Stay informed: Sign up for eNews Subscribe

The Hellenistic Far East Archaeology, Language, and Identity in Greek Central Asia

Read Chapter 1


Administering Bactria

From Achaemenid Satrapy to Graeco-Bactrian State

And from that amazing panhellenic expedition,

crowned with victory, everywhere acclaimed,

famed throughout the world, illustrious

as no other has been illustrious,

without any rival: we emerged,

a new world that was Greek, and great.


We: the Alexandrians, the Antiochenes,

the Seleucians, and the numerous

other Greeks of Egypt and Syria,

and in Media, and in Persia, and all the others.

With their far-flung realms,

with the nuanced policy of judicious integration.

And the Common Greek Language

which we've taken as far as Bactria, as far as the Indians.

-C. P. Cavafy, "In 200 B.C." (1931; trans. after Mendelsohn 2012, 171-72)


"Part of Bactria lies beside Aria toward the north, but most of it lies above and to the east of Aria. It is large and all-productive except for oil. Because of the excellence of the land, the Greeks who rebelled there grew so powerful that they conquered both Ariana and India as well, according to Apollodorus of Artemita. And so they subdued more peoples than Alexander-especially Menander if indeed he crossed the Hypanis River toward the east and advanced as far as the Imaus, for some were subdued by Menander himself, and some by Demetrios the son of Euthydemos, the king of Bactria. They took over not only Patalene but also the rest of the coast, which is called the Kingdom of Saraostos and Sigerdia. In sum, Apollodorus says that Bactria is the jewel of all Ariana; moreover, the Greeks of Bactria extended their empire as far as the Seres and Phryni,""that most prosperous Bactrian empire of the thousand cities."

The campaigns of Alexander the Great brought the Greek language and Greek rule to large territories of Central Asia and northwestern India. Cavafy's narrator in "In 200 B.C." uses Bactria and India to indicate the very ends of the earth, the farthest extent of the Hellenistic oikoumenē, the inhabited-for which read "Hellenized and civilized"-world. But Bactria has undergone something of a scholarly rehabilitation from the reputation it once held as the "Siberia of the Hellenistic world." Part of this rehabilitation is due to our better understanding of the political and economic ties that bound it to the world of the Near East. Although geographically distant from the centers of the various Near Eastern empires that asserted control over it (the Achaemenid Persian empire, and later the empire of Alexander the Great and his Seleucid successors), Bactria was neither economically nor politically peripheral. It possessed considerable natural resources and occupied a strategic position between the world of Iran and Mesopotamia and the world of the Eurasian steppe. The resources that made Bactria attractive to the Persians and then to the Greeks also meant that it could be used as a power base, whether as a springboard for a satrap's imperial ambitions or for the secession of an independent regional state.

In addition to its agricultural potential with irrigation, Bactria and its southern neighbor Arachosia were sources of manpower. Because of their geographical position, these territories also served as conduits for military forces and war elephants from northwestern India, which might serve under the command of the satraps of Bactria and Arachosia. The region continued to be a source of elephants for the rulers of the Seleucid empire as and when they were able to assert some authority there. In 305 B.C.E., Seleukos I received five hundred elephants under the provisions of a treaty with the Indian emperor Chandragupta Maurya but lost control of Arachosia under this same accord. A Babylonian astronomical diary records the satrap of Bactria supplying war elephants to the Seleucid king in 274/3 B.C.E. In the last decade of the third century B.C.E., Antiochos III conducted an extensive eastern campaign in which he, in the end, made little in the way of territorial acquisitions but did acquire elephants-one hundred and fifty in total-from both the Graeco-Bactrian king Euthydemos and the Indian king Sophagasenos, who ruled in the Kabul Valley.

The Badakshan Mountains of eastern Bactria were the location of probably the only exploited source of lapis lazuli in the ancient world, as well as deposits of other minerals and precious stones. There does not appear to have been any archaeological investigation of the Badakshan mines to date, but their products appeared as far away as Egypt, as early as the Predynastic period. At Shortughai, an outpost of the Indus civilization was established in around 2200 B.C.E., evidently to control access to the Badakshan mineral resources. As well as serving as a center for management of this resource, Shortughai was also notable for the working of lapis. The treasury at the Hellenistic city of Ai Khanoum contained some 75 kilograms of unworked blocks of lapis at the time of its abandonment. The "lapis factor" is therefore an important constant in the economic history of eastern Bactria.

A combination of archaeological and written sources (both literary and documentary) make Bactria an excellent case study for looking at how an empire might manage such a resource-rich region, incomplete though our knowledge of administrative systems must necessarily remain. Crucially, in order to be exploited effectively Bactria's resources required concerted effort and the mobilization of substantial labor forces. The land had agricultural potential but required irrigation. Lapis had to be mined from deposits high in the Badakshan Mountains. Soldiers had to be recruited or conscripted, and elephants had to be acquired and transported. Although there are only a few excavated sites of the Achaemenid and Hellenistic periods, the extensive field-survey project conducted in eastern Bactria in the 1970s provided data on settlement patterns and land use over a period of several millennia. More recently, a small number of administrative documents written on prepared skin have emerged without secure provenance onto the antiquities market. Three are in Greek and date to the late third or early second century B.C.E. A further group, deriving apparently from a single archaeological context, are in Aramaic and cover the period of the end of Achaemenid rule in the region and the conquests of Alexander the Great.

The sections that follow will discuss what these sources can tell us about the organized exploitation of agricultural and other resources in Bactria by locally based apparatchiks working in the ultimate service of external imperial powers. The presence and efficacy of such administrative structures is manifested in our sources both directly and indirectly-indirectly by the very existence of irrigation systems and mining operations that required concerted collective effort, such as might be made possible by a strong local administrative apparatus; and directly by the information contained in the documentary texts on revenue collection, payments to officials, and the central management of tasks such as building fortifications. On a more human level, among the Aramaic documents we have preserved the correspondence between a local governor and his superior, full of terse instructions, reprimands, and lengthy reiterations of previous orders. Resources could be redistributed locally or channeled to the imperial center, but they could also be mobilized quickly and efficiently in emergency situations, such as the need to supply the army column of the satrap-turned-king Bessos, in his flight from the army of Alexander the Great. My focus, in this chapter, is on the longue durée of Bactria's administrative history and on its external relations. Chapter 2 will give a smaller-scale, more detailed case study of the city of Ai Khanoum for which the discussion of the eastern Bactria survey in this chapter provides the necessary context. In chapter 4, the focus will again broaden to consider the relationship, economic and political, between Bactria and its "other" neighbors, the peoples of the steppe and pastoral and nomadic groups in the immediate hinterland of the settled river valleys of Bactria-Sogdiana. Like Bactria's links to the Near East, these connections are crucial to understanding the region's internal affairs.

The field-survey data, as already noted, give a valuable perspective on the longue durée in Bactria. The surviving documentary evidence in Aramaic, on the other hand, happens to cover an especially interesting microperiod in the administrative history of Bactria, the transition from Achaemenid to Graeco-Macedonian control. This provides an intimate perspective on a phenomenon widely attested throughout the Hellenistic world, the retention of the existing Achaemenid bureaucracy as a foundation for the administrative structures of the new Hellenistic states. Alexander the Great, it has frequently been argued, was the "last of the Achaemenids"; the Seleucid empire, too, "in its origin and its constituent elements, was a branch grafted directly onto Achaemenid stock." The Seleucid empire provides more copious evidence for exploring such questions of administrative continuity. Even allowing for the different types of evidence available for the two periods (primarily the Persepolis fortification tablets for the Achaemenid administration and Greek documents for the Seleucid), there is, as may be anticipated, much similarity between the two systems. The evidence from Bactria may be more limited, but my conclusions are essentially the same.

One of my major themes in the following discussion will therefore be not the development and imposition of mechanisms of imperial control but the retention of personnel and institutions, and the efficacy of such a policy in enabling the incoming regime both to efficiently manage and control resources and (potentially) to limit political resistance. Maintaining the existing bureaucracy, and keeping on the staff to run it, is not "lazy imperialism"-however calculating the laissez-faire attitude of its practitioners may or may not have been-but a sound political and economic strategy. Crucially for the present study, however, such administrative continuity may mean that control and exploitation of a territory by an external imperial power is less visible in the archaeological and documentary records than may be supposed.

Although I will introduce material-in particular from the archaeological field survey of eastern Bactria-from periods from the Bronze Age through to the Kushan empire, my focus will be on the periods of Persian control in the region, under the Achaemenid dynasty, and of Greek control, under both the external authority of Alexander the Great and his Seleucid successors, and the local control of subsequent independent Graeco-Bactrian dynasts. This corresponds to the latter part of the eastern Bactrian Iron Age, from around 800 B.C.E. through to the conquest of the region by Alexander the Great in 330-329 B.C.E., and the Hellenistic period, which in Bactria effectively ends with the northern nomadic invasions of the mid-second century B.C.E. The new data from the documentary texts now allow us to revisit and reengage with long-standing debates about the nature of the Achaemenid and Hellenistic imperial presence in Bactria. On one side of this debate have traditionally been the archaeologists (and in particular Jean-Claude Gardin), who emphasize the lack of any break in the archaeological record corresponding to an Achaemenid or Hellenistic takeover. On the other side stands the historical approach of Pierre Briant, who draws on historical textual evidence for substantial Achaemenid state intervention in the socioeconomic affairs of Bactria, and for the management of resources and resource-gathering systems such as irrigation. Resolving this debate is not a matter of simple compromise: archaeological and historical sources each obscure some aspect of the "true picture" of Achaemenid Bactria. This historical sources, for example, make no mention of the region's impressive irrigation works. The new documentary texts now also make it clear that the picture of continuity in material culture obscures the very real and effective mechanisms of Achaemenid imperial control.

From around the middle of the sixth century B.C.E., with Cyrus the Great's expeditions in Central Asia, Bactria begins to feature in our historical sources, which are mostly Greek and of whom Herodotos is the most prominent. This Greek historical lens is a problem in approaching the history of the Achaemenid empire as a whole, even though it has recently come to be balanced by an increased attention to contemporary sources-especially epigraphic and documentary-in other languages. A larger number of Greek and Latin historical works cover the period of the campaigns of Alexander the Great in the 330s and 320s B.C.E. and contain accounts of his military activities in Central Asia and the peoples and terrain he encountered there. The preserved sources draw on contemporary accounts but are later in date. For the period of Seleucid and Graeco-Bactrian rule in Central Asia, we possess far fewer historical sources, and the region's political history in this period must be reconstructed primarily on the basis of numismatic evidence.

For the purposes of the present discussion, I would like to extract a few themes or pieces of pertinent information from these historical sources, which may be compared and contrasted with the archaeological and documentary evidence, or which raise questions that we may turn to these data to answer. The first theme concerns the origin of officials in the Achaemenid satrapal administration. The higher echelons, such as the positions of satrap or local governor, tended to be occupied by a "dominant ethno-class" of Persians, with locals of the various satrapies in lower ranks. There does not appear to have been any attempt on the part of the Persians to spread or impose their religion or language beyond the ethnically Persian provincial elite. What this means in archaeological terms is that we should not automatically expect to see any appreciable presence or influence of Persian material culture in the provinces, whatever the degree of Persian political control. This, of course, is the view from the center: if Persian culture was not actively imposed, aspects of it may still have been adopted and adapted by local populations on their own initiative and on their own terms. The archaeological evidence from eastern Bactria indicates that, in this particular region, any such engagement with Persian culture on the part of local populations was limited.

The second point that I would like to draw out from the historical sources is the degree of authority exercised by local "big men," who might operate within wider imperial power structures but also independently. The agency of local governors or warlords can be seen most clearly in the fierce resistance encountered by Alexander and his army in Central Asia. Many such individuals mounted military opposition or retreated to a fortified hilltop, or both, and had to be rooted out one by one, at considerable costs in terms of manpower, provisions, and lost time. In Arrian's Anabasis of Alexander these men are usually referred to as huparchoi (hyparchs), a Greek term (one may translate approximately as"subrulers") that does little to clarify their precise status but does, perhaps, indicate rather well the extent to which their effective power, in practical, local terms, was out of proportion to their technically subordinate position within Achaemenid power structures. Hyparchs could have been agents for the implementation of imperial policy on a local scale in all sorts of areas, including irrigation schemes or the mustering of troops. As Alexander found, detached from such external obligations, they could also draw on local loyalties and resources to act as independent agents. Yet his policy, where possible, tended toward the incorporation of such powerful local figures: they were to be made to work for him.

At the level of the satrapal government, Bactria itself was also a possible personal power base within the Achaemenid empire or the world of the Hellenistic successor states. The most dramatic example of this use as a power base is the case of Bessos, the satrap of Bactria under Darius III, who after Alexander's defeat of the Persians at the battle of Gaugamela in 331 B.C.E. declared himself king, as Artaxerxes V, took Darius captive, and beat a retreat toward Bactria as a base from which to build his power. In the mid-third century B.C.E., with the independence of the Graeco-Bactrian state, under the Diodotids, from the Seleucid empire, Bactria again showed itself to be a formidable power base, from which a local ruler might challenge the authority of an overarching imperial power. The Aramaic documentary texts offer two particular insights into the potential of Bactria as a regional power base. They document the provisioning of Bessos's supply train and how the resources of Bactria were marshaled in his support. And they present us with the new figures of Akhvamazda, the probable satrap, and his governor Bagavant. Akhvamazda's ownership of personal lands and his administration of these along with affairs of state show the close relationships between local landowners with their own constituencies and the Achaemenid authorities.

The third, associated, preliminary point to be made from the historical sources is that attempts by outsiders to disrupt the status quo in Central Asia were often resisted violently, such resistance being marshaled by the kinds of local "big men" just discussed. Again, Alexander the Great provides the best example, with his foundation of the new settlement of Eschatē on the river Jaxartes (the modern Syr-darya) in 329 B.C.E. Although Alexander treated the river as a border with the steppe lands to the north and attempted to consolidate its position as such, this represented a fundamental break with the existing cultural and political state of affairs, where peoples on either side of the river were in constant interaction, and he inadvertently provoked a major rebellion. For an incoming power, there were political benefits to continuity in executive and administrative personnel. The retention of familiar institutions and social structures (even if these are now managed at the top level by different people) could also be very effective in reducing any "shock of the new" and potential for resistance or revolt among the colonized population.

As I have already suggested, this is the most important theme that emerges from our various forms of evidence from Achaemenid and Hellenistic Bactria, whether these be historical, documentary, or archaeological. Rather than a weakness or deficiency, the retention and productive use of existing structures instead of or in addition to introducing new systems of administration and exploitation is a strength, an active and effective strategy in efficient management of resources and imposition of control.


In the archaeological record from Central Asia, the Achaemenid period is still poorly represented. There are only a very small number of excavated "Achaemenid" sites, in the sense of sites whose sole or primary period of occupation falls in the period of Achaemenid rule or sites that betray any substantial influence of Persian or Near Eastern material culture. Although the corpus of available evidence is still at present very small, what evidence there is does rather serve to underscore the point made in this and many previous studies that the period of Persian rule in Central Asia is not one of substantial innovation in terms of material culture or systems of land management. This lack of innovation, however, does not suppose lack of imperial control. In addition, it is likely that our lack of substantial and identifiable archaeological remains of the Achaemenid period is, to some degree, to be attributed to accidents of preservation and practical constraints on fieldwork. Almost all the major historical cities of Central Asia (e.g., Bactra, Samarkand, Merv) bear two millennia or more of levels of occupation on the same site; these were generally abandoned only with the Mongol invasions of 1220-21 C.E. The same forces that inhibited investigation of the archaeology of Hellenistic Central Asia until the excavations at the Graeco-Bactrian city of Ai Khanoum-and gave the illusion of its nonexistence as a distinct and innovative period in the region's cultural history-have inhibited the investigation of Achaemenid Central Asia. (See the discussion above in the introduction to this volume.) The Achaemenid levels lie below even the Hellenistic levels of the major sites, and furthermore archaeologists tend not to have gone looking for Achaemenid Bactria to the same degree as they have gone to in looking for Hellenistic Bactria.

Where we can identify something of the character of the Achaemenid presence in Central Asia is in the data derived from archaeological field survey. From 1974 to 1978, an extensive field-survey project in eastern Bactria sought to provide a wider context-both chronological and geographical-for the DAFA excavations at Ai Khanoum. In order to restore to Ai Khanoum something of this context, I have chosen to begin with a discussion of the less well-known field survey before moving on to consider the scholarly celebrity that is Ai Khanoum in chapter 2. Although initially concerned, in part, to identify "Greek" impact and settlement on the "Bactrian" landscape, two factors that quickly became apparent to the personnel of the eastern Bactria survey were the considerable time-depth of settlement and agricultural activities in the region (as far back as the first part of the third millennium B.C.E.) and the remarkable continuity in the maintenance of irrigation systems, with canals tens of kilometers in length following the same courses through many centuries or, indeed, millennia. An initial survey area around the immediate hinterland of the city of Ai Khanoum was expanded in subsequent seasons to a larger area of eastern Bactria, between the foothills of the Badakshan Mountains and the Oxus River (Amu-darya).

Not only was the settlement history of the region long, but its connections with regions far beyond Bactria and Central Asia were long standing. These connections are best illustrated by the team's discovery, in 1975, of the site of Shortughai, 21 kilometers northeast of Ai Khanoum's position at the confluence of the Kokcha and Amu-darya. The settlement at Shortughai, in its first phase (ca. 2200-2000 B.C.E.), was an "outpost" of the Harappan civilization of the Indus Valley, with identical material culture, including ceramics, imported shell and carnelian, a seal depicting a rhinoceros with signs in the Indus script, and even the same standard brick sizes as used in Indus cities. The land around it was irrigated from the river Kokcha by a canal some twenty-five kilometers in length. The "lapis factor" may account in large part for Harappan interest in eastern Bactria at this period-once again, it must be stressed that even Shortughai was not the earliest archaeological site discovered in the region. In addition to beads, unworked pieces and chips of lapis were found at Shortughai alongside the flint microliths and drillhead used to work them. In Shortughai's second phase (ca. 2000-1800 B.C.E.), however, the picture changes entirely. The site becomes integrated into the local eastern Bactrian archaeological culture and yields no further evidence of long-distance connections to the south and east but does display evidence of connections to western Bactria and to the steppe. The constant is Shortughai's intimate relationship to and reliance on its immediate agricultural hinterland.

The precise status of the Harappan settlement at Shortughai (I hesitate to use the term "colony") is subject to debate, as is the nature of its connection to the Indus Valley, and its position on the "lapis route" to the West. But the relevance of this site to my main discussion, on the Achaemenid and Hellenistic administration of Bactria, is very clear. The dramatic changes in material culture between the site's two periods represent some fundamental realignment of its relationship to the civilization of the Indus Valley. Whether we can view this in terms of the mechanisms of imperial control and their withdrawal is uncertain. With or without the overt cultural presence of an external political power, however, the archaeological evidence indicates no fundamental interruption in occupation, economic exploitation, or agricultural production.

Later settlements in eastern Bactria-the Hellenistic city at Ai Khanoum, which probably overlay an earlier Achaemenid settlement or the grand Iron Age-Achaemenid fortification, the "Ville Ronde," three kilometers to the north at Kohna Qala-come and go in similar landscape, where the intensity of agricultural exploitation may increase at certain periods but the practices and structures that permit this exploitation remain. The question, of course, is whether these structures (in particular irrigation canals) persisted despite the region's changing political status, or whether the various outside political regimes were themselves instrumental in setting up and maintaining such structures. In the very early periods of eastern Bactria's archaeologically recorded history, the question is also whether some kind of state, whether local, regional, or external, was necessary for the development of irrigation works in the first place, or whether these works may themselves even have led to the emergence of a state.

The latter question demands a certain level of engagement with the intellectual history, specifically, of Soviet archaeological research in Central Asia as a coda to my broader survey of the question above in the introduction to this work. North of the Oxus (Amu-darya), in the former Soviet Central Asian Republics, discussion of the sociopolitical conditions that enabled the construction and management of large-scale canal irrigation systems has traditionally been framed in Marxist social-evolutionary terms. In brief, the model proposed in most such studies is that the organization of labor required to construct canals on a scale such as that in eastern Bactria supposes a centralized political authority capable of mustering such labor, and moreover that this political authority must have represented an evolution from an earlier, more egalitarian sociopolitical system to a more hierarchical one, which employed slave labor. I would like to make the rather simpler and more modest points that irrigation on this scale requires collectivization and mobilization of substantial labor forces, that a local state or external imperial power might play a role in organizing such labor, and that this state or empire had much to gain from efficient organization of agricultural production. We are not in a position to say much more about these issues in the Bronze Age, but for the Achaemenid and Hellenistic periods we now have the necessary documentary evidence to say something about how, by whom, and at what level (local, regional, empirewide) eastern Bactria's agricultural, mineral, and human resources were managed. The information to be gleaned from these documentary texts will be discussed in the sections that will follow. The remainder of this section will be devoted to an overview of the survey data from eastern Bactria, and what these data have been taken to signify.

The first point to be made about the canals that supported the settlements of eastern Bactria concerns their size and sophistication. As well as being very long (canals of 20 to 30 kilometers as early as the Bronze Age), they manage changes in elevation of as much as ten meters. By the Bronze Age, the sites associated with these canals already seem to display a hierarchy from smaller agrarian settlements to larger, fortified sites, although it should be noted that it has not always been possible to date such fortifications securely. In the Iron Age, the population increases, as does the number of fortified sites; it may be possible to date this growth more specifically to the latter part of the Bactrian Iron Age, which corresponds to the period of Achaemenid dominance. In the Hellenistic period, the plain of Ai Khanoum appears to have experienced an intensification in irrigation works and an increase in population corresponding to the period of Greek settlement. Although the Achaemenid takeover did not modify the local ceramic assemblage, the Graeco-Macedonian settlement did. Local Hellenistic ceramic forms maintain a very close correspondence to contemporary ceramics from the world of the Mediterranean littoral, just as the Greek inscriptions and documentary texts of Bactria and Arachosia are linguistically and palaeographically indistinguishable from their contemporaries in Hellenistic Egypt. (See further below.) As I have already noted, although the scale and intensity of settlement and irrigation wax and wane (the progression noted here is something of a simplification), there is otherwise no fundamental break in either the material culture assemblage or in the continued use of individual canals.

The Iron Age intensification of settlement and irrigation implies some regional organization in eastern Bactria more developed than previously:

We can, on the other hand, not advance any statements on the nature of this . . . authority, the limits of its authority, or its place in the political apparatus of Central Asia in the same periods; the Persian conquest, finally, does not necessarily signify the end of this relative autonomy: the coherent pursuit of the same programmes of irrigation until their completion under the Greeks leads us to suppose a certain permanence of local politico-administrative structures subordinated to a foreign authority which doubtless discovered very quickly the advantages of maintaining them.

The latter point-that incoming imperial powers might integrate preexisting local systems of administration and land management into their apparatus of government and exploitation-is, I would argue, the one most fundamental. Not only is it supported by the more recently discovered documentary evidence (as will be discussed in following sections), but it offers a point on which we may reconcile the long-standing debates between Gardin and his colleagues versus Briant. I have already outlined the basic thrust of Briant's arguments above. Bactria, he argues, should not be considered in isolation from other provinces, as a "special case" within the Persian empire: the documentary lacuna that had thus far obscured the mechanisms of the Achaemenid administration in Bactria was a lacuna in our evidence and did not prove the nonexistence of an ancient bureaucracy and its paperwork (as has since been confirmed). Long-term continuities may imply not that the Achaemenid state lacked influence on or failed to intervene in the local administration but rather that that state mobilized and used existing structures and systems. Here, it should be noted, Gardin and Briant are essentially making the same point.

One reason why Achaemenid imperial control in Bactria is not especially archaeologically visible is because much of the existing administrative and economic apparatus was retained. The retention of these systems does not imply the lack of impact of imperial control, but it does tell us that one of the ways in which this imperial control operated was to take advantage of existing structures. Where Achaemenid and Hellenistic imperial control does become visible is in the documentary record, to which I now turn.


Although documentary texts from Achaemenid Bactria have emerged only very recently, their existence and their contents provide confirmation of what scholars have been arguing for some time: that Bactria, a province of economic and strategic importance to the Achaemenid state, can have been administered only with the same care and using the same administrative registers (the Aramaic and Elamite languages) and tools as other provinces of the empire.

Now that we do have some documents from the satrapal archives, what do these add to our understanding of the Achaemenid administration of Bactria? The first thing to be said is that the documents currently known contain no references to irrigation or canals but do refer to private and public land under cultivation, the management of such land, and its agricultural products. The archaeological evidence already considered shows that this agricultural production, in Bactria, was made possible by large-scale irrigation works, but the execution of such works is not mentioned in any of the written sources presently at our disposal. Another point for which the archaeological and documentary evidence must be viewed as complementary rather than accumulative is geographical provenance. The best-documented survey regions are eastern Bactria, the area on which I have focused, and the valley of the Surkhan-darya, in north-central Bactria (which will be discussed in chapter 4). The documents, in contrast, most probably come from the satrapal archive in the city of Bactra, in central Bactria. Some of them relate to the administration of the area around the town of Khulmi, just to the east. Nevertheless, as will be discussed, there are numerous places in which the documents provide new information to bring to bear on existing historical questions.

The documents known at present comprise thirty texts on leather and eighteen wooden sticks, all written in ink. The language is Imperial Aramaic, the standard administrative register of the Achaemenid empire and the same as that of other Achaemenid Aramaic documents from Egypt. In terms of script, orthography, and official terminology and formulae, the Bactrian and Egyptian documents are all but identical. Any Iranian loanwords tend to be Old Persian-that is, from the language of the imperial center-not from local languages. The provenance of the Bactrian documents is not known, but they almost certainly came from in or near Bactra (modern Balkh), the capital of the satrapy of Bactria, and later also of the Graeco-Bactrian kingdom. Bactra has, unfortunately, not yet been extensively excavated, although material of the Achaemenid and Graeco-Bactrian periods is gradually coming to light. The chronological range of the dated documents is 353-324 B.C.E., with a single outlier dated on palaeographical criteria to the fifth century B.C.E. The texts therefore span the period of the final decades of Achaemenid rule in Bactria through to the seventh regnal year of Alexander the Great. They provide an invaluable record of the administrative and economic affairs of Bactria at a period of political transition.

Internal connections between the documents indicate that they came from a single archive and that this archive was that of the satrap at Bactra. An important subgroup of documents consists of eight letters and two economic notes addressed or otherwise connected to a man named Bagavant, who was the paḥtā (governor) of Khulmi, probably modern Khulm, about seventy kilometers to the east of Bactra (Group A). The letters, however, are rough copies-they contain many corrected and uncorrected errors, as well as traces of previous, erased writing-and therefore belong to the archive of the sender, Akhvamazda, not of Bagavant. The content and epistolary style (abrupt, demanding, lacking in courtesy phrases) of the documents indicate that Akhvamazda was Bagavant's superior. His title is not explicitly given, but that he was satrap of Bactria seems intuitive. In addition to the Akhvamazda-Bagavant correspondence, the collection contains many documents that have no mention of either official, reinforcing the notion of a satrapal archive belonging to a bureau that administered the various economic affairs of the province. Another subgroup of documents consists of letters between officials of more nearly equal rank, as may be seen from their forms of address. Two further groups comprise lists of provisions and agricultural products to be distributed to certain persons of groups of people and wooden tallies giving quantities of commodities given on credit, all dated to Year 3 of Darius III: that is, 333-332 B.C.E. Since these texts come from a central satrapal archive, it may be that these goods are being distributed to satrapal agents or employees through a named official agent and are not records of transactions between private individuals. The physical form of the tally sticks is of some interest: two identical texts were written on the interior faces of a split wooden stick, and each party kept one half, so that the details of the transaction could later be verified.

The Achaemenid imperial apparatus exercised considerable direct control over the economic and military affairs of the province. Agricultural produce is gathered, consolidated, and distributed. The kinds of products mentioned include livestock (chickens, geese, horses, camels), harvested grain, and processed foodstuffs such as yogurt, wine, and vinegar. Significantly, the documents differentiate between cattle and sheep roaming free at pasture and those confined in an enclosed, supervised space: an important confirmation of the importance of the pastoral economy in Achaemenid Central Asia and the ability of the state to supervise and profit from it. Such products are both collected and distributed to lists of named individuals and groups, including soldiers. Among the tasks with which Bagavant is charged are the transportation of grain to the satrapal granary, collection of vinegar from one of the satrapal estates to provision a convoy, and fortification of towns. One item of correspondence concerns Bagavant's resistance to an earlier order to build a wall around a town: the crops have not yet been harvested, he protests, and are at risk from locusts. Akhvamazda gives his permission for the work force to complete the harvest before starting on the fortifications. Although as already noted no documents refer to irrigation works and their maintenance, this last-mentioned letter does demonstrate the extent to which the Achaemenid apparatus was able to mobilize and direct labor.

As in the Greek documents that will be discussed in the next section, onomastics gives us some limited but potentially still useful evidence on the ethnic origins or assumed ethnic identities of various officials. The highest officials all have Persian names, as do the "scribes" of the documents-who may in fact also be executive officers rather than simply secretaries. Like the Bactrian Greeks of the Ai Khanoum treasury texts (on whom see further below), these Bactrian Persians may have had deeper roots in the region, whether as long-term residents or as the products of families that had been settled there for some generations and integrated with local communities. Some of the names in the documents do, however, bear traces of local Bactrian influence, in particular theophoric names derived from the deified river Oxus.

Still better evidence for the deep enmeshment of the Achaemenid administrative apparatus and its personnel in local affairs and socioeconomic structures, however, comes from the indistinct dividing line between Akhvamazda's administration of state property and his personal holdings. Bagavant is charged with managing both the satrap's own personal estates (for example, carrying out repairs on buildings there) and state lands, and it is sometimes difficult to tell which category a particular landholding falls under. Although nominally imperial servants, both men clearly also command strong local power bases, and Akhvamazda's own personal economic interests are bound up with his "official" remit and the resources at his disposal. The incorporation of powerful local "big men"-whether men of Persian origin who had built up Bactrian power bases or the scions of powerful local families of longer standing-was, I would argue, one of the great strengths of the Achaemenid administration in Bactria. The ultimate example is Bessos, the satrap of Bactria who acceded to the imperial throne after Alexander's humiliating defeat of Darius III. As Alexander pursued him into Central Asia, Bessos was able to draw on local resources and support. Among the papers from the satrapal archive, we appear to have one document that relates directly to local provisioning of Bessos's army (Naveh and Shaked 2012: C1 [Khalili IA 21], Nov.-Dec. 330 B.C.E.):

In the month of Kislev, year 1 of Artax[erxes] the King. (col. 1, l. 1)

Provisions in Maithanaka for Ba[yasa], when (l. 2)

he passed from Bactra to Varnu. (l. 3)

Artaxerxes was the throne name assumed by Bessos, and it is very tempting to identify the Bessos (Bayasa) of line 2 with the historical Bessos: it is certainly of the right date for this to be possible. He therefore appears twice in the same document: as the ruling authority by whom paperwork was dated and as the powerful individual who was able to draw on Bactria's administrative apparatus to supply his army.


On the 15th of Sivan, year 7 of Alexander (col. 1, l. 1)

the King. Disbursement of barley <from> Vakhshudata, the barley supplier (l. 2)

in Ariavant. (l. 3)

(Naveh and Shaked 2012: C4 [Khalili IA 17], 8 June 324 B.C.E.)

Six years later, similar documents were being dated by a different king. The Aramaic documents from Bactria cover a period of transition, from the last decades of Achaemenid rule through into the very early years of the new Graeco-Macedonian regime. In a very literal sense, the Aramaic documents do indeed record "business as usual": the continuing regulation of the province's economic life according to existing practices, within existing administrative hierarchies, and written down in the same language according to the same templates. In dating formulae, the Persian kings give way seamlessly to Alexander the Great, the "last Achaemenid." Since the documents currently known come down only as far as the 320s B.C.E., it is at present impossible to say if and when this Achaemenid Aramaic bureaucracy was modified or replaced. At least by the time of the first Greek documentary texts now known, in the late third or early second centuries B.C.E., Bactria was being administered in Greek.

At least initially, as was discussed above, it is also difficult to find any clean break in the archaeological record corresponding to the political changes of the late fourth century. The most attractive hypothesis is to view much of the third century as a lengthy period of transition, in the course of which the Greek colonial presence gradually made itself felt and undertook programs (whether or not these were centrally directed) for the intensification of settlement and exploitation of the agricultural territory in eastern Bactria. It is probable that this was accompanied by a longer period of crossover or transition from an Aramaic to a Greek bureaucracy, with multilingualism and different languages being used for different purposes on different media, but the evidence with which we may examine this transition is unfortunately very sparse. Although there are Aramaic texts from the regions south of the Hindu Kush into the first half of the third century B.C.E., these are public inscriptions by an external political authority (the Indian Maurya empire) and give us only indirect evidence for the continued use of Aramaic to regulate local affairs. But an ostrakon from Ai Khanoum, fragmentary and cryptic though it is, is in fact rather good evidence that administrative functions in Bactria continued to be performed in Aramaic in at least some contexts or at some levels well into the period of Graeco-Macedonian rule. This ostrakon dates to the second half of the third century B.C.E. (It is written on a Hellenistic ceramic vessel of that date.) The text is too badly broken to be read fully, but it appears to record a transaction having to do with grain. It may therefore be compared to some of the Greek "ostraka" from Ai Khanoum discussed in the next section, which give lists of payments to or by named persons. If this ostrakon does indicate some continuity in accounting practices at Ai Khanoum, we should also be alert to the possibility that some of the Greek titles used in the treasury texts may have been borrowed to refer to offices that developed from, or were even identical to, local Achaemenid precursors.

The majority of the Greek texts that will be considered in the next section come from the Graeco-Bactrian city of Ai Khanoum, the site that the eastern Bactrian survey was originally designed to contextualize. I would like to anticipate some of my discussion of this city in the next chapter by examining how the Greek settlers of Bactria-Alexander's demobbed soldiers-and their descendants modified their built environment in the course of the third century B.C.E. and what these renovations may mean.

Ai Khanoum is at present the only extensively excavated major settlement site of the Hellenistic period in Bactria, but its problematic publication history has left many questions about its history and chronology to remain unresolved. The ceramic data from the site, which received lengthier (although still not comprehensive) publication only some time after the excavations ceased, now offer a more reliable chronological framework within which to place various construction programs undertaken at Ai Khanoum. These data also make it very clear that there was Achaemenid-period occupation on the site, which predated the Graeco-Macedonian "foundation" and its structures.

The most substantial program of renovation at Ai Khanoum took place between ceramic periods IV and V, which, according to current datings correspond to (roughly) 260-220 B.C.E. and 220-200 B.C.E. This mid-to-late third-century date is significant for two reasons. First, it is the period by which Bactria had succeeded in asserting its independence as an autonomous kingdom. And second, it is well after the death of the first or even the second generation of the Greek settlement. As well as providing a medium through which the inhabitants of Ai Khanoum could reimage their connections to their Greek colonial origins, remodeling the urban landscape of the city also allowed them to make blunter, more assertive statements of their local autonomy, in the form of grander, more monumental architecture. If not a fundamental break in terms of government and administration, this marked a point when Bactria started to be administered and developed on its own terms rather than as part of a Near Eastern empire, whether this was that of the Achaemenids, Alexander, or the Seleucids. It was not the introduction of a new external imperial power that marked a new direction for Bactria but a change in its status from province, however devolved its provincial administration, to independent kingdom.


From the last dated Aramaic documents of the new (but not bureaucratically innovative) Graeco-Macedonian administration in Bactria in the 320s B.C.E., there is a gap of a century or more until the first preserved Greek documentary texts, of the late third or early second century B.C.E. The two documentary corpora give us rather different kinds of information, and the Greek material is on two different media (skin and ceramic vessels) and from several different locations. Another contrast with the Aramaic documents is that the majority of the Greek texts come from an excavated context and are thus securely provenanced, both to a site (Ai Khanoum) and to a building within that site (the treasury, in the administrative quarter).

The office at Ai Khanoum from which these texts derived (and any additional bureaux that existed within the administrative quarter) were on a scale rather different from that of Akhvamazda's central satrapal archive but perhaps more akin to Bagavant's regional governor's office. Although we cannot precisely identify the territory that Bagavant administered or assess contemporary land usage or other forms of economic exploitation there, the regional eastern Bactrian center at Ai Khanoum may be related to a wider archaeological context, the area of the eastern Bactrian survey, the data from which were discussed above. If the Aramaic documents give us the local viewed from the (Achaemenid) center, the Greek texts of the Hellenistic period give us a more regional perspective on economic affairs.

The city of Ai Khanoum and its excavated remains will be discussed fully in chapter 2; a glance at the city plan will provide enough information for the purposes of the present discussion. The economic texts on ceramic vessels derive from the treasury, a small court surrounded by magazines, accessed off the main courtyard of the palace. The treasury, although an integral part of this complex of buildings, was positioned in such a way as to make it accessible from the outside, directly from the main court, without requiring a visitor to pass through the rooms and corridors to the south. This was perhaps a more suitable location for an office that necessitated frequent comings and goings of people and goods.

Although the treasury also yielded some fragments of literary texts on perishable materials, I shall be concerned here with texts of an economic nature written on ceramic vessels, the majority in ink but a smaller number incised post-firing. The texts were written when the vessels were intact and still in use: that is, these are not "ostraka" in the sense of broken sherds of pottery reused as cheap writing material, although some complete vessels contained more than one piece of writing, with a canceling sign before each superseded text. For the most part, the texts relate to the contents of the vessels, which include coins, olive oil, and incense, and they describe the reception, quantification, and verification of these contents by several named parties. These include individuals with Greek and with local Bactrian-Iranian names: insufficient evidence, as yet, to say much about ethnic relations and identity at Ai Khanoum. Sums of money are counted in either Indian units of currency or Greek drachmas. The best-known text is dated to "Year 24" of an unnamed king, most likely the Graeco-Bactrian king Eukratides. In addition to the regnal year, archaeological and palaeographical criteria and as well the presence of Indian currency date these texts to the first half of the second century B.C.E., after the Bactrian Greek military expansion into northwestern India. I discuss not all these dipinti here, many of them preserving only a few letters, but only those that bear text sufficient to yield useful information.

Typically, these texts are records of transactions relating to coins or commodities that are concluded dia (through) or para (from) various individuals, whose official positions, if any, go unremarked. In most cases, the verb "to be counted" (arithmeisthai) is probably also to be understood. Otherwise, one or another individual esphragistai (sealed), probably to be understood both literally as of the sealing of a vessel and figuratively in the sense of a supervisor "signing off on" the matter. Two such officials may "seal," perhaps as a double-check. We therefore have a record, however brief and perhaps cryptic, of the names of some of the employees of the treasury, where most of these texts were found, as well as their roles in activities there. I summarize these here (with numbers referring to texts as given in Canali De Rossi 2004):

Aryandes: 327 (?), 330 "through."

Hermaios: 325 "counted," 326 "counted," 327 (?).

Hippias: 323 "through."

Kosmos (or Kosmas): 334 "through."

Molossos: 323 "sealed," 328 "through."

Nikeratos: 333 "from," 334 "verified through," 334 "sealed."

Oxeboakes: 324 "counted," 325 "counted."

Oxybazos: 324 "counted," 324 "sealed," 346 (?).

Philiskos: 330 "from," 331 "from," 332 "from," 347 (?).

Philoxenes: 349 ownership.

Sosipatros: 348 ownership?

Straton: 323 "sealed," 328 "from," 328 "through," 330 "through."

Tarzos: 328 "sealed."

Theophrastos: 332 "sealed."

Timodemos: 325 "from," (?), 341 (?).

Zenon: 324 "from."

The names Kallisthenes, Sinoph[.]tos, Xatrannos, and Oumanos occur in the list (no. 345), and Kallisthenes perhaps also as one "through" whom a matter was transacted (no. 353).

I qualify the treasury as the location where "most of" these texts were found, because a few similar pieces were recovered from elsewhere in the city. This displacement occurred because the population who occupied the site after its destruction in the mid-second century B.C.E. disturbed the contents of many of its buildings, so that it is not uncommon to find items that would be more at home in a treasury context elsewhere, in particular in the sanctuary of the city's main temple. This is the case with four separate texts on a single vessel recovered from the temple sanctuary, which record cash payments para (from) one individual that ērithmētai or ērithmēntai dia, "have been counted by" or simply "through" two others and "sealed" by a fourth. The sums are in Greek drachmas, or in Indian currencies with names derived from toponyms such as Taxila. Many such Indian coins were discovered at Ai Khanoum. Other, more fragmentary texts of the same formula record similarly substantial payments in Indian currency, as well as cases where the denomination has not been preserved.

These are quite considerable sums of money, suggesting, along with the regularity with which the precise figures of five hundred drachmas and ten thousand Indian coins recur (to the exclusion of any other amounts), that these are transactions of an official nature, not personal payments (for example) of taxes or for purchases. Whatever the ultimate sources of these sums-and the Indian coinage may suggest tribute or taxes exacted from Indian lands conquered by the Graeco-Bactrian kings-what is happening at the treasury is evidently that incoming revenue is being consolidated and recorded, according to fixed divisions. I propose that the regular figures chosen-as well as being convenient, round counting units-have much to do with the practicalities of storage and transportation. What seems less likely-although the possibility must be considered-is that these coins counted and verified (see below) in the treasury were the products of a local mint. Although blank bronze flans attest to the presence of a mint at Ai Khanoum, there is no firm evidence that silver coinage (such as drachmas) was struck there.

The texts already discussed relate to the transaction and supervision of various payments in commodities and coin by a series of individuals. These are the most common formulae among the treasury texts, although there are a couple of exceptions. One, which is written in a more cursive script than the others, may be a more lengthy financial account written on an ostrakon rather than a record of the same type as the texts already discussed. Although this text is fragmentary, clearly it records more than one smaller sum (44 drachmas, 7 drachmas, 8 drachmas, a possible total of 60 drachmas), has the names [He]rmaios and [Ar]y[a]ndes, in the genitive case (which could be taken with either para or dia), and contains a reference to tas anaphoras (revenues). There is also a list of personal names in the genitive case (Kallisthenes, Sinoph[.]tos, Xatrannos, Oumanos), with a possible total number of seventy-four at the foot of the list, suggesting another account of sums paid by or through named individuals, with a final total. A different format is followed in a better-preserved text, in which dokimos ("approved" or "assayed") silver, or "legal tender," is received through Kosmos, dedokismatai dia (verified by) Nikeratos, and sealed by this (explicitly stated) "same" Nikeratos. It may be that the verification performed by Nikeratos is to be connected with the presence of more than one currency standard within the treasury.

A small ceramic fragment excavated in the monumental entranceway to the city's administrative quarter bears a rectangular stamp with an image of an amphora and the legend epi agorano[mou] Khaireas ("during the administration when Khaireas was agoranomos"-although in view of the syntax as restored Khaireas may not be the same person as the agoranomos). Although it cannot be ruled out that this amphora was an import, it is simpler to propose a local origin, especially since Bactria was itself a wine-producing region. We therefore have the probable presence of an official called the agoranomos at Ai Khanoum, a figure known from elsewhere in the Hellenistic world as a regulator of weights and measures, and general supervisor of marketplace commerce.

Wine may well have come from the local area around Ai Khanoum, as may honey and even olive oil (despite Strabo's claim that Bactria does not produce oil), but some of the other commodities excavated in the treasury or named in its texts did not, including incense and cinnamon. The size of the treasury and its position within the city would not have permitted the storage or regulation of bulk agricultural produce such as grain. It therefore appears likely that the treasury's role was more specifically to manage "special" commodities, which were imported, produced only in small quantities, subject to state control, or produced under a state monopoly. These commodities were not just received and redistributed as complete units but were being assayed and repackaged. Some texts record the transfer and consolidation of contents between partially full vessels. Weights were also discovered in the treasury.

The notion that the treasury was a clearinghouse specifically for especially uncommon or valuable goods is confirmed by the archaeological finds of raw materials from the storage magazines. These contained a very large quantity of unworked blocks of lapis lazuli, some seventy-five kilograms packaged in a sack that had decayed, and a much smaller quantity of worked lapis. Other finds included ingots of silver and gold, and of raw glass, semiprecious stones, obsidian, vessels made of glass, metal, stone, and alabaster, and a few smaller worked "luxury items" such as pieces of jewelry containing semiprecious stones.

From the texts, it is possible to say something about the personnel who operated the treasury and their relative positions. This information may be compared and contrasted with that in the Greek documentary texts on skin from Bactria, which I will go on to discuss. The treasury texts appear to come from a comparatively short period of time and involve a rather small range of names. Individuals may appear in more than one text, in different roles, and in more than one role within a single text. In one, Nikeratos both verifies and seals-it is stated explicitly that this is the same Nikeratos-and the like is probably also true in another, where an Oxeboakes both counts and seals. We cannot tell for certain whether the same individual is always concerned or simply two men with the identical name, but the lack of an identifying patronymic or other indication of position in all cases suggests that it was self-evident to the record keepers who within their small bureau was involved (and accountable).

Do officials with Greek or Bactrian names appear in different capacities? The figure "from" whom always has a Greek name; the others may be Greek or Bactrian. Philiskos, whose name is Greek, appears as the agent "from" whom something in transacted in up to four texts. Local Bactrian-Iranian names include the theophoric "Oxus" names Oxeboakes and Oxybazos, a continuity in local naming practices from the Achaemenid period and probably earlier. But it is questionable how meaningful this Greek-Bactrian division is: the corpus of texts is very small, and a name's linguistic origin may not have any direct relationship to its bearer's descent or ethnicity.

This storeroom-bureau, with its own records on the vessels of products and payments themselves, must also have been accompanied by an archive at a higher level of the administration, with accounts on perishable media that have not been preserved. As with the Achaemenid Aramaic documents, we are therefore seeing only a small portion of one office of a much more extensive administrative apparatus, with internal functional as well as hierarchical divisions.

Where we do have contemporary texts preserved on perishable materials, however, is at a lower level of the administration, in the sense of one that relates to regulation of the legal and financial affairs of individuals and their dealings with the state bureaucracy. There are, at present, just three known such Greek texts, on prepared skin, from Bactria. Like the economic texts from the Ai Khanoum treasury-and like, it should be recalled, the Aramaic documents of a century or so before-these documents are written in hands indistinguishable from those of the contemporary eastern Mediterranean world, testifying to their origin and perhaps continued inclusion in a shared bureaucratic system or modus operandi. My interest here is primarily in the data on economic and administrative affairs that may be gleaned from these rather than their (considerable) interest in matters of Graeco-Bactrian chronology. In all cases, it should be recalled that the texts are fragmentary and do not always yield much sense. They certainly do not furnish a coherent map of Bactrian administrative hierarchies.

The first published document was a tax receipt dated to around the first half of the second century B.C.E.It bears a dating formula according to local Greek kings, with a Macedonian month name, and a location in a place called Asangorna. A guardian of the law (nomophulax) is the first official named: "Although the placing suggests that he should be an annual magistrate whose name is used for dating purposes, it may be that he is mentioned because of some responsibility he had for the legality of the transaction." Next comes a chain of command stretching upward from a logeutēs (tax gatherer), who acted sumparontōn (in the presence of) a figure whose name is not preserved, who was tou sunapestalmenou hupo (sent out likewise by) two further officials, all dia (through, by the agency of) one Diodoros epi tōn prosodōn (controller of revenues). The original tax gatherer, Menodotos-the subject of this extended formula-acknowledges receipt from an individual whose name is not preserved of tē ōnē ta kathēkonta (payments due in respect of a ["the"] purchase). The precise meaning of this is unclear: the editors of the text suggest that it relates to either a tax concession purchased from the government or a tax payable on a contract of purchase. Toward the end of the document there is a cryptic reference to something to do with a temple or sacred affairs, suggesting that we may here be dealing with financial affairs and state regulation of a temple. All preserved personal names are Greek, with the exception of the patronymic of the taxpayer, which appears to be Iranian.

Despite the many lacunae and uncertainties in this text, the administrative hierarchy is fairly clear. The text most probably derives from a local office of the financial administration, directed by Diodoros the "controller of revenues." Subordinate to Diodoros were a number of "tax gatherers" and other such officials, and they all formed part of a chain of command directing taxation and revenue from individual payers up through increasingly more senior and more centralized bureaux to their ultimate destination in the royal coffers. This, then, is the kind of collection and organization of revenue (albeit just one kind of revenue, from local taxation rather than also from luxury or state-controlled resources and imports) that took place lower down in the administrative pyramid than at the treasury office at Ai Khanoum. It is tempting, furthermore, to trace a fairly direct paper trail from records of individual payments (such as the Asangorna receipt) through to consolidated lists of payments from named individuals (such as nos. 327 and 345 from the treasury, which may record individual payers or the heads of local offices) through to the final packaging of revenue into units of five hundred drachmas, perhaps for shipment to a royal treasury at Bactra. This reconstruction is, of course, hypothetical, and I do not argue that the particular individual documents that I have discussed here stood in this relationship to one another.

The two further Greek administrative texts on skin were published only recently and date probably from the late third to early second century B.C.E. Both are unprovenanced in any meaningful sense, and so it is really impossible to say if they derive from the same source or location as the Asangorna text or each other. The first records the payment of a sum of one hundred drachmas of coined silver and is perhaps a contract or receipt relating to a loan of money. If so, this is one of the only texts relating to the transaction of financial dealings between private individuals within a legal framework. The text is said to have been drawn up in a place named Amphipolis, which, like Asangorna, cannot yet be located securely on the ground. The second text is briefer and more fragmentary still, dates to sometime in the second century B.C.E., and records something (a sum of money?) ha ekhei Arkhisēs epi phorai (which Archises has [received] for transport). It also contains a cryptic reference to "stone." As the editors note, it is probably pushing things too far to read this as a reference to lapis lazuli, but the finds from the Ai Khanoum treasury, as discussed above, alert us to the possibility of any form of semiprecious or worked "stone" that could be valued and collected as a commodity.


I would like to be quite cautious in assessing the extent to which all of these texts, even in combination, can be used to reconstruct the workings of the bureaucracy of Bactria in the Hellenistic period, still less the extent to which this bureaucracy was based on its Achaemenid predecessor. As was noted in the introduction to this chapter, the probability is great that the Achaemenid administration of Bactria, like that of other regions of the Achaemenid empire, supplied the template for later Hellenistic administrative structures. Continuity in personnel may be traced only as far down as the early years of the reign of Alexander but is likely to have persisted for much longer. The field-survey evidence, crucially, demonstrates that the essential workings of systems of land management and resource extraction were not interfered with. As was discussed above in the introduction to this chapter, the apparent absence of evidence for the Achaemenid and Hellenistic takeovers in eastern Bactria is in fact rather good evidence for a particular political and economic strategy, the retention of existing personnel and systems of economic production and exploitation in the service of a new political power. The survey data, along with the existence of administrative texts of the fourth and fifth centuries C.E. in the Bactrian language, raise the tantalizing possibility that one day we may even be able to trace something of Bactria's longer-term administrative history.

All this relates only indirectly to the cultural and social history behind these administrative structures. How much investment did the local elites of Bactria retain in the new political system? The historical sources, discussed in the introduction to this chapter, once again bring us down only as far as the early third century B.C.E., showing how local "big men" reacted to the new Graeco-Macedonian regime, whether by rebellion or collaboration. Names, as has already been noted, tell us nothing certain about people's identities or family backgrounds, and so we cannot say much about the social or ethnic origins of the officials named in the Greek documentary texts. To what extent did elites "acculturate" in response to the incoming Greek authority? The material culture of eastern Bactria, as already noted, received a marked influx of Greek or Mediterranean ceramic forms; but pots, notoriously, are not people. The urban sites of Bactria-the inevitable Ai Khanoum but also cities where excavation is ongoing, such as Bactra and Termez-supply plentiful data to begin to answer such questions, but it can be difficult to decide which aspects of these cities' material culture and architectural programs held contemporary ethnic resonance. Sometimes a Corinthian column is just a column. The following chapter will examine how a more holistic approach to Ai Khanoum's urban landscape-considering the institutions of the city in terms of their relationships to one another, and the city in the context of its immediate agricultural hinterland-may begin to furnish some answers to such questions.