The Syro-Mesopotamian Origins of Arslantepe's Administrative System

Sarah Kielt
Bryn Mawr College

The late fourth millennium period at Arslantepe, a site in the Taurus mountains of southeast Anatolia, is rich in evidence of administrative practices, in particular, the use of seals and sealings as accounting and controlling devices for a variety of goods. The majority of the evidence from Arslantepe comes from a local, regional tradition. Nevertheless, the excavators have connected the administrative system to southern Mesopotamia.

I will argue, using examples from earlier periods and sites, that the fourth millennium, BC administrative system at Arslantepe develops out a tradition from greater northern Mesopotamia. Identification of the system at Arslantepe as a northern one challenges current scholarship, which associates it with southern Mesopotamia. Southern Mesopotamia has acquired such status and prestige in modern scholarship, that it has been desirable to link sites in outlying areas with the south. I believe that in this case, that desire has altered our perception of the identity of the administrative system at Arslantepe.

In the fourth millennium, material culture relating to the Uruk period in southern Mesopotamia appeared in sites to the east in Iran, and in greater northern Mesopotamia. The nature of the interaction between southern Mesopotamia and these other areas is often characterized as a "core" and "periphery" relationship within an expansionist model. So-called "colony" sites in greater northern Mesopotamia have almost one hundred percent material from southern Mesopotamia, while other sites have primarily local material alongside a smaller percentage of southern Mesopotamian material.

This "Uruk Expansion" coincides with a dramatic increase in the repertoire and use of administrative devices in southern Mesopotamia. Cylinder seals and cylinder seal impressions, first seen a few hundred years earlier, occur in great numbers in this period. Tokens, information storage devices in use in the north since the late Neolithic period, and in the south at least since the Ubaid period, are now found enclosed in hollow clay balls which are impressed with cylinder seals and marked with numerical notation. In this period we also have the first use of clay tablets, first used with numerical notation, and eventually with the pictograms that were the first means of writing.

This boom in administrative techniques in southern Mesopotamia during the Uruk period is reflected in the material found at the Uruk-related sites in greater northern Mesopotamia. Cylinder seals and impressions, hollow clay balls, and even numerical tablets were found at "colony" sites like Habuba Kabira. At these sites, where the pottery, architecture, and administrative devices correlate perfectly to those found in the south, it is a fair interpretation to assume that these items are imports from the south. Other sites, like Arslantepe, had some local material, some southern Mesopotamian. In these cases, it is more difficult to know which administrative devices developed out of a local tradition and which were imported from the south.

The earliest levels at Arslantepe date to the end of the northern Ubaid period, approximately the end of the fifth millennium BC (Frangipane, 1997: 213). Monumental architecture is first noted in Late Chalcolithic levels (period VII), ca. 3700-3400 BC, calibrated (Alessio et al, 1983: 578).

The following period at Arslantepe, VIA, extends to roughly 3000 BC. A limited amount of southern-style pottery shows contact with southern Mesopotamia. Administrative activity dramatically increases at this time (Frangipane, 1997: 214). The architecture of this phase includes a large, palace-like complex, called Building IV, only part of which has yet been excavated. In this complex, a short hall off of the east side of the main corridor connects two storerooms. The southern storeroom contained mass-produced bowls, and sealings--clay lumps attached to containers or doors to restrict access to them. A recess within the western wall of the main corridor was filled with many thin strata of discarded material, including over 2,000 sealings. Adjacent to Building IV is Building I, identified by the excavators as a temple. A narrow room in the northeast corner of the temple held stratified levels of discarded material, including 130 fragments of sealings (Palmieri, 1983: 98).

From the wealth of information at Arslantepe, the excavators have discerned the system in which the sealings were used. Items such as bags, wicker containers, ceramic jars, doorlocks, and sacks were sealed with a lump of clay, which was impressed with a stamp or cylinder seal. The lumps of clay, or sealings, show an imprint on the reverse, of the wicker fibers, or the shape of the jar, and so on, so we know to what it was originally affixed. When the sealed item was opened, the clay sealing was archived; that is, it was collected in or near the place of the breaking. The next step in the system was to collect the sealings from the various places they were stored, count them, and dispose of them. The disposal was done carefully, presumably to prevent the sealings from interfering with future accounting procedures. Thus the sealings in the southern storeroom represent the initial phase of archiving, while the stratified sealings in the two narrow rooms represent the final phase of disposal.

The sealings, along with, perhaps, the mass-produced pottery found with them in the southern storeroom, make up the administrative system of Arslantepe. The question is, does this system have more in common with those that preceded it in greater northern Mesopotamia, or with what we know of the administrative system of southern Mesopotamia in the Late Uruk period? Certainly, the mass-production of pottery at Arslantepe is echoed in the thousands of beveled rim bowls found in the south. Some scholars link these bowls to a system of rations distribution. The pottery at Arslantepe, however, is wheel-made, while the beveled-rim bowls are cruder and mold-made. It is possible, despite this difference, that the idea of the mass-production of bowls came to Arslantepe from the south.

The sealings reveal more information about the administrative system. They bear impressions of both cylinder and stamp seals, though predominantly stamps. The cylinder seal is indisputably a southern innovation; it should be emphasized, however, that the cylinder seals at Arslantepe were cut in a local style, what may even be considered an "Arslantepe" style. This style is characterized by gauging, a curved back to the animals, and figures often depicted with three fingers or claws. A cursory comparison of impressions from Arslantepe with impressions from Susa or Uruk, in the south, show the distinction. The southern examples are drilled and modeled, and generally show a linear organization of space; while the Arslantepe seals are gauged and have a more circular organization of space.

A comparison between Arslantepe's seal impressions and those from sites in greater Northern Mesopotamia yields more striking similarities. I will now consider three sites whose administrative systems resemble that of Arslantepe. Tepe Gawra is located some distance from Arslantepe, near the Upper Tigris in northern Mesopotamia. Level VIII at Gawra is roughly contemporary with Arslantepe VIA (Rothman, 1997: 185). This level at Gawra shows evidence of economic centralization: in addition to craft centers, there was a central storehouse (Rothman, 1997: 185). Sealings with seal impressions were found in this storehouse, as well as in a building just to the south, interpreted as an administrative center (Rothman, 1994a: 113). The other seals and impressions found at this level were concentrated in workshop and religious areas (Rothman, 1994a: 113). This pattern of distribution has much in common with that at Arslantepe, where seals have been found associated with centralized administration and storage.

The sealings at Gawra, much like those from Arslantepe, are from sacks, jars, baskets, doors, tags, and bullae, which are clay lumps attached to string (Rothman, 1994b: 115). There are similarities in motif as well. If sealings from both sites are compared (see publications for illustrations), it can be noted that both include images of two animals on top of each other, an animal with two or three smaller animals around it, right-side up and upside-down animals, and filling elements. Although the three-clawed style and curved back seems, in this period, to be restricted to Arslantepe, the compositions are similar enough to argue for a common regional style.

Another example comes from Degirmentepe, about a thousand years earlier than Tepe Gawra VIII and Arslantepe VIa, approximately the end of the fifth millennium, BC. Degirmentepe is located near Arslantepe on the upper Euphrates. Degirmentepe was taking part, in some way, in a culture shared across many regions, the Ubaid culture. The early part of the Ubaid culture was restricted to southern Mesopotamia, but the later Ubaid reached as far as this site, in southeast Anatolia. It is important to recognize that this was not a southern Mesopotamian culture which had overtaken greater northern Mesopotamia. Instead, certain stylistic elements, such as architecture and pottery, accompanying technological innovations such as irrigation farming, spread north and south, at different rates, from an area probably centered in the central Mesopotamia region. Whatever the nature of the relationship between the various areas with Ubaid material, it is important to emphasize that the Northern Ubaid, of which Degirmentepe is a part, is distinct from the southern Ubaid, and does not represent a rupture from the Halaf period which preceded it (Breniquet, 1996). 

In this period, Degirmentepe consisted of tripartite houses and workshops, and a defensive wall. Twenty-four stamp seals were found, characterized by a gauged style with geometric and figural images (Esin, 1994: 60). In addition, approximately 450 sealings were found. These include jar and sack sealings, bullae, and door sealings (Esin, 1994: 66-69). Thus the usage for the seals appears to have been the same at Degirmentepe as at Arslantepe, 1,000 years later in the same geographic region.

Stylistically, parallels can be drawn to both Gawra and Arslantepe. At Degirmentepe there is an image of a pointy-headed human figure, which is comparable to a figure from Tepe Gawra, from approximately the same period, though with the three-fingered style seen in the later seals from Arslantepe. We can also compare the vultures at Degirmentepe with those at Arslantepe. The vultures at Degirmentepe are carved in a characteristic "Degirmentepe style, " with the interior space filled with hatched lines (Esin, 1994: 62). One, however, has the three claws, characteristic of the later Arslantepe style.

Seals and sealings were found across the site of Degirmentepe, but the majority of sealings were found together in one room, corresponding to the storage areas in Arslantepe's system of seal use and disposal.

The earliest example of a sealing system like Arslantepe's comes from Sabi Abyad, a site in the Balikh Valley in Northern Syria. It shows evidence of continuous habitation from the Late Neolithic period through the middle Halaf period. Level Six falls in the transition from Late Neolithic to Early Halaf. This level was destroyed by fire and thus both architecture and in situ finds are well-preserved. These finds include approximately 300 clay sealings. Their date, approximately 6000 BC (calibrated) make them the earliest clay sealings known. The use of the sealings, as well as their distribution at the site, show that they functioned in a system which resembled, on several levels, that of Arslantepe. The sealings were used on baskets, mats, ceramic vessels, stone bowls, and leather bags, much like the Arslantepe repertoire of use, with the exception of door sealings (Akkermans and Duistermaat, 1997: 1).

The majority of sealings were found in two rooms of Building II, a rectilinear building with small, cell-like rooms. This building has been identified as a storage facility. The sealings were found associated with small vessels, figurines, and hundreds of tokens (Akkermans and Duistermaat, 1997: 2). The placement of the broken sealings in a storage room parallel the archive phase of Arslantepe's sealing process.

The motifs on the seal impressions range from geometric designs to anthropomorphic designs. Although I would not argue for stylistic similarities between these early Sabi Abyad impressions and the late Ubaid or Uruk period seals, the motifs, such as the pointy-headed anthropomorphic figure, caprids, and filling motifs, are similar.

Considering the evidence of systems, like the one at Arslantepe, in use in the north since the late Neolithic period, and considering that there is nothing purely southern about Arslantepe's system, what is the evidence for saying, as the excavators have, that this system follows a southern Mesopotamian model? The main evidence cited is the cylinder seal impression from Arslantepe which depicts a figure on a sledge. It is carved in the local style, but the sledge has been said to be a clear pictorial import from the south (Frangipane and Palmieri, 1987: 299). It has been compared to a plaque showing a similar sledge. It must be noted, however, that the plaque is from the antiquities market; its provenience is unknown (Sürenhagen, 1985: 231). It cannot, therefore, be used as evidence of a southern Mesopotamian influence on the Arslantepe cylinder seal.

The other evidence cited in support of a southern Mesopotamian model is the centralization of the system at Arslantepe, with storehouses and a redistributive system. The redistributive system has not been clearly demonstrated or explained, however, and the storehouses have been seen in the north since the late Neolithic.

Despite the excavators' interpretation of Arslantepe as an autonomous site which had regional influence before contact with southern Mesopotamia, this view of autonomous development stills falls into the shadow of the influence of southern Mesopotamia. Our view of the sites in greater northern Mesopotamia in the late fourth millennium has been colored by our sense of the momentous history being made, literally, in southern Mesopotamia at the same time. In 1952, V. Gordon Childe pinpointed the shift from barbarism to civilization to southern Mesopotamia in the Late Uruk period, on the basis of the development of writing at the site of Uruk (Childe, 1952: 128). This early distinction has left a marked Uruk-centric bias in the ensuing research on this period. As Hans Nissen has observed, however, when writing first emerged it was merely another step in the development of an accounting system (Nissen, 1987: 287). It is therefore crucial to readjust our view of this period and see the developing urbanism in greater northern Mesopotamia, southwestern Iran, and southern Mesopotamia as the product of the preceding developments in all of these areas, rather than as a diffusion from southern Mesopotamia.

Works Cited

Akkermans, Peter M. M. G., and Kim Duistermaat
1997. "Of Storage and Nomads--The Sealings from Late Neolithic Sabi Abyad, Syria, " Paleorient 22/2.

Alessio, M. et al.
1983. "C14 Datings of Arslantepe, " in M. Frangipane and A. Palmieri (eds.), Perspectives on Protourbanization in Eastern Anatolia: Arslantepe (Malatya). An Interim Report on the 1975-1983 Campaigns, Origini 12/2: 575-580.

Breniquet, Catherine
1996. La Disparition de la culture de Halaf. Paris: Ed. Recherche sur les Civilizations.

Childe, V.G.
1952. New Light on the Most Ancient Near East, 4th Edition, London.

Esin, Ufuk.
1994. "The Functional Evidence of Seals," in Archives before Writing, P. Ferioli et al (eds.): 59-81.

Frangipane, Marcella
1997. "Arslantepe" in E.M. Meyers (ed.), The Oxford Encyclopedia of Archaeology in the Near East, Vol. I: 212-214. Turin: 125-147.

Frangipane, Marcella and Alba Palmieri
1987. "Urbanization in Perimesopotamian Areas: The Case of Eastern Anatolia," in L. Manzanilla (ed.), Studies in the Neolithic and Urban Revolutions: The V. Gordan Childe Colloquium, Mexico, 1986, B.A.R. International Series, 349, Oxford, British Archaeological Reports: 295-318.

Nissen, H.J.
1987. "The Urban Revolution of Mesopotamia-Reconsidered," in Manzanilla, L. (ed.), Studies in the Neolithic and Urban Revolutions, Oxford.

Palmieri, Alba
1983. "Arslantepe Excavations,1982," V Kazi Sonuçlari Toplantisi: 97-101.

Rothman, Mitchell
1994a. "Sealing as a Control Mechanism in Prehistory: Tepe Gawra XI, X, and VIII," inChiefdoms and Early States in the Near East: The Organizational Dynamics of Complexity, Monographs in World Archaeology No. 18: 103-120.
1994b. "Seal and Sealing Findspot, Design, Audience, and Function," in Ferioli et al (eds.), Archives Before Writing, Rome: 97-119.
1997. "Tepe Gawra," in E.M. Meyers (ed.), The Oxford Encyclopedia of Archaeology in the Near East, Vol. V: 183-186.

Sürenhagen, Dietrich
1985. "Einige Kulturelle Kontakte zwischen Arslantepe VIA und den frühsumerisch-hochprotoelamischen Stadtkulturen," in M. Liverani et al. (eds.), Studi di Paletnologia in onore di S. M. Puglisi, Rome: 229-236.



Return to Symposium Home Page