The ‘Hasanlu Gold Bowl’

Just over 30 years ago, the extraordinary vessel known as the “Hasanlu Gold Bowl” was discovered in the debris of Burned Building I-West, part of the major architectural complex belonging to period IVB on the High Mound. Found not far from the skeletal hand of the individual who had been fleeing with the piece when the building collapsed toward the end of the 9th century B.C., the bowl was raised aloft by its ex­cavator, Robert H. Dyson, Jr., almost 3000 years later (Fig. 1).

It was immediately evident that the new find represented a unique and important example of the ancient goldsmith’s craft, display­ing a high degree of technical mastery and a wide range of decora­tive motifs that, it was hoped, could provide a key to the religious and mythological traditions of the site. Within a year of its discovery, the bowl had been published in Expedition by Edith Porada (1959)­a penetrating study of style and imagery that remains a primary source of information on the piece. Over the intervening years, the bowl has been discussed in count­less scholarly books and articles. Most recently, the accumulated literature on the bowl has provided the opportunity for an important analysis by Marie-Therese Barrelet of the very nature of empirical data, scholarly inference, and ex­planation in our field (1984).

Nevertheless, as scholars of the ancient Near East continue to de­bate the bowl’s proper chronologi­cal place, the decipherment of its motifs, and the ethno-linguistic tradition behind its manufacture, the “meaning” of the bowl con­tinues to elude us. Thus, for this special issue of Expedition devoted to Hasanlu it seems timely to re­view where we stand, and perhaps also to point the way toward some fruitful avenues of future research, given shifts in analytic perspective and the accumulation of new data over the past 30 years.

Background

The history of discovery of the gold bowl is generally provided by Porada and has been repeated with a variety of details in subsequent publications, but some points are worthy of mention here. The bowl was discovered in association with a party of three men (Fig. 2). All were found sandwiched between a thick layer of burned material be­low and heavy bricky collapse above, in the southeast room (9) of Burned Building I-West (see Dyson, “Architecture,” Fig. 10). The leader of the party carried an iron sword and a gold-handled dagger. The second individual, who had been carrying the bowl in his right hand, wore a gauntlet marked by several rows of bronze buttons. The third man bore a star-shaped mace, dagger, and sword. Neither the buttons nor the wea­pons have permitted scholars thus far to determine if these men were among the invaders or the local defenders at the time of the destruction of Hasanlu in period IVB.

The bowl was found lying against the north wall, just beyond the fingerbones of its bearer. A low column of stone in the center of the adjacent southwest room is now understood to have been the center support of a stairway leading to the upper floor, and perhaps to the roof as well. Since the three men were facing away from the stairs as they fell, one cannot argue that they were racing toward the stairs to escape; they may, however, have been heading toward a window in the exterior wall that gave onto a back alley, as the building was being gutted by fire. While the absence of other objects in the debris might suggest that the men were on the roof rather than in an interior room, two stratigraphic considerations suggest that they were indeed on the second floor. First, they were thoroughly buried by collapsed brickwork; second, the volume of collapse, when added to the height of the standing walls found in excavation, implies the presence of a second story (R. Dyson, pers. corn.).

Because the bowl was found flattened, it is difficult to recon­struct its original shape exactly. On Hasanlu, where a metal workshop dating to Hasanlu IV has been found (de Schauensee 1988). Until such investigations have been car­ried -out, it seems reasonable to proceed within the hypothetical—but tenable—framework that the bowl is indeed a product of north­west Iran and could well have been made at Hasanlu itself. Date Archaeological and historical evi­dence suggests that Hasanlu IVB was destroyed toward the end of the 9th century B.C. (Dyson and Muscarella 1989). Clearly, the bowl had to have been made before this destruction. Just how much earlier depends on how one weighs typo­logical and stylistic comparisons with other, related material. Es­sentially, either the bowl was made sometime during the period IVB occupation (ca. 1000-800 B.C.), or it was older, a valued heirloom that continued to be used into the 9th century.

All of the arguments concerning the chronology of the bowl have been summarized in Barrelet (1984). Most scholars have fol­lowed Porada in placing its manufacture in the late 2nd millennium B.C. (1250-1000), contemporary with period IVC at Hasanlu (see Dyson, “Rediscovering Hasanlu,” Fig. 5; Dyson and Muscarella 1989). This date has been suggested on two grounds: 1) stylistic parallels with finds in Mesopotamia historically dated to the late 2nd millennium; and 2) the absence of themes and narratives denoting contact with the Assyrians, apparent in architecture and in some artifacts associated with period IVB (e.g., the silver beaker, Fig. 7). Never­theless, Muscarella (1971, 1988b) and Barrelet prefer a date in the 10th-9th century, contemporary with the bowl’s archaeological con­text, based upon iconographic links with other 1st millennium works. Specifically, Muscarella has argued that the closeness of composition and motifs between the Hasanlu bowl and 9th-century carved stone reliefs from Carchemish—for exam­ple, gods standing on the back of a couchant lion (Fig. 9), and a wo­man with mirror seated on a lion (Fig. 10)—must be seen as evidence of contemporaneity.

This is a problem that needs further and systematic study; its ultimate determination depends not only on the validity of the particular parallels drawn, but also on very basic underlying principles of how scholars weigh evidence in argument. At issue is how one independently considers stylistic details, iconography, and composi­tion, and how one decides to give greater weight to one or the other if the evidence does not concur on all counts.

I think it is important to empha­size the strength of some of the arguments for the later date, no matter where one stands in the end. First and foremost is the archae­ological context of the bowl’s dis­covery. Stylistically the bowl is very much at home with other works in metal and ivory found within the Hasanlu period IVB occupation, a fact that must be explained away to accommodate a date of manufacture in the 2nd millennium. If the bowl was made in the late 2nd millennium, this would fall within Hasanlu period IVC, which begins during the 12th century B.C., or during Hasanlu period V (ca. 1500-1200 B.C.), a phase that has major architectural and ceramic continuity with period IV (see Dyson and Muscatel 1989, Dyson 1978, Young 1967). Unfor­tunately, we have virtually no decor­ated works from either period IVC or from period V with which to compare the bowl. We need to know a great deal more about the artifactual and cultural continuities between the phases in order to determine whether the “climate” of periods IVC or V could have pro­duced or sustained the bowl.

Second, the chronological evi­dence provided by three swords or daggers depicted on the bowl is ambiguous. Porada rightly points out that these are types known from 2nd millennium tombs; how­ever, two of these types may also occur in later contexts. A single example of the type with a hori­zontal guard just below the hilt has been found on the Citadel at Ha­sanlu in the same building as the bowl (Fig. 11; see also Muscarella 1988b). The second dagger type depicted, with a crescent-shaped attachment between the blade and the hilt, poses a problem because the Late Bronze Age tombs in which this type is found have generally been re-opened and con­taminated with later, Iron Age, materials (see Schaeffer 1948:407; Dyson 1964). Thus, until close typo­logical analysis of the weapons depicted on the bowl is under­taken, followed by a comparison with the corpus of related pieces excavated and published in recent years, they should not be used as evidence for an earlier origin of the bowl.

The third, more complex argu­ment for a later date concerns conservative tendencies in artistic decoration under certain cultural and historical circumstances. One of the most compelling arguments for the 2nd millennium dating of the gold bowl is that it evinces none of the Assyrian elements ap­parent on the silver beaker and on other artifacts from the site. Since historical contact with Assyria is documented in the early 1st mil­lennium, the argument goes, the bowl must precede that contact—just as the silver beaker must post­date it. However, in a study of the effects of Assyrian “influence” upon Hasanlu, I have argued that Hasanlu was not equally receptive to the Assyrian imprint in all cul­tural domains. It is in the public/ political arena where we see Ha­sanlu responding to Neo-Assyrian stimuli—the external facades of buildings, sumptuary goods such as ivory, motifs of political power—while the interior organization of architectural space and the reli­gious system largely retain their local configurations (Winter 1977). If that is so, then the necessary precedence of the bowl in relation to the beaker vanishes, because its subject-matter would fall into the more conservative realm of the religious and so need not reflect Assyrian contact. Such an interpre­tation clearly leaves open the possi­bility of a 9th-century date for the gold bowl, contemporary with the beaker but representing a different, more conservative aspect of Hasanlu society.

In future studies, then, it would be highly desirable to have more systematic articulation of stylistic criteria in the establishment of relative dating sequences. Further analysis of archaeological compar­anda and of the complex cultural context for the Iron Age levels at Hasanlu, already in progress, would need to be considered, and, of course, we must have more excavation!

What?: Iconography and Narrative Reading of Motifs

We may turn now to the second area of needed study, the bowl’s decorative scheme. This section can be subdivided into three parts: (1) the identification of individual motifs; (2) the compositional relationships between motifs; and (3) the possible narrative “reading(s)” of the bowl as a whole.

Motifs

Porada laid the foundations for the interpretation of individual motifs, based upon visual and textual parallels to material from across the ancient Near East and stressing the “Hurrian” connections (see box on Hurrians). Subsequent studies have added associations and/or shifted emphases, and the whole range of attributions has recently been summarized for each motif (Barrelet 1984:43-46).

An array of three deities in chariots appears in the upper field, moving from left to right, meeting a delegation of three walking or standing men facing left (Fig. 6). The lead deity has been identified as a storm god by virtue of his attribute animals, the bulls pulling the chariot, from whose mouths issue streams of water or rain (Fig. 12). The chariots following are pulled by equids (mules?). The second deity may be identified as the sun god by the disk and rays or wings issuing from the top of his head, represented similarly at Car­chemish (Fig. 9) and elsewhere in the early 1st millennium B.C. The third deity may perhaps be iden­tified as the moon god on the basis of the association between the pair of bulls’ horns set on his head and the curve of a lunar crescent, as well as the frequent pairing of sun and moon gods (Fig. 9) and the equally frequent references to the importance of storm, sun, and moon in the Hurrian pantheon (Lambert 1978).

The leader of the three men confronting the divine array ex­tends a goblet or beaker before the lead chariot, while each of the two individuals following him brings a ram, presumably for sacrifice. This libation and sacri­fice is well known in the texts and representations of the Hittite and Neo-Hittite periods to the west, and of Elam to the south (see Dyson, “Rediscovering Hasanlu,” Fig. 13). The libation especially, as performed by priests and/or rulers, was part of standard ritual practice (Haas and Wilhelm 1974, Porada 1970). The sacrifice of sheep to the high gods, especially the weather god, is attested from Hurrian and Hittite contexts in the 2nd millen­nium as well as from Urartu in the 1st (Haas and Wilhelm 1974, Ziman­sky 1985). The three offerants thus add a human, liturgical component to the otherwise divine and myth­ological elements of the bowl’s decoration.

Motifs in the lower field have proved more elusive, but also more intriguing. Porada early identified what appears to be the principal scene of a heroic figure wearing kilt and some sort of boxing gloves(?), engaged in stylized com­bat with a half-human figure emerging from a rock enclave out of which issue three fierce serpents (Fig. 13). She associated it with the Hurrian myth in which the youthful storm god Tešub does battle with Ullikummi, a stone monster created by the aged god Kumarbi in an attempt to unseat Tešub and resume his former place as chief god. How this would relate to the cou­chant lion upon whose back the whole event seems to be taking place is uncertain, although there is the precedent at Carchemish for divine figures to be set upon such a creature (Fig. 9).

Other elements of the same myth, preserved on tablets from Boazköy, have been seen by some scholars in the motifs to the im­mediate left and right of the battle. (Bogazkoy was the capital’ of the Hittite Empire in central Anatolia at a time when close relations between the Indo-European Hit­tites and the neighboring Hurrian population can be attested; see box on Hurrians.) At left, a squatting female handing an odd-looking baby(?) to a seated male could well represent the preamble of the same story, when the infant Ullikummi was born and laid on the knees of his father, Kumarbi. To the right, the nude goddess on two rams may be identified as Istar/Šaušga, sister of the storm god in the Hittite/ Hurrian version, who tries unsuc­cessfully to seduce Ullikummi be­fore Tešub engages him in battle.

It is important to note, however, that within the Hurrian repertoire, there may well be other myth cycles whose events fit better the imagery on the bowl, but which are less well known. For example, we now know of an episode similar to that about Ullikummi, in which the god Kumarbi engenders a second mon­ster (called Hedammu) to combat Tešub (Siegelova 1971). The crea­ture’s name is written with the determinative sign for snake or serpent (significant for the three serpents attached to the mountain stronghold here?), and in this sequence the goddess Istar/Šaušga is successful in seducing him, there­by breaking his power. This very fragmentary text is enough to indi­cate that even better literary paral­lels for the imagery than the Tešub/ Ullikummi story may eventually turn up within the hurrian canon. The other motifs on the bowl are less easy to identify, much less to associate with a known narrative, but some parallels can be cited for individual elements. The figure of an unmanifest deity, or if it relates to either the storm god above or the narrative combat nearby, then there are numerous lines of inquiry to pursue: (1) Babylonian/Kassite representations of altars as divine symbols on carved boundary stones (kudurrus) of the late 2nd and early 1st millen­nium; (2) Hittite and Neo-Hittite reliefs of offering tables before divine or deceased figures; (3) the full range of Elamite libations and offerings depicted on seals; and (4) textual references to offerings to deities, manifest and unmanifest, across all of the above traditions.

What is interesting in this regard is that the stool/altar/table motif is placed to the right of the three swords, the goddess-on-lion, and the woman carried by eagle, and it may relate to them in meaning. It may as well be related to the scene of squatting woman, infant, and seated man (Kumarbi?) below. This leads us to the second part of the present section.

Composition and Relations Between Motifs/h2>

Upper and Lower Fields: Primary Focus

The upper field of the bowl’s decoration is clearly demarcated by the implied groundline of the gods and opposing males. It is as if the three deities receive the offering or worship represented by the beaker and sacrifical animals brought by the three men. At the same time, some visual unity between the upper and lower field is achieved through the device of the punctate circles that fill the space between the storm god and the libation “priest” and then fall about the serpents of the mountain enclave. These circles have been identified as water, appropriate to both the Ullikummi story in which the stone baby grew in the midst of the sea, and also the related Hedammu myth where the creature similarly resided in the sea. It is also apt in that, in the Ullikummi episode at least, Tešub’s prayer to the god of sweet waters (Ea) results in the cutting off of the stone monster from his base, permitting the hero’s ultimate victory (see Porada 1965).

This compositional relationship between the upper and lower fields raises another issue, the nature of the drawing by which the imagery of the bowl is seen and studied: although laid out as a flat, two-dimensional schema, the bowl is after all a three-dimensional cylin­drical object. If we may assume that the culmination of the scene in the upper field is the point of meeting of the two processions, divine and human, as emphasized by the special patterning of water as well, then the scene directly below it (i.e., the combat) would have been the primary visual focus of the lower field, facing the viewer as the “front” of the bowl (Fig. 16).

The Lefthand Zones

When the drawing is laid out to conform to the readable narrative in the upper field (Fig. 6), it is easier to see how the rest of the motifs in the lower field fall into place. While figures in the upper field are de­ployed in a band of consistent height, those of the lower field are subdivided into two zones to the left of the combat motif, with smaller scale figures. Thus, the eagle-vehicle, goddess and lion, swords, and man with stool/altar/ table are set above the Humbaba complex, and the woman, child, and seated man. All of these figures are shown seated or crouching, which helps diminish their scale. However, their position just below the celestial deities of the upper field should not be lost sight of.

This positioning calls to mind the standard organization of divine symbols on the more-or-less con­temporary Kassite kudurrus re­ferred to above, which sometimes have horizontal rows of symbols or deities clearly divided by register lines, but frequently also have them merely distributed in hierarchical zones. This association is parti­cularly enticing, as the established hierarchy on the kudurrus calls for the celestial gods (sun, moon) above, with the next highest rank­ing gods just below (e.g., Fig. 17).

And if we have identified the middle zone on the gold bowl as referring to the goddess (Kubaba) and god (Nergal), with mythologi­cal scenes (Cilgamesh and Ku­marbi) below, we have here also a hierarchical arrangement that unites the upper, middle, and lower zones.

In fact, this division into hierarchi­cal zones is not only typical of Kassite kudurrus, it is also apparent in the distribution of motifs on a well-attested cylinder seal belong­ing to Sautatar, ruler of the Hurrian kingdom of Mittani in the mid-2nd millennium (Fig. 18). This last is especially interesting, since it has been shown that the seal was an heirloom used secondarily by dy­nastic successors, and impressions of it have turned up on tablets found in the eastern Hurrian sphere at Nuzi, 50 to 75 years after the death of its original owner (D. Stein, unpublished ms.). Given the sharing of so many texts and motifs across the whole Near East in our period, one might hope to narrow the field of attributions by looking not only at individual motifs but also at compositional devices, and by posing the question of whether certain preferences in composition might not have a more restricted distribution, thereby allowing better identification with particular regional and/or ethno-linguistic traditions.

We have suggested above that the empty altar/table to the right of the swords might represent a deity. If it were a reference to the high god Ea, this would be appropriate on several counts. First, he would join the ranking of the gods on the level just below the celestials, as is con­sistently the case on the kudurrus and also seems to have been the case in the Hurrian pantheon (Lam­bert 1978). Second, since Ea plays a role in the Tešub-Kumarbi myth cycle, the motif would have a narrative link to the scene im­mediately below, as well as to the larger scale major scene of combat to its right. In this way, both motifs to the left of the principal combat scene in the lower field would relate to the narrative at the same time as they participated in the hierarchy of the lefthand zones.

The Righthand Zones

Our drawing of the bowl places the nude goddess and archer to the right of the combat scene. We cannot be certain that this is how the scene was intended to be read, and again we must stress that the bowl was round, so the archer/hero figure (Fig. 19) would also appear to be facing the eagle and its human burden in a continuous circle. Nevertheless, it is important to note that the archer here wears a similar headband and the same kilt as the hero of the combat (Fig. 6). This brings up the important question of a dress “code” as employed on the bowl to emphasize aspects of meaning and significance.

Porada has suggested that the hero fighting the stone monster in the lower zone is the same figure as the lead god in the upper zone, since we know Tešub to have been the storm god in the Hurrian pan­theon, as well as the hero of the myth cycle; he has merely changed his clothes, the kilt being more appropriate for battle (1965:103). One problem with this identifica­tion is that the storm god of the upper zone seems to be winged, or at least to be portrayed with rays issuing from his shoulders, but the kilted hero is not. If, however, this detail were to be argued as not inconsistent with the identity of the two figures, then the squatting figure to the left of the combat

scene who extends his beaker could well represent the very same charac­ter with yet another garment (his hair, beard, and headband are identical) at another stage of the narrative (Tešub imploring Ea for help, as the text records?). We would then be faced with the familiar device of a “consecutive narrative,” in which several epi­sodes in a single narrative are represented sequentially, often with principal characters repeated.

There is another possibility, how­ever, and that is that the differences in dress are meant to signify separ­ate sets of figures from differing realms. In such a way, all three of the celestial gods in chariots wear the same long wrap-around fringed garment with short sleeves; the three men approaching them all wear a slightly different garment, long but apparently not wrapped, with a shawl covering the right shoulder and no evidence of sleeves. The two kneeling figures battling a central third are dressed in short-sleeved tunics (mythical heros?). The goddess on a lion, male with beaker, squatting woman, and seated male all seem similarly dressed in long, vertically paneled and patterned garments, with an identical pattern used for the cloak/ unwrapped dress of the nude fe­male standing on two rams (all deities, but distinct from the celes­tial gods above?). And finally, the two full-size figures of the lower field, combat-hero and archer, are virtually identical in their hair, head­bands, and fringed kilts, the only exception being the projection of a snake’s head at the end of the archer’s headband. Are they separate-but-related youthful hero/ gods? Or could they represent the same individual? Within the context of the Tešub myth cycle already identified, this last would not be impossible. For, if the framing scenes around the central combat all relate to the same mythical narra­tive, then the hero’s repeated repre­sentation in triumph at the far right would fit well. Such an identifica­tion is particularly compelling since recent work on differences between the eastern and western branches of the Hurrian population in the 2nd millennium show that in the east, Istar/Šaušga was higher in status than in the west; she often stood next to Tešub in visual representa­tions (for example, on cylinder seals from Nuzi); and she was considered as his consort, not just his sister (Wegner 1981; Stein 1988). Since any Hurrian elements likely to be present at Hasanlu would be equally part of the eastern tradition, the juxtaposition here of the nude goddess (an iconography demonstrable for Istar/Šaušga) with the tall hero!god, victorious after his combat with the stone-monster, would make excellent composi­tional sense in the narrative flow of the lower field.

Again, such suggestions only point the way toward future study. One would have to reconcile the alternatives of the three possible figures of Tešub in different dress in the first system with the two figures of Tešub in identical dress in the second. Especially at issue would be the possibility of combining the two systems simultaneously, and/or the identity of the storm-god in the upper field were he not to represent Tešub (see below) .

The “Reading”

This leads us to a summary of the visual imagery on the bowl as it can be reconstructed thus far. Once again, one must picture the bowl as a slightly flaring three-dimensional object of impressive size (Fig. 16). What one would then see is several sets of independent, overlapping, and interrelated themes across multiple zones, employing both the height and the circumference of the bowl. The upper field is filled with two files of figures, converging to a point of dramatic interaction. If we consider this meeting to be the focal point of the upper field, then we may also conclude that what falls immediately below this point would be the focal point of the lower field. Indeed, what is represented seems to be identifiable as the culmination scene in a mythological cycle as­sociated with the Hurrian god Tešub.

To right and left of this focal point in the lower field seem to be dependent motifs belonging to the same narrative. At the same time, if the extra division into zones to the left of the focal scene can be read as related to the hierarchical order of the Hurrian world of gods and heros, then here too there is visual organization from top to bottom, as well as along the horizontal plane. Similarly, the placement of the nude goddess and hero god to the right of the focal scene in the lower field corresponds well to the disposition of the two men bringing sacrifical rams in the upper field (see Fig. 6).

In sum, the organization of the various motifs into a readable com­position, into overlapping sets of images that add up to a grand scheme of meaning and reference, clearly seems to have been carefully worked out by the designer of the gold bowl. Aspects of the divine, the liturgical, and the mythological have been combined into a tense and dynamic whole. There is much more to be done in the identification of individual motifs and in the analysis of the visual organization, based especially upon new texts translated within the past 30 years, as well as upon recent studies of representations on cylinder seals. And one would want also to pursue the design on the base of the bowl, with its four goats, one on each side of a square checkerboard panel (Fig. 20); this may seem merely decorative but could well yield some information when compared to other known vessels, particularly from Marlik Tepe (Negahban 1983). This last notwithstanding, the very fact that one may now see the various motifs on the bowl as compositionally and meaningfully interrelated and readable is no small progress in the face of the bowl’s enigmatic imagery.

Who?: Ethnic, Linguistic, and Cultural Associations

We now turn to the third and most difficult issue on which re­search is needed: the ethnic, lin­guistic, and cultural traditions of the makers and the users of the gold bowl. The very attempt to identify an underlying linguistic or cultural tradition for an art object raises major questions of interpretation. Specifically, with the population mixtures attested throughout the ancient Near East in the 2nd mil­lennium (see box on Hurrians), we must assume that the resultant inter­actions would produce a high degree of shared literary traditions, design repertoires, and even com­positional devices. In such cases, a particular iconographic element could no longer be assumed to belong exclusively to any one ethnic or linguistic population.

What is more, although it is often convenient to assume that a given ethnic group and its associated language would be mapped with identical boundaries, or that a single language group would manifest a single and distinctive culture, there are many occasions where this may be demonstrated not to be the case: instances in which a group that exhibits behavioral patterns and belief systems that would be recog­nized as a cultural unit may contain within it several different language groups; or conversely, instances in which a single linguistic group may be divided into several distinct ecological and cultural adaptations.

Does this mean we should not attempt to pursue the “Who?” of the Hasanlu gold bowl? No. But by emphasizing the complexity of the issue, we are stimulated both to articulate and to further refine the logical and methodological implica­tions of asking such questions, and so, hopefully, to come closer to the complexity of the specific historical situation before us.

What has been striking in our discussion thus far is the consistency of “Hurrian” literary and visual associations with the decorative scheme of the Hasanlu bowl. Most scholars have been careful to point out these elements without actually postulating the necessity of Hurrian manufacture—although they may speculate about a possible Hurrian background underlying its manu­facture (Porada 1959, Mellink 1966, Barrelet 1984).

In the present case, if we take as a point of departure the Hurrian literary and stylistic parallels that have been demonstrated for the bowl, it is not unreasonable to pursue the possibilities for a “Hurrian” context of manufacture and usage, as long as we are judicious in distinguishing evidence from as­sumption. And since it is generally agreed that the gold bowl is at home within the “local style” of Hasanlu IVB which cuts across several media and was associated with functioning artifacts, the “Who?” of the bowl may be said to be closely tied to the “Who?” of the site itself.

2nd Millennium Hurrians?

There is no written evidence from Hasanlu that would attest to the identity of the inhabitants. We know neither the ancient name of the site, nor the state to which it may have belonged, much less the lin­guistic and/or ethnic affiliation(s) of its population—neither preced­ing nor during the period with which we are concerned.

In the absence of textual sources from the site, we are thrown back upon an evaluation of the archae­ological record. On occasion, pot­tery types have been used as bench marks for identifying the presence of a specific ethno-linguistic group. When a type is limited in distribu­tion and clearly linked, it can be a most useful marker: for example, a particular red water jar is used in present-day Iran only by Turkish-speaking households in various areas of the country. The common buff- or orange-colored water jar used by the rest of the population in Iran today is, however, widespread across various subgroups (Persians, Kurds, Lurs), and would not serve as a marker of any specific group. One must therefore be able to isolate a particular ceramic type as significant using ethnographic or historical information.

For studies of the past, where the validity of a marker cannot be con­firmed by observation on the ground, a single variable such as pottery will probably not be suf­ficient to identify distinct ethnic groups. Rather, multiple variables will need to be isolated and cor­related, including distinctive sub­sistence patterns, associated animal bones indicating variant food prac­tices or prohibitions, and special­ized distributions of small finds. Such variables might then be inter­preted as indicating ethnic, linguis­tic, or cultural differences, rather than merely differences in social class or wealth.

In the case of the Hurrians, we may trace them back to an arrival in the Near East via the Caucasus sometime in the late 4th to early 3rd millennium B.C. The newcomers seem to have brought with them a highly burnished black or dark gray colored pottery, identified with the Kura-Araxes culture in the Cau­casus. This pottery appears sud­denly in Anatolia and Iran around 3000 B.C., associated with signi­ficant changes in architectural con­struction and metal technology, changes that suggest large-scale population movements rather than internal change or trade. Since Hurrian is known to be a member of a Caucasian language group, and since the first Hurrian texts and place names are later attested in areas where the Kura-Araxes cul­ture appears, the identification of this archaeological culture as Hurrian is at least strongly suggested, if not absolutely confirmed. Thus, scholars have seen the arrival of Kura-Araxes-type pottery in Anatolia and Iran during the early 3rd millennium, and in the Levant during the late 3rd millennium, as evidence of the Hurrian migrations. (Kura-Araxes pottery occurs in Iran to the north of Lake Urmia and along its eastern shore, as well as in central Iran; however, there is no evidence in Solduz for a related occupation.)

By the 2nd millennium, the presence of Hurrians is well documented in northern and eastern Mesopotamia. “Habur-ware” pottery (with dark painted horizontal bands on a buff ground), characteristic of both northern Mesopotamia and the Solduz valley in this period (Hasanlu period VI), was originally thought to be indicative of the distribution of Hurrian-speaking peoples of that period. It has now been effectively shown, however, that the distribution pattern for the pottery defines a socio-economic “interaction-sphere” of more than a single ethno-linguistic group (Kramer 1977). This does no preclude the possibility of a Hurrian population at Hassanlu in period VI, but does indicate that the presence of a single type of decorated pottery is not sufficient to confirm it. Thus for the time being we can neither prove nor disprove the presence of Hurrians during the early 2nd millennium at Hasnalu on the basis of pottery.

Indo-European Immigrants?

In the archaeological record characterizing Hasanlu periods V and IV-pottery assemblage, architecture, burial practices, clothing style-the changes from the previous period VI are so marked as to suggest a significant disruption, and event a shift of the dominant local cultural tradition/population. This shift, and particularly the new ceramic industry (differing from that of period VI in both technology and style, and characterized by burnished monochrome vessels in black, gray, red, and tan), was initially associated with the coming of the Indo-European Iranians onto the plateau (see box on Indo-Europeans; Young 1967).

Recent archaeological research in the central Zagros has led to a rejection of the hypothesis (Young 1985). In the region of west central Iran where Assyrian references do include Indo-European geographic names, and where Iranians, therefore, can with confidence be said to have been settled, the characteristic pottery is not burnished monochrome (so-called gray ware), but instead buff-colored and painted (for examples, that at Babajan in the 8th-7th century B.C.). In the most recent linguistic study on the subject, the presence of Iranian speakers in west central Iran is not considered certain before the 9th to 8th century. Thus, neither the archaeological nor the available linguistic evidence can be used to argue for the presence of Iranians in the late 2nd and early 1st millennium at Hasanlu, that is during periods V and IV (Dandamaev and Lukonin 1989).

Nevertheless, it is still possible that the cultural disruption represented by periods V and IV at Hasanlu is in some way related to the arrival of Indo-European elements from the east (Young 1985). It is also certainly the case that changes in material culture indicating a shift in the dominant population need not represent a total population replacement in the valley (Dyson 1976). Nor would such changes necessarily entail total linguistic shifts, much less the total transformation of the former inhabi­tants’ culture (Diepold, In press). Hence there could have been within the occupations of Hasanlu V-IV a significant mixture of people for­merly in residence, now dominated by or subsumed under those newly arrived.

It is, however, very possible that the archaeological record of Ha­sanlu V and IV does not mark the arrival of Indo-Europeans at all. Against the presence of this linguis­tic group at Hasanlu is the absence of Indo-European personal or place names in any Assyrian references to the area. In the absence of such references, there now remain no compelling reasons for an associa­tion of the burnished monochrome pottery of the Iron Age in Solduz with Indo-Europeans.

1st Millennium Hurrians?

Since there is neither local lin­guistic evidence to provide a clear indication of the identity of the inhabitants of Hasanlu V and IV, and the archaeological (that is, ceramic) evidence is ambiguous, scholars have attempted to ap­proach the problem from the per­spective of historical texts available from neighboring cultures. Working from Neo-Assyrian and later Urar­tian itineraries and geographical descriptions, Hasanlu has variously been included in the ancient lands of Mannaea, Gilzanu, and Armarili, or identified as the site of Mesta (Boehmer 1964, Meade 1979, Levine 1973-74, Salvini 1984).

Based upon the premise that names given to geographical pheno­mena and sites in an area generally reflect the language group of the local inhabitants, Boehmer has argued that the predominantly Hurrian place names associated with Mannaea in 9th-eentury Assyrian texts indicate a Hurrian population in that country. Now, Mannaea is presently understood to have been located to the east or southeast of Hasanlu; the uncertainty is whether or not Hasanlu to the northwest. For the present all we can say is that, with evidence for a Hurrian population to the south and east of Hasanlu at the end of the 9th century, as well as evidence at the Mesopotamian site of Nuzi for a Hurrian population to the west of Hasanlu in the later 2nd millennium, the possibility of some Hurrian presence in the inter­mediate area is relatively high (see Muscarella 1987).

Hurrians, Iranians, and the Bowl

Let us now come back to what can be said of the “Who?” of the gold bowl, based upon its decora­tive scheme. As noted above, the most striking parallels that can be cited for the Hasanlu gold bowl to date come to us from Hurrian sources: the episodes in the Kumarbi cycle of the conflicts between Tešub and Ullikummi and between Tešub and Hedammu as literary references; the relationship be­tween Istar/Šaušga and Tešub in the east Hurrian pantheon; and the disposition of motifs in upper and lower zones comparable to the seal of Šauštatar (Fig. 18). But it is important to keep in mind that even these parallels may simply be a function of the accidents of archae­ological recovery. On the one hand, there may well have been other myth cycles within the Hurrian repertoire that are presently un­known to us, but would fit better the imagery of the bowl; on the other hand, there may be other, less well documented traditions where the fit would be better if only we had the data. And there is an additional possibility that must be considered: that what had indeed originally been part of the Hurrians tradition had already been ab­sorbed by and accommodated into a later, Hurrians tradition. Thus, before we assign the bowl to the Hurrian sphere, it is important to pursue alternatives to the Hurrian parallels. For example, given the questions surrounding the identity of the peoples associated with the occupation of Hasanlu V-IV, it seems necessary to inquire into the larger world of possible Indo-European sources for the same imagery.

The Indo-European pantheon was also headed by a storm god, and Indo-European speakers among the Mittani and in the Indian subcontinent have been strongly associated with chariotry. What is more, the Indo-European Hittites of the Anatolian plateau in the 2nd millennium had already absorbed stories and images gathered from a number of surrounding traditions (hence the copies of Hurrian myths and ritual texts preserved at Boazkäy, and the Hurrian gods depicted at Yazilikaya), and the Indo-European Luwian peoples in southeastern Anatolia/northern Syria were mixed with a base population that was probably heavily Hurrian. It is precisely here, on reliefs from the sites of Malatya and Carchemish, that several paral­lels in imagery with scenes on the gold bowl may be noted (Mellink 1966). Scholars looking at the art of these two sites are therefore con­fronted with the same sort of dilemma as at Hasanlu: determining whether what we are seeing in the art of the late 2nd and early 1st millennium is to be attributed to a tradition brought in by the new­comers, to a tradition belonging to the earlier base population of the region, or, finally, to a tradition assimilated by one group through contact with another. The problem is exacerbated by the lack of characteristic features in art that can demonstrably be called “Hurrian” as opposed to generalized Near Eastern, since the Hurrians also seem to have absorbed and adopted the motifs and styles of peoples with whom they came in contact (see Barrelet 1984, Mus­carella 1988a).

Questions we must therefore ask ourselves are: first, whether we could be missing the relevant Indo-European mythological texts and visual materials, be they Hittite, Luwian, or Iranian, that would provide parallels to the bowl as viable as those in the Hurrian sources already noted; and second, whether what we are seeing are perhaps original Hurrian themes now in an Indo-European reper­toire. In the first case, attempts to substitute Avestan (later Iranian) sources for Hurnian in explaining the iconography of the bowl have not been successful (Duchesne­Guillemin in Barrelet 1984); also, whenever “Iranian” elements have been referred to, they are in the nature of minor details of repre­sentation rather than larger issues of narrative or motif. It is the second case that appears the most prom­ising, for in the classical Creek tradition, much of the Kumarbi cycle had been incorporated by the poet Hesiod into his account of Chronos and Zeus—a good “Indo-European” Greek myth.

With this case of cultural bor­rowing in mind, there are several tantalizing threads that need to be pursued. One such thread would be comparison with the details of an Indo-European myth of a hero who loses his herd to a three-headed serpent/monster, and who must then defeat the serpent in battle with the aid of a warrior god (Mallory 1989:131). A second thread would be to investigate fully the three-headed serpent motif, clearly quite popular in some later Judo-European artistic traditions (for example, on a silver-gilt mirror from Thrace of the mid-4th century B.C.; Mallory 1989:Fig. 12). And a third thread would be to seek parallels in cultural and structural phenomena, such as the promi­nence given to priests and warriors as classes in the Indo-European social order, as reflected also on the gold bowl.

Until such studies are under­taken, I feel we cannot resolve the question of whose literary and cultural tradition seems to underlie the imagery of the bowl. For the present, the Hurrian associations seem more viable than any Iranian or generalized Indo-European possi­bilities, so we should not underrate them until a well-formulated and better-grounded explanation pre­sents itself.

It is indeed unfortunate that the Hurrian artistic tradition itself seems so ill-defined. It is also un­fortunate that the only Hurrian literary parallels available to us for the bowl are from the 2nd mil­lennium, while contemporary 1st millennium visual parallels seem to come from ethnically and linguistic­ally mixed populations—a situation that in fact could have prevailed at Hasanlu as well. One major avenue yet to be pursued stems from the demonstrated linguistic relationship between Hurrian and Urartian (see box on Hurrians). As Mellink has noted, if the languages are so closely related, then one should expect to find relationships in myth­ology and in artistic expression as well (1972-75). Therefore, before we propose or deny a Hurrian base to the imagery of the gold bowl, it is essential to look at the Urartian sources and the Urartian repertoire, even if the material preserved to us is largely 8th century, slightly later than the bowl.

An Urartian Connection?

Several lines of evidence suggest the importance of Urartian sources for our understanding of the bowl. First, both chariotry and sheep sacri­fice, central motifs on the gold bowl, are significantly attested in Urartian sources (Zimansky 1985). Second, arrays of three rearing serpent heads on long necks are known from Urartian art, most particularly on Urartian royal hel­mets of the 8th century (Azarpay 1968: Pls. 10,16), and need to be carefully compared with the ser­pents that issue from the citadel in the combat motif of the lower zone on the bowl. Third, Urartian deities are just as frequently represented standing or seated upon their attri­bute animals as their Hittite and North Syrian counterparts, and these representations should be fully analyzed as parallels to the gold bowl’s nude goddess on two rams and the seated female on her lion. This last, in particular, finds no less excellent parallels in Urartian bronzes than on the reliefs of Carchemish, where seated female dei­ties hold something in an out­stretched hand, and are shown seated on the back of a lion (Tasyurek 1977:Fig. 4).

Fourth and most important, in all of the 3rd and 2nd millennium Hurnian sources we possess, the weather god Tešub is the top-ranking deity of the pantheon. What distinguishes the Urartian from earlier Hurnian pantheons is the introduction of the god Haldi as supreme deity (Diakonov 1981). Although he has clear aspects of a weather god, Haled is quite distinct from Tešub. Tešub himself is still included in the Urartian pantheon, along with sun and moon gods and a variety of others, but he has been reduced to a lower rank, and it is the youthful, heroic aspects of the god that are emphasized.

When we look at the gods of the gold bowl, it will be recalled that the deity in the lead chariot of the upper zone has been identified as a weather god, and by his leading position it seems reasonable to assume that he is of top-ranking status. Yet, in our discussion of the dress code for distinctive sets of figures on the bowl, we have noted the difference in dress between this god and the figure in the lower zone tentatively identified as Tešub, by virtue of his participation in an identifiable “heroic” narrative. If it is possible for the same figure to wear different garments at different stages of the narrative (and to sometimes be shown with shoulder emanations and other times not), then the distinction in dress is not a useful one. However, if, as I sus­pect, types of garments are re­stricted to sets or classes of characters, then if Tešub is below, we must ask the question, Who is above? The Urartian pantheon pro­vides an answer in ranking Haldi above Tešub. Now, Urartu is not the only 1st millennium tradition that distinguishes Tešub from the main storm god of the pantheon: on a stele from Tile Barsib, near Carchemish on the Euphrates, a list of deities written in hieroglyphic Luwian includes both the primary storm god Tarhunzas and Tešub (Hawkins 1981:166). This separa­tion may be a function of changes in the 1st millennium in general, or particularly within a Hurrian/ Urartian sphere; in any case, it must enter into discussions of the fit be­tween the iconography of the bowl and the religious system of a sug­gested host tradition.

If these comparanda turn out upon closer examination to be valid, then we must re-examine whether it is indeed necessary to seek explanatory parallels for the gold bowl from as far away in space as Anatolia, or in time as the mid-2nd millennium. Concurrently, we need excavation and analysis of archaeological materials in the Urartian homeland in eastern Turkey to compare with 10th-9th century materials from north­western Iran. Specifically, we would want to know whether sites located between Lake Van and Lake Urmia were part of a single “interaction sphere” that could be identified as Urartian in the late 2nd and early 1st millennium, as it certainly was later (see Kleiss and Kroll 1977).

Sometime around 810-800 B.C., Urartian armies marched through the Solduz valley, and Urartian rulers left inscriptions attesting to their political annexation of the region. Period 111B at Hasanlu has long been known to represent an Urartian occupation in the 8th-7th century B.C. (see Dyson, “Redis­covering Hasanlu“). It has thus been suggested that the Urartian invasion of the late 9th century may have been responsible for the destruction of Hasanlu IVB, although, again, we do not know 9th century Urar­tian material culture well enough to establish whether any of the armor or weapons found at Hasanlu could be identified as Urartian and/or belonging to the invaders (see Mus­carella, this issue).

Even if we assume an Urartian destruction of the Citadel, we still cannot know whether this would represent a “first encounter,” or whether conquest had been preceded by long-standing cultural relations. Presumably, the peoples identified as Urartian would have represented a closely related ethno­linguistic and cultural entity to the Hurrians, and would have shared significant portions of their myth­ological and religious traditions. Therefore, if the Soldua valley and Hasanlu had had significant cultural relations with, or were populated by “Urartians” prior to the end of the 9th century, then we cannot rest comfortably in a “Hurrian” hypo­thesis without factoring an Urartian component into the equation. It may even be necessary in the end to replace the Hurrian hypothesis with a “Urartian” one, for what, after all, is Urartu if not in some ways a 1st millennium political confederation comprised of descendants or close relatives of 2nd millennium Hurrians?

Finally, a great deal of progress has been made in the last 30 years in recovering and understanding the late 2nd-early 1st millennium art of Elam to the south, the Caucasus to the north, and Bactria to the north­east (Amiet 1986, Oganesian 1988). The recent discovery of an Elamite version of the epic of Cilgamesh at a Urartian site in Soviet Ceorgia merely underscores the complexity of the cultural mixes and move­ments of the period (J.D. Hawkins, pears. corn.). Before any final attri­butions can be made for the bowl, it must also be studied with regard to traits held in common with, or indebted to, these other artistic heritages, if only to better under­stand what the shared north-south Zagros tradition may have been without reference to either Hur­rians, Urartians, or Iranians.

Conclusion

Having defined the parameters of uncertainty surrounding the Ha­sanlu gold bowl, we can now begin to eliminate alternatives and move toward a picture of the best pos­sible interpretive “fit” for the vessel’s imagery and cultural back­ground. As stated above, there are many Hurrian literary and visual associations for the bowl; however, even if these are accepted as valid, there are several competing ways of seeing the bowl in cultural per­spective. In one view, a base popu­lation of Hurrian could well have been in the Solduz area, co-existing with an incoming group of Indo-Europeans speaking peoples during Hasanlu periods V and IV; the bowl would then reflect a substrate literary and religious tradition, possibly even one adopted and assimilated by the arriving group as well. In another view, Indo-Europeans who moved into the area might long since have ab­sorbed Hurrian myths through oral and literary traditions, and made them their own. In a third view, no Indo-Europeans arrived at Hasanlu at all; instead, the new population we see in periods V and IV repre­sents peoples related to the 2nd millennium Hurrians and 1st millen­nium Urartians, and the bowl is entirely at home in this context. In a fourth and final view, there is such a common pool of motifs in the period that we cannot specify the mythological or religious tradition underlying the decorative scheme of the bowl at all.

In the end, what is needed for a better understanding of the gold bowl is a clearer picture of the historical population movements and ethno-linguistic mixtures in northwest Iran during the late 2nd and early 1st millennium B.C., as well as clearer models for ethnic interaction in antiquity (see Kamp and Yoffee 1980). In addition, we need a far better understanding of the syntax, or structure, as well as of the individual motifs on the bowl. And we need to know a great deal more about the possibilities of establishing correlations between the structural properties of imagery, of language, and of a culture—that is, of identifying particular modes of representation with specific social, cultural, and ethno-linguistic groups.

Until some of these factors are better understood, the meaning of the gold bowl in the context of Hasanlu in the early 1st millennium B.C. will, I fear, continue to elude us. Nevertheless, despite the fact that the situation may be more com­plicated due to conflicting evidence than it was 30 years ago, I would argue that we have made some progress in deciphering the bowl. This progress reflects not only new excavated and textual data, but also new questions being asked by art historians and archaeologists of their material.

I myself am persuaded that the multiplicity of Hurrians sources—literary imagery, organization of the pantheon, visual subdivisions, narrative devices—is sufficient to keep alive the possibility of a Hurrian connection. If anything, the identification of Hurrian com­ponents on the bowl at the time of its excavation has been strength­ened, and we now have a clearer sense of the “reading” of the bowl visually. In addition, the possibility of a Hurrian or early Urartian presence in the Hasanlu area cannot at present be rejected. We must hand over to our archaeological, historical, and linguistic colleagues some of the problems of inter­pretation related to our uncertainty; and it may well be that we must first make significant progress on the “Who?” of Hasanlu, before we can answer questions concerning the “Who?” of the gold bowl. But there is also more to do for the art historian in pursuing systematic analyses of the bowl’s style and technique, its iconography and com­position, and its relationship to materials from surrounding tradi­tions. And we must keep in mind that we have been heavily depen­dent upon literary texts for parallels, but that visual representations can reflect an oral tradition as well: stories carried by bards who themselves can move between cultures and languages.

Hopefully, at the next 30-year review, we will have more answers to our questions and will be better able to see this most important of archaeological finds in its fullest, most complex social and historical context. For the present, the bowl continues to sing of a world of gods and men, heros and demons that resonates just beyond our ken.