Om Barbartemplets datering

Resumé

On the Date of the Temple at Barbar in Bahrain

During his first stay at Bahrain, in the winter of 1953-54, P. V. Glob discovered on a survey along the northern coast of the island near the village of Barbar« ... in the edge of a gravel mound of considerable size the top of a large stone block with carved holes, two square ones and a round one« (1). The same winter Glob made a sounding on the site and it transpired that the mound contained an impressive building dating from the 3rd millennium B.C. - a temple in provincial Sumerian style. In the following years most of the mound and its surroundings were excavated.

The investigation showed that three main phases in the development of the temple could be distinguished.

The earliest temple was built on a rectangular platform about 2 m. high, nearly 25 m. long and 16-18 m. wide, surrounded by a stone wall. On top of the platform there were along the edge remains of small rooms placed around an open courtyard. From the west a ramp or stairway led up to the temple, and in the southwestern corner two staircases descended to a square well. The water in this well has undoubtedly played an important part in the cult already in the time of the earliest temple, and the significance of the well is shown by the fact that it was also in use -in a slightly altered outer form- during the second and third phases of the temple. A deposit consisting of nearly a hundred conical clay goblets and a few weapons of copper was found in the fill of the platform. In a small recess under the ramp leading up to the temple a cylinder-shaped jar of limestone was found, covered by a stone slab secured with bitumen. In just such a position one might have expected to find a foundation tablet. But the jar was empty. The organic material that it originally contained had long since vanished.

With the second temple a considerable expansion of the plan took place. The central platform was retained, but it was now surrounded by a lower terrace, confined by an oval wall measuring ca. 70 m. in length. On the south facade of the temple a wide staircase led up to the more than two meter high oval terrace. From the central platform there was towards the east, via an 8 m. long ramp, access to a lower, oval courtyard with a circular, raised fire-place in the middle. The courtyard was covered with thick layers of ashes containing burnt bones of sacrificed animals. Towards the west there was a 15 m. long flight of steps from the central platform down across the oval terrace to a basin measuring 3 x 4 m. The steps as well as the basin were made of finely cut limestone ashlars and on both sides of the staircase there were traces of a double row of wooden pillars. The pillars had stood in sockets of limestone, and -like other woodwork in the temple- they had been coated with thin copper sheets. Along the outer side of the oval wall a narrow staircase led down to the well which was already in use in the time of the first temple. Copper models of spearheads and crescent-shaped shafthole axes were found in the fill of the central platform together with beads of marble, lapis lazuli and carnelian, decorated fragments of ivory and a single narrow sheet of gold. Furthermore, a copper axe, a copper adze (Fig. 5), and a small rattle (Fig. 4) of copper were found.

The third and most recent temple is bigger and more regularly shaped than the two older constructions. In the middle a new central platform was erected, this time a square one ca. 30 m. on either side, enclosed by a colossal wall of stone ashlars, which in places is still preserved in a thickness of 6 meters. Along the wall there were remains of ashlar-built rooms towards west and north. Most of the platform had, however, been laid out as an open, trapezoid space paved with limestone slabs. Two circular offering tables built of stone and with a low seat between them stood in the middle of the courtyard. East of the offering tables three stone altars were erected, and in the southwestern end of the courtyard stood a short row of nearly 1 m. high stone blocks, each of which was perforated by a round hole. Along the edge of the holes clear marks of wear were visible. They led Glob to suppose that the stones had been used when tethering the animals for sacrificing (3).

In the northeastern corner of the courtyard there was a square pit in the floor. Here the remains of a foundation deposit were found: three alabaster jars (Fig. 7), a beaker, a few weapons, a male figure, a small bird, and a considerable number of nails and fragmentary sheets -everything of copper. In the same area a bull's head of copper and several stone jars were found under the floor.

As in the preceding phase the central platform rested on a lower, oval terrace enclosed by an ashlar wall, which still stands to a height of more than 4 m. towards the south. The new terrace covered a somewhat larger area than the older one. It has presumably measured about 100 m. in east-west direction, and ca. 60 m. from north to south. The basin and the long flight of steps that were built in connection with the second temple, and the well southwest of the temple, were incorporated in the new architectural unit.

In its general plan as well as in many details the Barbar temple is closely related to the contemporary Sumerian temples at Khafajah and al-'Ubaid. These temples were both dedicated to Ninhursag, goddess of the earth, and as a hypothesis the name of this goddess has also been connected with the Barbar temple (4).

Already in 1954 - after the first sounding in Barbar -Glob realized that the temple should be dated to the 3rd millennium B.C. (5). In the intervening time both P. V. Glob and T. G. Bibby have elaborated the chronology of this period by demonstrating Indian and Mesopotamian connections to the Bronze Age finds from Bahrain (6). The following comments on the date of the Barbar temple must therefore be taken as a provisional supplement to what is already known.

The most reliable evidence for the dating of the earliest temple is provided by the clay goblets found in abundance in the fill of the platform, and by a painted sherd, found in a layer north of the first temple.

The goblets, which are 10-14 cm. high, have a conical upper part and a low foot, that in some cases is solid and in other cases hollow (Fig. 2). They are irregularly shaped, made of clay containing some lime and tempered with fine sand, and fired so that the surface as well as the core has a pale red colour. In Barbar the goblets are only known from the foundation deposit of the earliest temple. They are, however, variants of a type of conical goblets that in Mesopotamia is known from all the Early-Sumerian cities, and they are often -as in Barbar- found in temple deposits. In Mesopotamia they are in time confined to Early Dynastic I (8). In Nippur they even seem to concentrate around the middle of that period (9), while in Warka they possibly start a little earlier (10).

Of importance for the dating of the earliest temple is also the sherd seen on Fig. 3. It belongs to the transition between the body and the shoulder of a large, wheel-made jar, painted in black and plum-red. On the lower part of the body is a horizontal metope-decoration: a wide area filled in with black cross-hatching, limited by narrow black and red bands. On the transition between the body and shoulder of the jar there is a slightly incised groove, along which a horizontal black band runs on the shoulder of the jar. Inside this band the space is filled out with red paint. The core is reddish buff, tempered with fine sand, and the clay contains a large quantity of mica. While the above-mentioned goblets must be seen as locally made variants, the sherd undoubtedly comes from an imported jar of late Jamdat Nasr type. Polychrome jars of this form, and with the same decoration, have been found in Ur (11), Jamdat Nasr (12), Tell Uqair (13), Warka (14), Telloh (15), and in Khafajah (16). With a single exception (17) they all belong to the latest phase of the Jamdat Nasr period and the transition to E.D.I.

Even if it is tentatively assumed that several centuries passed between the manufacture of the Jamdat Nasr jar and the depositing of the sherd at Barbar, it must nevertheless be assumed that the local imitations of Early-Dynastic goblets were produced at a time when the type was still in use in Mesopotamia. If the opinion is accepted that the first of the Early-Dynastic periods -in any case in the Diyala area- has covered nearly 300 years, the date for the erection of the first temple at Barbar can hardly be put later than 2700 B.C., possibly a little earlier.

Two of the metal finds from the foundation deposit of the second temple have Mesopotamian and Persian parallels:

The small rattle of copper with sides pierced by triangular holes and a lug at either end (Fig. 4) corresponds with a rattle found in the area near the royal tombs in Ur (19). The main part of these have been referred to ED IIIa-b (ca. 2600-2370 B.C.). Three rattles found in a grave on Tepe Giyan (20) represent a slightly later variant of this type. On account of the pottery this grave has been dated to Giyan IV C, which is probably contemporary with the Ur III period in Mesopotamia (ca. 2113-2006 B.C.) (21).

The shafthole adze of copper belongs to a group of adzes, examined by Jean Deshayes (Deshayes' types B1-2a) (22). Adzes of this type are known from Tell Fara (ca. 2700-2500 B.C.), Ur (ca. 2500 B.C.), Susa (ca. 2500 B.C.), Tepe Gawra (ca. 2500-2250 B.C.), Assur (22nd-21st century B.C.), and Maïkop (second half of the 3rd millennium B.C.) (23).

It is obvious that the copper rattle and adze cannot alone form the basis of a precise dating of the erection of the second temple. But it is most probable that the time must lie around the middle of the 3rd millennium B.C. Support for this assump­tion may perhaps also be drawn from the three Mesopotamian oval temples in Kha­fajah, al-'Ubaid and al-Hiba, to which the second Barbar temple is related as far as the architecture is concerned. They were all built in ED II (ca. 2750-2600 B.C.) (24).

It is not unreasonable in this context to mention the fact that in the deposits of the second as well as the third temple a few painted sherds (Fig. 6) were found. They are similar to the pottery that in the last few years has been found on Umm an-Nar and in Hili by the Danish archaeological expedition in Abu Dhabi (25). These finds, which show connections to southeastern Iran and Afghanistan, seem to belong to the middle of the 3rd millennium B.C. (26).

Among the objects from the foundation deposits of the third temple, the similarity of the small male figure and the bull's head of copper to finds from Ur and Susa belonging to the middle or later half of the 3rd millennium B.C. has already been pointed out (27). It has also been mentioned that the alabaster jars, found beneath the floor of the third temple, were of Egyptian origin (28). Alabaster jars of Egyptian type are known from many Near-Eastern cities (Byblos, Mari, Ur etc.), and in some cases it is evident that they were also made in Egypt. As far as the Sumerian jars are concerned, doubt has, however, been cast on the place of their origin (29), and the small alabaster jar with a lid from Barbar, seen to the right on Fig. 7, is indeed of a very coarse material which is not to be found in Egypt. On the other hand there is no doubt that the two tall cylindrical jars, seen on Fig. 7, are Egyptian. They must on account of the rim profiles be assigned to the later part of the Old Kingdom (30).

There is one more detail in the third temple that seems to reflect Egyptian influence. Parallels to the above mentioned stone blocks with holes, which by Glob were interpreted as tethering stones for the sacrificial animals, are known from Egypt, where they have been used in connection with bull offerings. The stones are found in burial monuments from the VIth Dynasty (ca. 2320-2160 B.C.), e. g. in the tomb of Mereruka in Saqqara (31).

In the time of the third temple stamp seals of steatite occur. They are of the type that in the Bronze Age is found in great number along the west coast of the Arabian Gulf from Kuwait to Bahrain (32). In Barbar 9 of these seals (Fig. 8) were found, together with 6 seal impressions on small lumps of clay (Fig. 9) and an impression on the side of a red jar (Fig. 1). A thorough analysis of the chronological placing of these seals has not as yet been made, but it is obvious that within the nearly 500 seals known today various styles can be distinguished which both in time and space are widely spread. In Kuwait some of the seals have been found in association with Akkadian cylinder seals. A unique double-convex seal in a gold setting has on one side four gazelles cut in a style characteristic of the best local seals, on the other side a typical Early-Akkadian representation of two naked heroes (33). While some of the seals thus go back to at least Akkadian time, others were still in use in the Isin­Larsa period. An impression of a stamp seal of late type has for instance been found on a cuneiform tablet, dated to the tenth year of the reign of king Gungunum of Larsa (ca. 1923 B.C.) (34).

Most of the seals from the Barbar temple were found in the basin west of the temple. Also a sample giving a C-14 date of 2050 ± 100 B.C. originates in this place. Another C-14 date has been run on charcoal from the layers in which the seal impressions were found. They have been dated to 2080 ± 100 B.C. (35).

In conclusion, then, it can be stated chat the first temple at Barbar was built in the beginning of the 3rd millennium, and hardly later than ca. 2700 B.C. The construction of the second temple, which at the same time is the first oval temple, cannot be dated with certainty, but it is likely that the building derives from the middle of the 3rd millennium. On account of the Egyptian and Mesopotamian parallels to objects from the foundation deposit, the third temple can scarcely be put back later than around 2200 B.C. Stamp seals and C-14 dates suggest chat the temple was still in use towards 2000 B.C. P. V. Glob was therefore right when already after the first year's sounding in Barbar he wrote: »The discoveries made in the upper level of the Barbar temples, of which only a small area has yet been investigated, thus show connections with both Mesopotamia and India in the third millennium, to which period the upper complex must thus be ascribed, and most probably to the latter end of the period« (36).

Peder Mortensen