The Finds of Hasanlu

The Art of the Manneans


By: Edith Porada

Columbia University

With the collaboration of R. H. Dyson and contributions by C.K. Wilkinson


The excavations at Hasanlu in the Solduz valley of present-day province of West-Azerbaijan, in which the gold bowl was found, have thrown light on the prehistory of north-western Iran, especially in the late second and early first millennium B.C. The best known and richest period so far excavated, Period IV, is characterized by grey pottery, accompanied by black and red varieties. The black variety is very thin and brittle, burnished and fired in a reducing atmosphere; the cross-section of some of the sherds shows that the pottery was occasionally fired a second time in a reducing atmosphere to obtain the desired black metallic surface. The body of the vessel is sometimes fluted, guadrooned or may be decorated with grooves. The handles are often raised higher than the vessel rim and have a tab for the thumb, like beer-mugs of modern times. This fine blac pottery was used in the citadel as a kind of palace ware and was undoubtedly made in imitation of metal vessels. While it has been found in buildings, it has not to date been found in graves. The grey to grey-black pottery which has been found in graves of this period is much thicker than the black pottery and is of a less fine paste. An average grave in the cemetery area contains a combination of small bowls, a large storage jar and a long spouted jar for liquids. We show one of these spouted jars standing on the tripod on which it was found. The feet of the pottery tripod are in the shape of a bull's cloven hooves; in another example, they are formed like shoes with upturned toes. A spouted jar has been found in every excavated grave as well as in the living quarters. Obviously the jars served not only as funerary gifts but were also used for practical purposes; one can pour from them very well.

The period of grey pottery is preceded by an earlier stage of grey button-base pottery, Period V. Instead of spouted vessels, goblets like the one in Plate 27 [above right] were placed in the grave with the last drink for the deceased. They were accompanied by small storage jars and bowls and, as in the later graves, a quarter or two of goat or sheep. The light grey pottery of this period is related to that of central Iran both in its grey colouring and in its goblet shapes, which are similar to pottery from Sialk V, Giyan I and I Khurvin, a necropolis near Teheran. At the same time the shapes of associated painted buff-coloured pottery, also with tiny bases, connect the entire cultural layer of Period V at Hasanlu with levels in northern Mesopotamia which are dated about 1500-1200 B.C., the period during which the Hurrians were a dominant ethnic element. The painted vessels are, however, rare in Period V and appear to be a carry-over from the preceding level in which they are characteristic and the grey pottery is completely absent.

These ceramic comparisons indicate approximately the dating of the levels at Hasanlu, which has now been given greater precision through radio-carbon measurements. The time of the button-base pottery of Period V is given by four samples which range in date from 1217 ± 122 B.C. For the period of the grey ware eight samples have been assayed, mainly using fragments of wooden beams and columns of poplar wood from the buildings. The dates obtained should determine, therefore, the time of the cutting of the wood and the construction of the buildings. These dates range in time from 1033 ± 51 B.C. to 950 ± 55 B.C. [p. 108] [P-424] with an average of 1001 ± 20 B.C. The beginning of the period should thus fall around 1000 B.C. Only two samples of material which may be considered contemporary with the destruction of the citadel have as yet been measured: one, a sample of grapes [hence the fire took place in August or September of the year], measured 912 ± 69 B.C. [P-577], and the other, a sample of charred wheat, 811 ± 69 B.C. [P-576]. More samples must be assayed, but the present evidence provides an average date of 862 ± 49 B.C. for the general time of the destruction--a date which is significantly different statistically from the date of the construction. [1] A date in the late ninth century is also supported by a comparison of the artifacts found in the burned buildings at Hasanlu and those discovered elsewhere.

The description of the pottery of these two periods provides an indication of historical conditions in the area to which Hasanlu belongs. There is a major break around 1200 B.C. with the end of the painted buff ware and the abrupt appearance of the button-base grey ware. There is no break in the development of the grey ware between the late second and early first millennium B.C., although new forms probably reflect new influences. The spouted jar, which is a large container, replaces the goblet of earlier times; moreover, the concept of a spouted jar appears to occur in Sialk V earlier than at Hasanlu and probably indicates influence from central Iran at the latter site. Non-ceramic finds at Hasanlu show that in the ninth century B.C. strong Assyrian influence was also felt.

Assyrian military reports and administrative documents of the ninth century indicate that the area from Lake Urmia and its vicinity in southern Azerbaijan to the mountains of Kurdistan was called Mannai. Tentatively, therefore, we have called Hasanlu IV, with its characteristic grey pottery dated to the tenth and ninth centuries B.C., 'Mannean', assuming at the same time that the Manneans of that period were partly descended from Hurrians or a relative people of the second millennium. Such an assumption is supported by the place-names and personal names of the Manneans recorded in the Assyrian and Urartian annals. At the same time we recognize the possibility that other ethnic elements such as the Indo-European also may have been included by this time within the Mannean area. In the north and north-west the Manneans were neighbours of the powerful Urartians, whose centre was situated on Lake Van about 800 B.C. To the west of the Manneans were the Assyrians, separated from them by the Zagros mountains. In the south-east the Medes began to occupy the plain of Hamadan and emerged in the seventh century as a third power bordering Mannai.

The Manneans are first mentioned in the reports of the Assyrian king Shalmaneser III [858-824 B.C.], who made allies into the neighbouring countries in the east and west in order to secure his frontiers and to collect booty. At the time of the Urartian expansion under King Menua [810-781 B.C.] the Manneans are also mentioned in Urartian inscriptions. This is roughly the time of the destruction of the citadel of Period IV at Hasanlu.

Hasanlu, whose ancient name we do not know, was one of the small towns of the region, crowned by a high citadel. In the citadel, which was surrounded by a powerful fortification wall of mud brick set upon a stone foundation, were probably housed the religious centres of the town and the seat of the local lord or administrator. The scattered houses of the outer town, which were inhabited by craftsmen and probably also by merchants and farmers, and the cemetery [p. 110] which belonged to the town had no protection. In case of war the inhabitants fled to the citadel. The walls of the citadel were probably about nine metres high and were over three metres thick. Every thirty metres there was a fortification tower. Evenly spaced between the towers was a pair of piers which reinforced the mud-brick wall. This type of fortification wall is very similar to that of the Urartians. The main gate was on the western side of the citadel but has not yet been excavated. Within the fortification wall three major buildings have been unearthed: two in the south-western quarter and one in the north-west. The three buildings all share a common general plan consisting of an entry portico leading into a long, narrow reception room, behind which lies a columned hall. Either the portico or the reception room is flanked by a stairway leading to an upper floor. The columned hall is flanked by storage-rooms on both sides. These halls may be the prototypes of the later columned halls of the Achaemenid period. Only the column-bases remain. Each consists of a stone slab under a mud-plaster socle which preserves the original round shape of the wooden column. Charred fragments of the columns show that they were of poplar wood. The columned halls of all three buildings were roofed with poles, small pieces of timber, small wooden slats, mud plaster and probably reeds. In Buildings I and II a small rectangular area of paving with a sunken pithos drain suggests the possibility that there may have been an open area in the roof for purposes of light and ventilation. On the other hand these paved areas might also have been used for some ritual or more practical purpose. A bench ran along the walls, and in Building I there were three hearths.

Building I consisted of two wings, east and west, lying across an open court. Building I East has a portico leading to a large square room on the entrance wall of which a flat stone forms a platform. One is reminded here of similar raised daises in Assyrian throne-rooms and may assume that here, too, a reception room was intended. The entrances of the two wings of the building which face the court, as well as the entrances to Buildings II and III, are reminiscent in the position of the portico and the reception room lying behind it--both parallel to the faade--of the frontal section of northern Syrian places called bit hilani by the Assyrians. The joining of this architectural plan with the local columned hall is interesting. Building II is considered tentatively to be a temple. Such an identification is suggested by the stone platform set in the centre of the portico entrance, some of the small objects found in the building, and the tragic evidence of some forty skeletons, mostly very young women, who apparently were killed just inside the entrance to the columned hall in which they had sought refuge. In the storage-rooms, on either side of the columned hall in all three buildings, were large pottery containers and pottery funnels, probably for wine [crushed wine grapes have been identified among the plant remains], which is still made in this region.

Between Buildings I and II lies a two-room structure called the Bead House by the excavator because so many beads of white paste, carnelian and sea-shells were found in it. Many of the shells were brought back from the area of the Persian Gulf and Indian Ocean and indicate trade with the south. Broken bone cosmetic containers decorated with patterns of incised circles and shattered tripod bowls of grey basalt were also found. Together with these objects was a tiny fragment of gold foil bearing the figure of a winged Assyrian genius. This motif points to a connection of this small building with the religious installations of Hasanlu. In a single-roomed structure nearby, called South House by the [p. 112] excavator, a hearth and a large number of small grey-brown pottery bowls with pointed bases were found. These vessels may also have had a ritual use.

The ground-floor rooms which have been discussed may not have been the most important at Hasanlu. On the second floor, over the rooms around the columned hall, indicated by the stairways, the volume of collapsed brickwork, a section of collapsed wall six metres in height, and the scatter of objects in the debris above burned ceiling beams and plaster, there may quite possibly have been the most important rooms, just as they are in modern buildings near the site. At any rate, an important second-floor room must have been in the southeast corner of Building I above a little refuse room, for it was into this area with its fill of broken shallow black bowls and sheep and goat bones that the three soldiers and the golden bowl discussed above had fallen. In a similar manner a mass of objects including a bronze stand with figured decoration and a silver and electrum beaker had fallen from above the doorway at the rear of the portico of the west wing of the building.

The silver beaker was wrapped inside and out in a piece of material which may be seen in the impression of the weave on the partially corroded surface. The simplest explanation is that this material was meant to protect the beaker from [p. 113] tarnishing by oxidization. Since leather fills this purpose better than cloth, an alternative suggestion might be that the bowl was in a bundle of loot being readied for removal from the building.

The shape of the beaker is reminiscent of the beaker with concave sides from Marlik, of vessels from Luristan, and also of tall beakers found in Georgian excavations. [2] The decoration of the beaker, however, is unique.

Four raised ridges divide the surface of the beaker into five cylindrical fields, of which the upper and the lower ones are the narrowest and have a decoration of triple palmettes. In place where the electrum overlay is well preserved, one can see that it was hatched in different directions; this hatching gives the impression of tightly wound silver thread. The wide field shows the victory scene following a battle. Below there is an animal contest and possibly a hunting scene. One field has been left empty. Most interesting is the victory scene. The charioteer looks in the direction of the horses he guides, while the archer, with drawn bow, probably the main figure of the scene, guards a disarmed figure walking behind the chariot, who in turn is followed by an armed soldier leading a riderless horse. A second armed soldier brings up the rear. A second enemy appears to lie between the chariot body and the wheel, imploringly raising one hand.

The organization of the victory procession reflects the Assyrian style of rhythmically ordered strips illustrating battle scenes, especially as created by the artists of Ashurnasirpal II [883-859 B.C.]. [3] Several features of the chariot on the beaker can be paralleled in AshurnasirpalÕs reliefs: the heavy body of the chariot to which crossed quivers are attached on the side, the form and proportion of the wheel, even the bird's head in which the chariot pole ends. The fact that the body of the chariot here rests directly upon the axle, not in front of it, may be due of the artist's quest for symmetry.

In the hunting reliefs of Ashurnasirpal II, the king is occasionally shown shooting backwards with bow and arrow--a remarkable physical feat in a rapidly moving chariot without springs. In battle reliefs the Assyrian king is never shown in this position, probably because such a representation offended the Assyrian sense of realism. There can be found parallels in the Assyrian reliefs for the enemy who raises a hand in supplication and also for one hanging over the chariot wheel, [4] but not for the prisoner walking calmly behind the chariot and represented in a more dignified manner than are similar captives in Assyrian reliefs. Whether this representation on the silver beaker is purposeful or accidental cannot be decided here. The detail is one of many which differentiate the beaker of Hasanlu from Assyrian prototypes. The most characteristic features of the figures of the Hasanlu beaker are the straight hair tied with a band or diadem [except for the charioteer who wears a pointed helmet] and the straight beard, in contrast to the curly hair and beard of the Assyrians. Further characteristics of the figures of the Hasanlu beaker are the low receding forehead, strongly accentuated noses, large circular eyes and the rich decoration of the clothes. The outlines of the animals are marked on the inside with a series of concentric arches similar to those which indicate some of the joints of the figures on the gold bowl. The rest of the animal body is here indicated by hatching in irregular strips. Despite its crude exaggeration the decoration of the animal bodies is reminiscent of that on the gold bowl. Moreover, one can note in both bowl and beaker a related intensity of expression. Although the beaker belongs to a period which is two hundred years later, and has incorporated Assyrian [p. 114] influence in theme and composition, the liveliness of the representation, which is the most attractive trait of this metal art, has been retained.

It is not impossible that we may be able to associate with the style of the beaker a group of objects of a different nature also found at Hasanlu: large iron garment pins which end in bronze lions. Their use was determined only through their association in groups of one to three with individual bodies in the columned hall of Building II. The fragmentary lion reproduced below in our Plate 29 was purchased during the first season of the excavation, but had certainly been found at Hasanlu. It shows the thin ruff and tufts of hair on the mane also found on the lion represented on the silver beaker from Hasanlu. One might even point to the fact that the bronze lion's legs and body are marked by patterns of hatched lines also used on animals of the beaker, though none have the same regular herring-bone pattern. The tendency toward geometrization noticeable in Iranian art of the tenth to ninth centuries B.C., as seen for example in some of the Elamite tiles assigned to the same period, and also in seals and bronzes of Luristan, again manifests itself here but in a different, local conditioned manner.

Closer connection between objects made and used at Hasanlu and the style of Elam in south-western Iran may be manifested in a glazed square wall-tile which has the bearded head of a human-headed bull in place of a plain knob. Both parts of the tile were fashioned separately, and then joined together before glazing and firing. The head is separated from the tile by a ridge which forms a kind of collar. The head is hollow, allowing the piece to fit over a wooden peg driven into the wall. Once in this position, the tile was fastened by a much smaller peg driven through a hole in the side just below the ridge. The colours now visible in the head are red, blue-green and black, but of course they were modified by fire during the destruction of the town.

This wall-tile represents a combination of the glazed wall-tiles of Assyrian palaces and their protective human-headed bulls, which were placed at the main entrances. [5] In none of the Assyrian palaces, however, has a similar combination of tile and plastically sculptured human-headed bull been found. One can only refer to a comparable tile found in the Ishtar Temple at Ashur, which bears a sculptured calf's head. [6] On the other hand the shape of the horns, the thick nose, the beard which is not too long, all relate to a terracotta head from Susa. [7] Since the use of glazed wall-tiles probably came into fashion at Susa, the head from Hasanlu may be traceable to Elamite influence. Other tiles from Hasanlu, however, which show stylized lotus and palmette designs and other Assyrianizing patterns, are surely imitations of Assyrian prototypes.

An object for which no foreign parallels can be cited is a knife-handle with gold cloisonné found in the excavation of Building II. The gold outlines describe a [p. 116] bearded man with shoulder-length hair and short-fringed kilt. His right hand is raised in a gesture of greeting or worship. From the wrist of the left hand, which is brought around the body, hangs a scarf or cloth. As on the prehistoric knife-handle from Tepe Sialk, a human figure seems to be represented in a posture of reverence. In both instances one may guess at the ritual use of the knife without being able to give any proof. Earlier cloisonné works are unknown in Iran, although the small falcon from Susa was produced in a related technique. The execution of the knife-handle of Hasanlu, however, is not careful enough to indicate a beginning of cloisonné technique. Rather, one would think that an already established technique was here applied cursorily and somewhat awkwardly. The Hasanlu knife-handle and later cloisonné work from Ziwiye [8] indicate the path which this technique took from its point of departure somewhere in Iran toward the north. From there it may have spread to the northeast, to reappear a few centuries later in more elegant form in the magnificent armlets of the Oxus Treasure. The reason for the rarity of cloisonné objects is surely their precious material. Originally this colourful technique must have been widely distributed in Iran and the neighbouring countries.

Some of the finest metal-work discovered at Hasanlu is preserved in the form of rhyta. The Greek word rhyton is derived from the running of liquid; it actually refers only to a vessel from which a thin stream of liquid issues. Usually the word is used, however, for drinking-vessels the lower part of which is formed by an animal head that often, but not always, has an opening between the lips from which the liquid can flow. The bronze vessel from Hasanlu in the form of a ramÕs head has no such opening and should therefore better be called a cup or situla in the form of a ramÕs head, like the closely related situlae from Gordion. [9] All these situlae have a separate cup to hold the liquid placed within the animals head. In a second example, a bronze horseÕs-head rhyton, no cup has been preserved, nor is there a spout between the lips. Both vessels, like the earlier drinking-horns from which they descend, would have had to be carried by cup-bearers, since they have no base on which to stand.

In the situla representing a ram the eyes are inlaid with an opaque blue substance, as were the eyebrows, and probably also the nostrils. [10] Below the rim there is a silver band with a pattern in repoussé of gazelles and rosettes. The horns were originally covered with silver, as is shown by tiny pieces still preserved. These different materials would have created a fairly colourful impression, as suggested in the reconstruction in our plate.

In the horse rhyton no traces of another material have been found. This object may have been meant to impress by it noble, naturalistic outline rather than by any effects of contrasting colours. The indication of the mane by undulating strands of hair reminds one in more naturalistic form of horses on Urartian bronze psalia found in a grave at Altin Tepe near Erzincan. [11] The mane, which is combed onto the neck, and which forms a semicircle on the forehead, seems to be characteristic of Urartian horses as well as of the horse rhyton from Hasanlu. It is also visible on a somewhat later horse from the Urartian fortress Taishebaini at Karmir Blur in Russian Armenia. [12] The Urartian horse is much more stylized than the softer naturalistic horse rhyton of Hasanlu.

A lively representation of animals in relief, more precisely of one preserved animal in relief, is found on the fragment of a vase made from a composite blue substance identical with or very similar to the inlays on the ram's-head rhyton already described. The rampant goat seen on the fragment is one of a pair [p. 118] flanking a central palmette-tree. The second goat is represented only by a fore and a hind leg. A winged figure may be indicated by the border of a wing on the upper left edge of the fragment.

Like all objects from Hasanlu, this one also has its own character, although connections with Assyrian representations are obvious. The closest similar motifs are found on the painted pottery from Ashur. In the Assyrian representations, however, natural trees which grow on mountains are clearly differentiated from palmette-trees, which seem to be artificial structures and do not issue from a mountain. Here, on the fragment from Hasanlu, the palmette-tree grows from a mountain which is indicated by the scale pattern below. Furthermore, the moufflons or ibexes of the Assyrian representations usually do not touch the tree; if they do so, they only place one leg on the tree. On the Hasanlu fragment, however, the animal places the three legs which are preserved upon the branches of the palmette-tree. In fact the branches almost seem to bend under the weight of the goat and remind one of goats eating the leaves of trees, as they may be observed doing today in Iran. It is instructive to compare the representation from Hasanlu with the repoussé gold plate from Ziwiye, where there is also an ibex next to a tree. The ibex from Ziwiye does not even touch the tree--indeed, does not seem connected with it--while this connection is made very clear on the Hasanlu fragment. Again one is struck by the liveliness of the expression in this and in the other works of Hasanlu discussed here. This liveliness seems to be a criterion of the style of this region during the end of the second and beginning of the first millennium B .C. Whether, as suggested by the find of a beaker from Marlik with similar lively, factual renderings of animals, the style extended considerably beyond Hasanlu, through northern Iran, is not yet certain.

The seals found at Hasanlu give a picture of the external connections of the town in the tenth and ninth centuries B.C. Most of them were made in Assyrian style, although probably only a few were actual imports. A distinctive and probably local style is represented by a tall cylinder [13] of the same blue composition already seen in the vase fragments and rhyton inlay. The slender animals with their uncertain postures in relation to the ground-line correspond to a general trend of seal representation in Iran in this period. Several stamp seals were brought from Palestine and Syria. A connection with these regions is also indicated by certain sea-shells and by the discovery of a ritual lion bowl of the blue material mentioned above, the base of which is in the shape of a hand. [14] This type of vessel is widely distributed in the West. The establishment of this contact with the West is very important, because it continued in the following centuries and is represented in the art of the Treasure of Ziwiye discussed in the next chapter. [p. 120]







1. Carbon-14 dates for Hassanlu are publishd in E.K. Ralph, 'University of Pennsylvania Radiocarbon Dates III,' American Journal of Science, Radiocarbon Supplement I [1959], pp. 49-51 and R. Stuckenrath, 'University of Pennsylvania Radiocarbon Dates VI,' Radiocarbon 5 [1963], pp. 85-90. The published dates are given in the old half-life of 5568± 30 years, but have not been recalculated in the above statement for a half-life of 5730± 40 years, adopted by the Carbon-14 Conference of 1962 as the best present estimate. Only samples receiving equivalent laboratory processing have been used in the above presentation.

2. Two beakers said to come from Luristan show forms somewhat similar to that of the silver beaker from Hasanlu and also have raised ridges which divide the surface of the beaker into several horizontal bands. One is illustrated by G. Goossens, Bronzen uit Loeristan [Koninklijke Musea... Brussels, 1956], Pl. 17 [formerly in the Graeffe Collection]; the other is reproduced in Survey IV, Pl. 68 B [formerly in the Holmes Collection]. Tall beakers of somewhat similar shape were found in Beshtasheni in Georgia, of which representative examples were illustrated by Schaeffer in Stratigraphie, Fig. 285.

3. The characteristics of the Assyrian chariot from the time of Ashurnasirpal can be seen in the reliefs reproduced in R. D. Barnett, Assyrian Reliefs, Pls. 24-27. See also, the article by B. Hrouda, 'Der assyriache Streitwagen,' Iraq XXV [1963], Pl. XXIX: 3, 5, 6 and p. 156.

4. An example of an enemy with his hand raised in supplication, as if to ward off the arrow which has already pierced his breast, is seen in the relief of Ashurnasirpal, Barnett, Assyrian Reliefs, Pl. 24; an enemy hanging over a chariot wheel is seen in ibid., Pl. 25-Both postures are combined in the rendering of the enemy on the chariot of the Hasanlu beaker.

5. For the glazed knobs and tiles of Assyrian palaces, see Andrae, Coloured Ceramics, Pls. 31-36. For the human-headed bulls and lions at the entrances to the main halls of the palace of Ashurnasirpal II, see A. H. Layard, Monuments of Nineveh [London, 1849], Pls. 3. For the human-headed bulls at the entrance to the throne room of Sargon II [721-705 B.C.], see Frankfort, Art and Architecture, pp. 76-77, Fig. 31.

6. W. Andrae, Die jüngeren Ischtar = Tempel in Assur [WVDOG 58, 1935], Pl. 37 b, text p. 93.

7. A glazed terracotta head of a human-headed bull or bull-man from Susa was published in MDP XXIX [1943], p. 64, Fig. 53:3. It seemed to correspond in size to the head from Hasanlu.

8. Godard, Ziwiyè, p. 104, Fig. 90.

9. For a discussion of these situlae or cups, see R. S. Young, 'Bronzes from Gordion's Royal Tomb,' Archaeology II [Winter 1958], pp. 227-231.

10. The blue material used for inlays on the ram rhyton, which is at present being analyzed, appears to be similar in quality to the 'Egyptian blue' objects from Persepolis analyzed by F.R. Matson and published in Persepolis II, pp. 133-135. The blue colour in most of these man-made materials is formed from various copper compounds and not from lapis lazuli as is often stated in the absence of analytical data.

11. The Urartian psalia with horse heads from Altin Tepe are published by Tahsin Ozgüç, 'Excavations at Altintepe,' Belleton XXV [1961], p. 287, Figs. 16, 17.

12. The horse head from Karmir Blur was published by B. B. Piotrovskii in Iskusstuo Urartu [Hermitage Museum, Leningrad, 1962], Pls. XXVI-XXVII.

13. A drawing of the impression of this cylinder seal was published in Archaeology 13 [Summer 1960], p. 128.

14. For a discussion of the 'lion bowl from Hasanlu', see the article with that title by M. N. van Loon in Expedition 4 [Summer 1962], pp. 14-19.