Iskandar Tepa. Preliminary report for archaeological excavation in season 2018 (south Uzbekistan)

STUDIA HERCYNIA XXIV/1, 145–158 Iskandar Tepa. Preliminary report for archaeological excavation in season 2018 (south Uzbekistan) Ladislav Stančo– Odiljon Khamidov– Shapulat Shaydullaev– Petra Mrvová– Tatiana Votroubeková– Tomáš Bek– Matěj Kmošek ABSTRACT Following up on the previous extensive surface surveys in the valleys and steppe of the Kugitang Mountains Piedmont in 2016 and 2017, the Czech ‑Uzbek archaeological expedition conducted small ‑scale trial exca‑ vations at the site of Iskandar Tepa, which had been identified as asmall rural Greco ‑Bactrian settlement. The main aim of the project was to confirm the dating and to reveal its possible function in the fortification system of the period in the Bactro ‑Sogdian borderlands. This report offers basic information on the field work and its results in the 2018 season. KEY WORDS Hellenistic; Bactria; Sherabad District; Loylagan Valley; Excavations. INTRODUCTION AND TOPOGRAPHY OF THE SITE While working on the extensive surface survey of individual valleys in the steppe landscape of the Kugitang Piedmonts in the north ‑western part of the Sherabad District, southern Uz‑ bekistan, the Czech ‑Uzbek team lead by Ladislav Stančo and Shapulat Shaydullaev, detected on the 12th September 2017– among other previously unpublished sites– also asmall, but topographically well pronounced settlement of the Greco ‑Bactrian period (Stančoetal. 2017, 130–132). Special credit should be given to all members of the team, but especially to Anna Au‑ gustinová and Tobiáš Kolmačka, who brought the first significant surface ceramic finds from this site to the attention of the expedition. The presence of fragments of the typical table ‑ware shapes, so ‑called fish ‑plates, among other pottery forms, led us to acloser examination of the site and to its preliminary dating to the Greco ‑Bactrian period (Stančoetal. 2017). The site is situated in the close proximity of the Loylagan Valley, north of the town of Sherabad (Pl. 4/1), between the modern villages of Ishtara and Loylagan, ca. 2 km to the north of the latter, and close to (and within sight of) the asphalt road connecting the two aforemen‑ tioned villages (Pl. 4/2). This road forms aborder between the Baysun and Sherabad Districts of the Surkhan Darya province, thus the site belongs to the northernmost margin of the latter. The site is located at an altitude of ca. 830 m.a.s.l. and is situated on the summit of anarrow promontory extending eastwards from an elevated ridge ca. 20 m above the bottom of aflat, narrow (ca. 500 m) and elongated valley, which itself stretches for ca. 8 km from north ‑west to south ‑east. Taking into account the most densely distributed pottery scatter upon the summit of the hillock along with linear traces of– or rather presumed debris / remains of– stone architecture, we considered this part to be the original settlement core. This part measures ca. 50 × 30 m only, while the entire area of the summit that forms an elongated E ‑W oriented oval, measures ca. 146 STUDIA HERCYNIA XXIV/1 260 × 40 m. It should be mentioned that abundant pottery fragments were also found all over the steep northern and southern slopes of the promontory down to the bottom of the valley hinting at 1) the intensive use of the site, and 2) substantial erosion moving the finds far from their original place of deposition, caused at least partially by the modern use of the summit for agricultural and grazing purposes attested by numerous finds of recent metal objects, as well as oral tradition. At this point, the name of the site deserves attention. Lacking alocal name at first, the site received aworking nick ‑name of ‘Iskandar Tepa’ meaning simply the hill of Alexander, which we kept even after we had learned that this site might be with some probability identical to asite called Tura Tepa that was mentioned briefly in asurvey report of the 1990s. Very scarce information on Tura Tepa and especially its dating to the ‘possibly Kushan and Early Medieval period’ does not allow us to be sure of this identification (Bobokhojaevetal. 1990, 26, 33, 35). The placing of the site on avery general map presented in the publication (Bobokhojaevetal. 27, fig. 1) does not help to resolve the question, since the Tura Tepa ‑mark on the map is placed to the north of that of Kulal Tepa, which is in fact situated to the north ‑east of Iskandar Tepa. Sverchkov, who obviously did not verify Bobokhojaev’sinformation, even placed the site of Tura Tepa to the south of the village of Khatak, but at the same time the text speaks of it as being situated ‘next to Kulal Tepa’ (Sverchkov 2005, 18). ARCHAEOLOGICAL RESEARCH EXCAVATION Immediately after its discovery in September 2017, the site became subject to atrial small ‑scale excavation (on the 16th of September) that has been briefly described in areport and hence there is no need to repeat the information except for the statement that the Hellenistic dating was only corroborated by the abundant ceramic material from the single test trench (Stančoetal. 2017, 130–132). The trench, on the other hand, did not reveal any traces of architecture. As asecond step, the surface of the site was briefly surveyed on May 2, 2018 by amember of our team, Tomáš Bek, using ametal detector in order to gain chronologically sensitive metal objects, including coins in case there were any. The single ‑detector survey lasting for four hours allowed only for sampling with no ambition to cover all the surface or to gain afull picture of the situation. As aresult, 11 metal objects were detected, which are listed in Tab.1 below (nos. 283–294). Asubsequent preliminary observation of the finds revealed that 7 out of the 11 objects might be dated to the Hellenistic period, including three coins (nos. 284, 287, 289), although only no. 287 has so far been securely determined as acoin of DemetriosI. Following the trial excavations, the Czech ‑Uzbek team started aregular investigation of the site core and detailed surface survey of the surroundings of the site in the early Autumn of 2018. The fieldwork lasted from September 14 till October 7, i.e. 21 working days. Ladislav Stančo, Odiljon Khamidov, Shapulat Shaydullaev, Petra Mrvová (Cejnarová), and Tatiana Votroubeková took part in the field work, assisted by six local workers from the village of Loylagan. As an excavation method, achessboard structure of the square trenches was applied with trenches measuring 4 × 4 m each (Fig. 1). In the end, 10 such squares were opened in the central part (Pl. 4/3, 4/4), while two additional trial trenches were excavated in the eastern (two squares: 10C and 11C) and western (one square) parts of the promontory in order 1) to determine the extent of the settlement, and 2) to uncover places with significant stone accumulation resembling structures. The process and the final results of the excavation works were documented by photography, photogrammetry from the ground, and by drone. 147L. STANČO– O. KHAMIDOV– S. SHAYDULLAEV– P. MRVOVÁ– T. VOTROUBEKOVÁ– T. BEK– M. KMOŠEK Code Material Object 283 copper alloy object 284 copper alloy coin 285 copper alloy arrow head 286 copper alloy object 287 silver coin 288 copper alloy coin 289 copper alloy coin 290 copper alloy coin 291 copper alloy decorated sheet 292 copper alloy ring 294 copper alloy pin 501 lead lump 502 copper alloy arrowhead 503 lead bar 504 copper alloy coin 505 copper alloy coin 506 copper alloy sheet 507 gold/silver decorated object (earring?) 509 copper alloy sheet ‑ perforated 510 brass button 512 copper alloy arrowhead 513 copper alloy arrowhead 514 lead stick 516 copper alloy coin 517 copper alloy sheet 519 copper alloy belt implement 520 copper alloy arrowhead 521 copper alloy coin 522 copper alloy coin 523 copper alloy object 524 copper alloy coin 525 silver coin 526 copper alloy applique 527 copper alloy coin 528 copper alloy fingerring 529 copper alloy coin 530 copper alloy coin 531 copper alloy sheet (mirror?) 532 copper alloy arrowhead 533 copper alloy coin 534 copper alloy hemisphere (sheet) 535 copper alloy arrowhead 536 copper alloy alloyed object Code Material Object 537 copper alloy bell 538 silver coin 539 copper alloy coin 540 copper alloy arrowhead 541 copper alloy head of implement 542 copper alloy coin 543 copper alloy sheet 544 copper alloy coin 545 copper alloy perforated sheet 546 copper alloy coin 547 copper alloy sheet 548 copper alloy coin 549 copper alloy sheet (perforated) 550 copper alloy coin 551 copper alloy coin 552 copper alloy coin 553 copper alloy coin ? 554 lead metal fragment (Button?) 556 copper alloy fingerring 557 copper alloy coin 558 copper alloy metal fragment 559 copper alloy coin 560 copper alloy coin 561 copper alloy coin 562 copper alloy coin 563 copper alloy/silver coin 586 copper alloy sheet 587 copper alloy sheet 588 copper alloy sheet (?) 589 copper alloy fingerring 1001 stone saddle Quern (fragment) 1002 stone saddle Quern (fragment) 1003 bone bone (human?) 1004 ceramic / terracotta ibex head 1005 Bone bone 1006 stone saddle Quern (fragment) 1007 copper alloy bronze sheet 1008 bone bone (human?) 1010 stone saddle Quern 1012 stone saddle Quern (fragment) 1013 bone bones (22 pieces) 1014 stone saddle Quern Tab.1: List of small finds from Iskandar Tepa. 148 STUDIA HERCYNIA XXIV/1 Fig. 1: Site -plan of the central part of Iskandar Tepa excavations (drawn by T. Votroubeková). 149L. STANČO– O. KHAMIDOV– S. SHAYDULLAEV– P. MRVOVÁ– T. VOTROUBEKOVÁ– T. BEK– M. KMOŠEK Generally speaking, the preservation of the cultural layers of the site was very poor, at least in terms of their volume. In the central part (Squares nos. 1B, 1D, 1E, 1F, 3B, 98D, 98E, 99D, 99E), the subsoil was reached in most of the excavated trenches, but the depth of the trenches and thickness of cultural layers varied between 10 and 30 cm only, rarely slightly more. The only significant features unearthed were rounded pits dug into the subsoil, interpreted as intended for the placement of large storage jars or pithoi, the lower parts of which were preserved in most cases. As arule, aca. 10–15 cm thick erosive subsurface layer had to be removed to reach either agreyish dusty layer or an accumulation of small stones and pebbles that do not form any discernible structure or its remains. In some parts, especially in Square 2E, this pebble/ stone layer seems to have formed acompact stratum resembling aplatform (Fig. 2). Besides afew more regular (ashlar ‑like) stones, saddle querns were discovered in this stone accumu‑ lation. What we call aplatform here might very well be the only remains of substructures belonging to the unpreserved building above. The data on the above ‑mentioned pits with well ‑preserved lower parts of storage jars (pithoi) are summarized in Tab.2. Typically, the upper diameter of the preserved portion of the storage jar body reaches ca. 65–72 cm, while the preserved depth of the jars varies between 30 and 50 cm, with the exception of no. 4. In this case, the pithos was dug in amuch shallower pit– or might be dug in adifferent period than the remaining ones. The bottoms of the storage jars are in some cases (Tab.2: 4) covered with agypsum ‑like layer or analogical fragments of broken pithoi are accompanied with gypsum ‑like (lime ‑stone like) blocks (Tab.2: 1, 5, 6; Fig. 3–4).1 Such blocks or layers have analogies in similar contexts at the site of Kurganzol, where these are– along with another pottery form– interpreted as forms for making milk products such as soft cheese (Sverchkov 2013, 108–109, fig. 91: 5–6). In 1 The size of the gypsum ‑like blocks corresponds well with the sizes of the pithoi. In the case of pithos no. 1, there are two blocks measuring 35 × 26 × 8 cm and 34 × 23 × 9 cm, pithos no. 5 was accompanied by asimilar gypsum block of 37 × 20 × 13 cm. The pit no. 6, where the existence of apithos is only hypothetical, agypsum block was found ca 60 cm below the surface and measured 13 × 17 × 7 cm. Fig. 2: Stone platform, square 2E, detail (photo by L. Stančo). 150 STUDIA HERCYNIA XXIV/1 No. Context no. Square no. Diameter of the storage jar Preserved depth (max.) Wall thickness Note 1 01F 01F 65–68 cm 30 cm 1.5–1.8 cm Small stones stacked around the body of the jar 2 99D003/006 99D 69 cm 34–39 cm 1.8–2 cm Pit dug into the subsoil; no stones around the body 3 99D004/007 99D 98D 98E 99E 67 cm 40–49 cm 1–2 cm Pit dug into the subsoil; stones stacked around the body of the jar 4 99E004/005 99E 56–57 cm 18–20 cm 1.8–2.2 cm Pit dug into the subsoil; no stones around the body; bottom covered with a white limy / gypsum layer 1–3 cm thick 5 01F005/007 01F 72 cm 50 cm 2 cm Pit is dug into the subsoil 6 01F006/008 01F 60–90cm 60–75 cm ?? The pit might have a different purpose; filled with stones, animal bones, pottery fragments, limy / gypsum objects Tab.2: Overview of the storage jars (pithoi) excavated in the core of the site. our case, we do not share this interpretation since large storage jars, the more so those firmly embedded in the floor, do not seem to be suitable for such processes. Anyway, samples of these blocks were taken in order to analyse their composition and to determine their function or process of origin. The preservation of the storage jars including their depth gives aclue to the explanation of the preservation of the site as awhole. First, we should expect the storage jars to be placed in the interior of (a) building(s). In the interior, one would expect either afloor or alevel of functional ground, anyway afeature to be easily recognised in the archaeological record. None has been unearthed during the excavations, thus we surmise that these were heavily affected by erosion caused by wind and water. Second, the lack of stone debris both on the summit and under the hill, points to the use of other building materials, most probably mud ‑bricks with wood as aminor material. We cannot exclude the usage of yet lighter structures of amobile character, such as tents, but the amount of pottery, especially of the storage vessels testifies rather to more permanent settlement structures. To sum up, we believe that on the hillock of Iskandar Tepa, there once stood small mud ‑brick buildings, which were inhabited by people involved in small ‑scale processing and the storage of grain or other food, be it for themselves or for other people living in the neighbourhood. Of these small structures, only tiny parts remain to be studied today, i.e. those that were dug into the ground such as storage jars. The fact that no structures– even the foundations of buildings– survived to the same level un‑ derground as the pithoi means that the structures were very modest. Judging from this and from the site topography– especially in comparison with other contemporaneous sites, the existence of fortifications on the hillock can be excluded. 151L. STANČO– O. KHAMIDOV– S. SHAYDULLAEV– P. MRVOVÁ– T. VOTROUBEKOVÁ– T. BEK– M. KMOŠEK SMALL FINDS Throughout the autumn field season, the metal detector, operated by M. Kmošek, L. Stančo, or L. Damašek, was employed in the excavated units and related soil ‑waste in order to gain data directly from the contexts, but also on the slopes of the hillock and at the bottom of the valleys to the north as well as to the south of the site. As aresult, 74 out of atotal of 85 small finds from the site are metal objects, among them 32 coins, eight arrow heads, and three finger rings (Tab.1: 501–1014). Metal artefacts were analysed on their surface by aportable spectrometer Olympus Delta Professional. Three objects are made of silver (coins), one of gold (ear ring), while the majority– 60 finds– are made of copper alloys. It should be noted that Fig. 3: Gypsum -like layer in the bottom part of the storage jar (photo by L. Stančo). Fig. 4: Gypsum -like blocks from the storage jar no. 1 (photo by T. Votroubeková). 152 STUDIA HERCYNIA XXIV/1 almost all of these metal objects were found outside of the trenches strata, i.e. on the surface of the site. Thus, we lack the proper context information on these, and therefore they have to be dealt with per se typologically without implicit correlation with the excavated units. We interpret them very cautiously. Aselection of these small finds is shown in Fig. 5. The full publication of the metal finds including in ‑depth analysis of the elemental composition is currently under way. Besides the coin of Demetrios Imentioned above, one more coin of the same ruler was found, as well as two of EuthydemosI. The earliest coin found at the site so far belongs to Diodotos II (Stančoetal. forthcoming). Among only eleven non ‑metallic small ‑finds, there are six fragmentary stone ‑made saddle‑ ‑querns (Fig. 6). Their presence is typical for rural sites across southern Central Asia prior to the Kushan period, when the new and more effective rotary querns occurred (Stančo 2018). The fact that the latter are completely absent at the site corroborates our assumption that the period of occupation did not exceed the Greco ‑Bactrian period. POTTERY ASSEMBLAGE CHARACTERISTICS The ceramic collection from the Iskandar Tepa 2018 excavation (Fig. 7), which comprises 2449 sherds with an overall weight of 126,653 grams, has been described and analysed separately and is currently being published elsewhere. Aselected part of the pottery material has been subject to detailed analyses: the petrographic, mineralogical, and chemical composition of the ceramics was characterized using thin ‑section petrographic analysis through optical micros‑ copy (OM), powder X ‑ray diffraction (XRD), and wavelength dispersive X ‑ray fluorescence (WD ‑XRF) (Ferrerasetal. forthcoming). SITE INTERPRETATION The elevated position above the fertile valley and proximity to one of the few passages through the mountains of Kugitang allow for the preliminary interpretation of this site as an outpost or watch ‑post serving the strategic needs of the rulers of the Greco ‑Bactrian kingdom along with the fortresses of Uzundara (which is actually visually interconnected being only afew kilometres away), Kurganzol, Kapchigay, and the Darband Wall. Due to the aforementioned thin cultural layers preserved at the site, we may assume that the settlement was inhabited for afew decades at the most. Consequently, all the archaeolog‑ ical material gained from the site– including the pottery assemblage– belongs to arelatively short period of time roughly determined by the discovered coins of Diodotos II, EuthydemusI, and DemetriosI. Considering the settlement development in the valley (see description and location of Kulal Tepa above), one can speculate about earlier defensive needs of the inhabitants of Iskandar Tepa that gave way to the rather more complacent mood of the people settling here in the Kushan period, who were apparently protected by different means or felt generally safer from harm than their Greco ‑Bactrian predecessors. ARCHAEOLOGY OF THE ISKANDAR TEPA VICINITY The above ‑mentioned site of Kulal Tepa (clearly visible on the map, Pl. 4/2) represents the only settlement site in the close vicinity– ca. 300 m from the top of Iskandar Tepa to the northeast– that was obviously inhabited for alonger period of time in Antiquity. Almost square in its 153L. STANČO– O. KHAMIDOV– S. SHAYDULLAEV– P. MRVOVÁ– T. VOTROUBEKOVÁ– T. BEK– M. KMOŠEK ground plan (80 × 80–90 m) and oriented according to the cardinal points (with aslight incli‑ nation towards the northeast in its south ‑north axis), the site was founded at the very bottom of the flat valley, giving up any defensive advantages of the neighbouring hillocks and being, at the same time, well suited for agricultural activity as well as for awater ‑bringing system. Surface pottery finds reveal the active period of the site to be in the Kushan and Kushan‑ ‑Sasanian period, which is attested also by the newly detected coin finds from around the site Fig. 5: Small finds selection (drawn by M. Kmošek). 208 STUDIA HERCYNIA XXIV/1 Pl. 4/2: Iskandar Tepa and its environs (by L. Stančo). 209PLATES Pl. 4/3: Aerial view of the excavation in the central part of Iskandar Tepa. Pl. 4/4: Central part of Iskandar Tepa excavations, photogrammetry (by T. Votroubeková).