Terra multiplex et varia natura. On the settlement patterns of Bactria in the Hellenistic Period
Abstract
9
Full text
Terra multiplex et varia natura. On the settlement patterns of Bactria in the Hellenistic Period Jakub Havlík ABSTRACT The proposed study aims to investigate the character of settlement transformation in ancient Bactria (upper Amu Darya basin) during the period following the conquest of Alexander the Great. This transformation is addressed by the comparison of settlement development in the two regions of Bactria, its north -western part in present -day Uzbekistan and the so -called Eastern Bactria in present -day Afghanistan, based on archaeological evidence. Attention is devoted mainly to the quantitative analysis of settlement sites attributed by previous researchers to the Achaemenid and Hellenistic periods on the one hand, and the evaluation of settlement hierarchy and spatial distribution changes on the other. The conducted analysis illustrates amassive settlement abandonment posterior to the fall of the Persian Empire. However, the results of the study suggest that the Hellenistic eastern Bactria, commonly associated with ahigh level of involvement of the new elites coming from outside, also exhibited many traits of structural continuity with the preceding period represented by ageneral settlement dispersal and reutilization of both, previous fortified centres and irrigation networks. On the other hand, interest in fortification and the settling of new areas at higher altitudes are clearly characteristic of the Hellenistic Bactria as awhole. KEYWORDS Central Asia; Bactria; Achaemenid Empire; Seleucid Empire; Hellenism; Settlement Archaeology; Settlement patterns. INTRODUCTION The fall of the Achaemenid Empire and the subsequent period of Graeco -Macedonian rule strongly influenced the socio -political situation in avast area stretching between the Eastern Mediterranean and far -distant regions of Central Asia. Despite its only marginal link to the Mediterranean, Bactria, the eastern outpost of the Achaemenid Empire, which developed in the context of alocal idiosyncratic cultural milieu, came after Alexander the Great’sconquest into immediate contact with Greek culture and became part of the newly arising Hellenistic world. The relative lack of ancient written sources (See Holt 1988; 1999) on the Achaemenid and Hellenistic period Bactria1 promotes archaeology to be afundamental source of knowledge on this region. The impact of the Hellenistic rule on the local material culture has been 1 For the purposes of this paper, Bactria is defined as the upper Amu Darya basin (cf. Gardin 1985; Leriche 2007). Delimited by Hissar, Pamir, and Hindukush and by the Karakum desert, the upper Amu Darya basin is aregion well defined not only in terms of geography, but also by alongstanding cultural development that dates back to prehistoric times (Dyakonov 1954, 123; Kuz’mina 1976; Gardin 1985). The precise historical meaning of the term Bactria has not been satisfactorily determined so far (Stride 2007, 99); however, it could evolve over time or be polyvalent: ‘Bactria can mean geographical region, administrative area or simply the oasis of Bactra’ (Leriche 2007, 125). On the recent contributions to historical geography based on both written and archaeological sources, see Rtveladze 2002; Rapin 2013; Gorshenina– Rapin 2021, 197–202. STUDIA HERCYNIA XXV/2, 9–41
10 STUDIA HERCYNIA XXV/2 extensively examined in various forms, 2 nevertheless it is settlement archaeology, 3 which has the great potential to reveal essential data for the reconstruction of the historical development of the Hellenistic Bactria, its internal organisation and land use. The study of ancient settlement patterns and forms within the context of the cultural landscape was undertaken rather rarely during the twentieth century in Bactria. The emphasis was often placed on the detection and excavation of isolated sites, with only limited interest in their wider surroundings.4 The study of spatial and functional relations and patterns beyond asimple placement on the map was conducted rather seldom.5 An important milestone in such studies is represented by the French survey of eastern Bactria led by J.-C. Gardin between 1974–1978 (Gentele 1989; Lyonnet 1997; Gardin 1998), oriented towards the examination of both the close and wider hinterland of Ai Khanum. The current state of archaeological research in Central Asia allows, in selected areas, an analytical evaluation of the large data 2 Cf., among others, papers in Kouremenos– Chandrasekaran– Rossi (eds.) 2011 or Lindströmetal. eds. 2013. 3 Settlement archaeology as ‘astudy of social relationship using archaeological data’ (Trigger 1967, 151). 4 On the reconnaissance missions in Bactria, see for example Rtveladze 1974; 1976; 1987 (Southern Uzbekistan); Fischer 1967; Kruglikova 2005 (Northern Afghanistan). 5 E.g., Pugachenkova– Rtveladze 1990, 62–82. Fig. 1: Bactria. Location of sectors II and IV. Based on Gardin 1985 (basemap: Esri).
11JAKUB HAVLÍK assemblages, which is, however, employed so far only rarely. In his unpublished dissertation, S. Stride examined in detail the archaeological geography of the Surkhandarya province in southern Uzbekistan from the Bronze Age to the Late Middle Ages and demonstrated the– up to that point unexploited –potential of such an approach. However, despite the recent rapid progress of modern landscape studies in southern Central Asia,6 the state of research on the settlement patterns of the Hellenistic period Bactria is, still far from sufficient.7 This paper addresses the question of settlement transformation between the Achaemenid and Hellenistic period Bactria (ca. middle of the 6th century– 4th century BC and the end of the 4 th century– middle of the 2 nd century BC respectively). The variance of this transformation is demonstrated in this study by the example of the two sectors of Bactria (Fig. 1), as defined by Gardin (1985, 39), Sector II (so called ‘Eastern Bactria’, i.e., the basins of Kunduz and Kokcha in present -day Northern Afghanistan) and Sector IV (Surkhan Darya and Sherabad Darya basins, corresponding roughly to present -day Surkhandarya province of Southern Uzbekistan). Synthesizing the results of archaeological investigations conducted so far in the area, the main focus of this paper is on asynchronic and diachronic quantitative comparison of the settlement structure of the periods in question. The structural transformation is demonstrated by the changes in the settlement hierarchy and the distribution of settlement sites within the designated size categories (Tab. 1; see below). In addition to quantitative characteristics, the identified qualitative traits (especially positioning, elevation, and signs of fortification) are taken into account. By means of this evaluation, attention is paid to the continuity and discontinuity pattern and possible transfer of settlement strategies between the Achaemenid and Hellenistic periods. 6 Stride 2005; 2007; Salvatori– Tosi eds. 2008; Buławka 2017; Mantellini 2018; Stančo– Tušlová eds. 2019. 7 The settlement structure and its transformation after the Greek conquest of Bactria has been– based on archaeological evidence– put in account only briefly as apart of broader general studies (Staviskiy 1977; Pugachenkova – Rtveladze 1990; Stride 2004) or alimited number of summarizing articles (Koshelenko– Serditykh 1987; Leriche 2007; Lindström 2021; Martinez -Sève 2021; Stančo 2021a). Class Site size Number of sites I 0.002–0.47 ha 208 II 0.49–1.3 ha 57 III 1.4–2.5 ha 30 IV 2.7–4.36 ha 22 V 4.9–7.8 ha 11 VI 8.4–12.5 ha 9 VII 14.0–20.0 ha 8 VIII 25.7–49.4 ha 8 IX 100.0–160.0 ha 3 Tab.1: Site size classification according to Jenks optimization method. The calculation was performed on the whole dataset of sites throughout Bactria attributed to the Achaemenid and/or Hellenistic periods known to the author.
12 STUDIA HERCYNIA XXV/2 The proposed study is based on the analysis of the compiled archaeological dataset8 containing 544 sites from all over Bactria known to the author 9 dated from the Achaemenid and/ or Hellenistic period. The evidence comes from archaeological investigations, both surface surveys and excavations of various extent, undertaken in the studied area since the first half of the twentieth century. Instead of the rather intuitive site typologies employed so far, 10 Jenks optimization method is employed in this study in order to achieve amore objective classification of the studied sites in terms of their size (Tab.1). This basic and common cluster analysis method determines the best arrangement of values into apreselected number of different classes.11 All the known sizes of the 356 sites (out of the 544 sites gathered altogether)12 were used for the calculation, which allows for agood synchronic and diachronic comparison of both periods. Nine classes (I–IX) enable amore detailed classification of numerous sites of area up to 1 ha. These small sites represent in total more than 72% of the classified assemblage.13 By its very nature, this classification is used as aworking tool for the purposes of this study and does not represent any attempt to establish anew settlement typology at this point. The aim of this paper is not to present acomplete image of the historical settlement of Bactria with all its particularities, but to outline selected general trends and contribute to abetter understanding of the Hellenistic period’ssettlement. The proposed comparison provides aclear illustration of an unequal settlement development in the two regions concerned and gives supporting evidence of arelatively high degree of structural continuity between the Achaemenid and Hellenistic eastern Bactria. 8 The dataset originates in the updated database employed in the author’smaster’sthesis (Havlík 2018). 9 Throughout this article, asingle area of past human habitation activity delimited in space (usually asingle tepa) is considered asettlement site. Some of the sites considered asingle settlement in the source literature are therefore in this study perceived as multiple sites. This perception does not contradict the assumption that these sites could at some point form asingle settlement unit. However, to confirm such ahypothesis, further analysis is required. 10 See Staviskiy 1977, 44; Pidaev 1978, 17 (Kushan period); Koshelenko– Serditykh 1987 (Hellenistic period); Stride 2004, 68–69, 292–294. However, these typologies are based above all on site surface area, and usually apply terms such as ‘avillage’ or ‘acity’ regardless of evidence of aparticular function. See the criticism of such afunctional attribution by M. Negus Cleary (Negus -Cleary 2013), who points out that– especially in Prehistoric and Ancient Central Asia– size and urban function of settlement are not necessarily related together. 11 This method is also known as ‘Jenks Natural Breaks Classification Method’. Required ‘natural’ arrangement of values into different classes is achieved by minimization of each class’saverage deviation from the class mean, while maximizing each class’sdeviation from the means of the other groups (Jenks– Caspall 1971; Conolly– Lake 2006, 142). 12 Where the size of the settlement changes over time, the larger size was taken into account for the purpose of the calculation. Wherever possible, the site was measured using the GIS. The projected coordinate system WGS 1984 UTM, Zone 42N offers in this case insignificant deviations. Where the direct measurement was made impossible by the current state of preservation, published values were employed. 13 Sites of unknown size have been taken into consideration in graphs as well. ‘Sites’ identified as artefact scatters that indicate the presence of loci of activity or sites destroyed by subsequent activities also fall into this category (e.g., Tušlová 2019, 71–79).
13JAKUB HAVLÍK LIMITS OF THE STUDY Ancient Bactria extends into the territories of four present -day states (Afghanistan, Turkmenistan, Uzbekistan, and Tajikistan), which experienced different political and economic development as well as employing differing archaeological approaches. In view of that fact, the state of research differs remarkably in the individual regions: Archaeological investigations of eastern Bactria14 took place in arather sparsely populated, much better -preserved cultural landscape, where remains of past human activities were well distinguishable and where the prehistoric irrigation network was still partly in operation during the 20th century (Gentelle 1978). In Surkhandarya province in southern Uzbekistan, as well as in other parts of Soviet and post -Soviet Central Asia, archaeologists have often faced alandscape changed beyond recognition by years of mechanized agriculture intensification (see Stride 2004, 130–132; Mantellini– Berdimuradov 2019). The restructuring of fields and irrigation systems and the foundation of new settlements led to the destruction of many archaeological features, settlement sites among them, all over the Surkhandarya province lowlands. This destruction took place prior to the initiation of systematic extensive reconnaissance surveys in the 1960s–1970s (cf. Rtveladze 1974; 1976; 1987). This situation affects an archaeological image of the past occupation of the area in question and makes astatistical comparison of both regions rather complicated. On the other hand, the longstanding favourable political situation in Uzbekistan has allowed the activities of numerous archaeological projects in the last forty years when research in Afghanistan has been possible only under very difficult circumstances.15 There is agreat qualitative difference between the evidence coming from particular sites: typical tepa sites (tells) are multi -period in their character as arule. Only 27% of the studied sites throughout Bactria are not superimposed by later (i.e., post -Hellenistic) occupation layers, while only 12.5% of sites from the dataset were excavated to some degree.16 This situation has agreat impact on data validity and also the accuracy of chronological attribution. Surface surveys in Central Asia offer only rarely more precise dating than those at the level of archaeological cultures or wide historical periods. Sites contemporaneous in terms of material culture revealed by the survey may not necessarily have been occupied at the same time. The image of settlement development is therefore inevitably flattened. Stratigraphical superposition poses adifficulty for this study also in terms of the estimation of the spatial extent of the studied sites and the assessment of their possible morphological traits. In the case of small sites with no signs of fortification, no major differences can be expected throughout the occupation span. However, in large multi -period sites, an appropriate appraisal of the character of their individual phases becomes problematic without an extensive excavation. Following Gardin’sapproach (Gardin 1995; 1998), in the case of multi -period sites where the precise morphology of aparticular period has not been confirmed by an excavation to date, the characteristics of the current site were taken into account for the calculation. Regarding the attribution of archaeological sites to the particular periods in question, it has already been discussed extensively (Francfort 2005, 316–324; Lhuillier 2018) that there have been many uncertainties with the precise identification of the ‘Achaemenid’ period within 14 Archaeological data on this sector are almost completely limited by the results of asurvey led by J.-C. Gardin between 1974 and 1978. See, however, also Fischer 1961; 1967 and Kohl 1978. 15 See for example Stančo 2019, 29 for an evaluation of the progress in site recognition in Sherabad oasis during the last fifty years. 16 The degree of excavation stretches from trial trenching to uncovering the whole settlement. Remarkable differences can be observed, especially in the case of aged projects, or in the quality of particular excavations as well.
14 STUDIA HERCYNIA XXV/2 the conservatively developing local cultural complex of Yaz II–III (Middle to Late Iron Ages). Pottery assemblages of these phases were sometimes automatically and wrongly labelled as corresponding to the Persian rule in Central Asia (Lhuillier 2018, 258).17 While according to the current state of research, the Yaz III phase definitely included the Achaemenid period (see the elucidating discussion in Lhuillier 2018, 258–260; cf. Mokroborodov 2015), pottery shapes of this phase were highly probably still in use even after the Seleucid takeover of the area (cf. Lyonnet 1997, 148–149; Lyonnet 2012). At this point, the only way to compare the Pre -Hellenistic settlement situation (i.e., the situation directly preceding the Greek rule in Bactria) with the settlement of the Hellenistic period for the purposes of quantitative analysis, is to confront the assemblage of sites yielding the material culture of the Yaz III phase18 with those sites which yielded the material culture of the Hellenistic period. Well aware of the outlined precautions, the term ‘Achaemenid period’ is used throughout this study as corresponding to the Yaz III phase in terms of material culture. With respect to the described limits, the ideas developed in this article should be understood as possible tendencies of the examined settlement transformation, reflecting the current state of archaeological research. The large -scale comparison of site assemblages as awhole within the wider chronological frame, employed predominantly throughout this study, indeed simplifies the highly complicated processes. Still, being aware of this fact, this approach has the potential to identify general trends in the past settlement strategies. SECTOR II: VALLEYS AND PLAINS OF EASTERN BACTRIA THE PRE ‑HELLENISTIC SETTLEMENT STRUCTURE The so -called eastern Bactria (Pl. 1/1) consists of agroup of plains and mountain valleys, the extent of which naturally delimits the individual settlement zones situated predominantly on the alluvial plains along the Panj left bank tributaries, including among others the Kunduz and Kokcha Rivers. In total, 78 sites were identified and attributed by B. Lyonnet (1997) to the Iron Age PhaseIII.19 They are rather sparsely distributed on the irrigated plains and in the mountain valleys. Based on the presence of aseries of small sites along the river streams flowing through the narrow valleys, acertain degree of regional -level interconnectedness can be supposed. An exceptionally dense settlement distribution emerges only on the plain of Dasht -iQala, dominated by the fortified site of Kohna Qala near the confluence of the Panj 17 Ceramic material from agreat proportion of the studied sites has not been revisited recently, nor even published at all, which can clearly influence the general image of the period’ssettlement considerably– afact of which the author is well aware. 18 Cf. Masson 1959 and the Iron Age Phase III in Lyonnet 1997, 112–119. The original chronology of Masson was replaced in the recent past by alower chronology in favour of amuch larger timespan of Yaz I–III cultures (for asummary see Lhuillier 2018, 258–259). 19 The sites were gathered in accordance with the attribution of B. Lyonnet (1997, 102–119), who distinguished three -phase -based dating with Phase III roughly corresponding to the Achaemenid period, and not the broader two -phase -based dating proposed by Gardin (1998, 21–25). The discrepancy between the 68 or 70 sites attributed to Phase III by B. Lyonnet (1997, tab.XVI:2) and the 78 sites proposed for sector II in this article is due to the revised counting of sites, where– unlike the source– one single tepa corresponds to an individual site. Similarly, there is adiscrepancy between the number of sites of the Hellenistic period. The final number also includes the results of the survey of P.L. Kohl (1978) and excavation of H.-P.Francfort (2013) in the Dasht -iQala plain.
15JAKUB HAVLÍK and Kokcha Rivers. The majority of the sites in sector II (see Fig. 2) are small, having an area up to 0.47 ha (class I; 39.7%),20 while there are 16 sites (20.5%) exceeding 1 ha in size and only 7 sites (8.9%) exceeding 4 ha (class Vand above). Within the identified Iron Age Phase III occupation, only avery small proportion (7 sites– 9%) of sites bear signs of fortification in sector II.21 All of them exceed an area of 4 ha. The two largest ones, Kohna Qala (ca. 30 ha) and Bala Hisar of Kunduz (ca. 29 ha), fall even into class VIII.22 On the other hand, the two remaining large sites (Dasht -iQala no 448 and site no 413: classes VII and VIII respectively) did not reveal any sign of fortification and seem to be in this period rather open villages or continuous settlement areas.23 The seven fortified sites are located, each in adifferent plain or amountain valley. The only exception is the smallest fortified site, Choq Tepe (1.7 ha; Gardin 1998, no 485) located about 12 km from Bala Hisar of Kunduz. Being placed on dominant positions within each plain, the fortified sites seem to be situated rather aside from the settlement clusters or on its edges. Based on Gardin’sdata (Gardin 1995; 1998), this organization occurs in the case of Kohna Qala, Bala Hisar of Kunduz, and Kurgan Tepe in the Taluqan plain, while the settlements of Khwaja Hafiz, Kafir Qala, and Deh Nahr -iJadid Ado not seem to be accompanied by such agroup of smaller sites at all.24 As mentioned above, the most numerous cluster of the Phase III period sites in this sector is located on the plain of Dasht -iQala (see Pl. 1/1), where in total 28 sites flourished.25 27 unfortified sites are concentrated along the channel system, apart of which had been in operation already in the Bronze Age (Francfort 1989, esp. 57–58; Gardin 1998, 45–46), and which witnessed extensive development during the Iron Ages (Lyonnet 1997, 108–115; Gardin 1998, 45–46, fig. 3.9; Francfort 2013; Gardin 2001). This settlement area was situated virtually under the control of Kohna Qala, asemi -circular double -walled settlement with acitadel in the central part, located on the promontory just above the shore of the Panj River. The distance between the minor settlements and the fortified Kohna Qala, which is considered the administrative centre of the plain before the foundation of Ai Khanoum (Gardin 1998, 45–46; Martinez -Sève 2015, 22), stretches between 0.5 and 19 km. 20 The 28 sites of unknown size could be mostly characterised as the smallest. Predominantly, it is the case of sites marked by Gardin as maison isoleé, fermé isoleé, or site amorphe (see Gardin 1998, 41–45, fig. 3:2–3), which would make the two smallest classes an even bigger majority within the assemblage. The exception is site no 499 near Ishan Tup (Gardin 1998, 80), apottery scatter stretching over an area of ca. 16 ha corresponding to activities of an uncertain character. 21 Note that none of these sites have been investigated by excavations, and all of them were affected by alater occupation in some way. Gardin (1995; 1998), however, associated the fortification of these settlements with the Pre -Hellenistic periods of occupation. 22 In Pre -Hellenistic Bactria, only Bala Hisar of Balkh (ca. 100 ha within the circular wall) falls into class IX. See Lhuillier 2018, 261, 266 for possible extramural occupation in this period with references. 23 Zone d’habitat in the case of n o 413 (Gardin 1998, 42), plusieur aire d’habitat continu (hameau ou village) in the case of the site Dasht -iQala 448 (Gardin 1998, 50). 24 Being one of six similar-sized unfortified sites distributed linearly along the ancient water channel (Gardin 1998, 83, fig. 10:24), the site of Choq Tepe differs again from the general pattern. 25 Counting together 26 sites mentioned by B. Lyonnet (1997, fig. 37, tabl. XVI(2)) and tepas no 106 and 111 excavated by H.-P.Francfort in 1977 (Francfort 2013, 159–161).
16 STUDIA HERCYNIA XXV/2 SETTLEMENT TRANSFORMATIONS IN THE HELLENISTIC PERIOD Unlike the elaborate chronology of Ai Khanoum, which has been recently reworked and modified in favour of alater, Seleucid, dating of its foundation as acity (Lyonnet 2012; 2013), 26 the nature of the data coming mainly from surface surveys does not allow one to outline settlement dynamics within the Hellenistic period in great detail (Lyonnet 1997, 148–149). The sites attributed by Gardin’steam to the Hellenistic period (Pl. 1/3) belong rather to the 26 According to Lyonnet (2012, 158–159), an earlier Greek presence, especially in the area of the citadel, cannot be ruled out, however not in the extent and character of the later city. Fig. 2: Histogram showing the sizes of sites in sector II (eastern Bactria) in the Achaemenid period. Based on the attribution to the Iron Age Phase III (Lyonnet 1997). Fig. 3: Histogram showing the sizes of sites in sector II (eastern Bactria) in the Hellenistic period.
17JAKUB HAVLÍK Graeco -Bactrian period– most likely after the eastern campaign of Antiochus III– which is indicated by the presence of grey -black ware (Lyonnet 1997, 148; cf. Lyonnet 2012; 2013; Martinez -Sève 2021, 229), and evidence of settlement processes related to the end of the fourth century BC is missing or remains only badly detectable at all.27 According to Lyonnet, it is highly improbable that the whole region was deserted after Alexander’sconquest. Instead, as already mentioned above, she formulated the hypothesis of acontinual usage of the traditional local pottery simultaneously to the newly introduced shapes in the Hellenistic period (Lyonnet 1997, 149), which makes apart of the settlements of this period virtually invisible for archaeologists. Concluding the results of the survey, it is possible to observe acontinuity with the Hellenistic (Graeco -Bactrian) period in the case of 46 of the 78 sites distinguished for the preceding period (Fig. 6; Pl. 1/2). At some point between the end of the fourth century and the course of the third century BC, 32 sites (41%) were probably abandoned. Half (16 sites) of the abandoned sites fall under the smallest class of size up to 0.45 ha, or under the category ‘unknown size’28 (12 sites; 37.5%). There is only one fortified Phase III site, where the occupation continuity is most probably broken: Khwaja Hafiz A(Ball 2019, no 594A; Gardin 1998, no 546). The other fortified sites of the preceding period appear to remain in function in some way. The settlement clusters in their vicinities were affected only partly by this abandonment (cf. Gardin 1998, 42, 78–79). Material collected by Gardin’steam in the most extensively inhabited settlement zones (i.e., north of Kohna Qala and north of Kunduz) could indicate there an unbroken continuity of occupation (cf., Lyonnet 1997, fig. 37–38; Gardin 1998, 42–44, 78–79, fig. 3.9–3.10, 10.22). On the other hand, the most significant abandonment was observed by Gardin in the case of rather sparsely distributed sites in the valley of Taluqan, along the Rud -iShahrawan, or on the plain of Imam Sahib. In these areas, however, new sites flourished nearby in the Hellenistic period (Gardin 1998, 68–71). During the Graeco -Bactrian period (Fig. 3; Pl. 1/3), this region witnessed arapid increase of both settlement density and extent of the settled territory, which has already been well outlined by the French archaeologists working in the area (Gardin 1998, esp. 112–114, 162–163; Francfort 2013, 177; cf. Martinez -Sève 2021, 227–235). In terms of number, the increase in settlement sites is breath -taking. Instead of the 78 sites of the preceding period, more than four times as many Hellenistic period sites have been recorded. In total, 332 individual sites scattered across eastern Bactria. However, more than half of these sites 29 are concentrated on the plain Dasht -iQala, in the vicinity of Ai Khanoum. This ‘carpet settlement’ consists mostly of very small sites interpreted by Gardin usually as isolated houses and building clusters (maisons isolées, groupes de maisons), or less frequently as farmsteads or isolated hamlets (fermes ou hameaux isolés),30 which characterises the prevailing nature of the settlement landscape sprawling between several larger, usually fortified settlements. However, it should be stressed (cf. Gardin 1998, 9–14), that three full seasons (1974–1976) were devoted solely to the survey of Dasht -iQala plain, while amany times wider area was surveyed within only two seasons (1977–1978). This fact inevitably leads to abias in favour of the immediate Ai Khanoum vicinity. 27 Material identified with an earlier phase of the Hellenistic period was– besides Ai Khanoum– detected only in the case of site no 34 on the plain of Archi (Lyonnet 1997, 148). 28 Cf. footnotes 13 and 20. 29 In total 173 sites, counting together all the sites described by Gardin (1998, 41–45, fig. 3:10) including sites of extramural ‘semi -rural’ zone C of Ai Khanoum (Gardin 1998, 42, fig. 3:4A). 30 In the case of the majority of these sites, information about their precise size is missing (cf., Gardin 1998, 42–43).
24 STUDIA HERCYNIA XXV/2 K. Abdullaev (2001, 28–29) briefly mentioned alate 4th century BC date also for the lowest strata of the stratigraphic cut at Payon Kurgan (0.9 ha– class III) located not far from Kurganzol. Leriche (2013, 140–141) proposed the foundation of this highly probably fortified settlement in general to the Seleucids and related it functionally to Kurganzol as apart of aline of control points along the road to Sogdiana (see also Sverchkov 2005a, 15; Sverchkov 2005b, 57). The third of the fortified sites in this area is Uzundara (Dvurechenskaya 2019b), located on the Susiztag ridge overlooking the piedmont steppe valleys where both Kurganzol and Payon Kurgan are located. N. T. Dvurechenskaya (2018b; 2019a) proposed dating its foundation to the end of the 4th century– early 3rd century BC.Again, it appears that it was precisely the early Seleucids who were responsible for the initiation of its construction as evidenced not only by the ceramic material dated by Dvurechenskaya (2018b; 2019a) from the end of the 4th century– early 3rd century BC, but also by the discovered coins of Antiochus I(Dvurechenskaya 2019b; cf. Stančo 2021a, 266–269, fig. 11.11). Regarding the very limited evidence of unfortified settlements attributed to the Early Hellenistic period, V. V. Mokroborodov (2007; 2021) excavated afarmstead at the site of Gisht Tepa (0.1 ha; class I) in the Pashkhurt valley of the Kugitang foothills. While the first phase is attributed to the late 4th century BC, the second phase of this unfortified settlement is said to be contemporary to the early 3 rd century BC material of Kampyr Tepa and Old Termez (Mokroborodov 2021). According to Stride (2004, 175, 186, no 432, 602), there are two more presumably unfortified sites (Surpa Tepa II and Qira Qira) located in the lowlands of the Upper Surkhan Darya valley, which could be attributed to the ‘Early Hellenistic period’ based on the pottery assemblage collected by surface survey. SETTLEMENT STRUCTURE OF THE HELLENISTIC PERIOD The image of the settlement situation of the Hellenistic period as awhole (Pl. 1/7) indicates in the studied part of northern Bactria aclear distinction from the situation of the Yaz III phase. In total, 36 sites of different sizes were recognized, that is, 65.5% in terms of number compared to the preceding period. Only at 5 (13.8%) of them is it possible to observe ageneral continuity between the Achaemenid/Yaz III and the Hellenistic (Graeco -Bactrian) periods: Khaytabad Tepa (Leriche– Annaev 1995, 11–14; Leriche 2013, 139–140), Jandavlat Tepa and Yalangoyok Ota Tepa in its vicinity (Stančo 2019, 362–364), Talitagora in Termez district (Annaev 2018), and the fortified settlement of Dabil Kurgan (see Mokroborodov 2021),46 all of them located in the southern part of Surkhandarya province. However, the rest of the sites were founded on new, though, not always distant locations.47 evidence for the presence of material with much later analogies than of the Early Hellenistic period (Sverchkov 2006; Sverchkov 2008, 149–164. Cf., Houal 2021, 117), Kurganzol is in this study considered as occupied also after the Early Hellenistic period. 46 According to S. Bolelov, the site was abandoned after the Greek conquest (archaeological material does not witness the occupation of the late 4th/early 3rd century BC) and resettled no earlier than during the 3rd century BC, presumably in the Graeco -Bactrian period (Bolelov 2004). 47 It has already been stressed (Rtveladze 2002, 136; Stride 2004, 313; Rtveladze 2017b, 21; Stančo 2019, 364), that in several locations it is possible to observe acertain trend in the founding of the new Hellenistic period settlements in the close vicinity of deserted sites of the preceding period, such as in the case of Shor Tepa– Kampyr Tepa, Talashkan Tepa I– Talashkan Tepa II, the Yaz III period sites of Bandykhan and the newly founded site of Yalangtush Tepa (cf., Sverchkov– Boroffka 2007, 129).
25JAKUB HAVLÍK The settlement situation differs from the Achaemenid period not only in terms of the number and general location of settlement sites, but also in the way of their distribution, the probable character of the individual sites, and the settlement hierarchy in general. The most distinctive phenomenon is the decline of coherent settlement clusters attested in the preceding period. The settlement structure transformed into arather nucleated pattern, with relatively large distances (usually no less than 5 km) between individual sites. This phenomenon occurs not only in the foothill area, where settlement distribution follows the watercourses of minor rivers, but also in well -irrigable lowland areas. Acontinuation of the trend in settlement clusters could be assumed in the Sherabad oasis in the vicinity of Jandavlat Tepa. In the Hellenistic period, in total six smaller unfortified (class I–III) sites flourished at distances of up to 9.6 km48 as the crow flies from the supposed oasis centre (see Stančo 2019, 362–363). Despite this obvious sparse distribution, the presence of such acluster makes Sherabad oasis, as far as we know, the most densely occupied part of sector IV in terms of number, which only underlines the general low settlement density of Hellenistic period northern Bactria.49 In the upper Surkhan Darya basin, two new centres emerged at adistance of about 25 km from each other: Khalchayan and Dalverzin Tepa. In addition to Khalchayan– Khanaka Tepa (9.1 ha; class VI), the second mound, Karabag Tepa (3.6 ha; class IV), was occupied during the Hellenistic period at adistance of no more than 0.8 km. The extent of the Hellenistic period Dalverzin Tepa probably did not exceed the area of the citadel (4.9 ha; class V) of the later vast Kushan town (Pugachenkova– Rtveladze 1978, 12–19; Pugachenkova– Rtveladze 1990, 69–7), however, Hellenistic pottery was detected in multiple places in the ‘Lower Town’, even in the area of the later city walls. The character and precise extent of the presumably unfortified settlement in this area remain unclear (Pugachenkova– Rtveladze 1978, 26–29). It is possible that Dalverzin Tepa took over the function of both Kyzyl Tepa and Dunyo Tepa at distances of ca. 12.5 and 7 km, respectively, after their abandonment at the beginning of the 3rd century. Since the nearest site situated downstream is Khaytabad Tepa, at adistance of around 70 km, Dalverzin Tepa seems to demarcate the southern boundary of the Upper Surkhan Darya basin in the Hellenistic period. These three well -defined, rather large settlements were accompanied by only three other sparsely distributed smaller ones (classes I–II) in this area.50 Acomparison of the settlement hierarchy (Fig. 5) with the preceding period clearly suggests acertain trend of hierarchy ‘equalization’, caused in region IV by asignificant decrease in the number of small sites, however, also by aslight increase in the number of larger settlements of classes V–VI. The overall settlement image is also transformed by aparallel proportional increase of sites, where signs of fortification were recognized (cf. Leriche 2007, 133; Leriche– Rtveladze– de Pontbriand 2014, 234). In total, there are 17 settlements in this region, which were presumably fortified during the Hellenistic period, most of them inves48 On the Talashkan Tepa II at adistance of ca. 16.5 km, considered as a‘border fort’ of the oasis taking over the function of Talashkan Tepa I, see Stančo 2019, 364. 49 However, as an exception, pairs of sites occur elsewhere, consisting of one larger site and one rather small ‘satellite’ site in its vicinity, such as in the case of the citadel of Old Termez and the ca. 1 km distant site of Chingiz Tepa (Leriche 2010, 157), apossible interconnection of which indicates the presence of apottery scatter between them, on the location of ‘Termez– Camp Militaire’ (see Stride 2004, no 131e). Asimilar constellation occurs in the case of Khanaka Tepa– Karabag Tepa at Khalchayan (Pugachenkova 1966), or Dalverzin Tepa– Jurabay Tepa (Stride 2004, no 583). This situation could indicate an originally more complicated pattern in lowland areas, the remains of which are not preserved until today. 50 Jurabay Tepa (Stride 2004, no 583), Kurgan Tepa (Stride 2004, no 285), and Surpa Tepa II (Stride 2004, no 432).
26 STUDIA HERCYNIA XXV/2 tigated by excavations. This proportion (47.2%) is more than double that of the Achaemenid period (21.8%). Besides the lowland locations, where fortified settlements already occurred during the 4th century BC, new walled settlements emerged at several strategic locations such as crossings of the Amu Darya in Old Termez and Kampyr Tepa (on the crossing points, see Rtveladze 2002, 39–66), or mountain valleys of the eastern Kugitang Tau and Baysun Tau piedmonts. It seems that these hilly areas were occupied only very sparsely during the Achaemenid period. Given the fact that the foothills offer only limited potential for irrigated agriculture (cf. Stride 2004, 137, 147), it is possible that exploitation of this region in the form of seasonal pastoralism left only limited or no traces. Outside the Pashkhurt valley,51 only two Yaz III/Achaemenid period sites were reported. Rather small settlements of uncertain extent (Sultankul Tepa and Mazar Tepa) were located along the upper stream of the Sherabad Darya, nevertheless, Sverchkov (2005a, 13–14) does not mention the Hellenistic occupation of either of them. During the Hellenistic period, at least ten newly established sites flourished in the piedmonts (cf. Stančo 2021a). Besides small, fortified settlements with asurface area of up to 1 ha which can also be identified as forts or fortresses, such as the already mentioned Kurganzol or Payon -Kurgan (only Uzundara is exceptionally large with its walls encircling ca 2.2 ha; cf. Dvurechenskaya 2019b), and the Iron Gates fortification system52 (Rakhmanov– Rapen 2003; Rapinetal. 2006; Stančo 2021a, 262–279), there is, in the Hellenistic period also clear evidence of small foothill settlements with no signs of fortification such as Gisht Tepa (Mokroborodov 2007; Mokroborodov 2021) or Daganajam Tepa (Stančo 2021a, 256; Stančo 2021b). In the case of these sites, other than primarily amilitary function must be supposed. DISCUSSION The comparison of settlement development in the two described regions offers insight into the nature of the transformation which affected Bactria approximately between the end of the 4th century BC and the middle of the 2nd century BC.The proposed analysis demonstrates that while in the Achaemenid period both examined regions bear similarities in terms of the number of settlement sites and in the proportional distribution of size classes, in the Hellenistic period, settlement patterns differ distinctively in each of them. This diversification apparently reflects the change in settlement strategies and the shift in land use which followed the arrival of Greek rule in the area. The Pre -Hellenistic settlement pattern appears to be defined predominantly by relatively dense clusters of very small settlement sites, presumably groups of farmsteads, which appear to be, in terms of number, adominant element of the Achaemenid period cultural landscape. When counted together with sites of unknown size, 53 small unfortified sites represented in both regions almost identically 75–76% of the sites detected so far (Fig. 7), which corresponds to aconsiderable degree of settlement dispersal, comparable also to the Iron Age in central Bactria (Balkh and Khulm oases, see Sarianidi 1977, 108–111, 116–128, ris. 5–8; Kruglikova 2005, 51 On the past settlement situation in this area, see Stančo 2016; Mokroborodov 2021. 52 On the newly detected possible components of the fortification system in the wider Iron Gates area (i.e., single towers, long walls and enceintes), see Bel’sh 2020; Stančo 2021b. These structures, however, need afurther evaluation in terms of their chronological attribution, function, and mutual relations. 53 Cf. footnotes 13 and 20.
27JAKUB HAVLÍK 326–331, 362) or in Margiana (Salvatori 2008). Among the larger sites, there is arelatively high proportion of fortified ones (Pl. 1/1), which– in combination with their usually dominant position in relation to the minor sites– already led several researchers to perceive them to be local (micro -regional) centres (e.g., Gardin 1998, 110–111; Stančo 2018; Stančo 2019, 360–362; Wu 2018, 207–212), to which sometimes aproto -urban or urban function has been attributed (Rtveladze– Khakimov 1973; Gardin 1998, 142–144; Šajdullaev 2002, 332–333). In comparison with contemporary sites in northern Bactria, the size of eastern Bactrian fortified sites is remarkable, in some cases reaching as much as 30 ha.54 Such an extent– unseen among northern Bactrian sites until the Hellenistic and Kushan periods (cf. Leriche 2007)– represents aremarkable difference between the Amu Darya left and right bank regions in the Achaemenid period.55 The internal dynamics of the complex transformation that followed are hard to define based on the current state of archaeological research. An inevitably flattened image of the settlement pattern prevents one from identifying partial processes beyond its changes, which could have occurred at astroke or in the form of agradual development. This is the case of one of the main features of the described settlement transformation, ahigh level of discontinuity in the occupation of, especially the smallest, Pre -Hellenistic sites (cf. Fig. 6). This break of settlement continuity seems to be striking in the western oases of Bactria, where– in terms of number– more than 90% of the sites were affected.56 At some point, the settlement as 54 Kunduz Bala Hisar and Kokhna Qala (cf. Ball 2019 no 631, 931; Gardin 1998, 41, no 490). 55 Sites of Bala Hisar of Balkh, Altyn Dilyar Tepa, and Cheshme Shafa (Ball 2019, no 38, 99; Marquis 2018), all situated in Central Bactria, however, significantly outstrip in size even those of eastern Bactria. 56 Similar figures as in the Surkhan Darya and Sherabad Darya basins (Gardin’ssector IV) could also be presented for central Bactria (sector III); see Sarianidi 1977, 110–112, 115–131 and Kruglikova 2005, 326–323, 362–363, who do not mention any Hellenistic period successor to the extensive Iron Age settlement clusters of Altyn -Dashli and Naibabad oases. See also Ball 2019, maps 4–8. In this part of Bactria, continuity between the Achaemenid and the Hellenistic period can be seen only at Fig. 6: Graphs showing the comparison of occupation continuity rates between the Achaemenid (Yaz III) and Hellenistic period. Sectors II and IV.
28 STUDIA HERCYNIA XXV/2 awhole started to develop along anew trajectory in the Hellenistic period, being tied with several fortified centres occupied continuously or reoccupied (cf. Leriche 2007, 132; Leriche 2013, 139–140; Rtveladze 2002, 134–136). Evidence of massive settlement abandonment could be easily associated with Alexander the Great’scampaign and the drastic measures taken against the strong resistance of the local population (see Koshelenko– Serditykh 1987, 54; Leriche 2007, 128; Leriche 2013, 139–140; cf. Litvinskij 1998, 49). However, without firm proof that this decrease occurred simultaneously in relation to asingle historical event, the attested break in settlement continuity must be placed within amuch wider time span. The origins of this discontinuity could be tentatively traced back to abadly known period between Alexander’scampaign and asupposed consolidation of the situation under the rule of the early Seleucids. This turbulent period, shaped by the initial conquest and its immediate consequences, repeated rebellions of Greek settlers and subsequent struggles among Alexander’scommanders after the king’sdeath 57 must have had the potential to profoundly affect the capabilities of local communities to keep land use unchanged, and could eventually have led to an attested shift in settlement development. The supposed consolidation of the state of affairs under the early Seleucids seems to be confirmed by the numerous new foundations attested for this period by excavations not only in the case of many new sites in northern Bactria (see the overview above), but also by the early building activities in Ai Khanoum related right to this period (Lyonnet 2012; 2013; Martinez -Sève 2015, 26–35). As already well outlined and argued by L. Martinez -Sève (2015, esp. 25–36 with references), awidespread Seleucid involvement can with all probability be related to the coregency or direct rule of AntiochusI. Being awidespread policy related to the city foundations throughout the Seleucid Empire (cf. Kosmin 2014, 192–194. For Central Asia see Martinez -Sève 2015, 27), hypothetic forced population relocations in Bactria (to Ai Khanoum or its vicinity?) could also be avalid explanation for the abandonment phenomenon. In eastern Bactria, the phenomenon of settlement abandonment manifests itself in aless radical way. Amuch higher degree of occupation continuity, evidenced here in the case of ca. 59% of sites in terms of number (Fig. 6; Pl. 1/2), promotes this region to aspecific position within Bactria. It is questionable if this statistic gives evidence for more stable or more favourable conditions which could eventually also lead to the choice of location for the foundation of Ai Khanoum.58 However, it is possible that the impression of awider continuity reflects rather later intensive populating of the region in the Graeco -Bactrian period, when its exceptional role is unquestionable (Martinez -Sève 2015, 35–40; Martinez -Sève 2021, 222–229). The major, indisputably urban, settlement of Ai Khanoum served as apolitico -economic point of attraction for all the surrounding densely inhabited areas (Gardin 1998, 46, 112–114). Besides the spatial overlap of the occupation phases at the level of individual sites, we may ask ourselves which other indices suggest structural continuity. It is remarkable that despite the enormous increase in the absolute number of sites in eastern Bactria, the proportional distribution within the designated size classes does not show any significant differences in comparison with the situation in the Achaemenid period (Figs. 2, 3, 7). Asimilar phenomenon can also be observed in the case of the proportion of sites that bear signs of fortification several sites such as Bala Hisar of Balkh and its close vicinity (Marquis 2018; Lhuillier 2018) or Dilberjin (Dolgorukov 1984). 57 On the impact of Alexander’scampaign on the local population, see Holt 1988, 60–69; Holt 2005, 105–114; 125–136; Iliakis 2021. On the settlers’ rebellions, see Koshelenko 1979, 181–221; Holt 1988, 89–91; Iliakis 2013). 58 The excavations of a? rural settlements? in the plain of Dasht -iQala (e.g., Francfort 2013) gives only very limited evidence of the character of abandonment at the level of individual settlements.
29JAKUB HAVLÍK (Fig. 8).59 The same as in the Achaemenid period, fortifications are mostly associated with the largest sites (Pl. 1/1, 1/3).60 The number of settlements multiplied and anew centre of so far unseen importance was formed, but the structure as awhole seems to follow the older trends to aconsiderable degree. This mode of continuity most likely reflects the intensive agricultural exploitation of plains along the rivers, which took advantage of the water channel network developed in all probability to ahigh level as early as during the Iron Age (Gardin 2001). The rural settlement boom in the wider Ai Khanoum hinterland lies in the very adoption of an already working irrigation system, the existing capacities of which were exploited in an intensified way as evidenced by only amoderate increase of the estimated area of irrigation in the Hellenistic period (cf. Lyonnet 1997, 112–115, 150–152, tab.XVI(2), XX; Gardin 1998, 109, 113–114, tab.22–23. See also Leriche 2007, 131; Martinez -Sève 2021, 227). 61 As stated by Gardin (1998, 154), the channels which seem to be newly developed in the Hellenistic period consist rather of those less profitable, originated mostly in the completion of Rud -iShahrawan, an extensive long -term project initiated long before the Greek and probably even the Achaemenid rule in the area (Gardin 1998, 113–114, 151–153; Gardin 2001). Whether completed on the order of new elites coming from outside or conducted under the control of alocal (native) authority, this particular achievement required local specialists skilled in such atask (Gardin 2001, 171), and clearly testifies to adisplay of ideological transfer– continuation of the older local land- -use strategy in the Hellenistic period. 59 Even though the absolute number of fortified sites tripled, their proportion within the assemblage did not change dramatically: it decreased from 8.9% (7 sites) in the Achaemenid period to 5.7% (19 sites) in the Hellenistic Period (see above). 60 Cf. Gardin 1998, tab 18. Gardin’stable does not include Ai Khanoum, Bus -iShan (Ball 2019, no 148) and Aliabad (Ball 2019, 29). 61 According to Lyonnet (1997, tab.XVI(2), XX), the estimated area of irrigated land increased only by 4.7–6.8% between the Phase III and the Hellenistic period. Fig. 7: Pareto charts showing the comparison of the sites’ proportional distribution within the defined size classes.
30 STUDIA HERCYNIA XXV/2 On the other hand, rural settlement, consisting mostly of isolated farmsteads specialised in agricultural production was dominant among many regions of the Greek homeland of the Late Classical and Hellenistic periods (see the discussion with references in Whitley 2001, 382–391; McHugh 2017). Employment of afamiliar strategy of territory exploitation in the region, where avery similar practice already took place in the Pre -Hellenistic period, would seem to be alogical way to manage the country. The Hellenistic period explosion of dispersed settlement structure, characterised by small farmsteads and villages in the wider hinterland of anewly established large urban centre, represents awidespread trend documented for example in the territories of Northern Levant, which in the 3rd century BC became one of the key regions of the Seleucid Empire.62 This phenomenon, connected with the decline of many older nucleated tell settlements (cf. summary in Hannestad 2013, 252; Kosmin 2014, 198), is well attested, for example, in the Amuq valley, where the major city of Antioch was founded by Seleucus I(Casana 2007). The settlement dispersal scheme tied to the new urban foundation clearly corresponds to the Hellenistic period transformation of eastern Bactria and especially the Dasht -iQala plain adjacent to Ai Khanoum. Unlike Northern Levant, in eastern Bactria there is no evidence of acomplete abandonment of the old centres. 63 Regardless of the assumed power shift to Ai Khanoum (Gardin 1998, 46), the older large, fortified settlements were in the Hellenistic period incorporated into the newly evolved structure, in which both aspects, the preceding local and the newly coming imperial, seem to have coexisted in some way.64 In this case, in eastern Bactria, we are possibly witnessing asuccessful fusion of traditional and foreign aspects of land administration. It is hard to find any evidence in the north -western part of Bactria for the described phenomena characteristic of eastern Bactria in the Hellenistic period (i.e., the steep rise in site number, foundation of amajor urban centre, comparatively higher degree of occupation continuity including proportional distribution of different size classes, and unbroken use of awater channel system). The number of the Hellenistic period sites in the north -western part of Bactria does not exceed that of the Achaemenid period, no clear urban centres comparable to Ai Khanoum were established, and the general degree of continuity is remarkably low at the level of individual sites (Fig. 6). 65 In terms of settlement hierarchy, the smallest size classes were still predominant in terms of number (Fig. 5), however, in terms of proportional distribution, acertain ‘equalisation’ trend occurred (Fig. 7). This phenomenon– caused by the numerical decrease of the smallest sites as well as the foundation of new sites such as Dalverzin Tepa or Khalchayan– Khanaka Tepa– is noticeable especially in the lowlands (Pl. 1/7). Here, sparsely distributed, mostly fortified, sites of larger classes, accompanied rather rarely by smaller ones, are shaping anucleated settlement pattern, which differs distinctively 62 T. J. Wilkinson (2003, 128–150; Wilkinson– Ur– Casana 2004), however, sees this phenomenon only as apart of along -standing trend beginning in the Fertile Crescent as early as in the early 1st millennium BC (see also discussion in Casana 2007, 210–212). 63 As for Kohna Qala, despite the fact that the Hellenistic period sherds are less numerous than sherds of the preceding periods, according to Gardin (1998, 46) its complete abandonment in the Hellenistic period cannot be confirmed and some kind of activities on the site in this period must be assumed (Cf. Martinez -Sève 2021, 229). 64 The described transformation of the settlement structure– at least in this case– does not correspond to the general pattern proposed by P.J. Kosmin (2014, 195–199) for the settlement policy employed by the Seleucids during the formation of their realm: No ‘technology of forgetting, with exorcism of the pre -Seleucid terrain’ seems to be employed here (cf. Kosmin 2014, 195). 65 Such aconclusion could be drawn in the case of Central Bactria as well. See references in footnote no 56.
31JAKUB HAVLÍK from that of contemporaneous eastern Bactria. The occurrence of rather small (classes I–III), often strongly fortified settlements– fortresses spread throughout the hilly piedmonts and along the Amu Darya which could have had only limited models in Pre -Hellenistic Bactria66– is characteristic of the territory of present -day Surkhan Darya. COMMON FEATURES IN THE SETTLEMENT DEVELOPMENT Despite the outlined differences, there are clearly common features between the two compared regions as well. Using these characteristics, it is possible to trace general tendencies in the settlement transformation and the nature of the Hellenistic period settlement as awhole. These trends seem to be foreshadowed to acertain extent by the general continuity/discontinuity pattern, as well as the character and spatial distribution of the identified Early Hellenistic settlements in northern Bactria. Besides the described remarkably high abandonment rate of smaller unfortified Achaemenid period sites, the emphasis laid on the fortification is clearly characteristic. In region IV, walled settlements represent almost half of the whole number of the Hellenistic period sites (i.e., 17 of 36). In eastern Bactria, this trend manifests itself by the increase in the absolute number of supposedly fortified sites, i.e., from 7 to 19. Nevertheless, the proportion of such sites does not greatly change (from 8.9% to 5.7%; see also Fig. 8). The increase in the total number of fortified sites could be considered acontinuation of such atrend initiated already in the Iron Age (Gardin 1998, 142–143; Šajdullaev 2002) and continued in the Achaemenid period (Lhuillier 2018, 264–267; Wu 2019, 211). Nevertheless, right in the foundation of such fortified settlements and military installations it is possible to presume government involvement, allowing us to outline amaster plan designed by new external elites. The emergence of aconsiderable number of new fortified sites testified by excavations in northern Bactria can be clearly traced back to the Early Hellenistic (Seleucid) period (see above; Pl. 1/6–7). These activities, indicating the need for strongpoints in the landscape, were 66 The closest parallel could probably be perceived to be the small fortified settlement at Talashkan Tepa I; see Šajdullaev 2002, 279–291; cf. Lhuillier 2018, 264–267). Fig. 8: Graphs showing the comparison of proportional distribution of sites featuring evidence of fortification in sectors II and IV between the Achaemenid and the Hellenistic periods.
32 STUDIA HERCYNIA XXV/2 highly probably linked to seizing control over the territory. The same reasons could have motivated new rulers to the continuous occupation of the already fortified lowland sites such as Kohna Qala (Gardin 1998, 42, 46) or Khaytabad Tepa (Houal, 2021, 25) and reutilization– new fortifying– of settlements fortified to asmaller extent so far, such as Kyzyl Tepa phase 2 (Wu– Sverchkov– Boroffka 2017) or Takht -iSangin (Druzhinina 2016).67 The attribution of the Lower Town fortifications of Kyzyl Tepa to the Early Hellenistic period suggests an intentional modification of earlier structures in order to enhance their defensive capabilities, possibly due to the transformation of the settlement into arefuge or amilitary camp (Wu– Sverchkov– Boroffka 2017, 317; Wu 2018, 200–201). The eastern Bactria survey also gives evidence of apossible similar reutilization of settlements dated to the earlier phases of the Iron Age (namely, Phase II), such as Kunsay, Kurgan Tepe A, or Chim Kurgan (cf. no 9; no 47, and no 137 in Lyonnet 1997 and Gardin 1998; see Pl. 1/3). The massive enceintes of these sites– even if unoccupied and probably deserted during the Phase III (i.e., the Achaemenid(?) period)– must have offered aconsiderable defensive potential for any passing army. The positioning of newly founded settlements reveals acertain newly enhanced landscape preference for locations close to natural pathways, crossing points through physical barriers such as mountain ranges and rivers (cf. Leriche 2007, 133; Martinez -Sève 2021, 239–241). Especially remarkable in the Hellenistic period is the evidence of an intensified occupation of higher altitudes. In the Achaemenid period, most of the settlements were concentrated at altitudes between 400 and 600 masl– corresponding to the lowland alluvial plains– and sites located at altitudes above 800 masl were only very rare. In the subsequent period, low altitudes were still preferred, however, the occurrence of numerous sites in mountainous areas manifests anew interest in previously only sparsely inhabited parts of Bactria, distant from lowland core territories, i.e., Kugitang Tau and Baysun Tau piedmonts (see above; cf. Stančo 2021a; 2021b) or valleys in the northern Hindukush foothills such as Rustaq, Kalafgan, and Farkhar (cf. Lyonnet 1997, 110–114; Gardin 1998, 92–96; Martinez -Sève 2021, 239–243). Some of them are located as high up as 1600 masl, as in the case of the Uzundara fortress or the Kalafgan settlement group. Areas that have revealed no signs of aprevious occupation were also newly settled, such as the valleys of Ishkamesh (Gardin 1998, 101), Baghlan and Aliabad (cf. also Ball 2019, maps 4–8).68 The distribution of walled sites along river crossings and main natural pathways through the massifs gives evidence for an intensified need for control over the transit routes and movement of people along them, or at least for an intentional manifestation of such acontrol and display of power. Such aspatial pattern was widely applied throughout the Seleucid Empire (Kosmin 2014, 190–192) and, as outlined above, its establishment in Bactria can be connected directly to the Seleucid administration. As evidenced in the case of many sites such as Kampyr Tepa, Old Termez or Uzundara, this pattern was successfully retaken, adopted, and further developed by later Graeco -Bactrian and– in many cases– to acertain degree even by Kushan rulers in the first centuries AD. It seems that settlement reorganization, which could be attributed right to the early Seleucid rulers (Antiochus I? Cf. Martinez -Sève 2015, 26–33) shaped the character of later, Graeco -Bactrian settlement to aconsiderable degree. 67 A. Druzhinina (2016, 58), however, does not suppose adefensive purpose of the newly erected walls of Takht -iSangin ‘lower town’. She understands them rather as amonumental building demarcating aborder of the settlement. 68 The supposedly Graeco -Bactrian site of Irganakh located in the mountain pass west of Kunduz would also fit this distributional pattern (Ball 2018, no 442; Martinez -Sève 2021, 241. Cf. Fischer 1961, 14; Fischer 1967, 348, 354). Its dating is, however, based only on the isolated find of abadly preserved column base.
33JAKUB HAVLÍK Despite the relatively numerous fortified sites identified in the mountainous areas, recent research in northern Bactria indicates that the settling of these regions was not motivated exclusively by military (defensive) interests as is often suggested (Sverchkov 2005b; Leriche 2007; Dvurechenskaya 2019b). The presence of small unfortified settlements– farmsteads or small hamlets?– in foothill regions (Mokroborodov 2007; 2021; Stančo 2021b) could point to an agricultural or most likely agropastoral exploitation of marginal lands at higher altitudes. This need for control could also be motivated by the very economic and, of course, administrative interest in achannelling of potential trade69 and transhumance taking place along the routes, where also smaller unfortified settlements were located. The data from the Hindukush valleys of eastern Bactria (Gardin 1998, 101) indicate asimilar situation, while the discovery of the– so far unique– site of Torbulok interpreted by the Tajik -German team as asanctuary (Lindström 2017) reveals the potential of mountainous areas as aplace of ritual. Even in the case of fortified sites, functions other than military can be traced archaeologically, such as in the case of Uzundara and asupposed marketplace before its gate (Dvurechenskaya 2018b, 175). QUESTION OF AN UNEQUAL SETTLEMENT DEVELOPMENT In the face of the presented comparison, afundamental question of the eminent discrepancy of settlement development in the two compared regions arises. Based on the current state of research, the influence of an environmental change does not seem to be aplausible explanation. Even in the north -western part of Bactria, new settlements were often established at only asmall distance from the old ones (cf. Rtveladze 2002, 136; Stride 2004, 313), which testifies to arather stable environmental situation. An unequal settlement development seems to have been brought about rather by an intentional decision or asocio -political process. Such asuccessful decision was clearly the very foundation of Ai Khanoum, awell -chosen strategic location which (cf. Bernard– Francfort 1978; Martinez -Sève 2015) made it– with necessary investments– afocal point of the whole Bactrian scale. It is well known that even though the intensified agricultural exploitation of its hinterland seems to have developed to the form documented by Gardin’ssurvey (Gardin 1998, 38–47) no earlier than during the 2nd century BC (Martinez -Sève 2021, 229), Ai Khanoum was established in its final extent defined by the city ramparts– i.e., as an ambitious large -scale project– as early as during the Seleucid period (Leriche 2007, 141–144; Martinez -Sève 2015, 31). From the very moment it was founded, the city was meant to be an exceptional central settlement. The different settlement strategy employed by the Seleucids and later by Graeco -Bactrian kings in the north -western part of Bactria does not necessarily indicate alongstanding instability of this region as could be evidenced by the extraordinarily low level of settlement continuity with the previous period. More likely, the network of rather small forts and nucleated settlements could reflect an interest in the control of strategic routes along the Surkhan Darya and Sherabad Darya Rivers without the intention of its intensified (agricultural) exploitation (Leriche 2007, 133), at least not in the same way as in the East. However, as mentioned above, the almost missing evidence of the smallest site classes presence can also reflect severe damage caused by modern mechanized agricultural activities. The view of present -day 69 Though the relatively poor material culture revealed at sites such as Kurganzol (Sverchkov 2008; 2013) does not bring up evidence on thriving commercial activities. The stratified data coming from large and medium -sized settlements in the lowland, which would of course be more suitable as commercial nodes, are on the other hand still missing, being buried beneath later occupation strata.
40 STUDIA HERCYNIA XXV/2 Pugachenkova– Rtveladze 1978 = Пугаченкова, Г.А.– Ртвеладзe, Э.В.: Дальверзинтепе– кушанский город на юге Узбекистана. Ташкент. Pugachenkova– Rtveladze 1990 = Пугаченкова, Г.А.– Ртвеладзе, Э.В.: Северная Бактрия -Тохаристан. Древность иАнтичность. Ташкент. Pugachenkova 1966 = Ругаченкова, Г.А.: Халчаян. Кпроблеме художественной культуры северной Бактрии. Ташкент. Pugachenkova 1987 = Ругаченкова, Г.А.: Шортепе. История Материалной культуры Узбекистана 21, 21–44. Rakhmanov– Rapen 2003 = Рахманов, Ш.– Рапен, К.: Железные ворота. In: Труды Бойсунской Научной Экспедиции 1. Ташкент, 22–32. Rapinetal. 2006 = Rapin, Cl.– Baud, A.– Grenet, F.– Rakhmanov, Sh.A.: Les recherches sur la région des Portes de Fer de Sogdiane. Bref état des questions en 2005. История Материалной культуры Узбекистана 35, 91–112. Rapin, Cl. 2013: On the Way to Roxane. The Route of Alexander the Great in Bactria and Sogdiana (328–327 BC). In: Lindströmetal. (eds.) 2013, 43–82. Rtveladze– Khakimov 1973 = Ртвеладзе, Э.В.– Хакимов, З.А.: Маршрутные исследования северной Бакгрии. In: Из истории античной культуры Узбекистана. Ташкент, 10–34. Rtveladze 1974 = Ртвеладзе, Э.В.: Разведочное изучение бактрийских памятников на юге Узбекистана. In: В.М. Массон (ed.): Древняя Бактрия. Предварителние сообшения об археологических работах на Юуге Узбекистана. Ленинград, 74–85. Rtveladze 1975 = Ртвеладзе, Э.В.: Кхарактеристице памятников Сурхандаринской области ахаменидского времени. Советская археология 1975/2, 262–266. Rtveladze 1976 = Ртвеладзе, Э.В.: Новые древнебактрийские памятники на Юге Узбекистана. In: В.М. Массон (ed.): Бактрийские древности. Ленинград, 93–103. Rtveladze 1987 = Ртвеладзе, Э.В.: Новые Бактрийские памятники на Юге Уэбекистана. История Материалной культуры Узбекистана 21, 56–66. Rtveladze 2002 = Ртвеладзе, Э.В.: Александр Македонский вБактрии иСогдиане. Ташкент. Rtveladze 2017a = Ртвеладзе, Э.В.: Ай -Ханум иКампыртепа. Культурные отношения, датировка. In: М.Д. Бухарин (ed.): Scripta antiqua. Вопросы древней истории, филологии, искусства иматериальной культуры. Альманах 6. Москва, 117–125. Rtveladze 2017b = Ртвеладзе, Э.В.: Кампыртепа– Александрия Оксианская. Город крепость на берегу Окса еллинистическое ипостеллинистическое время (конец VI в. до н. э.– Iв. до н. э.). Материалы Тохаристанской ЭкспедицииX. Археологические исследования Кампыртепа. Ташкент. Sagdullaev 1980 = Сагдуллаев, А.С.: Разкопки древнебактрийской усадьбы Кызылча 6 (по итогам работ 1974–1976 гг.). Советская археология 1980/2, 228–234. Sagdullaev 1987 = Сагдуллаев А.С.: Усадьбы Древней Бактрии. Ташкент. Salvatori, S.– Tosi, M. eds. 2008: The Bronze Age and Early Iron Age in the Margiana Lowlands. Facts and Methodological Proposals for aRedefinition of the Research Strategies. Oxford. Salvatori, S. 2008: The Margiana Settlement Pattern from the Middle Bronze Age to the Parthian -Sasanian Period. AContribution to the Study of Complexity. In: Salvatori– Tosi eds. 2008, 57–74. Sarianidi 1977 = Сарианиди, В.И.: Древние земеделци Афганистана. Материалы Советско -Афганской экспедиции. Москва. Shaydullaev– Annaev 1989 = Шайдуллаев, Ш.– Аннаев, Т.: Два новых пункта раннежелезного века на юге Узбекистана. Общественные науки вУзбекистане 1989/5, 31–34. Stančo, L. 2016: Archaeological Survey in the Surroundings of Kayrit (South Uzbekistan), Preliminary Report for Season 2015. Studia Hercynia 22/2, 73–85. Stančo, L. 2018: New Data on the Iron Age in the Sherabad district, South Uzbekistan. In: Lhuillier– Boroffka eds. 2018, 171–188.
41JAKUB HAVLÍK Stančo, L. 2019: The Oasis in time and space. Dynamics of the settlement pattern. In: Stančo– Tušlová eds. 2019, 353–386. Stančo, L. 2021a: Southern Uzbekistan. In: Mairs ed. 2021, 249–285. Stančo, L. 2021b: In the shadow of the Wall. Hellenistic settlement in the Baysun and Kugitang piedmonts. In: J. Havlík – L. Stančo – J. Kysela (eds.): Seen from Oxyartes’ Rock. Central Asia Under and After Alexander. Third Meeting of the Hellenistic Central Asia Research Network, 14–16 November 2018. Studia Hercynia XXV/2, 64–95. Stančoetal. 2019 = Stančo, L.– Shaydullaev, S.– Khamidov, O.– Augustinová, A. Damašek, L.– Bek, T.– Kmošek, M.: In the footsteps of Euthydemus. Preliminary report for archaeological survey in the Baysun District (South Uzbekistan), Season 2018. Studia Hercynia 23/1, 141–172. Stančo, L.– Tušlová, P.eds. 2019: Sherabad Oasis. Tracing Historical Landscape in Southern Uzbekistan. Praha. Staviskiy 1977 = Ставиский, Б.Я.: Кушанская Бактря. Проблемы истории икультуры. Москва. Stride, S. 2004: Géographie archéologique de la province du Surkhan Darya (Ouzbékistan du Sud / Bactriane du nord). Unpublished PhD thesis, Université Paris I– Pantheon -Sorbonne. Paris. Stride, S. 2005: An Archaeological GIS of the Surkhan Darya province (Southern Uzbekistan). The Silk Road 2/2, 30–34. Stride, S. 2007: Regions and Territories in Southern central Asia. What Surkhan Darya Province tells us about Bactria. In: Cribb– Herrmann eds. 2007, 99–117. Sverchkov– Boroffka 2007 = Сверчков, Л.М.– Бороффка Н.: Археологические исследования вБандихане в2005 г. In: Труды Байсунской Научной Экспедиции. Археология, История иЭтнография 3. Ташкент, 97–141. Sverchkov, L. 2005a: Archaeological Monuments of the Boysun District. In: History and Traditional Culture of Boysun. Review of the Boysun Scientific Expedition 2, 10–20. Sverchkov, L. 2005b: Boysun. Trial historical reconstruction. In: History and Traditional Culture of Boysun. Review of the Boysun Scientific Expedition 2, 56–65. Sverchkov 2006 = Сверчков, Л.M.: Опыт синхронизации керамических комплексов эпохи эллинизма (Кампыртепа, Термез, Джигатепа, Курганзол). In: Э.В. Ртвеладзе– Дж.Я. Ильясов (eds.): Материялы Тохаристанской экспедиции 5. Археологические исследования Кампыртепа иШортепа. Ташкент, 105–124. Sverchkov 2013 = Сверчков, Л.М.: Курганзол — крепость Александра на юге Узбекистана. Ташкент. Sverchkov, L.M. 2008: The Kurganzol Fortress (on the History of Central Asia in the Hellenistic Era). Ancient Civilizations from Scythia to Siberia 14, 123–191. Šajdullaev, Š.B. 2002: Untersuchungen zur frühen Eisenzeit in Nordbaktrien. Archäologische Mitteilungen aus Iran und Turan 34, 243–339. Trigger, B.G. 1967: Settlement Archaeology– Its Goals and Promise. American Antiquity 32/2, 149–160. Whitley, J. 2001: The Archaeology of Ancient Greece. Cambridge. Wood, R.2011: Cultural convergences in Bactria. The Votives from Temple of the Oxus at Takht -iSangin. In: Kouremenos– Chandrasekaran– Rossi eds. 2011, 141–151. Wu, X.– Sverchkov, L.– Boroffka, N. 2017: The 2010–2011 seasons of excavations at Kyzyltepa (VI–IVth centuries BCE), Southern Uzbekistan. Iranica Antiqua 52, 283–362. Wu, X. 2018: Exploiting the Virgin Land. Kyzyltepa and the effects of the Achaemenid Persian Empire on its Central Asian frontier. In: Lhuillier– Boroffka eds. 2018, 189–214. Jakub Havlík Institute of Classical Archaeology Faculty of Arts, Charles University Celetná 20, CZ-11000 Prague 1 [email protected]
152 STUDIA HERCYNIA XXV/2 Pl. 1/1: Sector II (eastern Bactria) in the Achaemenid period. Based on the attribution to the Iron Age Phase III (Lyonnet 1997). Distribution of sites within the defined size classes. Map by J. Havlík. Basemap: Esri. Pl. 1/2: Sector II (eastern Bactria). The Achaemenid period settlement abandonment. Note that the map does not correspond to any particular period of time but reflects the general continuity or discontinuity of occupation between Iron Age Phase III (Lyonnet 1997) and Hellenistic period sites. While the largest fortified sites appear to be occupied continuously, the settlement discontinuity affected mainly the sites of the smallest classes. Map by J. Havlík. Basemap: Esri.
153PLATES Pl. 1/3: Sector II (eastern Bactria) in the Hellenistic period (corresponding principally to the Greco- -Bactrian period). Distribution of sites within the defined size classes. The yellow symbols represent the evidence of occupation in both Iron Age Phase III (Lyonet 1997) and the Hellenistic period. Map by J. Havlík. Basemap: Esri.
154 STUDIA HERCYNIA XXV/2 Pl. 1/4: Sector IV (oases of Surkhandarya) in the Achaemenid period. Based on the attribution to the Yaz III period. Distribution of sites within the defined size classes. Map by J. Havlík. Basemap: Esri. Pl. 1/5: Sector IV (oases of Surkhandarya). Archaeological sites, in which Early Hellenistic period (late 4th–first half of the 3rd century BC) strata have been attested. The map represents the distribution of the sites based on the current state of research, not attempting to depict the transitional period settlement situation. Map by J. Havlík. Basemap: Esri.
155PLATES Pl. 1/6: Sector IV (oases of Surkhandarya). Achaemenid period settlement abandonment. Note that the map does not correspond to any particular period of time but reflects the general continuity or discontinuity of occupation between Yaz III and the Hellenistic period sites. Kyzyl Tepa, occupied highly probably only until the Early Hellenistic period is represented as ‘abandoned’ in this comparison. The settlement discontinuity affected most of the sites in the sector mainly the smallest classes of sites. Map by J. Havlík. Basemap: Esri. Pl. 1/7: Sector IV (oases of Surkhandarya) in the Hellenistic period (corresponding to the 3rd–2nd centuries BC). Note that yellow symbols refer to the evidence of occupation in the Yaz III and Hellenistic period, but not necessarily uninterrupted. Map by J. Havlík. Basemap: Esri.