Migrations & Identities 1.1 (2008), 19–43 ISSN 1753-9021 (print) 1753-903X (online) © LUP Greek identity and the settler community in Hellenistic Bactria and Arachosia Rachel Mairs The present study examines the construction and assertion of Greek identity in the Hellenistic period (c. 323–30 bce) colonial settlements of Bactria and Arachosia (modern Afghanistan), exploring what it meant to the Greek immigrants and their descendents to be ‘Greek’, and how they chose to assert this identity through public display, in architecture, ritual and inscriptions. Much of the archaeological and epigraphic material from the region immediately strikes the modern, classically-trained observer as conspicuously non-Greek, and there is evidence to suggest that this is also how it would have appeared to a contemporary observer from the Greek lands of the Mediterranean littoral. The elements of ‘Greekness’ that the settlers chose to maintain and discard are, however, in themselves suggestive of a flexible and evolving sense of ethnic belonging, which sought to assert Greek cultural and civic identity only in very specific arenas. In other areas, I argue that the features we regard as representing affiliation with non- Greek cultural traditions could equally well, to the local settler community, have held a more neutral ethnic force, which might readily be reconciled with a strong sense of Greek identity. The assertion of an ethnic identity by any group is a selective matter; people choose which cultural traits (language, artistic style, religion) to emphasize and de-emphasize in the articulation of that identity. The little-explored and prob- lematic corpus of archaeological, epigraphic and literary evidence from the Hel- lenistic settlements of Bactria and Arachosia provides a rich body of material for exploring this process in action, through the establishment and evolution of a numerically small, but politically dominant, migrant community. What is a paisano? He is a mixture of Spanish, Indian, Mexican and assorted Caucasian bloods. His ancestors have lived in California for a hundred or two years. He speaks Eng- lish with a paisano accent and Spanish with a paisano accent. When questioned concerning his race, he indignantly claims pure Spanish blood and rolls up his sleeve to show that the soft inside of his arm is nearly white. His color, like that of a well-browned meerschaum pipe, he ascribes to sunburn. He is a paisano, and he lives in that uphill district above the town of Monterey called Tortilla Flat, although it isn’t a flat at all. (John Steinbeck (1935), Tortilla Flat) In the summer of 329 bc, Alexander the Great founded a city on the River Jaxartes in Sogdiana, one of many new ‘Alexandrias’ along his route through the lands of the former Persian Empire (Fraser, 1996). Alexandria Eschate (‘Alexandria the Furthest’) marked the limits of his campaigns in Central Asia, a remote outpost of Greek settlement facing the unconquered and unexplored steppe across the The settler community in Hellenistic Bactria and Arachosia mairs.indd 19 18/04/2008 14:09:43 20 Rachel Mairs river. These campaigns (c. 330–327 bce) were perhaps the most difficult and protracted of Alexander’s military career. Faced with a hostile local population accustomed to the more laissez-faire approach of the Persian authorities (Bri- ant, 1984), Alexander’s attempts to impose direct political control and military occupation provoked an uprising, which was made all the more troublesome by extremities of climate and terrain, and shifting networks of local allegiances (Holt, 1988). Demoralized, Alexander’s Greek and Macedonian troops began to revolt, a situation compounded by attempts to compel some of their number to establish a permanent military settlement in the region. There are many episodes in Alexander’s eastern campaigns that bear a chill- ingly modern resonance, particularly when viewed from a postcolonial per- spective. Some commentators have even adopted the language of genocide to describe the actions of the Macedonian forces in this Hellenistic heart of dark- ness: ‘For large areas of Asia the advent of Alexander meant carnage and starva- tion, and the effects were ultimately as devastating as that of the Spaniards in Mexico. The conquerors created a desert and called it empire’ (Bosworth, 1996, 48–49). Perhaps the most surreal of these atrocities occurred shortly before the foundation of Alexandria Eschate, on the long and difficult march through the central Asian deserts. Almost three thousand miles from the Aegean, Alexan- der’s army suddenly came upon a city populated by Greeks. Ancient and modern historians differ in their interpretation of the circumstances, but the sources are agreed that the Graeco-Macedonian army massacred the inhabitants (see, most recently, Parke, 1985 and Hammond, 1998, who provide full discussions of the classical sources). These Greeks were descendents of the Branchidae, guardians of the oracle of Apollo at Didyma, near the Greek city of Miletos on the coast of Asia Minor. In 494 bce the Branchidae had betrayed Didyma to the Per- sians and the shrine was sacked; they were subsequently resettled in the east of the Persian Empire. For Alexander’s army, the slaughter of the Branchidae was therefore a highly symbolic act of revenge, an opportunity to right yet another of the longstanding wrongs of the Graeco–Persian conflict. Although vengeance is the principal motive emphasized by our extant sources, one troubling dimension to the massacre is noted by the later Roman historian Quintus Curtius Rufus. The Branchidae, Curtius informs us, ‘had not altogether forgotten the customs of their country, but had already become “bilingual”, and their mother tongue was corrupted little by little from contact with a foreign language’ (7.5.29; Biville, 2002, 100). The Branchidae welcomed Alexander’s army as compatriots; but to the Graeco-Macedonian forces, already experien- cing profound culture shock in the adverse conditions of Central Asia, matters were not quite so clear-cut. The Branchidae were ‘neither brethren nor barbar- mairs.indd 20 18/04/2008 14:09:43 The settler community in Hellenistic Bactria and Arachosia 21 ians’ (Holt, 1988, 74). The old betrayal of the shrine at Didyma was sufficient to ensure their extermination, but there is also a strong suggestion of Greek revulsion at this community’s rejection not just of Greece, but of Greekness. Now bilingual and barely maintaining their Greek lifestyle, the Branchidae per- haps represented what Alexander’s soldiers and settlers did not want to become. Whatever the prejudices that motivated this fear, it was a realistic one. A wave of military settlement under Alexander and his eastern successors, the Seleucid dynasty, established a numerically small but politically dominant Greek community in Bactria, the region around the River Oxus north of the Hindu Kush (Holt, 1999). During the course of the third century bce, Bactria gradually asserted its independence under a succession of local Greek kings. South of the Hindu Kush, Arachosia became part of the Indian Maurya Empire, under the terms of a treaty of 303 bce between Seleucus I and Chandragupta Maurya (Bernard, 1985, 85–95; Karttunen, 1997, 261–263; Foucher and Bazin- Foucher, 1942–1947, 313ff.), but inscriptions reveal the presence of a significant Greek-speaking community for a long period thereafter (Bernard et al., 2004; Fraser, 1979; Carratelli and Garbini, 1964). In the second century bce, the Bac- trian Greeks began to expand militarily into Arachosia and northern India, where numismatic evidence attests to their presence as late as the first cen- tury ce (Bopearachchi, 1991[Q]). Figure 1. The Hellenistic world mairs.indd 21 18/04/2008 14:09:43 22 Rachel Mairs A century and a half after the massacre of the Branchidae, a Greek king of Bactria began to mint coins with legends in the Indian Kharoshthi and Brahmi scripts, in addition to the traditional Greek. Agathokles (fl. c.190–180 bce),1 a Greek basileus (king), became on his Indian coins an Indian raja (Holt, 1988, 1–7). The official face of the rulers of Bactria in the period after Alexander had hitherto been an overwhelmingly Greek one; the few surviving inscriptions are in Greek, and many of the key public buildings of the colonial city at Ai Khanoum (dis- cussed further below) follow Greek architectural traditions. The most striking non-Greek elements, such as the city’s main temple and the grand administra- tive quarter, are reminiscent of the architecture of the Persian Empire rather than of local forms (Mairs, forthcoming 2008[Q]). Only here and there do we find evidence of the more complex and variegated ethnolinguistic picture that lay behind the official Greek face of Hellenistic Bactria: local Iranian names (Grenet, 1983); possible traces of a local language committed to writing (Rapin, 1992, 105); the presence of Greek and indigenous religious practices alongside one another at the same temple sites (Francfort, 1984; Litvinskii and Pichikian, 1994). The majority of our evidence derives from the single site of Ai Khanoum, and it is to be expected that further excavation in the region will allow us to modify our assessments somewhat, perhaps in line with the infinitely more complex cultural and ethnic picture that emerges from later, Kushan-period Bactria (Staviskij, 1986). It is nevertheless striking, from the evidence currently at our disposal, that sometimes extreme efforts were taken to publicly assert and defend a monolithically Greek identity, in the face of a more diverse pattern of cultural expression and individual behaviour that we can, for the most part, only guess at. It is not surprising that a Graeco-Bactrian king should choose to make a public statement of his authority in a language other than Greek. What is significant is that it happened so late. The present study will examine some of the ways in which we might access the patterns of behaviour underlying the material and epigraphic record in the Hel- lenistic period (from roughly the late fourth century bce until the first century ce) colonial settlements of Bactria and Arachosia, and explore what it meant to the Greek immigrants and their descendents to be ‘Greek’, an identity which, the evidence leads me to argue, was very actively claimed and defended. The construction of Greek identity was a dynamic process, asserted through pub- lic display in architecture, ritual and inscriptions. Much of the archaeological and epigraphic material from the region might immediately strike the modern, 1. Graeco-Bactrian chronology is a controversial topic. Unless otherwise stated, all dates are taken from Bopearachchi, 1991, Table 5. mairs.indd 22 18/04/2008 14:09:43 The settler community in Hellenistic Bactria and Arachosia 23 classically-trained observer as conspicuously non-Greek, and there is evidence to suggest that this is also how it would have appeared to a contemporary obser- ver from the Greek lands of the Mediterranean littoral. The elements of ‘Greek- ness’ that the settlers chose to maintain and discard are, however, in themselves suggestive of a flexible and evolving sense of ethnic belonging, which sought to assert Greek cultural and civic identity only in very specific arenas. In other areas, I argue that the features we might regard as representing affiliation with non-Greek cultural traditions could equally well, to the local settler community, have held a more neutral ethnic force, one that might readily be reconciled with a strong sense of Greek identity. Ethnic identity and the archaeological record The questions we ask of the archaeological evidence from the Hellenistic world are, inevitably, conditioned to a great extent by our familiarity, intellectual or personal, with more recent colonial and settler societies. Explicit or implicit modern exemplars are frequently used to ground analyses of the social func- tioning of Hellenistic ethnicity, in all its increasingly complicated cultural and genetic permutations. John Steinbeck’s Tortilla Flat, quoted at the begin- ning of this article, avoids the weighty social concerns of many of his other works, but the introduction of Monterey’s paisano community is neverthe- less, for those with a predilection for such comparisons, strikingly Hellenistic. There are obvious dangers in projecting onto the ancient world our know- ledge, or assumed knowledge, of modern societies. But, from the viewpoint of an archaeologist or ancient historian, the level of detail to be gleaned from such case studies provides an attractive counterpoint to the more fragmentary ancient evidence. Specifics of time and place might prevent the direct impos- ition of modern anthropological details onto an ancient framework, but if our interpretations must inevitably reflect a desire to give some order, even a narrative structure, to the source material, then we can do worse than use examples from the social sciences – or even works of fiction – to suggest some of the forms of everyday social behaviour that underlie the archaeological record (Bagnall, 1997, 238). The tendency among archaeologists and ancient historians to look to the social sciences for comparative data and material to orient our theoretical approach to the ancient evidence has in many cases been accompanied by a keen aware- ness of the fundamental differences between the two data-sets (see, for example, Will, 1985[Q], discussed in Bagnall, 1995; important theoretically-informed mairs.indd 23 18/04/2008 14:09:43 24 Rachel Mairs studies on ancient ethnicity include Goudriaan, 1988; Hall, 1997, 2002; Jones, 1998; Smith, 2003; Van Soldt et al., 2005). Even when the questions of how and whether we should compare cultural and ethnic interaction in ancient and modern societies are not directly addressed, discussions of the Hellenistic period in particular often have a distinctly modern colonial flavour. Modern colonial societies provide a rich but problematic repertoire of potential com- parisons for the Hellenistic world. Pertinent issues, such as intermarriage and elite acculturation, can be vividly illustrated with recourse to case studies with which a modern audience can be expected to have a certain familiarity, even if this familiarity derives largely from the mass media or fictional accounts. Green, for example, imagines local peoples hellenizing and vying to join the gymna- sium, ‘like Indians under the British Raj angling for the entrée to European club membership’(Green, 1990, 116). The parallel is extended into a number of different areas: the acceptance of English as a lingua franca, and the appetite of numerous educated Indi- ans for such plums of power as they could grab within the system as it stood (along with the social mores of club or cantonment), in no way mitigated the deep-abiding resent- ment of British rule, much less made any inroads against India’s own longstanding cultur- al and religious traditions… Further, just as a surprising number of Englishmen, despite their own rigid caste system and xenophobic assumptions, were fatally seduced by the lure of Eastern mysticism, so the Indo-Greeks, in a very similar situation, capitulated to some highly un-Greek local influences before they were done. Indian legends and Indian scripts invaded their coinage. Even if the notion of portraying the Buddha in human form was a Greek innovation, their sculpture and reliefs and architecture absorbed far more than they imposed. (Green, 1990, 320, cf. 350.) In one regard, this analogy serves its purpose well, highlighting the range of motivations and strategies exercized by local elites in their selective adoption of the cultural traits of the dominant power. On the other hand, little insight is given into the social processes underlying acculturation in the opposite direc- tion, and the question of whether the two societies discussed are even directly comparable is not addressed. The best approach, in using more recent historical or anthopological data to shed light on the functioning of ancient societies, is therefore to practise a degree of theoretical introspection and to be explicit in one’s methodologic- al strategies. The work of Jones (1998), in particular, has laid the ground rules for how we define and approach ethnic identity in the archaeological record, drawing on the sociological school established by Fredrik Barth (Barth, 1969). Although his ideas have subsequently been much discussed and refined, Bar- th’s major contribution was the definition of ethnicity as a constructed identity mairs.indd 24 18/04/2008 14:09:44 The settler community in Hellenistic Bactria and Arachosia 25 predicated upon common culture and descent, but bearing a variable relation- ship to an individual or group’s objective cultural traits and genetic origins. Which specific traits (such as language, artistic style, religious practice and physical appearance) are considered to carry ethnic connotations is a vari- able that can differ considerably from one group to another. The implications of this for the analysis of archaeological material are profound; in attempting to locate expressions of ethnic identity in the archaeological record, it is not the form or style of individual artefacts that should be the focus of our atten- tion, but rather the ways in which material culture was used and perceived. Differences in the forms of material culture used in different social and pol- itical spheres (religion, the home, mortuary contexts, public institutions) are often especially revealing. Identifying the precise points and spheres of inter- action between different cultural traditions, and – where the evidence permits – aspects of non-material culture (such as language use) also represents a pro- ductive approach. Although I do not, of course, claim that California in the early twentieth century presents us with a model for Hellenistic society, Steinbeck’s paisano can be used to illustrate some of these points, both about the nature of an ‘ethnic’ as opposed to a ‘cultural’ identity (or descent group), and about the differing criteria for membership of an ethnic group. The author’s definition, crucially, does not give us a straightforward answer to the question of what a paisano ‘is’, but focuses rather on how his identity is constructed. There is a conflict between the paisano’s claim that he is ethnically Spanish, and the way in which an outsider might categorize him on the basis of his skin colour and non-European descent, as well as his low socioeconomic status. The fic- tional paisano displays his natural skin colour in support of his claim to Span- ish ethnicity; an outside observer, on the other hand, might be more inclined to prioritize factors such as accent or known ancestry. Modern examples abound of groups that lay an especial emphasis on one or more other physic- al or cultural traits as being the defining criteria of their ethnic identity (lan- guage and religion being among the most popular), to the exclusion or lesser prioritization of others. The overt statement of such criteria is common. That such classification and reinforcement is not merely a product of modern intro- spection can be demonstrated by the fifth century bce Greek historian Hero- dotus’s itemization of the constituent components of Greekness (hellenikon): ‘our common blood, common tongue, common cult places and sacrifices and similar customs’ (8.144.2; Hall, 2002, 189). This has become a much-cit- ed definition of Greek identity. As noted by Hall (2002, 189ff.), however, we should be cautious in treating it as either definitive or universally accepted by mairs.indd 25 18/04/2008 14:09:44 26 Rachel Mairs contemporaries. What Herodotus’s definition usefully demonstrates, however, is the common strategy employed by ethnic groups to construct a communal identity from the cultural characteristics that members of the group are per- ceived to share. My aim, in the following discussion, is to demonstrate how this concept of ethnicity as a constructed identity predicated upon the highly selective use of cultural traits to reinforce group membership can provide us with a practical and constructive approach to archaeological data. The evidence from Hel- lenistic Bactria and Arachosia is notoriously sparse and problematic. In many respects, the goal of presenting a comprehensive picture of Greek colonial soci- ety in the region, down to the minutiae of daily life, is an unrealistic one; the exercise would be still more difficult for those groups that did not share the level of political and social power enjoyed by the Greeks. What is within the scope of the evidence is to identify some of the key cultural processes at work. The gradual evolution of the Greek settlers’ notion of their own identity is, it should be stressed, just one of these. The analysis of the construction of ethnic identity in modern societies is much aided by the opportunity to contrast observable behaviour with group members’ perceptions of that behaviour. With the archaeological record, this is a more difficult strategy to successfully implement. In cases in which we have a substantial contemporary corpus of textual evidence, it is possible to compare a group’s material culture with their own statements about their ethnic identity. It is also beneficial, however, to consider the ancient record more holistically, looking for patterns that cut across boundaries of genre between text and arte- fact. In the case of the evidence from Hellenistic Bactria and Arachosia, this boundary is an especially porous one. The few surviving written texts demand to be considered as pieces of material culture in their own right; they include inscriptions on jars containing trade goods, and on objects dedicated at temples. The following case studies will attempt to utilize the full range of archaeological and textual evidence at our disposal, to discern ‘ethnic’ behaviour in the Greek settlements of the region. The first example is taken from the Greek city of Ai Khanoum, in Bactria, where artefacts and architectural forms from the site reveal a diverse range of cultural influences. My argument, developed in greater detail elsewhere (Mairs, 2007; forthcoming, 2008), is that this apparent cultural chaos is in fact carefully ordered, and reveals the role played by ethnic identity in the city’s civic organization and day-to-day life. The second case study, considered in the following section, will examine one specific category of evidence – Greek inscriptions – over a wider geographical area and longer time period. Despite the interpretative difficulties presented by mairs.indd 26 18/04/2008 14:09:44 The settler community in Hellenistic Bactria and Arachosia 27 these inscriptions, many of which lack a secure archaeological context, I sug- gest that here too we can identify some of the strategies adopted by contempor- ary populations to define and assert Greek identity. Case Study: Ai Khanoum The city of Ai Khanoum lies at the junction of the Oxus (Amu Darya) and Kokcha rivers, in what is now north-eastern Afghanistan. Although it is occa- sionally identified as ‘Alexandria-Oxeiana’, its ancient name in fact remains uncertain (Rapin, 2003). Little archaeological material has been recovered from pre-Hellenistic strata, but much has been found from the Hellenistic city, from the period between its foundation in the late fourth century bce and its destruction by invaders in the mid-second century bce. It was excavated by the Délégation archéologique française en Afghanistan (DAFA) in the 1960s and 1970s, and was extensively pillaged during the subsequent conflict in Afghani- stan (Bernard, 2001; for a critical review of publications on the site, see Fuss- man, 1996). The site’s long excavation history, and the spectacular nature of the finds made there, has given it a prominence in the modern academic literature that was probably not matched by its ancient status. Ai Khanoum was the site of a large (about 1.5 square kilometres) and important city, with fortifications and major public buildings. It commanded a strategic position on the River Oxus, controlling trade routes and access to natural resources (such as lapis lazuli) in its regional hinterland. It was, however, only one of a number of major Graeco- Bactrian settlement sites in the region, and by no means the most important. The political capital was at Bactra (present-day Balkh), where excavation has, to date, been much more limited in extent (there have been several excavation seasons since 2002, but these have not yet been fully published: for updates see www.dafa.org.af). Other contemporary settlement sites in the region have also been subject only to more limited excavation (for example: Termez (see Leriche et al., 2001, Part I); Sakhsanokhur (see Litvinskii and Mukhitdinov, 1969); and Takht-i Sangin (see Litvinskii and Pichikyan, 2000)). It is therefore important not to treat Ai Khanoum as representative of the region as a whole. What it does present is a useful case study in which we can examine the operation of public ‘ethnic behaviour’ within a single settlement site. The material culture and architectural remains from Ai Khanoum present an eclectic range of cultural influences. The city’s major public institutions, in par- ticular, were built according to very different architectural models: the adminis- trative quarter combines Greek-style columns with a plan reminiscent of Persian mairs.indd 27 18/04/2008 14:09:44 28 Rachel Mairs palaces; there is a typically-Greek semi-circular theatre built into the hillside; and the city’s major temple is built in a form descended from a Mesopotamian prototype. There is evidence of strong commercial and cultural contacts with the Mediterranean world (for olive oil, see Bernard, 1980, 442; for ceramics, see Gardin, 1990; for Greek dramatic and philosophical works on papyrus, see Lerner, 2003), the Central Asian steppes (see examples in Francfort, 1984) and, especially in later periods, with the Indian subcontinent (Rapin, 1996). As I have argued elsewhere (Mairs, 2007, and forthcoming, 2008), the most constructive response to the apparent contradictions in cultural behaviour at the site is to adopt an ‘ethnic’ model. Ethnic identity, as discussed above, is constructed around a selected group of cultural traits. At Ai Khanoum, effort was made to assert a strong Greek public and civic identity, but not every form of public behaviour was necessarily considered to have strong ethnic connota- tions. Although a mixture of Greek and local Iranian names occur in the eco- nomic texts from the city’s treasury (Grenet, 1983; Rapin, 1992), the language of the administration was Greek, and Greek is also the language used in the few public inscriptions. The large theatre (Bernard, 1981, 113) and the discovery of Greek literary and philosophical texts (Lerner, 2003; Rapin, 1992, 115–130) are testimony to the presence of Greek intellectual culture. The city also had a gymnasium (Veuve, 1987)[Q], the quintessential public institution of the Greek world, which was the focus for cultural and scientific as well as athletic pursuits. A Greek inscription from the Ai Khanoum gymnasium bears a dedi- cation to the gods Hermes and Herakles, traditionally associated with Greek gymnasial culture (Robert, 1968). Finally, the colonnaded tomb-shrine of the city’s founder, Kineas, bears an inscription quoting from the famous text of ethical precepts set up at Delphi, the cult site par excellence of the Greek world (Bernard et al., 1973). An epigram appended to the inscription emphasizes that this text was copied down at Delphi itself, thus stressing Ai Khanoum’s direct connections to a wider Greek world (Robert, 1968). The areas at Ai Khanoum in which we find public behaviour that is less overtly Greek are perhaps surprising. They include the city’s major temple, known in the modern literature as the Temple with Indented Niches because of its distinctive architectural plan (most concisely and accessibly described in Downey, 1988). This plan is found at a number of other sites throughout the Near East, and has its ultimate origins in Mesopotamia (discussed in Mairs, forthcoming, 2008). The temple sanctuary yielded a diverse selection of artefacts, and evidence of an equally diverse range of cult practices (Francfort, 1984). These include burned offerings, libations to the earth and the dedication of statues. Only fragments of the central cult image are preserved, and, in contrast to most of the rest of mairs.indd 28 18/04/2008 14:09:44 The settler community in Hellenistic Bactria and Arachosia 29 the material from the sanctuary, these suggest a statue in Greek style. A thun- derbolt motif on the statue’s sandal has led to the argument that it represents a form of the Greek god Zeus, perhaps syncretized with another local deity (Bernard, 1969, 338, 340–341; Grenet, 1991). The lack of a more traditional Greek temple at Ai Khanoum demands that we treat the Temple with Indented Niches not as an eccentric aberration, but as an integral part of the city’s public life. It was located on the main street, only a short walking distance from the shrine of the founder, the main administrative quarter and the theatre, and we cannot assume that it served an entirely separate, isolated constituency. Yet it is clear that people chose to display very different forms of cultural behaviour within the temple’s precincts than they did in some of the city’s other major institu- tions. To an external observer or analyst, this might appear to be contradictory, even hypocritical behaviour. However, it represents precisely the disjunction between objective cultural characteristics and perceived ethnic identity noted above. At Ai Khanoum, there were places in which it was important to assert a strong Greek identity through conspicuous adherence to Greek cultural trad- itions or the emphasis of connections to the wider Greek world. The material from the Temple with Indented Niches, however, demonstrates that these same individuals might engage in practices with very different cultural origins. The size, status and location of the temple suggest that these practices can only have been performed openly, with no perceived contradiction on the part of the agents with their behaviour elsewhere. The evidence from Ai Khanoum there- fore allows us to identify some of the cultural characteristics that were consid- ered central to Greek ethnic identity in the city, and those places in which the variety and alteration of practice were not considered to represent a threat to individual or communal ethnic identity. Case Study: Greek inscriptions from Hellenistic Bactria and Arachosia The evidence from Ai Khanoum presents us with an important opportunity to examine the functioning of ethnic identity in the Graeco-Bactrian state in microcosm through the day-to-day life of a single settlement. As I have argued above, the layout and distribution of material culture at the site reveals a func- tional division between those places in which it was appropriate to make overt statements of Greek identity, and those in which greater flexibility in cultural display was permissible. As previously suggested, it is the use of material cul- ture, rather than any one specific form, that provides the best index to ethnic behaviour. mairs.indd 29 18/04/2008 14:09:44 30 Rachel Mairs Although we currently lack the raw data to conduct a similar depth of analy- sis on other sites from Hellenistic Bactria and Arachosia, one category of evi- dence – Greek inscriptions – permits us to adopt a wider regional perspective. An inevitable caveat (but one nevertheless worth stating) is that formal inscrip- tions present us with an elite viewpoint, representing the concerns of the small segment of the population that was wealthy enough to commission inscrip- tions, or educated enough to read them. The Greek epigraphic material from the region is collected in Canali de Rossi (2004), with two subsequent discover- ies published in Bernard, Pinault and Rougemont (2004). If we omit the briefer, less formal texts from Canali de Rossi’s compendium, such as brick marks and texts on ceramic vessels, Greek inscriptions are currently known at a relatively small number of sites: in Bactria, from Ai Khanoum, Takht-i Sangin, Zhiga- tepe and Kuliab; and in Arachosia, from Kandahar and its vicinity. It is my intention to present a brief survey of the evidence before discussing its ethnic implications. Full texts and translations of the inscriptions can be consulted in Canali de Rossi (2004), as well as in the individual publications cited below, with the most recent reports in [Q]SEG LII (2002 [2006], 1514–1529). Arachosia The city of Kandahar (ancient Alexandria-in-Arachosia) has to-date yielded the only Greek inscriptions from southern Afghanistan. The earliest (Fraser, 1979) dates to around 275 bce; the block on which it was inscribed had been reused in a house of a later period, which means that we lack any information on its original archaeological context beyond that which can be inferred from the contents of the inscriptions itself. The text is fragmentary; it states that it was set up by the son of a man named Aristonax (his given name is missing) ‘in this [temple] sanctuary’. The words ‘my saviour’ and ‘wild animal’ suggest that it is a thanks-offering for the dedicator’s rescue from an animal attack; whether a deity or a more corporeal saviour is being commemorated is unknown. The text also contains a probable reference to the city of Alexandria-in-Arachosia itself, although only the letters ‘A-L-E-X’ are preserved. The next two inscriptions date to the 250s bce, and are remarkable for the insight they give into the status and vitality of the Greek language under Indian rule. As noted above, Arachosia and its Greek settler population had passed under the administration of the Indian Maurya Empire by the terms of a treaty of 303 bce. The emperor Aśoka (fl. c. 269–232 bce) – grandson of Chandragupta, the treaty’s original signatory – was notable as the first Indian ruler to publicly espouse Buddhist ethical ideas in a series of decrees inscribed on pillars and mairs.indd 30 18/04/2008 14:09:44 The settler community in Hellenistic Bactria and Arachosia 31 rock-faces across the Maurya Empire (Allchin and Norman, 1985; Thapar, 1997). At Kandahar, instead of the Indian Prakrit dialects used elsewhere, these edicts were set up in Greek and in Aramaic. Aramaic was also used for two briefer Aśokan edicts, discovered in eastern Afghanistan near the modern border with Pakistan (Dupont-Sommer, 1969; Ito, 1979; Mukherjee, 1984); its presence in the region is a legacy of earlier Persian imperial control (for Hellenistic and Achaemenid period Aramaic texts from Bactria, see Rapin, 1992; Shaked, 2004, 105). The first Greek decree (Schlumberger, 1964), inscribed on stone blocks that might originally have formed the wall of a public building, is a direct trans- lation of two Aśokan edicts known from inscriptions elsewhere in the Mauryan Empire. Portions are missing from the beginning and end, but 22 lines of text are preserved. It describes the Emperor’s remorse after the bloody conquest of Kalinga in eastern India, and the new ethical code he wishes his subjects to fol- low: humility, compassion, vegetarianism, respect for one’s elders and kindness towards one’s dependents are among the virtues emphasized. The second Greek edict, cut into a rock-face by the side of the main road into Kandahar, is much briefer and is accompanied by an Aramaic text (Schlumberger et al., 1958). The contents of the texts in the two languages run parallel to each other, but are not direct translations; nor does either text directly translate a known Indian original. Once again, the decree describes Aśoka’s conversion and efforts to impose a new Buddhist ethical code throughout his empire. As pronouncements of an outside power, the Greek Aśokan edicts present us with only limited information on the contemporary Greek community in the city. We have no evidence that indicates how Aśoka’s efforts at proselyt- ism (which included sending embassies to western Greek courts: Thapar, 1997, Appendix, MRE 13) were received. They do, however, provide an important index to the vitality of the Greek language in Kandahar in the middle of the second century bce. Not only are the decrees composed in grammatically and syntactically ‘good’ Greek, but it has also been suggested that their language reveals an attempt to translate the Aśokan edicts into Greek cultural vocabu- lary (Schlumberger, 1964). In the longer, monolingual inscription, there are only two instances in which the translator chose to transliterate Indian words rather than find Greek equivalents: bramenai and sramenai, terms for members of priestly castes or religious orders. Schmitt (1990, 50) suggests that this could reflect the fact that the Greek-speaking inhabitants of Kandahar were already familiar with such terms, rather than a failure on the part of the translator to find a good Greek translation. (For numerous examples of bilingualism, ‘trans- lationese’ and code-switching in inscriptions and texts from elsewhere in the ancient world, see Adams et al., 2002.) mairs.indd 31 18/04/2008 14:09:44 32 Rachel Mairs Whatever the inroads made by Indian culture and religion at Kandahar, the Greek inscriptions reveal that the Greek military settlement of the late fourth century bce had left a Greek-speaking (and/or Greek-reading) population pow- erful and numerous enough to be courted in their own language. The fourth and final Greek inscription from Kandahar comes from a later period, when the region was once again under Greek political control following its conquest by the Greek kings of Bactria. It emerged from the antiquities market, and therefore lacks a secure provenance or archaeological context (Bernard, Pin- ault and Rougemont, 2004). From internal clues, it can be deduced that it was originally mounted on the wall of a tomb. The inscription is in verse and its language is educated and archaising, in conscious imitation of the great Greek epic poet Homer. The first letters of each line combine to spell out an acrostic: ‘Through/by Sophytos, son of Naratos’. Sophytos tells how his formerly wealthy family was ruined by an unspecified misfortune. The young Sophytos, who had received a good Greek education, went out into the world as a merchant, amassed great wealth and returned home to restore his ancestral tomb and the good name of his family. The language, tone and cultural reference-points of the inscription are almost ostentatiously Greek. The names Sophytos and Naratos, however, are not of Greek origin; they have been identified as Greek transcrip- tions of Indian personal names (Pinault in Bernard, Pinault and Rougemont, 2004). What are we to make of this apparent adoption of Greek cultural norms by a person apparently of non-Greek descent? We know too little of Sophytos’s family background and his repertoire of forms of cultural behaviour (did he speak an Indian language in addition to his Greek?) to draw any firm conclu- sions about his ethnicity. But, as will be discussed further below, the cultural cues in his inscription enable us to make some points about the ways in which a contemporary inhabitant of Kandahar might attempt to project a strong Greek public identity. Bactria Two inscriptions from Bactria have already been briefly considered above: the inscriptions from the gymnasium and from the shrine of Kineas at Ai Kha- noum. Also from Ai Khanoum come a number of much briefer texts from the city’s cemetery, identifying those interred there (Canali De Rossi, 2004, num- bers 360–362). They include the (Greek) names Lysanias, Isidora and Kosmas, as well as more anonymous designations such as ‘the little one’. Unlike in Arachosia, where we have inscriptions only from the city of Kanda- har and its immediate vicinity, in Bactria Greek inscriptions have been found at mairs.indd 32 18/04/2008 14:09:44 The settler community in Hellenistic Bactria and Arachosia 33 sites other than Ai Khanoum: a fragmentary funerary epigram of a man named Diogenes from the indigenous Bactrian temple site of Zhiga-tepe in the Bactra Oasis, mentioning Hades, the Greek god of the dead (Bernard, 1987, 112–113); and a votive to the deified River Oxus from the Temple of the Oxus at Takht-i Sangin, by a man with the Iranian name Atrosokes (Litvinskii et al., 1985). An unprovenanced inscription, published in the same edition as the inscription of Sophytos from Kandahar (Bernard, Pinault and Rougemont, 2004), is said to come from the site of Kuliab, just north of the Oxus in modern Tajikistan. It is a dedication of an altar to the Greek goddess Hestia, ‘in the grove of Zeus’, by a man named Heliodotos, in honour of the local Greek king Euthydemos and his son Demetrios. Unfortunately, we lack archaeological evidence from Kuliab to provide this inscription with a good local context. Ethnic identity in the inscriptions At first glance, the Greek inscriptions from Hellenistic Bactria and Aracho- sia present us with some obvious clues to the ethnic identity of their authors and intended audience. The use of the Greek language is in itself significant; although we must assume that the linguistic landscape of the region was con- siderably more complex, a very few Aramaic texts are the only exceptions in an otherwise uniformly Greek epigraphic picture. Furthermore, the literary style of the Greek used is in several cases (Sophytos, the Aśokan edicts) highly sophisti- cated. Human and divine names are also potentially significant. The majority of the personal names (Aristonax, Diogenes, Kineas, Klearchos, Lysanias, Isidora, Kosmas, Triballos, Straton) are Greek in origin. The exceptions are Atrosokes (Iranian) and Sophytos and Naratos (probably Indian). If we leave the Aśokan edicts from Kandahar out of consideration, as decrees of an external political power, then the deities mentioned are also mostly Greek: Apollo, the Muses and the Fates (Sophytos), Hermes and Herakles (Ai Khanoum gymnasium), Hades (Diogenes, Zhiga-tepe), Hestia, Zeus and Fortune (Heliodotos, Kuliab). The only exception is the deified River Oxus. References to cult sites are gener- ally brief and relatively uninformative: ‘in this sanctuary’ (the inscription of the ‘son of Aristonax’); ‘in the sanctuary of Kineas’ (the Delphic inscription of Klearchos); and ‘in the grove of Zeus’ (Kuliab). To what extent, however, is the overwhelmingly ‘Greek’ picture that emerges from the inscriptions of the region a product of the nature of the evidence? In one or two cases, knowledge of the wider archaeological context enables us to see that matters were not quite so simple. The gymnasium and Delphic inscriptions mairs.indd 33 18/04/2008 14:09:44 34 Rachel Mairs from Ai Khanoum contain no hint of the city’s considerable cultural diversity. Two of the other Bactrian Greek inscriptions come from the excavation of cult sites, where the architecture and artefacts show only limited Greek influence (for Zhiga-tepe, see Pugachenkova, 1979; for Takht-i Sangin, see Litvinskii and Pichikyan, 2000). These cases should lead us to question some of the instances of inscriptions for which we have less evidence for their original context. We have no archaeological evidence from the ‘grove of Zeus’, for example, men- tioned in the Kuliab inscription of Heliodotos; might its description in a Greek inscription subtly misrepresent its actual nature? Like the Aśokan edicts from Kandahar, which make some effort to render Indian terms using a Greek cultural vocabulary, examples abound from other regions of the Hellenistic world of a Greek name masking a more complicated reality. In Egypt, documents on papyri preserved in the region’s dry climate present us with an enviable degree of insight into intermarriage and accultura- tion between Greek settlers and their descendents, and the indigenous Egyptian population. One interesting phenomenon is the practice of dual-naming, where the same individual might use a Greek name or an Egyptian name according to context (Clarysse and Thompson, 2006, 218ff.). Often, circumstantial evidence from another document is the only factor that enables us to identify when this is taking place. The Greeks also had a long tradition of referring to foreign gods by the name of a Greek equivalent, rather than transcribing the indigenous name. Syncretism between a Greek and a local deity was relatively common; in Hellenistic period Commagene (eastern Turkey), the Greek god Zeus was fused with the local form of the god Ahura-Mazda to form the composite deity Zeus-Oromazdes (Sanders, 1996). In Hellenistic Bactria and Arachosia, the use of non-Greek names in Greek texts might suggest that the former practice (the adoption of dual personal names) was not a common strategy. The use of Greek names to refer to non- Greek gods, however, is strongly suggested by a number of pieces of evidence. The material from the Temple with Indented Niches at Ai Khanoum is testi- mony to a certain level of flexibility in religious practice; we should expect this to be accompanied by a similar degree of flexibility in the way these religious practices might be referred to. It is significant that the only occurrence of a non- Greek divine name in the Greek epigraphic material is that of the deified River Oxus. We might suggest that the god ‘Oxus’ was simply untranslatable, or of sufficient local relevance and importance not to require any ‘translation’ for local Greeks. This inscription comes from the pedestal of a small Greek-style statuette representing a satyr-like figure, showing that, even where a local name was retained, Greek iconography might still be introduced. It is possible – in mairs.indd 34 18/04/2008 14:09:45 The settler community in Hellenistic Bactria and Arachosia 35 fact, probable – that the other Greek divine names might also be the prod- uct of the Greek settlers using familiar terms to refer to local gods. The name of the goddess Hestia, to whom Heliodotos’s altar at Kuliab is dedicated, also appears in Greek script on ceremonial drinking-horns from the site of Nisa, in the Parthian Empire, where it is probable that a local goddess is in fact intended (Bernard, 1991). In this regard, these inscriptions therefore present us with some important negative evidence. They have a tendency to portray contemporary local society as more monoculturally Greek than it in fact was, which is in itself indicative of a process of ethnic ‘rationalization’ of a complex multicultural society. In a more positive light, the epigraphic evidence also indicates some of the means by which Greek identity was actively perpetuated and asserted. The inscrip- tion of Kineas from Ai Khanoum, with its emphasis on the city’s connection to the great Greek cult site at Delphi, has already been briefly considered above. Several of these inscriptions also suggest that Greek education and intellectual culture were highly valued as markers of Greek ethnicity, at least among the economic and social elite. The papyrological evidence makes contemporary Egypt our best source of information on Greek education in the Hellenistic period (Cribiore, 1996, 2001; Morgan, 1998; Thompson, 2007). By far the most commonly-attested author in literary papyri and schooltexts is Homer, who occupied a position equal to or greater than that of Shakespeare in modern English education. Many schooltexts from Hellenistic Egypt also contain lists of proverbs and moral precepts, expected to instill in students a sense of common Greek ethical values. Those who reached the highest levels of the Hellenistic educational system would study rhetoric, and the Egyptian city of Alexandria became a centre where great philosophers, poets and scientists worked under royal patronage. The Greek inscriptions from Hellenistic Bactria and Aracho- sia present a number of indications that their authors had received a level of education far beyond that required for basic literacy. The case of Sophytos is the most striking. He tells the reader directly that in his youth he cultivated the virtues of Apollo and the Muses, patrons of Greek intellectual culture. The literary style of the inscription also betrays him as a man familiar with the Homeric epics, and not above the flowery style and ostentatious use of recher- ché vocabulary that are the hallmarks of many works of Hellenistic literature. Even if Sophytos’s origins remain uncertain, there can be no doubt that he presents himself in a cultural mode that is thoroughly Greek. By emphasizing his education and cultivation of traditional Greek virtues, he could even be attempting to demonstrate that he possesses many of the key criteria for ethnic group membership. mairs.indd 35 18/04/2008 14:09:45 36 Rachel Mairs The archaeological and epigraphic evidence from Ai Khanoum also provides testimony to the strength of Greek intellectual culture in the region, and sug- gests that it might be an important component of Greek identity. The inscrip- tion from the sanctuary of the city’s founder, Kineas, displays ethical maxims from Delphi. Texts on perishable materials from the treasury preserve intrigu- ing fragments of dramatic and philosophical texts (Rapin, 1992, 115–130; Lerner, 2003). The city had a theatre, and a water-spout from a public fountain is in the shape of a mask used by a comic actor (Leriche, 1987). Two sun-dials were also excavated, showing the presence of at least some individuals with scientific training or interests (Veuve, 1987)[Q]. The use of Greek names and terminology to conceal a more diverse cultural millieu, and the conspicuous display of Greek literate education and culture, suggest that setting up a Greek inscription was in itself an action with profound ethnic connotations. The inscriptions from Bactria fall into three broad categor- ies. First, funerary inscriptions, including the epigram of Diogenes from Zhiga- tepe and the briefer series of names from the cemetery at Ai Khanoum. Secondly, religious dedications, specifically from important cult sites where we also have evidence of royal patronage. The Temple of the Oxus at Takht-i Sangin, the site from which the dedication of Atrosokes to the Oxus was recovered, also yielded states[Q] representing Graeco-Bactrian kings (Bernard, 1990, 57). Although we have no archaeological data from the Grove of Zeus at Kuliab, it was evidently an appropriate place for Heliodotos to dedicate an altar on behalf of the king and his son. The third group includes inscriptions that also have a religious or funerary element, but whose primary focus I would argue to be civic, associ- ated with the Greek community and city of Ai Khanoum (the Delphic epigrams and the gymnasium inscription). These three broad categories are not perhaps, in and of themselves, of any great significance; they are all contexts in which it is to be expected that important individuals might choose to set up a public inscription, for whatever purpose. In the case of Bactria, and in particular Ai Khanoum and its hinterland, however, it is also possible to make some points about where people do not set up inscriptions. The Temple with Indented Niches yielded no Greek inscriptions; nor did Ai Khanoum’s main adminis- trative quarter yield any written material, with the exception of the economic and literary texts from the treasury (Rapin, 1992). The places in which Greek public inscriptions do occur – the gymnasium and the tomb-shrine of the city’s founder – are precisely those which, as argued above, formed the focus for the inhabitants’ sense of their Greek identity. The brief inscriptions from the cem- etery also suggest that it was considered important to project a Greek identity in death (Mairs, 2007). In Arachosia, the picture is tempered by the fact that for mairs.indd 36 18/04/2008 14:09:45 The settler community in Hellenistic Bactria and Arachosia 37 most of this period the region was not under Greek political control; but here too it is significant that Greek inscriptions could be used as a key vehicle for the assertion of a Greek identity in important religious and funerary contexts, and as a medium for appealing to the community by an outside power. The obvious danger in drawing all of these inscriptions together and trying to make them tell us something more in conjunction than they each do separately, is that the archaeological picture from the region is fragmentary. As the recent discovery of the inscriptions of Sophytos and Heliodotos demonstrates, there is always the potential for important new material to emerge from fresh excava- tions or from the antiquities market. The current excavations at the city of Bactra are perhaps the best candidate to revolutionise our knowledge of the nature and distribution of Greek inscriptions in Hellenistic Bactria. In comparison with many cities from other regions of the Hellenistic world, Ai Khanoum, despite its extensive excavation, has yielded relatively few inscriptions. Archaeological field surveys in the city’s hinterland have also failed to uncover any more epi- graphic material. The other known inscriptions from Bactria occur in smaller but relatively important settlement and cult sites. On this basis, we might argue that the reason we do not have many Greek inscriptions from Bactria is that the provincial cities and countryside simply did not produce many. Given the concentration of the known inscriptions in centres in which we have evidence of a strong civic identity or royal cult patronage, we might also posit that the city of Bactra, as the royal capital, will yield more inscriptions, demonstrating how the royal authorities used the Greek epigraphic tradition to reinforce their own power and the community’s sense of Greek identity. Conclusion Making sense of the archaeological and epigraphic material from Hellenistic Bactria and Arachosia, with its complicated intersections of cultural traits, demands a level of engagement with the social processes behind its produc- tion. The principal stratagem adopted in the present study has been the use of the concept of ethnic identity to structure our approach to the evidence; this focus must not detract, of course, from consideration of the region’s wider social and cultural dynamics (cf. Erdösy, 1995, 95). Ethnic identity, as discussed above, is constructed around a selective repertoire of cultural traits. Knowledge of modern anthropological case studies can give us important insights into the process by which these criteria for ethnic group membership can be selected and asserted. The greatest difficulty in analysing the archaeological evidence is mairs.indd 37 18/04/2008 14:09:45 38 Rachel Mairs in deciding which specific traits were considered important in any given society. Contemporary textual evidence may be of some assistance here. But it is also possible to analyse architectural remains and material culture in ways that reveal variations in cultural behaviour across a particular site or region, and allow us to consider the implications such variations might have for ethnic identity. Where such an interpretative strategy can be successfully implemented, we have an opportunity to develop more nuanced analyses of the Hellenistic world, and other ancient societies in which the material culture presents an apparent conflict between two or more cultural forces. If we admit that the relationship between cultural practices (such as religion or language use) and a sense of identity can be a matter of negotiation, rather than adherence to strict rules of conduct, then this at least provides us with a constructive way of approaching the evidence, even if things remain far from straightforward. It also serves to make us intensely suspicious of any suggestion of conflict between an individ- ual’s projected identity and cultural practice; this is nearly always a matter of rationalization, or ‘neutralization’ of behaviour, rather than direct conflict. It will hardly do to suggest that the populations of Hellenistic Bactria and Arachosia existed in a constant state of postmodern multicultural angst. On the contrary, strategies were employed to allow individuals and groups to make sense of the social and cultural forces at work in their society. The reason why people adopt such ‘ethnic’ behaviour involving boundary maintenance and the public reinforcement of communal cultural attributes is that, for their own pur- poses, it works. The purpose of the exercise is not to enforce cultural norms, but to regulate them in a way that responds to a group’s practical needs. This could still suggest, to an external observer, a degree of contradiction or hypocrisy, but for the agents themselves, the construction and reinforcement of an ethnic identity allows them to overcome – or conveniently ignore – these cultural con- tradictions. For people who considered themselves Greek, for example, would it have mattered that the temple at which they worshipped followed an architec- tural plan we describe as Mesopotamian (Mairs, forthcoming, 2008)? I would suggest that this apparent stylistic contradiction would have been of little or no import to those secure in their notion of their own Greek identity. An outsider, of course, might have perceived this differently. The development and evolution of a sense of Greek ethnic belonging among the settler community of Hellenistic Bactria and Arachosia and their descend- ents leads us to revisit the case of the Branchidae, the community of Greek descent settled in the east under the Persian empire. Alexander the Great’s Greek and Macedonian troops found Central Asia a hostile environment, and were alienated by its dramatic cultural and climatic differences from their mairs.indd 38 18/04/2008 14:09:45 The settler community in Hellenistic Bactria and Arachosia 39 homelands in the eastern Mediterranean. The Branchidae still considered themselves ethnically Greek, but the army of Alexander found their culture unfamiliar and degenerate. In the material culture of the subsequent Greek set- tler communities of the region, we begin to lose this sense of stark encounter with the ‘other’, as the criteria of Greek identity, and with this the bounds of the alien, are gradually and subtly reset. It is probable that the culture of the Hellenistic-period Greeks of Bactria and Arachosia would have seemed as for- eign to a contemporary visitor from the western Greek world as the Branchi- dae had seemed to their own ancestors. Yet the population of the region were apparently satisfied that they fulfilled the criteria of Greek ethnic belonging; their inscriptions show that they were able to rationalize a culturally diverse range of practices through a Greek cultural vocabulary, and that even individ- uals with Indian names might vaunt their Greek education. Public expression of the region’s cultural and linguistic diversity emerged only relatively late in its history, yet it is clear that this represents the culmination of a much longer process of cultural and ethnic evolution. References Adams, J.N., Janse, M. and Swain, S. (2002), Bilingualism in Ancient Society: Language Con- tact and the Written Text, Oxford, Oxford University Press. Allchin, F.R. and Norman, K.R. (1985), ‘Guide to the Asokan inscriptions’, South Asian Stud- ies, 1, 43–50. Bagnall, R.S. (1995), Reading Papyri, Writing Ancient History, London and New York, Routledge. Bagnall, R.S. (1997), ‘Decolonizing Ptolemaic Egypt’, in P.G.P. Cartledge and E.S. Gruen (eds), Hellenistic Constructs: Essays in Culture, History and Historiography, Berkeley, CA, University of California Press, 225–241. Barth, F. (1969), Ethnic Groups and Boundaries: The Organization of Cultural Difference, Bergen and London, Universitetsforlaget and Allen & Unwin. Bernard, P. (1969), ‘Quatrième campagne de fouilles à Aï Khanoum (Bactriane)’, Comptes Rendues de l’Academie des Inscriptions et Belles-Lettres, 313–355. Bernard, P. (1980), ‘Campagne de fouilles 1978 à Aï Khanoum (Afghanistan)’, Comptes Ren- dues de l’Academie des Inscriptions et Belles-Lettres, 435–359. Bernard, P. (1981), ‘Problèmes d’histoire coloniale grecque à travers l’urbanisme d’une cité hellénistique d’Asie centrale’, in 150 Jahre Deutsches Archäologisches Institut 1829-1979, [Q], 108–120. Bernard, P. (1985), Fouilles d’Aï Khanoum IV: Les monnaies hors trésors, Questions d’histoire gréco-bactrienne, Paris, de Boccard. Bernard, P. (1987), ‘Le Marsyas d’Apamée, l’Oxus et la colonisation séleucide en Bactriane’, Studia Iranica, 103–115. mairs.indd 39 18/04/2008 14:09:45 40 Rachel Mairs Bernard, P. (1990), ‘L’architecture religieuse de l’Asie centrale à l’époque hellénistique’, in Akten des XIII Internationalen Kongresses für Klassische Archäologie, Mainz am Rhein, Verlag Philipp von Zabern, 51–59. Bernard, P. (1991), ‘Les rhytons de Nisa: à quoi, à qui ont-ils servi?’, in P. Bernard and F. Grenet (eds), Histoire et cultes de l’Asie centrale préislamique: Sources écrites et docu- ments archéologiques, Paris, Éditions du CNRS, 31–38. Bernard, P. (2001), ‘Ai Khanoum en Afghanistan hier (1964–1978) et aujourd’hui (2001): un site en péril’, Comptes Rendues de l’Academie des Inscriptions et Belles-Lettres, 971–1029. Bernard, P., le Berre, M. and Stucki, R. (1973), ‘Architecture: le téménos de Kinéas’, in P. Ber- nard (ed.), Fouilles d’Aï Khanoum I (Campagnes 1965, 1966, 1967, 1968), Paris, Klinck- sieck, 85–102. Bernard, P., Pinault, G.-P. and Rougemont, G. (2004), ‘Deux nouvelles inscriptions grecques de l’Asie Centrale’, Journal des Savants, 227–356. Biville, F. (2002), ‘The Graeco-Romans and Graeco-Latin: a terminological framework for cases of bilingualism’, in J.N. Adams, M. Janse and S. Swain (eds), Bilingualism in Ancient Society: Language Contact and the Written Text, Oxford, Oxford University Press, 77–102. Bopearachchi, O. (1991), ‘Les derniers souverains indo-grecs: une nouvelle hypothèse’, in P. Bernard and F. Grenet (eds), Histoire et cultes de l’Asie centrale préislamique: Sources écrites et documents archéologiques, Paris, Éditions du CNRS, 235–242. Bopearachchi, O. (1991), Monnaies gréco-bactriennes et indo-grecques: catalogue raisonné, Paris, Bibliothèque nationale. Bosworth, A.B. (1996), Alexander and the East: The Tragedy of Triumph, Oxford, Clarendon. Briant, P. (1984), L’Asie centrale et les royaumes proche-Orientaux du premier millénaire (c. VIIIe – IVe siècles avant notre ère), Paris, Éditions Recherche sur les Civilisations. Canali De Rossi, F. (2004), Iscrizioni dello Estremo Oriente Greco, Bonn, Rudolf Habelt. Carratelli, G.P. and Garbini, G. (1964), A Bilingual Graeco-Aramaic Edict by Asoka: The First Greek Inscription Discovered in Afghanistan, Rome. Clarysse, W. and Thompson, D.J. (2006), Counting the People in Hellenistic Egypt, Volume 2: Historical Studies, Cambridge, Cambridge University Press. Cribiore, R. (1996), Writing, Teachers and Students in Graeco-Roman Egypt, Atlanta, GA, Scholars Press. Cribiore, R. (2001), Gymnastics of the Mind: Greek Education in Hellenistic and Roman Egypt, Princeton, NJ, Princeton University Press. Downey, S.B. (1988), Mesopotamian Religious Architecture: Alexander through the Parthians, Princeton, NJ, Princeton University Press. Dupont-Sommer, A. (1969), ‘Une nouvelle inscription araméenne d’Asoka trouvée dans la vallée du Laghman (Afghanistan)’, Comptes Rendues de l’Academie des Inscriptions et Belles-Lettres, 158–173. Erdösy, G. (1995), ‘The prelude to urbanization: ethnicity and the rise of late Vedic chief- doms’, in F.R. Allchin (ed.), The Archaeology of Early Historic South Asia: The Emer- gence of Cities and States, Cambridge, Cambridge University Press, 75–98. Foucher, A. and Bazin-Foucher, E. (1942–1947), La vieille route de l’Inde de Bactres à Taxila, Paris, Les Éditions d’Art et d’Histoire. mairs.indd 40 18/04/2008 14:09:45 The settler community in Hellenistic Bactria and Arachosia 41 Francfort, H.-P. (1984) Fouilles d’Aï Khanoum III: Le sanctuaire du temple à niches indentées, 2 – Les trouvailles, Paris, de Boccard. Fraser, P.M. (1979), ‘The son of Aristonax at Kandahar’, Afghan Studies, 2, 9–21. Fraser, P.M. (1996), Cities of Alexander the Great, Oxford, Clarendon Press. Fussman, G. (1996), ‘Southern Bactria and northern India before Islam: a review of arch- aeological reports’, Journal of the American Oriental Society, 116, 243–259. Gardin, J.-C. (1990), ‘La céramique hellénistique en Asie centrale: problèmes d’interpretation’, in Akten des XIII Internationalen Kongresses für Klassische Archäologie, Berlin 1988, Berlin, Philipp von Zabern, 187–193. Goudriaan, K. (1988), Ethnicity in Ptolemaic Egypt, Amsterdam, J.C. Gieben. Green, P. (1990), Alexander To Actium: The Hellenistic Age, Berkeley, CA, University of Cali- fornia Press. Grenet, F. (1983), ‘L’onomastique iranienne à Aï Khanoum’, Bulletin de Correspondance Hel- lénique, 107, 373–381. Grenet, F. (1991), ‘Mithra au temple principal d’Aï Khanoum?’, in P. Bernard and F. Grenet (eds), Histoire et cultes de l’Asie centrale préislamique: Sources écrites et documents archéologiques – Actes du colloque international du CNRS (Paris, 22-28 Novembre 1988), Paris, Éditions du CNRS, 147–151. Hall, J.M. (1997), Ethnic Identity in Greek Antiquity, Cambridge, Cambridge University Press. Hall, J.M. (2002), Hellenicity: Between Ethnicity and Culture, Chicago, University of Chi- cago Press. Hammond, N.G.L. (1998), ‘The Branchidae at Didyma and in Sogdiana’, Classical Quarterly, 48, 339–344. Holt, F.L. (1988), Alexander the Great and Bactria: The Formation of a Greek Frontier in Central Asia, Leiden, E.J. Brill. Holt, F.L. (1999), Thundering Zeus: The Making of Hellenistic Bactria, Berkeley, CA, Univer- sity of California Press. Ito, G. (1979), ‘Aśokan inscriptions, Laghmân I and II’, Studia Iranica, 8, 175–183. Jones, S. (1998), The Archaeology of Ethnicity: Constructing Identities in the Past and Present, London, Routledge. Karttunen, K. (1997), India and the Hellenistic World, Helsinki, Finnish Oriental Society. Leriche, P. (1987), Fouilles d’Ai Khanoum V: Les remparts et les monuments associés, Paris, De Boccard. Leriche, P., Pidaev, C., Gelin, M. and Abdoullaev, K. (2001), La Bactriane au carrefour des routes et des civilisations de l’Asie centrale: Termez et les villes de Bactriane-Tokharestan: Actes du colloque de Termez 1997, Paris, Maisonneuve & Larose. Lerner, J.D. (2003), ‘The Aï Khanoum philosophical papyrus’, Zeitschrift fur Papyrologie und Epigraphik, 142, 45–51. Litvinskii, B.A. and Mukhitdinov, K. (1969), ‘The ancient fortification of Saksanokhur (Southern Tadzhikistan)’, Sovyetskaya Arkheologiya, 160–178. Litvinskii, B.A. and Pichikian, I.R. (1994), ‘The Hellenistic architecture and art of the Tem- ple of the Oxus’, Bulletin of the Asia Institute, New Series 8 (Studies from the Former Soviet Union), 47–66. mairs.indd 41 18/04/2008 14:09:45 42 Rachel Mairs Litvinskii, B.A. and Pichikyan, I.R. (2000), The Hellenistic Temple of the Oxus in Bactria (Southern Tadzhikistan), Volume 1: Excavations, Architecture, Religious Life, Moscow, Vostochnaya Literatura. Litvinskii, B.A., Vinogradov, Y.G. and Pichikyan, I.R. (1985), ‘The votive offering of Atrosokes, from the Temple of the Oxus in Northern Bactria’, Vestnik Drevnei Istorii, 85–110. Mairs, R.R. (2007), ‘Ethnicity and funerary practice in Hellenistic Bactria’, in H. Schroeder, P. Bray, P. Gardner, V. Jefferson and E. Macaulay-Lewis (eds), Crossing Frontiers: The Opportunities and Challenges of Interdisciplinary Approaches to Archaeology, Oxford, School of Archaeology, 111–124. Mairs, R.R. (forthcoming, 2008), ‘The “Temple with Indented Niches” at Ai Khanoum: eth- nic and civic identity in Hellenistic Bactria’, in O. van Nijf and R. Alston (eds), Cults, Creeds and Competitions in the Post-Classical Greek City, Leuven, Peeters. Morgan, T. (1998), Literate Education in the Hellenistic and Roman Worlds, Cambridge, Cambridge University Press. Mukherjee, B.N. (1984), Studies in Aramaic Edicts of Aśoka, Calcutta, Indian Museum. Parke, H.W. (1985), ‘The massacre of the Branchidae’, Journal of Hellenic Studies, 105, 59–68. Pugachenkova, G.A. (1979), ‘Zhiga-tepe (finds 1974)’, in I.T. Kruglikova (ed.), Ancient Bac- tria 2: Materials from the Soviet-Afghan Archaeological Expedition, Moscow, Nauka, Rapin, C. (1992), Fouilles d’Aï Khanoum VIII: La trésorerie du palais hellénistique d’Aï Kha- noum, l’apogée et la chute du royaume grec de Bactriane, Paris, de Boccard. Rapin, C. (1996), The Legend of Shakuntalâ and the Indian Treasure of Eucratides at Ai Kha- num, New Delhi and Manohar, Centre de Sciences Humaines. Rapin, C. (2003), ‘Le nom antique d’Aï Khanoum et de son fleuve’, in O. Bopearachchi, C. Landes and C. Sachs (eds), De l’Indus à l’Oxus: Archéologie de l’Asie centrale, Lattes, Musée de Lattes, 115. Robert, L. (1968), ‘De Delphes à l’Oxus: inscriptions grecques nouvelles de la Bactriane’, Comptes Rendues de l’Academie des Inscriptions et Belles-Lettres, 416–457. Sanders, D.H. (1996), Nemrud Dagi: The Hierothesion of Antiochus I of Commagene: Results of the American Excavations directed by Theresa B. Goell, Winona Lake, IN, Eisen- brauns. Schlumberger, D. (1964), ‘Une nouvelle inscription grecque d’Açoka’, Comptes Rendues de l’Academie des Inscriptions et Belles-Lettres, 126–140. Schlumberger, D., Robert, L., Dupont-Sommer, A. and Benveniste, E. (1958), ‘Une bilingue gréco-araméenne d’Asoka’, Journal Asiatique, 246, 1–48. Schmitt, R. (1990), ‘Ex Occidente Lux: Griechen und griechische Sprache im hellenistischen Fernen Osten’, in P. Steinmetz (ed.), Beiträge zur Hellenistischen Literatur und ihrer Rezeption in Rom, Stuttgart, Franz Steiner, 41–58. Shaked, S. (2004), Le satrape de Bactriane et son gouverneur: Documents araméens du IVe s. avant notre ére provenant de Bactriane, Paris, de Boccard. Smith, S.T. (2003), Wretched Kush: Ethnic Identities and Boundaries in Egypt’s Nubian Empire, London, Routledge. Staviskij, B.J. (1986), La Bactriane sous les Kushans: Problèmes d’histoire et de culture, Paris, Maisonneuve. Thapar, R. (1997), Aśoka and the Decline of the Mauryas, Delhi and Oxford, Oxford Univer- sity Press. mairs.indd 42 18/04/2008 14:09:46 The settler community in Hellenistic Bactria and Arachosia 43 Thompson, D.J. (2007), ‘Education and culture in Hellenistic Egypt and beyond’, in J.A.F. Delgado, F. Pordomingo and A. Stramaglia (eds), Escuela y Literatura en Grecia Anti- gua, Cassino, Edizioni dell’Università degli Studi de Cassino, 121–140. Van Soldt, W.H., Kalvelagen, R. and Katz, D. (2005), Ethnicity in Ancient Mesopotamia: Papers Read at the 48th Rencontre Assyriologique Internationale, Leiden, 1–4 July 2002, Leiden, Nederlands Instituut voor het Nabije Oosten. Veuve, S. (1987), ‘Cadrans solaires gréco-bactriens à Aï Khanoum (Afghanistan)’, Bulletin de Correspondance Hellénique, 106, 23–51. Veuve, S. (1987), Fouilles d’Ai Khanoum VI: Le gymnase: Archiecture, céramique, sculptur, Paris, de Boccard. mairs.indd 43 18/04/2008 14:09:46 mairs.indd 44 18/04/2008 14:09:46
学霸联盟