the only Hurrian capital clearly matched with an actual archaeological site—since doubts exist regarding the possible capital(s) of that confederation of Hurrian states known as the Mitanni kingdom.
Together with the Sumerians and the Akkadians, the Hurrians gave rise to one of the major new types of the earliest urban civilization, about 5000 years ago. What characterized Hurrian culture was its dependence on the vast mountain hinterland which begins in northern Syria. Urkesh was the
Weal Archaeologist 60:2 (1997)
Biblical Archaeologist 60:2 (1997)
on the major communication routes which the resources of the Anatolian plateau Is, timber, stone). At the same time, Urkesh j privileged climatic niche, with abundant i water table allowing for an extremely pro-ing regime. The new insights that the Urkesh possible on archaic Hurrian civilization are nee for the history of the ancient Near East, on was oriented towards the northern moun-(the Outer Fertile Crescent), and as such .tinctive type of early urban civilization, dis-n the two other major types of urban cu lture Semitic, as at Ebla). Its mythology exer-jence on later Hittite religion, which preserved xts and borrowed specific deities and myths, cumented by the material found in our exca-importance of the queen and of the dynastic ining to the royal family) are also to be traced Tonography. Hurrian onomastics extended vn to Palestine in the Amarna period. So ird millennium ceramic ware (Earlv Tran-)wn as Khirbet Kerak in Syro-Palestinian te first musical score, as preserved in Ugarit, nmit to memory the tune of a Hurrian reli-
View of the AK building (looking North) with the two small "vaults" (b3 and d3 on the top plan, p. 91) on the right. On the other side of these vaults, excavators expect to find the well-preserved remains of a palatial reception area.
Conclusion
What specifically have we established so far in Urkesh, and what can we expect in the foreseeable future? We are, unfortunately, not too sanguine about finding tablet archives or important artifactual inventories within building AK; even if it is the royal palace, we may have to be resigned to just discards, like the seal impressions, however informative. The reason is that we have found no trace yet of a massive destruction at the site, and especially not within the presumed palace. For our misfortune, the city does not seem to have been sacked in antiquity. There is, in fact, no clear-cut mention of it in the records of Akkadian conquests'2—just as there is, on the other hand, no mention of it in the archives of Ebla. This is all the more remarkable in that, by contrast, Tell Brak/Nagar, which is not all that far from Mozan, had come under direct Akkadian rule and was in direct contact with Ebla. It would appear that a sharp cultural and political boundary divided the Khabur plains into a Northern and a Southern half, in spite of the total absence of geographical barriers between the two. This might suggest that, in the third millennium, the Hurrian urban sphere extended only along the piedmont
Plan of the AK building as we have excavated and projected it after the 1996 season. Sector E is understood as a monumental entrance, but this is still tentative. Sectors G-K are purely hypothetical, and serve only as a possible scenario for our next season of excavations in the" summer of 1997.
region of the Tur-Abdin, in what may be called the urban ledge of the vast rural hinterland in the mountains to the north.13 When ancient documents employ the title, "King of Urkesh and Nawar," it is to the northern highland that the term Nawar refers, in our view, rather than being an equivalent of Nagar, the ancient name of Tell Brak. The title "King of Urkesh and Nawar" would then refer to the main city and its hinterland, somewhat like the title of "King of Mari and Khana."14
Urkesh may well have been the major center of such a Hurrian urban ledge, and it would have dated back to the beginning of urban history. Such antiquity is suggested by the depositional continuity in the areas excavated so far at Mozan. The major architectural evidence we have to date from Mozan consists of an inner city wall, a temple, and the presumed palace dating respectively, and of course approximately, to about 2700,2450 and 2200 bce. Only the presumed palace can be shown to be associated with an explicit Hurrian presence. However, all stratigraphic indications point in the direction of a strong depositional continuity, and on this basis it seems possible to infer a fundamental Hurrian nature to the beginning of the settlement.
The antiquity of Urkesh is also suggested, as we have seen, by the fact that the city enjoyed a central position in Hurrian myth. When the god Silver, in the story with which we began, eventually made his way to Urkesh to seek out his father, he failed to find him because he was away,
A Hurrian myth tells of the city god Kumarbi, the father of the city of Urkesh, whose son. Silver, fails to find him at home upon a visit to his city. Kumarbi had trekked off to the highlands; his roaming the mountains may be depicted in this Urkesh seal impression.
roaming in the mountains. Whether or not one of our seal impressions does represent Kumarbi roaming in the mountains we cannot say for sure—nor will we try to follow him there. There is enough work ahead of us in his home city!
Acknowledgments
Our work at Mozan/Urkesh has been possible because of the constant support of the Syrian Directorate General of Antiquities and Museums: without interruption, the Directorate has continued to foster that enviable climate of collegial respect and organizational efficiency that makes Syria such a haven for archaeological activity. We wish in particular to thank the Director General, Dr. Sultan Muhesen, and the Director of Excavations, Dr. Adnan Bounni, under whose tenure the excavations in building AK have taken place. The local officials of the Directorate in Hassaka and Qamishli, Mr. Jean Lazar and Mr. Ali Ali, have also been invaluable with their day to day assistance. The excavations of Building AK have been supported through grants from National Endowment for the Humanities, the National Geographic Society, the Ambassador International Cultural Foundation, the Ahmanson Foundation, the S. H. Kress Foundation, the L. J. and M. L. Skaggs Foundation, the Oriental Institute of the Pacific, the Cotsen Family Foundation, Syria Shell Petroleum Development B.V., and various donors. The composite drawings of the Urkesh seals were made by Cecily J. Hilsdale, and the stippled drawings by Pietro Pozzi. This article is closely based on a lecture we gave at the Louvre on April 4,1997. '
blical Archaeologist 60:2 (1997)
Biblical Archaeologist 60:2 (1997)
93
Tablet
ind some forty cuneiform tablets, ken, all dating to the late Akkadian it interesting one is a school tablet, en in antiquity, but could be 3m three pieces found within a short (other on the floor of the building. This e tablet was discarded by its user who
a scribal apprentice. This leads us to fence of a scribal school connected with it' text preserves a section of the lexical sary of Sumerian profession names, st has been found at Abu Salabikh in nci at Ebia. A comparison of the three ilarities (the sequence and in some are identical) and differences (the
a greater incidence of determinatives
>f the lion under the feet the goddess Ishtar (Boehmer decoration of her throne (1965:384,387,389) is quite om the Urkesh lion reclining at the foot of the
presented composition, aiming to articulate clearly it is inconceivable that the position of the bull he part of the artist.
Ih sides of the face are shown in the frontally depicted nd humans (both male and female) in Early Dynas-969:pls. 114-16) and possibly the figure of Enheduanna 59:pl. 130). A shell seal from Man renders a bearded ; (Parrot 1956:pl. LXV: 329). The heavy framing of le is very different from the long, ornamented braid princess at Urkesh.
properly dynastic but does indicate the extension of
the power of Sargon. See Winter 1987.
5 This is also true in the case of the soldier grasping a bearded enemy in the Akkadian stela from Telloh (Moortgat 1965:pl. 135) and on a variety of Akkadian clyinder seals including presentation, animal combat, and scenes depicting battles between the gods. The gestures on the Telloh stele are paralleled on a recently published Akkadian seal (Bleibtreu 1996).
6 In Akkadian art, the instances of the compositional integration of the text and the seal design are few, if the vertical framing aspect of Akkadian inscription boxes is not taken into consideration. One early Akkadian seal with a motif under the inscription which is part of the theme of the main composition belongs to the scribe Kalki (Boehmer 1965:717). Abeau-tiful seal of a scribe from the reign of Sharkalisharri has the inscription framed over the backs and between the horns of two water buffalo. In this case we have a compositional integration but one which does not carry an overt cultural meaning. In some Urkesh seals the text and its visual equivalent are integrated both in terms of the composition and on a level which conveys a heightened cultural meaning.
7 Her method of churning is different from that shown in the Akkadian Etana seals (e.g., Boehmer 1965:701).
8 Specific iconographic elements, especially in the king's seals, have parallels in Akkadian iconography, e.g., the god with his foot raised or the warrior with a pointed hat (Buccellati and Kelly-Buccellati 1996b:75-77).
9 While there are at least eight seals with deities or other figures holding a child, the overall concept is quite different from the integrated themes in the Urkesh dynastic program (Boehmer 1965:483,555-60, and one unpublished seal in the collection of the Biblical Institute of the University of Freibourg, Switzerland).
10 Compare ashti~nshen-iff-we
wife brother+my+of "the wife of my brother"
shen-iff-we-ne-n tiza-nn-a shir-a-sshe
brother-my-of heart-pleasing-relative
"pleasing to the heart of my brother."
11 Frankfort 1952. The correlation of the architectural model and the term as used in the Assyrian royal inscriptions still remains conjectural, but seems quite likely, and is in any case largely adopted in the literature. The meaning of the term remains uncertain, though it appears to be of Hi trite etymology (s.v. hilänu in The Assyrian Dictionary of the Oriental Institute of the University of Chicago and W. Von Soden, Akkadisches Handwörterbuch), and it almost certainly would not have applied to the Urkesh building which is older in date and belongs to a different linguistic tradition.
12 Except for a single possible reference, proposed by Steinkeller, which is fragmentary and in an uncertain context, see Buccellati and Kelly-Buccellati 1996b:71, n. 27.
13 This was nicely shown in a satellite picture published a few years ago in the Biblical Archaeologist 58:3 (1985).
14 This follows a pattern which remains at home in the political tradition of royal titulary in Syria and Palestine; see Buccellati 1967:140-42.
Bibliography
Auerbach, E.
1997 S. v. Eshnunna. The Oxford Encyclopedia of Archaeology in the Near East. New York: Oxford University.
Kha bur \bdulAziz
Tell Brak
Plains
Si it jar
TellAjaja   # j ^
t ▼
Map of the Habur region with the "Hurrian Urban Ledge" indicated as a band stretching along the piedmont area of the Tur Abdin. While most scholars assume that the title "king of Urkesh and Nawar" refers to the two ancient cities corresponding to modern Tell Mozan and Tell Brak, we assume instead that Nawar refers to the mountainous hinterland as approximately comprised within the circle.
Bleibtreu, E.
1996  Wiener Zeitschrift für die Kunde des Morgenlandes 86:53-63. Boehmer, R. M.
1965  Die Entwicklung der Glyptik Während der Akkad-Zeit. Berlin: Walter Degruyter Buccellati, G.
1967  Cities and Nations of Ancient Syria. Studi Semitici 26. Roma.
Buccellati, G. and Kelly-Buccellati M [B & B-K]
1988 Mozan 1. The Soundings of the First Two Seasons. Bibliotheca Mesopotamia 20. Malibu: UNDENA.
1995a Mozan, Tall. Pp. 386-93 in Reallexikon der Assyriologie 8 (5/6).
1995b The Identification of Urkesh with Tell Mozan (Syria). Orient-Express 3:cover page, 67-70.
1996a The Royal Storehouse of Urkesh: The Glyptic Evidence from the Southwestern Wing. Archiv für Orientforschung 42-43:1-32.
1996b The Seals of the King of Urkesh: Evidence from the Western Wing of the Royal Storehouse AK. Wiener Zeitschrift für die Kunde des Morgenlandes 86:65-100, pi. 1-7.
1996c Evidence for a Royal Palace at Tell Mozan/Urkesh. Orient-Express 3:72-76.
1997a S. v. Mozan, Tell. The Oxford Encyclopedia of Archaeology in the Near East. New York: Oxford University.
1997b The Courtiers of the Queen of Urkesh: Glyptic Evidence from the Western Wing of the Royal Storehouse AK. Subartu 4/2:195-216.
Frankfort, H. 1952  The Origin of the Bit Hilani. Iraq 14:120-31.
Hofmer, H. A.
1990 Hittite Myths. Translated by H.A. Hoffner, edited by Gary ML Beckman. Society of Biblical Literature: Writings from the Ancient World 2. Atlanta: Scholars Press.
Kelly-Buccellati, M.
1990a Trade Metals in the Third Millennium: Northeastern Syria and Eastern Anatolia. Pp. 117-30 in Resurrecting the Past: A Joint Tribute to Adnan Bounni, edited by P. Matthiae, M. Van Loon, and H. Weiss. Nederlands: Historisch-Archaeologisch Instituut.
1990b Three Seasons of Excavation at Tell Mozan. Pp. 119-32 in Tall al-Hamidiya 2, edited by S. Eichler, M. Wafler, D. Warburton. Orbis Biblicus et Orientalis: Series Archaeologica 6. Freiburg and Gottingen: Universitatsverlag Freiburg Schwiez.
1990c A New Third Millennium Sculpture from Mozan. Pp. 149-54 and pi. 26 in Essays in Ancient Civilization Presented to Helene /. Kan-tor, edited by A. Leonard and B. Williams. Studies in Ancient Oriental Civilizations 47. Chicago: Oriental Institute.
1996 Nuzi Viewed from Urkesh, Urkesh Viewed from Nuzi: Stock Elements and Framing Devices in Northern Syro-Mesopotamia. Studies on the Civilization and Culture of Nuzi and the Hurrians 8:247-68.
Margueron, J.
1982 Recherches sur les palais mesopotamiens de l'age du bronze. Paris: Librairie Orientaliste Paul Geuthner.
1987 Les palais syriens a l'age du bronze. Pp. 127-58 in Le systeme palatial en Orient, en Grece et a Rome, edited by E. Levy. Actes du
ical Archaeologist 60:2 (1997)
Biblical Archaeologist 60:2 (1997)
95