Invisible Exports in Aratta: Enmerkar and the Three Tasks more

29 Invisible exports in Aratta: Emmerkar and the three tasks by Irene Good Long-distance exchange is a subject of great interest to scholars of the ancient world. Interaction between different groups is a fundamental developmental factor in early state formation. Although economic texts offer concrete quotidian data, literary texts can shed light on the more social aspects of long-distance exchange. Textiles are a principal medium of exchange, not only because they can be transported as a ‘low bulk-high value’ commodity, but also because they carry deep social meanings; yet textiles are underrepresented aspects of the archaeological record. The story of Enmerkar and the Lord of Aratta illustrates the importance of textiles as a medium of exchange, and provides a backdrop for archaeological textile research. In a well-known Sumerian tale, Enmerkar, ruler of Uruk, wished to initiate trade with the land of Aratta to procure precious materials for rebuilding the temple of their patroness deity Inanna, at a time when established trade relations did not exist between the two polities (Fig. 29.1).1 Uruk therefore sent an emissary to the Land of Aratta to procure the precious stones and metals needed for this important project. The lord of this land first demanded that Enmerkar prove himself worthy of an exchange relationship by successfully performing three impossible tasks. The first task was to bring grain to Aratta, but in a net rather than a sack. Enmerkar sent grain in a net. He cleverly allowed the grain to sprout first, thereby enabling the grain to transported as a cohesive bundle. The second task was to make and bring to Aratta a sceptre made out of an unknown substance--not wood, nor metal nor stone. Enmerkar made a sceptre out of felt.2 The third task was to bring this lord a champion fighting dog. The dog was not to be black, white, yellow, red or spotted. Enmerkar brought the Lord of Aratta a champion fighting dog dressed in a little cloth coat. This story is instructive as a fable; yet it also offers insight for students of social history into the cultural, political and economic dynamics in western Asia in the 3rd millennium BC. There was a dissonance between these two polities, Uruk and Aratta, which was caused at least in part by a critical difference in their respective economic approaches.3 The Mesopotamian style was a market-based commerce, and the Iranian style was a more prestige-conscious exchange, evidenced by the behavior of the Lord of Aratta. It has been suggested that this discordance may even have been the catalyst for the advent of maritime trade in the second half of the 3rd millennium BC. Each of Enmerkar’s three impossible tasks in some way relates to textiles and fibers. The first challenge required a specific type of textile: netting, not woven cloth. The second challenge was met by using a felted material. The third challenge was resolved by an unexpected use of woven cloth.4 This story illustrates, through the medium of textiles, clever challenges and Fig. 29.1: Map of western Asia showing ancient geographical regions. The area designated ‘Turan’ is within the area proposed to be ancient Aratta [see Steinkeller 1984, 265 (Fig. 2)]. responses in a negotiation of access to precious commodities and of power. The underlying message is clear: not only are textiles a testament to inventiveness and creative thinking, but they also create currency for exchange between polities whose cultural barriers have otherwise hindered economic relations. But what can we see of this in the archaeological record? Where was Aratta? The location of Aratta and the environment of the Hilmand Basin The description of the landscape along the journey from Uruk to Aratta, taken from no less than four separate accounts,5 offers us clues about the location and also about the local environment of Aratta. We know from these stories that Aratta was a polity that had direct access to precious stones and metals within its borders, lapis lazuli in particular, and was graced with a city with a royal court. According to the Sumerian texts that mention Aratta, it is a place far away from Uruk. Between Uruk and Aratta were the seven mountains of Anshan, and then an empty plain to cross over once beyond the mountains. Of the precious stones to be brought from Aratta, lapis lazuli is of central importance, as it is only found in a few distinct deposits in the world. Several scholars have suggested that the location of Aratta would have access to the Badakshan lapis mines of northeastern Afghanistan.6 Certainly the access to exchange between Mesopotamia and this area was active during the mid-3rd millennium BC, as evidenced by the movement of chlorite vessels and lapis lazuli7 – particularly semi-processed stone at Shahr-i Sokhta in Iranian Seistan – and the culturally related sites Shortugai and Mundigak in northern Afghanistan, a region designated ancient Turan.8 The westernmost of the protourban settlement sites of this polity is Shahr-i Sokhta, and is therefore reasonable to be seen as the most likely center of what Sumerian texts refer to as Aratta (Fig. 29.1). Hansmann details the trek from Uruk to Aratta9 and argues convincingly that Seistan best fits the geographic description, and that Shahr-i Sokhta is the likely candidate for its capital.10 Much of the alabaster and lapis in processed and semi-processed state has been identified in the craftmen’s quarters at Shahr-i Sokhta. 170 The Afghan and Iranian Seistan region as a whole, and the Hilmand Basin in particular, has privileged access from the vantage point of lands to the West to numerous precious mineral and metal ore deposits including silver, from the highlands of the Pamir and Hindu Kush.11 The Dasht-i Lut Desert poses a formidable obstacle to travel from Mesopotamia to this region rich in minerals. However, by traversing its southern edge, along southern Anshan and the Makran, one may reach the highlands of Baluchistan, Seistan and Afghanistan. These regions are not only rich in accessible metal ores and other mineral resources, but also provide access to the Indian subcontinent and offer watercourses to sustain life in the foothills along the desert. The environment of Seistan is essentially a mosaic of dry lowland interspersed with small stands of riparian forest along major watercourses, and halophytic grasslands run along the deltaic fans from the piedmont runoff. This whole region is deprived of marine moisture because of the surrounding highlands and mountain ranges. The environment has not changed significantly over the past 6000 years.12 However local changes have been dramatic because of the deflection of watercourses, for example in the Rud-i Hilmand.13 The water table in this region is extremely shallow and rivers are very unstable, changing course frequently.14 There is, therefore, an abundance of archaeological sites from the Bronze Age through the Islamic period, well preserved due to the dry, salt-rich lithosol alluvial soils, exposed after the rivers have shifted course. The herbaceous vegetation around Lake Hilmand is typical of desertic and salt steppe, mainly comprised of reed marsh along the periphery of the lake, such as Juncus maritimus and Scirpus mucronatus.15 Basketry and reed impressions found in plaster from Shahr-i Sokhta attest the extensive use of reeds in antiquity. Salt crusts developed over the site at Shahr-i Sokhta, acting as a natural desiccant. These unusual edaphic conditions helped to preserve many organic materials. Textiles were recovered from several different contexts, from all four periods of occupation at Shahr-i Sokhta, spanning over a millennium.16 Most of the textiles recovered were from Periods II and III, but at least one textile fragment came from Period I. There is also a textile fragment from Period IV. Irene Good Shahr-i Sokhta emerged at the beginning of the 3rd millennium BC with signs of longdistance contact from the start, as evidenced by the proto-Elamite tablet found in the deep trench XDV, along with proto-Elamite seal impressions. Connections with southern Turkmenistan are also in evidence from ceramics. Although Period I is not thoroughly known as yet, it can be stated that the area of the site occupied at that time was about 10–20 ha. During Period II, there was a shift in settlement to the eastern part of the mound, known as the Eastern Residential Area. The site expanded to about 45 ha by this time. It is during Period II that craftworking was in such great abundance. The activities include semiprecious stoneworking and shell drilling, in particular. By Period III, the site had expanded its original size tenfold, with continuous settlement organization. Graves attest to a certain degree of social stratification, and there is evidence for the accumulation of wealth, particularly among females. Craft working seems to have been directly correlated with the accumulation of wealth. Cultural affinity or contact with other sites at fairly long distances is increasingly evident by ceramics which are related to types from Mundigak and the Quetta Valley. Wheel-made Buff Ware is introduced. Trends initiated during Period II continue during Period III. Period IV saw the decline and abandonment of the Eastern Residential Area. The period is known only from the Burned Building, erected in the southern boundary of the Eastern Residential Area towards the end of the previous period. Fire destroyed the structure, and what occurred afterwards was nothing more than squatter occupation. This pattern of abandonment seems to follow a general pattern of de-urbanization at the end of the 3rd millennium BC, as evidenced on other parts of eastern Iran and in Turkmenistan. An examination of the textiles and small finds from Shahr-i Sokhta related to textile manufacture has allowed a contextual view of textile production within the economy of 3rd millennium Seistan.17 Textile production was local and quotidian, evidenced by a wide distribution of spindle whorls and of combs at the site throughout all phases of occupation. Intact textile fragments are all plain woven, some of goat hair and some of sheep’s wool, and yet they represent a wide variation in thread Invisible exports in Aratta: Enmerkar and the three tasks spin and gauge, as well as in density of weave, demonstrating a wide variety in usage. Some are fragments of weft-faced slit tapestry; others are of warp-faced strap bands. Many textiles were in the form of cordage and nets of vegetable fiber. Textile and basketry information has also been gleaned from plaster and ceramic impressions. A number of textile fragments evidently were of the same form, yet derived from different loci; thus, demonstrating the presence of several cloths of the same cloth type. Fiber Sample 2b 3 4 5 12 21a 21b 22 24b Period (phase) IV III III II (6) II (6-7) II (6-7) II (6-7) II II II(6) 1B 1 II II II (5-6) II(6) II I III II II II II II II Prov. Textile Type Fiber Type 171 IUB G 112 SC XEU 1 XIR CCXIII XIE LVIII XIE/ XDYLXIX XIE/XDY LXIX 1B 1 RYL 2 XIL 4 LVIII 8 1B 4 1B 4 XIA XIL CCXXX 1B 4 G 129 SC XIE LVIII 1 XIE LVIII 1 XIE LVIII 4 RYL E. wall RYL 3 CCIV loose fiber net net fiber net textile textile twine sticks inside net rope cord fiber cord plied threads loose fiber mat textile rope/cord rope/skein rope/skein fibers rope/skein basketry unproc. Crotalaria Crotalaria Crotalaria cf. Crotalaria Crotalaria undet. bast undet. bast Corchorus unproc. Crotalaria Crotalaria Crotalaria Crotalaria cf. Crotalaria undet. bast Linum Crotalaria Scirpus Linum Crotalaria unproc. Crotalaria Crotalaria Crotalaria Crotalaria Scirpus Shahr-i Sokhta vegetable fiber samples. Vegetable fibers are generally more challenging to identify than animal fibers. Most plants contain fiber cells, usually bundled in the leaf or stem (bast), but sometimes as seed components. Fibers from an animal are selfcontained, whereas vegetable fibers are derived from a small part of a plant, and they may or may not be completely processed. Many other plant cells may be present in the sample, such as cork or outer (cambial) cell layers. These extraneous cells may or may not be diagnostic, though their presence informs us of the level of fiber processing. Phytoliths, parenchymal cells and other cellular extranea may assist in identification. Fiber cells are, however, the principal microscopic component to be reliably used in identifying the species of plant from which they derived. Twenty-four of the more than 100 textile and fiber specimens examined from Shahri Sokhta are of vegetable fiber. They are as follows: two samples of flax fiber were found, as well as 16 of sunn hemp and two of jute (Fig. 29.2). Scirpus reed was also in the vegetable fiber inventory and of the vegetable fiber samples examined only these four types were present. Scirpus reed was used for basketry and matting. It was undoubtedly procured locally, as it is today, for the same purposes.18 Jute (Corchorus) netting from a trash pit in sector XIE was identified. This plant and Linum usissitatum may have been grown locally, though it is unclear whether Linum was locally produced as a fiber. Flax seed has indeed been identified in modest quantities in the botanical remains from Shahr-i Sokhta,19 at Pirak in neighboring Baluchistan,20 at Tepe 25 26 27 28a 29 35 36b 37 58 69 70 72 73 74 75 Hissar on the Damghan Plain,21 and at late 3rd millennium BC Shortugai.22 The linen fibers from Shahr-i Sokhta were scutched down to fiber ultimate (Fig. 29.3), and were from highly spun, two-plied threads, qualitatively unique in the Shahr-i Sokhta textile inventory. A bast fiber of great interest, known as sunn hemp, has been identified in the form of net fragments in 16 out of the 24 vegetable fiber samples from the Shahr-i Sokhta textile inventory. This tropical plant, Crotalaria juncea L., is a fiber native to South Asia. This fiber must have been imported to Shahr-i Sokhta, as palaeoenvironmental and botanical studies of Seistan do not support the idea that it could have been grown locally. The sunn hemp samples were derived from several types of remains: five were loose fiber samples, five were of rope or skein fragments, and four Fig. 29.2: Table showing vegetable fiber fragments from Shahr-i Sokhta. ‘Unproc.’ = unprocessed, ‘Undet.’ = undetermined (adopted from Good 1999). 172 Fig. 29.3: Photomicrograph in polarized light of Linum usissitattum from Shahr-i Sokhta fiber sample # 35, from Period II, XIL 4. Irene Good were from netting. These samples therefore represent different modes of workmanship. The fiber is found only in the unprocessed or semi-processed state (i.e., extraneous cells are present and fiber bundles are intact) as rope or skein fragments, and in a more thoroughly processed state, it is found as netting (Figs. 29.4, 29.5, 29.6). Local reeds (Scirpus sp.) were useful for basketry and matting. They were not, however, adequate for making nets, not having the required flexibility or tensile strength. Flax was present, but not apparently used for making nets; and in fact linum was underrepresented in the vegetable fiber inventory. These facts lead us to suspect that the bast from the sunn hemp plant was an important fiber at Shahr-i Sokhta, having multiple uses, both as semi-processed cordage for packing material and as more highly processed nets. They may have even come into the site as nets and cordage to begin with, carrying commodities. The nets made of these fibers may have originated as containers for some other export, possibly from Makran or from Sind.23 Fig. 29.4: Net fragments of sunn hemp fiber from Shahr-i Sokhta, Period III, XEU 1. Enmerkar’s Net Fig. 29.5: SEM of net fragment sample fiber ultimate, identified as Crotolaria juncea Fig. 29.6: SEM of modern comparative specimen of C. juncea bast fiber bundle, showing ‘S’ helical twisting of ultimates. Long-distance exchange is a critical aspect of the social history of textiles. The fiber evidence from Shahr-i Sokhta just reviewed brings us a new relevance to the tale of Enmerkar and the Lord of Aratta, specifically regarding his first impossible task. The so-called ‘invisible exports’24 known from Mesopotamian texts (slaves, textiles and foodstuffs), have traditionally been notoriously difficult to detect in the archaeological record. Nets and cordage, though humble and often underappreciated, are crucial components of this movement of goods, in local exchange as well as far across borders.25 We can now attend to trace evidence of perishable artifactual remains such as these; as they are potential indices of imported commodities which are archaeologically difficult to recognize. Far from the land of Enmerkar, the presence of sunn hemp nets and cordage at Shahr-i Sokhta allows us to appreciate the significance of the archaeological recovery of textile fibers. Even more directly than texts, archaeological fiber evidence can reveal some of the ‘invisible exports’ that moved across western Asia during the 3rd millennium BC. Invisible exports in Aratta: Enmerkar and the three tasks 173 Notes 1 For full translation from the Sumerian, see Cohen 1973. This and related myths date to the Ur III period (between 2112–2004 BC), though they refer to the reign of Enmerkar, who according to the Sumerian King List, reigned ca. 2950 BC: see Jakobsen 1973, table II. Vanstiphout points out that Enmerkar belonged to the second dynasty after the flood, and was responsible for re-building the city of Unug (Uruk). Thus the Sumerians set this story in the ‘beginning times’: see Vanstiphout 2003, 1, 4, n. 11, 12. The translation is of a felted fibrous substance – ‘the remnants of cooked hides’ ,Vanstiphout 1991, 225. According to a more detailed interpretation of the text, he describes this material as shining, cured over ten years inside a hollow reed(!); perhaps as a sort of refined fiber-gelatin material. The important feature in the context of the story is that it is manmade, not natural, see Vanstiphout 2003, 81. Steinkeller notes that the reference is obscure, personal communication, 2002. Possehl 1986. Furthermore, the order of these three elements (netting, animal fiber/felt and cloth), follow the order of their invention: first, cordage and netting is made from vegetable fibers; then animal skins and their fibers are utilized; then finally loom-woven textiles are produced. As outlined in the tale of ‘Enmerkar and the Lord of Aratta’, ‘Enmerkar and Enushkeshdana’, ‘Lugulbanda and Enmerkar’, and finally ‘Lugulbanda and Mount Hurum’, see Vanstiphout 2003. 6 Tosi 1970; 1974; Hansman 1978; Pinnock 1988. 7 Lamberg-Karlovsky 1988; Lamberg-Karlovsky & Tosi 1989; Tosi 1970; 1974. 8 Amiet & Tosi 1983; see also Good 2004. 9 Hansman 1972; 1978; see also Steinkeller 1982, Fig. 2. 10 See Madjizade 1986 for an alternate view. 11 Pigott & Stech 1986. 12 De Marco & Dinelli 1977; Costantini & Tosi 1978. 13 Raikes 1983. 14 Fischer 1973. 15 Zohary 1973, 471; Costantini 1983; Costantini & Tosi 1978, 179. 16 Good 1999. The dates are: Period I (3200–2800 BC); Period II (2800–2400 BC); Period III (2400–2000 BC); Period IV (2000 BC). 17 ibid. 18 Costantini & Tosi 1978, 179. 19 Costantini 1977, 169–170. 20 Costantini 1979, 331. 21 Costantini & Dyson 1990, 57–58. 22 Wilcox 1991, 149. 23 Makran was the place of origin for the abundant Indian chank shells (Xancus pyrum), found worked in the Craftman’s Quarters at Shahr-i Sokhta, and may indeed have been contained in sunn hemp nets. 24 See Crawford 1975. 25 Indeed, Pinnock notes that at Ebla, a small cache of lapis lazui chunks were found bundled together, apparently in a container (a cloth sack? a net?!) that had completely disintegrated, Pinock 1988, 108. 2 3 4 5