Handbook of ancient Afro-Eurasian economies. Volume 1: Contexts

von Reden, Sitta (Ed.) Book — Published Version Handbook of ancient Afro-Eurasian economies. Volume 1: Contexts Provided in Cooperation with: De Gruyter Brill Suggested Citation: von Reden, Sitta (Ed.) (2020) : Handbook of ancient Afro-Eurasian economies. Volume 1: Contexts, ISBN 978-3-11-060774-1, De Gruyter Oldenbourg, Berlin, https://doi.org/10.1515/9783110607741 This Version is available at: https://hdl.handle.net/10419/270829 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by-nc-nd/4.0/ Sitta von Reden (ed.) Handbook of Anc ient Afro-Eura sian Ec onomies V olume 1: C ontexts Hand book of Anc ient Afro-E ura sian E c onomies V olume 1: Contexts Edited by Sitta von Reden In cooperation with Mamta Dwivedi, Lara F abian, K athrin Leese-Messing, Lauren Morris, and Eli J. S. W eaverdyck This project has received funding from the European Research C ouncil (ERC) under the European Union ’ s Horizon 2020 research and innovation programme (grant agreement No 7 42645). ISBN 978-3-11-0604 51-1 e-ISBN (PDF) 978-3-11-06077 4-1 e-ISBN (EPUB) 978-3-11-060494-8 This work is licensed under a Creative C ommons Attribution-NonCommercial-NoDerivatives 4.0 International License. For details go to http:/ /creativecommons.org/licenses/by-nc-nd/4.0/. Library of C ongress C ontrol Number: 2019946355 Bibliographic information pub lished by the Deutsc he Nationalbib liothek The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data are available on the Internet at http:/ /dnb.dnb.de. © 2020 Sitta von Reden and the chapters ’ contributitors. P ublished by Walter de Gruyter GmbH, Berlin/Boston This book is published with open access at www.degruyter.com. C over illustration: Remember/ Помни , by Nicholai R oerich (1924), Roerich Museum (New Y ork) © Bridgeman Images (Berlin). T ypesetting: Meta Systems Publishing & Printservices GmbH, W ustermark Print and binding: CPI books GmbH, Leck www.degruyter.com Acknowledgements This handbook is based on the interdisciplinary research project “ Beyond the Silk Road: Economic Development, Frontier-Zones, and Inter-Imperiality in the Afro- Eurasian Weorld Region (300  to 300  ), ” and made possible through the award of an Advanced Grant by the European Research Council from 2017 to 2022 (ERC- ADG 742645). Further particulars about our research can be found on our website www .basar.uni-freiburg.de. We wish to thank the European Research Council and the anonymous readers of the proposal for their financial support and helpful com- ments. We are also grateful to our scientific advisors, Anthony Barbieri-Low , Peter Eich, Enno Giele, Herrmann Kulke, Rachel Mairs, Bryan Miller, Richard Payne, Hi- manshu Prahba Ray , Armin Selbitschka, Michael Speidel, Soeren Stark, Miguel John V ersluys, and Josef Wiesehöfer, for their advice and encouragement especially dur- ing the early stages of this project. At the University of Freiburg we benefitted from the help and expertise of Marco Ferrario and Milinda Hoo, while Peter Palm, Berlin, drew all the maps for this vol- ume. Special thanks go to Alison Weaverdyck who corrected the English of the non- native speakers, and to Mark Marsh-Hunn who added last-minute support. Clara Hillebrecht has been our incessant assistant during the entire period of research for the first volume, while she, Benedict Lerbs, and Stephan Neitmann have helped to prepare the manuscript for publication. Last but not least we thank Florian Hoppe at de Gruyter Press for suggesting and encouraging the publication of the results of this project in the form of a handbook, and for his patient advice in any editorial matter. Mamta Dwivedi Lara F abian T sang Wing Ma K athrin Leese-Messing Lauren Morris Sitta von Reden Eli W eaverdyck Open Access. © 2020 published by De Gruyter. This work is licensed under the Creative C ommons Attribution-NonCommercial-NoDerivatives 4.0 License. https:/ /doi.org/10.15 15/97831106077 41-202 C ontents Acknowledgements v List of C ontributors xi T ransliteration and Orthography xiii Abbreviations xv Sitta von Reden Introduction: Ancient E conomie s and Global C onnections 1 Part I: Empires Sitta von Reden Introduction 13 Sitta von Reden 1 The Hellenistic Empires 1 5 Lauren Morris 2 C entral Asian Empires 53 Mamta Dwivedi 3 Early Historic South Asia 95 Kathrin L eese-Messing 4 The Qin and Han Empires 129 Ursula Brosseder 5 The Xiongnu Empire 195 Lara Fabian 6 The Arsa kid Empire 205 Eli J. S. W eaverdyck 7 The R oman Empire 241 Ab breviations Ancient literary works from the Greek and Latin corpora are cited in full upon first reference, and then abbreviated according to the conventions of the O xford Classical Dictionary , third edition. Corpora of epigraphic and numismatic evidence are abbre- viated using the same conventions. Papyri, ostraca and tablets are cited according to the Checklist of Greek, Latin, Demotic and Coptic papyri, ostraca and tablets (https:/ /papyri.info/docs/checklist). Ancient transmitted works from the Chinese tradition are cited according to Si- nological norms, such that Shiji 15.685 refers to Shiji chapter 15, page 685 in the Zhonghua shuju edition. These sources are listed under the title of the work in chap- ter bibliographies. Open Access. © 2020 published by De Gruyter. This work is licensed under the Creative C ommons Attribution-NonCommercial-NoDerivatives 4.0 License. https:/ /doi.org/10.15 15/97831106077 41-206 Sitta von Reden Introduction: Ancient E c onomies and Globa l C onnections The purpose of this three-volume handbook is twofold. First, it aims to provide a tool for interdisciplinary research on ancient economies during the imperial period of the third century  to the third century  . Second, it aims to suggest ways of approaching the connectivity of the Afro-Eurasian region from a new economic perspective. It is widely acknowledged that the expansion of relationships between the Afro-Eurasian empires in antiquity was accompanied by the movement of large amounts of goods: fine textiles, leather items, pearls, ivory , dies, spices, drugs, un- guents, animals, and much more. The visibility of such items in places far away from their origin leaves no doubt. Yet the mechanisms by which these goods were mobilized in their areas of production or extraction, and the exchange systems through which they spread into distant locations, are far less certain. We argue that the notion of Silk Road trade based on nineteenth-century perceptions of caravan trade, national economies, and markets is ill-suited to analyzing the nature and dynamics of the connectivity of ancient empires. 1 The chapters of this handbook aim to globalize ancient history without presuming a context that make ancient inter-imperial economic connections a precursor of modern globalization. 2 Over the last 15 years, scholarship has seen a proliferation of comparative re- search on ancient empires. 3 The question of connections across Eurasia, in contrast, has suffered relative neglect or has been locked within the flawed notion of the Silk Road. This handbook attempts to shift the problem of connections into a framework that has been developed in world history and world systems theory . It starts from the uncontroversial fact that while ancient imperial courts and historiographers in- vented empires as ‘ one ’ and universal, they were neither culturally homogenous nor fully self-sufficient. 4 The complicated levels of interdependence of imperial and local economies, as well as the diversity of social and ecological landscapes within which exchanges took place, make it hard to approach empires as socio-political ‘ containers ’ engaging with other such containers via international trade. The con- cept of inter-imperiality , in contrast, brings into focus local economic and ecological heterogeneity , peripheries, as well as imperial coevolution and global (inter)depen  Christian 2000; Rhezakani 2010; Chin 2013. Hansen 2012, 7 ‒ 8 and Galli 2017 for the commercial background of the nineteenth-century Silk Road concept.  Osterhammel 2011; Jennings 2011 for different approaches to globalization in historical perspec- tive.  Morris and Scheidel 2009; Scheidel 2009; 2015; Bang and Bayly 2011; Gehler and Rollinger 2014; Monson and Scheidel 2015.  Barkey 2008; Burbank and Cooper 2010; Düring and Stek 2018. Open Access. © 2020 Sitta von Reden, published by De Gruyter. This work is licensed under the Creative C ommons Attribution-NonCommercial-NoDerivatives 4.0 License. https:/ /doi.org/10.15 15/97831106077 41-001 2 Sitta von Reden dency structures. 5 It focuses attention on regions and landscapes at the edges and in between empires as nodes of exchange and interaction. 6 Much recent historical and archaeological research has established ancient imperial peripheries and fron- tier zones as distinct regions of economic activity and connectivity . 7 Frontier zones are strongly affected by , and affect, imperial development and the fiscal-military politics of nearby or overarching empires. At the same time, frontier zone economies have their own logics, develop special economic opportunities as a result of their location, and in turn have important effects on imperial consolidation. 8 In the con- text of inter-imperial exchange, they are not peripheral but central. The focus on frontier zone processes allows us to analyze long-distance trade within the Afro- Eurasian inhabited world, giving due recognition to the fact that this trade was mediated and often shaped by local or at most regional exchange networks with profound imperial impact. The chronological frame we have chosen for the handbook aims to slice out a historically meaningful period in ancient world history . 9 The centuries between 300  and 300  form a distinct period of dynamic empire formation and trans- formation in the Afro-Eurasian zone. 10 The Hellenistic Empires forming in the after- math of the conquests of Alexander the Great at the end of the fourth century  for the first time created imperial connections between the Greek mainland, Egypt, the Red Sea, and western and Central Asia as far as northern India and the Pamir mountains. At roughly the same time, the Qin and Han dynasties formed the first multiethnic empire in East Asia, while the Mauryans developed what might be called imperial corridors through the Indian subcontinent. In a process of imperial coevolution, the Xiongnu in the Inner Asian steppe promoted greater degrees of political cohesion alongside the Qin and Han, while the Arsakids (Parthians) emerged as a new imperial power independent of the Seleukids, spreading from the Central Asian steppe into western Asia in the course of the third century  . The end of the period is marked by significant imperial transformation across Afro-Eura- sia, which profoundly affected the nature of connectivity . Explicit textual references to long-distance commercial trade along overland routes reaching from China via Sogdiana into Central Asia to markets in northern Mesopotamia and the Mediterra- nean increase in the fourth century  . 11 Before that time, similarly explicit evidence concentrates on the maritime routes between India, the Arabian Peninsula and Egypt, and between the Persian Gulf and Syria via the Euphrates and Palmyra. 12  L. Doyle 2014a; 2014b; see also Turchin 2009 .  Sahadeo 2011.  Cherry 2007 ; Düring and Stek 2018, 3 ‒ 6 with further bibliography .  Rodseth and Parker 2005; Reger 2017 ; Boozer 2018.  Bentley 1996; McKeon 2012; Brosseder 2015, 200 ‒ 203.  Turchin 2009 within a different argument.  Hansen 2012, 19 ‒ 21; Kolb and Speidel 2017 .  Young 2001; Evers 2016; Cobb 2018. Introduction: Ancient Economies and Global C onnections 3 This discrepancy is not accidental, we argue, but the consequence of geopolitical, imperial, and religious change. From the time of the fourth century  onward, the Roman Empire and the integration of its frontier zones changed significantly . By the fifth and sixth centuries, the imperial world of the Mediterranean and western Asia had transformed into what has been called a “ commonwealth of kingdoms, ” or “ empires of faith. ” 13 They marked the beginning of new social and religious orders, a greater mobility of people, and new forms of identity emerging from religious rather than imperial ideologies of belonging. 14 In East, South, and Central Asia, the imperial context changed likewise, though in different ways, with the fall of the Han dynasty , the decline of the Kushan Empire, and the subsequent fragmentation of Central Asia. Yet, the mobility of religious groups and Buddhist forms of ritual expression and art drove new common identities, exchange, and interaction be- tween Indian, Central Asian, and Chinese communities. 15 The imperial world of the previous period was conducive to global connections, too, but in very different ways. Imperial formations strongly influence local economies and the flow of resour- ces and agrarian surplus. Empires are defined as multiethnic political entities, usu- ally created by conquest, held together by a fiscal-tributary regime, and dominated by a political center that exerts some form of power over its constituencies and far distant peripheries. 16 Despite their relatively loose control over regions and local populations, empires create ‒ like more tightly integrated states ‒ a certain degree of institutional stability through administrative and material infrastructures, protec- tion, access to adjudication, and sometimes common coinages. 17 However, being less well integrated than modern nation-states, imperial polities are better ap- proached in terms of overlapping local, regional, and trans-regional exchange net- works rather than as national market economies governed by effective state policy , regulation of coinage, and cohesive structures of contractual law . 18 At the same time, the ability of hegemonic emperors to concentrate resources through oppress- ive fiscal-military regimes leads to the growth of nodes of consumption in the form of spectacular imperial courts and royal cities, large imperial armies, and a gradual increase of administrative institutionalization as a result of tribute extraction and  Fowden 1993; Sarris 2011.  Beautifully described by Brown 2018.  Neelis 2010; 2012.  Classically formulated by M. Doyle 1986, and comparative studies in, e.g., Alcock et al. 2001; Burbank and Cooper 2010; Cline and Graham 2011; Bang and Bayly 2011; Hurlet 2008; Gehler and Rollinger 2014; Stek and Düring 2018.  Allsen 2011.  Ristvet 2018 for the term network empire. The development of more integrated market econo- mies in the course of empire consolidation and greater degrees of state integration has been dis- cussed intensely in connection with the Roman Empire and its long-term development from the second century  to the third century  ; see, e.g., Eich and Eich 2005; Scheidel 2015. 4 Sitta von Reden imperial governance. Effective extraction of tribute, as Shmuel Eisenstadt has fa- mously argued, mobilizes resources and agricultural surplus that otherwise are locked into the claims and consumption regimes of local aristocracies, religious in- stitutions, and other rural or urban hierarchies. 19 In many imperial formations, the disposable economic surplus may not increase through increased productivity and technological change, but through the mobilization of local resources and their sub- sequent entry into larger circuits of exchange. It has been suggested therefore that the organizing principles of the economies of tributary empires are likely to have differed from the political economies of nation-states. 20 Production regimes were influenced by fiscal extraction and imperial consumption rather than by market forces. And capital accumulation was not achieved by capital investment and pro- ductive credit, but by what Subrahanjaman and Bayly have called “ portfolio capital- ism ” : entrepreneurial strategies of making agricultural surplus disposable through market institutions, which were mobilized by fiscal demand. 21 The emphasis on particular principles of imperial economies does not dispute the potential of ancient empires to increase productivity , to stimulate market ex- change, credit, and private economic initiative, leading to better economic perform- ance. It does not dispute the interest of local governments and private economic agents in improving management structures, technology , and institutional infra- structures. But it calls for the need to understand such developments as processes demanding explanation, rather than simply being taken for granted in a grand nar- rative of market development and global commercial trade. It is our contention that frontier zone development had important feedback effects on the empires in the Afro-Eurasian region, that it stimulated inter-imperial connectivity , as well as local, regional, and imperial economic development. This handbook proceeds in three steps in order to develop these various lines of thought in the framework of interdisciplinary research. In the first volume we discuss historical, evidentiary , and historiographic contexts that are indispensable for understanding trans-imperial connectivity . These contexts provide an introduc- tion to historical discussions and a foundation for comparative analyses. Moreover, they develop transparent and reflexive lines of inquiry about how evidentiary and historiographic discrepancies have affected historical understanding, particularly in a trans-regional and interdisciplinary setting. One of the big challenges of inter- disciplinary work is the question of how to cut across disciplinary boundaries. Col- laborative volumes tend to produce parallel demonstrations with little dialog be- tween individual contributions. Single-author publications, by contrast, have a basis in one field combined with limited expertise in others. In order to overcome some of these obstacles, the chapters of this volume have been written by an inter-  Eisenstadt (1963) 2010, xvi, 33 ‒ 68.  Bang 2007 ; 2008.  Bang 2007 , 25; Subrahanjaman and Bayly 1988. Introduction: Ancient Economies and Global C onnections 5 disciplinary team that has collaborated on a daily basis. By providing in-depth ana- lyses, each of which being targeted at readers who are external to that particular field, we aim to provide a foundation for more informed interdisciplinary dialog. The first section of the volume provides contextual information about the Afro- Eurasian world zone in the form of historical overviews. The pioneering studies of ancient global trade, still resonating in more recent work, attributed to particular empires specific functions in the movement of goods from East to West: China and India, with their wealth in precious resources consumed as luxuries elsewhere, were the suppliers; the elites of the Hellenistic and Roman Empires were the consumers; and the poorly documented polities and mobile pastoralist spaces of Inner and Cen- tral Asia were attributed the sad role of having been mere transit zones. 22 With our discussions of these empires, we aim to open the debate about their functions by outlining each empire as a heterogeneous socio-political space that filled multiple roles in exchange circuits of various kinds and scales. Ancient empires, moreover, are known to us through very uneven sorts and volumes of evidence. The second context presented in this volume is, therefore, a series of chapters that discuss the different types of evidence and the methodologies of interpretation that have been brought to bear on economic history across the region. Discrepancies between the evidentiary bodies lie behind many of the false assumptions that underpin modern accounts of ancient Afro-Eurasian connectivity . Working with this material requires bringing together very separate textual tradi- tions and archaeologies in order to overcome imbalances in the data, a process that requires great care. The final part of the first volume situates ancient economic histories and Silk Road studies within a third context: that of particular historio- graphical and disciplinary traditions. Not only have different scientific communities approached economic history differently , but these histories have played different roles in processes of national identity formation and the national self-positioning in local or imperial pasts. Tracing several discourses concerning the study of Afro- Eurasian economic history helps to explain the divergent nature of contemporary conversations about this space, while also contextualizing the work of previous gen- erations of scholars and increasing its legibility and reconsideration within our own frameworks. In the second vol ume we identify economic struct ures and developments that allowed resources, goods, and capital to concen trate as well as to spread into fron- tier zones. Over the last generations , economic history , archaeology , and anthro- pology have seen a large amo unt of theoretical work which has increased our un- derstanding of economic de velopment in vastly different social and ecological settings. We will use these insights in order to develop models of explanation and  Chase-Dunn and Hall 1977 , 149 ‒ 187 ; Cameron and Neal 1986, 20 ‒ 43. Examples of more recent work include Fitzpatrick 2011 and many studies on Indo-Roman trade; Chakravarti 2015 for discus- sion. 6 Sitta von Reden methods of analysis that are suitable for comparative research at a global scale. We start from nodes of acqui sition and consumption that are v isible in the archaeo- logical and textual record s in the form of househ olds, temples, cities , and capitals, as well as armies that were stationed in some cases permanently in imperial fron- tier zones. Taki ng into consideration the diver sity of social systems, ecologie s, and evidence across the ancient Afro-Eurasian world region, our investigation will be divided into “ actors ” and “ tools ” tha t stimulated extraction, concentration, a nd circulation. As actors we d efine administrative systems, social grou ps, and individ- uals as much as institution s and landscapes. The se actors deployed or acted upon certain tools of extraction, concent ration and circulation: money , markets, te chnol- ogy , legal syste ms, and physical infrastructures. Through these broa d categories of analysis we hope to compare and explain different kinds and s cales of economic development across the Afr o-Eurasian empires without unde rprivileging regions that have left notoriously difficult evidence. In the third volume, we will analyze frontier zones and borderlands as sites of social and imperial encounters as well as network formation. 23 Much recent work has been devoted to analyzing frontiers and borderlands as distinct spaces. They are recognized as specific sites of imperial encounter and interaction in the form of negotiation and appropriation, but also resistance and violence. 24 Combining recent research foci and the expertise of the researchers collaborating on this handbook, the volume will focus on a particular selection of frontier zones, such as the Egyp- tian Eastern Desert, Syria, the Ponto-Caspian zone, the Hindu Kush, forested Central India, and the Hexi corridor. While the second volume considers forms of economic development in frontier zones, the chapters of the third volume take a more com- plex approach to borderland processes. Borderlands and frontiers often form along and across open boundaries between agrarian, pastoral, and coastal communities, as well as along and across different ecologies that require different economic and social strategies. 25 Such dividing lines suggest, both theoretically and empirically , more complicated scenarios for the movements of goods than is captured by the idea of trade from empire to empire. The second part of this volume will thus turn to the nature of exchange networks within and across imperial frontiers, and their connections to local, regional, and imperial hinterlands. Although, given the struc- ture of available datasets, formal network approaches are only occasionally applica- ble, we nevertheless favor network perspectives as they are less loaded with eco- nomic presumptions than market concepts. 26 Networks and markets are two spatial  Boozer 2018, 209 ‒ 214 for a discussion of the concepts of frontier zone and borderland, often used interchangeably . Brosseder 2015 for the different anthropological models and network theories that have been applied to long-distance exchange in recent archaeological and anthropological research.  Ristvet 2018.  Stek and Düring 2018, 357 ‒ 358.  Brosseder 2015, 209 ‒ 210. Introduction: Ancient Economies and Global C onnections 7 expressions of exchange that relate small intensive and large extensive circuits of exchange together, often under particular local social, military , religious, and ad- ministrative circumstances. 27 There is no doubt that much exchange in frontier zones took the form of trade via marketplaces, as markets provide a unique degree of condensation of interaction, as well as great opportunities for institutional and fiscal control. Yet the concept of the market ‒ so closely related to the debates over national economies ‒ underrepresents the social dimensions of exchange, the heter- ogeneity of its motivations, and its social complexity . 28 The profound importance of inter-imperial exchange for empire formation and transformation will emerge against the backdrop of these complications in a combination of micro-perspectives and their relevance to understanding larger circuits of exchange connecting vast geographical distances. Ref erences Alcock, S. E. 2001. Empires: P erspectives from archaeology and history . C ambridge: Cambridge University Press. Allsen, T . T . 2011. “ Premodern empires. ” In J. H. Bentley (ed.), The Oxford handbook of world history , 361 – 379. Oxford: O xford University Press. Bang, P . F . 2007 . “ Trade and empire: In search for organizing concepts for the R oman economy. ” Past and Present 195, 3 – 54. – . 2008. The Roman bazaar: A comparative study of trade and markets in a tributary empire . Cambridge: C ambridge University Press. Bang, P . F ., and C. A . Bayly, eds. 2011. T ributary empires in world history . Cambridge: C ambridge University Press. Barkey, K. 2008. E mpire of difference: The Ottomans in comparative perspective . C ambridge: Cambridge University Press. Bentley, J. H. 1996. “ Cross-cultural interaction and periodization in world history. ” American Historical Review 101.3, 7 49 – 770. Boozer, A. L. 2018. “ The archaeology of imperial borderlands: A view from Roman Egypt and Sudan. ” In B. S. Düring and T . D . Stek (eds.), The archaeology of imperial landscapes: A comparative study of empires in the ancient Near East and Mediterranean world , 206 – 239. Cambridge: C ambridge University Press. Brosseder, U. 2015. “ A study on the complexity and dynamics of interaction and exchange in Late Iron Age Eurasia. ” In J. Bemmann and M. Schmauder (eds.), Complexity of interaction along the Eurasian steppe zone in the first millennium  , 199 – 332. Bonn: V or- und frühgeschichtliche Archäologie, Rheinische Friedrich-W ilhelms-Universität Bonn. Brown, P . 2018. “ C ommerce, diplomacy, and cultural contacts along the Silk Road in Late Antiquity. ” In N. Di C osmo and M. Maas (eds.), E mpires and exchanges in Eurasian Late Antiquity: Rome, China, Iran, and the steppe, ca. 250 ‒ 750 , 96 – 107 . Cambridge: Cambridge University Press.  See the contributions in Knappett 2013; Leidwanger and Knappett 2018.  Classically , Granovetter 1985; also Bang 2008. 14 Sitta von Reden landscapes, and inter-imperial connections we investigate in the next two volumes. Yet the variable degrees of academic attention that the spaces we cover have re- ceived in global scholarship have led to discrepancies in the ways these chapters proceed. Given that the histories of the Hellenistic and Roman Empires have fre- quently been written and re-written, these two accounts were challenged by the goal of communicating established narratives without eliding controversies and al- ternative approaches. In other cases, the structures of the imperial spaces are more elusive, and the histories that are given here are rather to be read as contributions to ongoing disciplinary debates. Ref erences Scheidel, W . 2015. “ Introduction. ” In W . Scheidel (ed.), State power in ancient China and Rome , 3 – 10. Oxford: O xford University Press. Burbank, J., and F . C ooper. 2010. E mpires in world history. Power and the politics of difference . Princeton NJ: Princeton University Press. Sitta von Reden 1 The Hellenistic Empires I Introduction The Hellenistic Empires that emerged in the aftermath of the Macedonian conquest of the Achaemenid Empire (ca. 550 – 330  ) laid the foundation for the connectivity of Asia, Egypt, and the Mediterranean for many centuries to come. Although often referred to as kingdoms, they are best understood as empires dominating a vast and heterogeneous political and economic space. 1 The ideological goal of their kings was world dominion, yet the degree of their control over regions, resources, and local polities was uneven, and their styles of rule responsive to various micro- and macro-traditions. 2 Looking at the Hellenistic kingdoms as parts of a dynamic imperial space alters our perspectives on their territoriality , frontier zones, and temporality . From the time of Alexander ’ s campaigns onward, the Hellenistic Empires were in constant transformation. They gained some coherence through fiscal-administrative practic- es, military regimes, and political ideologies. But they remained fragmented due to their mutual competition and by the continuity of local power regimes, hierarchies, and economic practices that influenced local land and people. The Hellenistic impe- rial world was also polycentric, both in terms of the continuity of local power cen- ters and in terms of the distribution of royal capitals that were built in various parts of the Hellenistic Empires. 3 Given the polycentricity of the Hellenistic Empires, what constituted imperial cores and peripheries is arguable. The cities of the Aegean and  Ma 2013; Thonemann 2013; Fischer-Bovet forthcoming for chapter-length accounts of the Helle- nistic Empires. Companion volumes such as Erskine 2003; Bugh 2006; Weber 2007 cover many aspects discussed in this chapter. Particularly influential studies include Koenen 1993; Ma 1999; Manning 2003; Capdetray 2007 ; Kosmin 2014, 2018. This chapter is indebted to all these works.  Macro-traditions refer to long-term imperial memories and models (Persian, Hellenistic, Roman, etc.), and micro-traditions to local memories and continuities. Their complex relationship is at the center of research on the Hellenistic period, see numerous articles by Pierre Briant published since the 1970s, some being translated in Briant 2017a, 429 ‒ 610; Koenen 1993; Sherwin-White and Kuhrt 1993; and, more recently , Manning 2010; Moyer 2011; Ma 2013; Strootman 2014a, 2014b; Strootman and V ersluys 2017 ; Fischer-Bovet forthcoming.  Kosmin 2014, 93 – 120. The cores of the Seleukid Empire were Syria and Babylonia; Egypt had its core in the delta and Lower Egypt, but rival courts existed temporarily in Cyprus and Cyrene. The Antigonid West had old Macedonia at its core, but Athens maintained some symbolic centrality in the Greek world. Note: I am grateful to Peter Eich, Christelle Fischer-Bovet, and Andrew Monson for helpful discus- sion and comments on this chapter. Special thanks to Jennifer Gates-Foster, Christelle Fischer- Bovet, and Milinda Hoo for letting me see forthcoming work. Open Access. © 2020 Sitta von Reden, published by De Gruyter. This work is licensed under the Creative C ommons Attribution-NonCommercial-NoDerivatives 4.0 License. https:/ /doi.org/10.15 15/97831106077 41-003 16 Sitta von Reden Map 1: C ores and peripheries of the Hellenistic Empires during the early third century  . © Peter P alm. The Hellenistic Empires 17 Asia Minor, for example, temporarily formed an outer circle of the Ptolemaic Em- pire, thus being peripheral to its core. Yet culturally and symbolically , they were the very core of the Hellenistic world, as kings defined themselves as Macedonians and Greeks. Many Greek immigrants in the new Seleukid and Ptolemaic Empires came from these regions. The same holds true for Thrace and Southern Syria in that both lay between two Hellenistic spheres of influence. They were geographically at the border of two Hellenistic Empires, but economically and culturally they were by no means marginal. The societies of the Hellenistic Empires had several cultural layers. Neither the conquerors nor the conquered populations were culturally or ethnically pure enti- ties moving in and out of foreign domination. What we call Greek, Persian, Egyp- tian, or Iranian are convenient but misleading shortcuts for a transcultural mix of traditions, languages, material culture, social practices, and political systems that converged in special circumstances. Representations in Greek iconography or lan- guage were not necessarily produced by Greeks. Public inscriptions in one particu- lar or several languages, and artifacts in either one or a mixture of styles, were statements about the intended addressees, affiliations, and self-positioning within a repertoire of imperial possibilities. Vassal or client kings who adopted Greek coin- ages (Hellenistic or Athenian, even), or used Greek as the official language, did not necessarily speak Greek among themselves, or regarded themselves as Greeks, but placed themselves deliberately or by necessity within a particular imperial orbit. I present two examples. In the centuries preceding the Macedonian expansion into Asia and Egypt, the dynastic predecessors of Philip and Alexander had deliberately positioned them- selves as Greeks. They constructed mythical descent from Herakles, adopted Greek as the language of official communication, and issued coinages based on the Athe- nian weight standard. When Alexander ’ s successors established themselves as kings over Asia and Egypt, they continued to emphasize their belonging to what we now call classical Greek culture. They collected Greek literature, adopted Athenian law and legal practice, listened to Greek philosophers, and made Greek education a focus of court culture and cultural patronage. Yet the Macedonians had never lived in city-states ( poleis ), nor had they shared typically Greek civic values so strongly associated with Greek poleis , such as equality and freedom. Nevertheless, poleis with civic institutions became major instruments of administrative and ideo- logical control of the Hellenistic Empires, especially in Asia, but also in Egypt and Macedonia itself. 4 Another example is Egypt. In the eyes of both ancient and modern observers, Egypt seemed and seems rather resilient to cultural change. Temple iconographies and inscriptions were highly conventional and formulaic well into the Ptolemaic period. They emphasized conquest as the condition of peace, the return of the di-  Ma 2013; Clancier and Gorre forthcoming. 18 Sitta von Reden vine images (statues) as a sign of benevolent rule, and pharaonic support of temple building activity as the condition for general prosperity . The Ptolemies, too, were represented in these terms, having expelled the Persian foe, returned the statues to the Egyptian temples, and supported the indigenous cults and temples. 5 Yet not only the Ptolemies, but also the Persians, had respected the Egyptian gods, and just like the Ptolemies, they had transformed the country administratively and culturally in cooperation with local priesthoods. 6 When the Macedonians conquered Egypt, they encountered several political, administrative, and cultural traditions to which they added further. And when the Romans conquered Egypt 300 years later, once again, being Egyptian or being Greek were fairly artificial administrative labels slid- ing over much more complex social realities. 7 Because of the diverse motivations and social practices that caused Hellenism to spread, it is also problematic to use the term Hellenization as an active pursuit of the Graeco-Macedonian ruling elites. The fact that the Greek Alexander historians lamented Alexander ’ s adoption of Persian rituals and intermarriage with Bactrian and Persian aristocracies shows that Greeks had cultivated other expectations of the conquest than what had happened in its aftermath, namely interaction and en- tanglement, rather than segregation and revenge. 8 Despite coercion and control, the Hellenistic ruling apparatus incorporated and responded to long-lasting local cul- tural practices that continued to resonate in the emerging Hellenistic tradition. 9 II The Mac edonian T akeover The story of Alexander III ’ s campaign is too well known to be in need of much elaboration here. A war against Persia had been planned under Alexander ’ s father Philip II, and the Corinthian League formed by the Greek cities in mainland Greece and the Aegean had agreed to it. Macedonian imperial ambitions had developed rapidly from the mid-fourth century onward. Philip ’ s armies had created military corridors in Thrace, Thessaly , and Illyria. He had married into several royal families to secure these and other territories. And there had been attempts to negotiate coop- eration and confederations especially with the important cities of Corinth, Thebes, and Athens. The Corinthian League gave their support to the war, propagated as a pan-Greek revenge against the Persian invasion during the Persian Wars of the early  Briant (2003) 2017c.  Lloyd 2000, 383 – 384.  Bowman 1996, 121 – 186.  Badian 1965.  Moyer 2011; Strootman 2014a, 2014b; Koenen 1993 for transcultural ideas of Ptolemaic kingship; numerous agrarian and administrative practices went back to Persian, Pharaonic, or Greek prece- dence; Briant (1994) 2017b; Manning 2003; Thonemann 2013. The Hellenistic Empires 19 fifth century  . Yet Philip was assassinated before the campaign started, probably in reaction to his seventh marriage that could have impaired Alexander ’ s dynastic succession. Alexander ascended to the throne in 336  . In 334, his army crossed the Helle- spont that separated Macedonian Thrace from Persian Lydia. Within two years, the Greek cities in Asia Minor were ‘ liberated ’ from the Persians, the Phoenician cities in Syria captured, and Egypt and Libya transformed into a Graeco-Macedonian satrapy (Persian administrative unit). Alexander ’ s troops pushed into Babylonia, Persis, and Media where the former Persian residences were located. They plundered the rich treasuries of Babylon, Susa, and Persepolis, installed Greek satraps , burned down the citadel of Persepolis, and pursued the Persian king Darius III who had fled to the Median capital of Ekbatana. The rule of the last Persian king, Darius III, had been anything but stable, much to the advantage of Alexander, who cooperated with disaffected local elites and their armies. One of the pretenders was Bessos, a relative of Darius and royal satrap of Bactria. Though Bessos claimed the empire for himself, his ambitions served Alexander ’ s interests. It was Bessos who finally killed Darius when he took refuge from Alexander in Ekbatana. Yet, rather than negotiating with the usurper, Alexan- der aligned himself with his opponents by offering royal burial to Darius. Bessos fled to Bactria, where he proclaimed himself King of Kings, Artaxerxes V . Losing the support of his own generals, Bessos was delivered to Alexander and executed in Ekbatana in 329  . The conquest of Bactria and Sogdiana still proved difficult and violent, despite some cooperation of local military elites. It seems that it was only through marriage into a leading Bactrian family that Alexander eventually gained control over the powerful central Asian kingdoms. 10 Marrying a Bactrian princess was not the only gesture by which he achieved convergence to Persian political practice. Already in Egypt, Alexander had had pro- claimed himself son of Ammon Zeus by the oracle of Siwa. This combined Macedo- nian ideas of divine kingship with Persian concepts of the King of Kings being the substitute of god, as well as integrating the Egyptian notion of the pharaoh en- dowed with divine K a . It had great appeal to local populations, kings, and priests, whose claims over land and income were thereby confirmed. However, many Greeks resented the idea of a divine kingship. This was partly because it implied a degree of hierarchy irreconcilable with civic equality , but more importantly it involved a degree of military expansionism not all Alexander ’ s generals were willing to sup- port. Alexander ’ s rejection of Darius ’ s peace offer in 330  had already provoked criticism among a military exhausted by warfare. Internal discontent increased as Alexander made his armies march farther east and began to adopt elements of Per- sian dress and court ritual. The link between expansionism and resentment is amply  Although Arrian calls her a captive of war (Arrian Anabasis [Arr. Anab .] 4. 19 ‒ 20), later her father Oxyartes was installed satrap of Paropamisos (6. 15). 20 Sitta von Reden proved by the increase of conspiracies against Alexander during the far-eastern campaigns. In their course, a whole level of generals was exchanged for personnel that were more loyal. These new generals were the contenders for Alexander ’ s suc- cession and transported the idea of a universal empire to the next military genera- tion. In 327  , the armies turned south and, crossing the Hindu Kush, marched down to the western Indus valley . The region seems to have had an ambiguous status under the Persians as either satrapy or vassal kingdom. 11 A Macedonian sa- trap was installed here, while the area south of the Hydaspes, under control of an Indian king called Poros in Greek sources, became a vassal kingdom after several indecisive battles in 326  . Alexander and his armies were able to develop further Macedonian corridors in northern India along the Indus valley down to the city of Patala (see ch. 3, map 1, this volume), occupying two more short-lived satrapies bordering the Thar Desert. In 325  , Alexander was ruler of a space that, in terms of satrapal administration, was more or less coextensive with the Achaemenid Em- pire. It excluded, however, northern parts beyond the Lesser Caucasus, and the steppe regions along the western and eastern coasts of the Caspian Sea. No Hellenis- tic king ever attempted to expand beyond these ecological boundaries. 12 III Ec ologies and C limates When Alexander died in 323  , the Macedonian Empire stretched across a vast and heterogeneous ecological space (see map 1). Fertile alluvial plains were inter- laced with mountain ranges and desert zones, neither easily subjected to social and political control, nor very suitable for economic exploitation. In the west, the empire included the densely populated Mediterranean micro-ecologies of the Aegean and Asia Minor characterized by Greek poleis, for which dry farming, private property rights, a specific relationship between town and countryside, and political regional- ism were typical. In the south, it stretched into the populous country of Egypt, which for millennia had been dominated by the Nile ecology that produced im- mensely rich agricultural yields under the condition of artificial irrigation. Tradi- tionally , this land was controlled through institutional landownership by the tem- ples and the pharaoh. In the north, the empire included the fertile Pontic region (never conquered by Macedonian armies), partly settled by Greek-speaking popula- tions since the seventh century  . Its northernmost frontier was adjacent to and in some connection with populations of the steppe belt, a socio-ecological zone that  Briant 2002, 754 – 757 .  There were expeditions along the Iaxartes and into the steppe of Inner Asia in the early years of Seleukos I, but these seem to have been explorations rather than attempts at conquest, Kosmin 2014, 72 – 73. The Hellenistic Empires 21 was traditionally alien to the Greek way of life. In the east, it approached the natural border of the Pamir and Himalaya, while farther south it reached southwest down into the Indus Valley . The ecological micro-systems of the Asian landmass are too numerous to be surveyed here. Yet with six of the seven administrative complexes (Great Satrapies) of the Achaemenid Empire located in Asia, these offer a useful guide to understand- ing the Asian landmass as an imperial as well as economic landscape. 13 Though highly diverse and disconnected in themselves, they nevertheless seem to have formed networks of populations that were grouped together into administrative en- tities, arguably long before even the Persian expansion. 14 These were, firstly , Persis and eastern Iran, the heartland of the Achaemenid dynasty . Secondly , Media spanned across large parts of the Zagros Mountains, forming the western border of Sogdiana and Bactria. This was a land of mountains and valleys along the axis of the mountain range, with one major transverse river valley forming a major line of communication between Babylon and Ekbatana. Thirdly , Lydia stretched across Anatolia with its Mediterranean and Black Sea temperate climate, on the one hand, and pocketed micro-ecologies ranging from fertile river valleys to high mountain ranges in eastern Anatolia, on the other. Fourthly , Mesopotamia and Babylonia were situated along the rivers Euphrates and Tigris, which watered wide and very fertile alluvial plains providing great opportunities for cereal agriculture under artificial irrigation. Their economic potential, both in terms of agrarian fertility and strategic position as a contact zone between the Persian Gulf and Syria, was fully exploited by the early Seleukids. 15 It made Syria and Babylonia the core region of their empire, both politically and economically (see map 2 and 3). There was, fifthly , Arachosia, with the core region in southern Afghanistan, including the Gedrosian Desert in the south, Drangiana in the northwest, and the lower Indus valley in the east. Much of this region was covered by arid desert zones, but Arachosia itself was agriculturally fertile, and though Drangiana was mostly a poor desert zone, it had some fertile plains along the river Etymandros. Finally , Greater Bactria comprised the kingdoms of Bactria (roughly the area of central and northern Afghanistan), Sogdiana in the north, and Aria in the southwest. 16 It is a region of semiarid mountain zones, inter- laced with fertile river valleys that are potentially very productive if irrigated suc- cessfully . The great potential and power of Bactria must be seen in its location adja- cent to the steppe and its strategic position as a communication line of roads and infrastructures linking Mesopotamia, Iran, and northern India. 17 Administratively ,  Also Morris, ch. 2, this volume.  Jacobs 1994; Jacobs 2011; and Morris, ch. 2, this volume. See Aperghis 2004, 35 – 51 for a bold survey of local geographies and estimated population densities.  Fabian, ch. 6, this volume for the importance of this route for the Arsakids ruling this space thereafter.  Morris, ch. 2, this volume.  Mairs 2014 with Strabo Geography (Strab.) 11. 11. 1 on the fertility of Bactria; for the connections of the Steppe with Gandhara, Morris, ch. 2, this volume. 22 Sitta von Reden the Far Eastern satrapies reached across the Hindu Kush down to the upper Indus valley , once again connecting mountain ecologies with fertile agricultural areas. 18 As Morris argues below , the Hindu Kush posed a far less severe socio-ecological dividing line than the Alexander historiography wants us to believe in its attempt to stylize Alexander ’ s Hindu Kush crossing as a truly heroic achievement. Neverthe- less, the satrapies of the northwestern Indus valley were soon ceded to the king of the new Mauryan Empire that formed shortly after the Macedonian conquest. Though these satrapies were lost, relationships between the Mauryan and Seleukid kings remained friendly , and the boundary uncontested. IV Empires in the Ma king The formation of the Hellenistic Empires after Alexander ’ s conquests and death con- tinued to be an immensely violent process. The ambition of becoming the ruler of the entire empire that Alexander had left gave rise to six Wars of the Successors (321 – 272  ) and remained the goal of all contenders who eventually occupied parts of it. 19 The story of these wars is a series of indecisive victories and defeats, changing military alliances, dynastic marriages, assassinations, and intrigues. Yet this story helps to understand Hellenistic history as prefigured by internal competi- tion and ever-in-flux zones of conflict and internal frontiers. Because of the continu- ous competition and warfare between the successor kings and allies among the con- stituent polities, Hellenistic imperialism was essentially inward looking. 20 After Alexander ’ s premature death, the succession was unresolved. His mental- ly handicapped half-brother Philip III Arrhidaios and his yet unborn half-Bactrian son, Alexander (IV), were the two dynastic heirs, but unfit for government when the succession was to be settled. However, as succession in the Macedonian monar- chy was largely dynastic, the two heirs were appropriated for rival claims to legiti- mate succession. Their assassination in 317 and 309 respectively , despite making little difference to actual politics, changed the rules of the game and remade legiti- macy to mean a combination of military success, support of armies, and local ac- ceptance.  Kulke and Rothermund 2018, 13.  Meadows 2012; Hauben and Meeus 2014; Strootman 2014b; Fischer-Bovet 2019 with special em- phasis on the Ptolemaic ambitions in southern Anatolia. For an emphasis on the alternative territo- rial designs of the Seleukids, Tuplin 2009; Kosmin 2014.  Bugh 2006. The main sources for the history of the Wars of the Successors are Diodoros ([Diod.] 17 and 18); several Lives of Plutarch (Plut.), and the Alexander historiographies of Arrian, Curtius Rufus, and pseudo-Kallisthenes. The Hellenistic Empires 23 There were several contenders for succession. Among the older generation of Alexander ’ s father, the most important ones were Antigonos Monophthalmos ( “ the one-eyed ” ), Perdikkas, Antipatros (followed by his son Kassandros), and Polyper- chon. All belonged to the Macedonian aristocracy , had institutional ties of friend- ship with Philip II and Alexander, and occupied top ranks in Macedonia or the imperial army . A second group of aspirants were the generals of Alexander ’ s genera- tion. These were, to name but the most prominent, Lysimachos, who had become satrap of Thrace after Alexander ’ s death; Eumenes, in charge of the northeastern Anatolian satrapies ; Ptolemy , satrap of Egypt; and Seleukos, military governor ( chiliarch ) of the troops in Macedonia. Perdikkas seems to have been the strongest contender at first, as he was elected commander of the imperial army in the first meeting of the generals after Alexan- der ’ s death. He suggested they hold off on further action until Alexander ’ s child was born. This plan would have put him in the position of trustee of the dynastic successor as soon as the baby was born. Yet opposition formed around Antipatros, who claimed dynastic succession for Arrhidaios, and the infantry armies. Antipatros was also an ancestral enemy of Olympias, Alexander ’ s mother, who had forged a marriage between Alexander ’ s sister Kleopatra and Perdikkas, with dynastic inten- tions in mind. Possession of Alexander ’ s body was another strong symbol for legiti- mate succession, and armed conflict became inevitable when Ptolemy captured it for burial in Alexandria. Rebuffed by this arrogant act, Perdikkas marched against Ptolemy , while Antipatros pursued Perdikkas into Asia Minor. Yet before war broke out, Perdikkas was killed by members of his own army , leaving Ptolemy in posses- sion of Alexander ’ s corpse, and Antipatros the new leader of the imperial army . However, his ambitions were cut short by his own unforeseen, though natural death just two years later. Antipatros ’ s position now passed on to his sons, Kassandros and Polyperchon. Polyperchon revived the Corinthian League to gain the support from the Greek cit- ies. But Kassandros joined forces with Antigonos, military governor of Asia, against Polyperchon and the Macedonian Perdikkas-Olympias alliance. Olympias, still striv- ing for the succession of Alexander IV , had Arrhidaios and his wife murdered in 317  . But she was executed by Kassandros in retaliation. The death of Olympias, Philip III, and his wife allowed Kassandros to strengthen his position in Macedonia, which was consolidated by his marriage to Thessalonike, a daughter of Philip II and heiress of the kingdom. 21 Meanwhile, Antigonos fought several wars to solidify his position in Asia. In 316, he controlled much of Asia, and according to one ancient source, was proclaimed King of Asia in Persepolis. 22 Yet his position created a new opposition joining Lysimachos of Thrace, Ptolemy of Egypt, Kassandros of Macedo-  In principle, any male offspring of a Macedonian king was entitled to the inheritance of the kingdom; see Ogden 1999; Strootman 2014b.  Diod. 19 . 14. 1. 30 Sitta von Reden or Macedonian roots – a process that can best be observed in Egypt. 44 Throughout the Hellenistic period, the friends formed a hierarchical and exclusive group of cul- tural Hellenes that reproduced their status through Greek education, Greek forms of entertainment, music, games, and athletic contests, as well as the orchestration of the ruler cults. The culture and learning of the courtiers, their ceremonies, and rituals spread to local metropoleis and towns where Greek and local elites emulated it as a means of integrating themselves into the wider group of the dominating ethno-class. The second characteristic of Hellenistic monarchy was the royal interaction with local constituencies: cities, temples, and populations. This interaction was partly coercive and tributary in nature, as the next section will show . But in order to create stable structures of authority and legitimacy , it also took more benevolent forms. Gift-giving and offers of protection were the most important means by which the kings created relationships with, and loyalty among, different social bodies. Gifts were sent to sanctuaries at Delphi, Olympia, or Delos in order to enhance royal prestige within the symbolic centers of the Hellenistic Empires. Gifts in the form of subsidies or tax relief were given to cities who reciprocated with crowns, statues, and cultic honors. Local temples received financial contributions ( suntaxeis ), tax relief, and support for construction work, as well as grants of income from land and customs dues, in return for ritual services that stabilized the kingdom. Negotiations and concessions are best known from the multilingual stelai (the most famous being the Rosetta Stone) that the Egyptian priests erected at the end of ritual gatherings ( sunodoi ) between kings and priests. Other recipients of gifts were high officials entering the administrative apparatus, and provincial military officers who were rewarded with land and gift estates ( doreiai ). In return, the kings gained loyal fol- lowers, economic prosperity , and increased tax income. 45 Ruler cult was a third characteristic of Hellenistic monarchy and closely con- nected to the kings ’ role as benefactors. Religious cults had been devoted to kings in Greece, Persia, and Egypt before, yet the immediate model was Alexander. 46 Al- exander had become a religious figure in Egypt immediately after his visit to the oracle of Siwa (which had confirmed him as a son of Zeus), while numerous Greek and non-Greek cities offered him cults spontaneously on the occasion of a victory , sometimes on the demand of generals and the king ’ s friends. The successor kings immediately latched onto this ritual. A cult for Demetrios and Antigonos was set up in Athens when they liberated the city from Kassandros. Ptolemy II introduced a cult for his deceased father and his father ’ s wife Berenike, who were worshipped as Savior Gods ( Soteres ) from about 278  . Soon afterward, he set up a cult for his  Fischer-Bovet 2014.  There are some indications that large gift estates formed distinct tax districts for whose income the estate owner was responsible.  Jones 2010; Mitchell 2013. The Hellenistic Empires 31 own deceased sister-wife, Arsinoe. Not only was this cult henceforth celebrated in all Egyptian temples with revenues of one of the largest taxes in Egypt (the so-called apomoira ); but Arsinoe being associated with the mother goddess of Aphrodite equalling the Egyptian goddess Isis, became the most popular cult-figure in the entire Ptolemaic world. 47 In dating formulas, Ptolemy put his own cultic name Phil- adelphos ( ‘ sister-loving [god] ’ ) alongside the title of the Alexander priest who was eponymous for each year, and thus both circulated together as part of all official documents throughout the country . 48 In Seleukid Asia, essentially the same thing happened. Kings received cults by cities spontaneously , while they were divinized officially after their death. From the late third century onward, living kings started to have themselves divinized, as we learn from dating formulas and official letters requesting that priests for the living king be elected. 49 In the Antigonid kingdom controlling much of the Greek world in the mainland and Aegean, no official ruler cult was established, but spontaneous cults sprang up in Greek cities in honor of individual kings. Ruler cults and the divine status of the kings and queens consolidated the cos- mological role of the Hellenistic monarchies. It elevated the kings and queens to god-like figures who were in control of natural fecundity , procreation, prosperity , and the Nile inundations. The king (pharaoh ) as a divine figure also legitimized local Egyptian priesthoods ’ claims to income from land and people. 50 In combina- tion with the king ’ s and queen ’ s role as benefactors, the ruler cult formed a major means of communication between local cities and royal centers. The kings and queens were far away , but their sacred districts, priesthoods, and statues were present, as were the libations, sacrifices, and festivals taking place in their names. 51 The celebrations of festivals and cults were powerful means of building identity and community in the Hellenistic world. Local social life was attached to their figures and rendered the experience of empire a matter of joyful festivity and song. The power of cults also helped to build dynastic continuity , a major problem in any monarchy , but in particular in the Hellenistic Empires where multiple marriages created ambiguous lines of succession. The problem of dynastic succession was greatest in the Asian and Egyptian empires, while in Antigonid Macedonia, the vio- lent fight over the succession of Alexander seems to have put an end to the conflicts over succession once and for all. In the Asian and Egyptian empires, however, the Seleukids and Ptolemies established new royal dynasties with Alexander as their founder. This dynastic origin was widely disseminated through coinage minted in Alexander ’ s name and carrying his image. Eponymous priesthoods were established  Koenen 1993.  Chaniotis 2003.  Chaniotis 2003, 473; Lanciers 1993.  As for example expressed in the Pithom stele , Mueller 2006, Appendix.  Chaniotis 2007 , 139 – 140. 32 Sitta von Reden in his name, while numerous stories developed about his divine origin and heroic status acquired by conquest. Local Egyptian myth-making gave him a double de- scent as the son of Philip II of Macedonia and of Nektanebo, the last native pharaoh of Egypt. 52 The dynastic cult was most strongly developed in Egypt. Dynastic festi- vals were celebrated regularly , and dynastic succession expressed explicitly on early Ptolemaic coinage. 53 For fear of dynastic trouble, Ptolemy II even married his own sister Arsinoe shortly before her death. Brother-sister marriage was practiced hence- forth roughly every other generation in the Ptolemaic dynasty , although this did not solve dynastic frictions. Multiple marriages aiming to confirm territorial possessions and diplomatic friendship fostered internecine warfare and alienated the royal houses internally . Both ancient and modern commentators regard the dynastic problems of the Hellenistic royal houses as one of the main reasons for the Hellenis- tic monarchies ’ incapacity to respond successfully to military and political challen- ges in the long-term. V .  Administration The administration of the empire was linked to the person of the king but gradually developed into an institutional apparatus governing in his name. The intimate rela- tionship between the king and his administration had conceptual, ideological, and economic dimensions. Conceptually , the administration was just another aspect of the king ’ s pragmata (affairs). Ideologically , the empire was the king ’ s personal household, just as regional administrations were regarded as parts of this house- hold. 54 In principle, all land was royal land. It could be ceded or given to others, but some of it was cultivated by the king ’ s own royal tenants. Like a good landlord, the king was personally interested in his land and agrarian development. Economi- cally , benefactions to local elites and temples, land donations, and land develop- ment created incentive structures for the mobilization of local resources and people. This form of rule was not just a Macedonian invention. At the time of conquest, the Macedonian satraps took over the administrative systems of the Achaemenid Empire and only gradually changed these structures as a matter of practice. The Achaemenids, on their part, also had not interfered radically with preexisting ad- ministrative practices, authority structures, and agrarian systems. 55 Yet they eventu- ally established a centralized system of foreign authority to which the local constitu- encies responded. They had created roads and land registers, standard weights,  Alexander Romance I, 1 – 30; Stephens 2003, 64 – 73.  For dynastic cult, Hölbl 2001, 77 – 124, 160 – 177 in place of numerous other studies. For the dis- tinction between dynastic and ruler cult, see Pfeiffer 2008.  Gehrke 2008, 48, 57 – 59 .  Briant 2002, 411. The Hellenistic Empires 33 measures, and some coinages, and probably also the practice of differentiating be- tween royal land and other types of land. 56 Keeping the diversity of local structures in mind, some general principles of the Hellenistic administration can be observed. Two qualifications need to be made, nevertheless. First, due to the excellent papyrological evidence from rural Egypt, 57 we are much better informed about this than any other part of the Hellenistic World, including Alexandria and the Greek cities with their exceptional historiographical tradition. Egypt was as typical or atypical for the rest of the Hellenistic world as any other part of it, but it had developed particularly cohesive state structures and a highly developed administrative system over 2,000 years. Its agricultural rhythm was determined by the Nile inundation, its agrarian income dependent on changing field sizes, and the local administration heavily involved in the maintenance of the dyke system. All of this created specific administrative structures, similar perhaps only to those of Babylonia for which we lack comparable evidence. The second problem is administrative language. Having to find Greek terms for local institutions and offices, the conquerors had at their disposal a limited linguistic repertoire that had developed in the Greek and Macedonian context. Administrative terminology thus obscured substantial socio-economic differences and created a degree of coher- ence and familiarity that will have been greater in linguistic than in practical terms. To judge from the Egyptian evidence, the basis of Hellenistic administration was the district. All of Egypt was divided into what the Greeks called nomoi , region- al divisions dating back to the Old Kingdom. The nome reflected the continuity of local agrarian organization regardless of the shift of domination in the political cen- ters. 58 At the top of the nome was a nomarchos ( ‘ head of the nome ’ ) who supervised agricultural production and the dyke system, as well as fulfilling some judicial func- tions. An oikonomos was in charge of the nome income in cash and in kind, while a checking clerk ( basilikos grammateus ) keeping records worked at his side. They had their subordinates in the topos (tax district ) and the komos (village). The tax income, carefully pre-assessed on the basis of a sowing schedule and plot sizes each year, was collected, recorded, and countersigned in storehouses ( thesauroi ) to which the Ptolemies added banks ( trapezai ) with very similar accounting systems. Yet while the in-kind income was collected by the nome administration itself, all monetary taxes were auctioned off to tax farmers. Though not collecting the taxes themselves, they guaranteed a fixed total sum and kept a sharp eye on its efficient collection. The dioiketes stood at the center of the royal administration. He was in charge not only of finances, but also of the coordination of the economy and the  Briant 2002, 410 – 421.  V on Reden, ch. 8 C, this volume.  Derow and Bagnall 2004, 285 – 286; Thonemann 2013 for Attalid Asia Minor; Capdetrey 2007 , 227 – 267 for the Seleukid administration. 34 Sitta von Reden correspondence of the empire. The royal mints and the treasury ( to basilikon ) were also under his control. 59 People could submit legal matters, often emerging from tax collection practices but also from other conflicts, in the form of petitions directly to the king. The petitions were decided on in the king ’ s name and sent back for execution to the heads of the nome . Alongside the nome administration, the Ptolemies developed a parallel struc- ture that resulted from their military presence and the cleruchic system. Rather than keeping their soldiers under arms, they gave them land lots ( kleroi ) that provided them with an income for themselves and their families. The allotments varied in size according to military rank and achievement, giving a military social order to the immigrant population of the nome . Military cleruchs belonging to the same garri- son under an eponymous commander ( hegemon or phrouarchos ) formed communi- ties that served many informal social and economic purposes as well. At the top of the military hierarchy in the nome was a strategos, who not only had military func- tions, but supervised the cleruchs , their land, and filled judicial functions, much in the same way as the nomarch . Both must have worked closely hand in hand at first, but in the course of the third century , the strategoi took over the role of the nom- archs , and eventually replaced them. This is quite a remarkable fact, as the strategoi were direct royal appointees, whereas nomarchs had been subordinate to the dioike- tes, who had replaced the former satrap and thus been part of the satrapal structure that the Ptolemies took over. There must have been a military superstructure in the capital as well, but not much detail is known. 60 In the Seleukid Empire, we encounter similar administrative principles and in- stitutions with two important exceptions. Both led to rather different administrative hierarchies and geographies. 61 First, Egypt under Achaemenid rule had been a sin- gle satrapy to which the Ptolemies added just some regions in Cyrene, Cyprus, Syria, and temporarily , Asia Minor. 62 Thus, the royal economy of Egypt was equivalent to that of one satrapy , while in the Seleukid Empire, the king was on top of several satrapies . So in Asia, we find dioiketai cooperating with strategoi at a regional (sa- trapal ) level and acting in conjunction with oikonomoi and phrouarchoi, who in Egypt were subordinates of the dioiketes . 63 There is evidence, nevertheless, for ad- ministrative heads ( hoi epi t ō n pragmat ō n ) and subdivisions of the satrapies ( hyp-  V on Reden 2007 , 79 – 152; Capdetrey 2007 , 306 – 331.  Fischer-Bovet 2014.  See most comprehensively , Capdetrey 2007 ; also Ma 2000, 122 – 139; and Aperghis 2004, 263 – 295 for a survey of the most important administrative documents.  Syria and Phoenicia, Cyprus and Cyrene were closely integrated into the Egyptian administra- tion, while the cities of Asia Minor maintained their own fiscal institutions and coinages, see Bagn- all 1976.  Aperghis 2004, 269 – 274 with examples from Palestine (when under Seleukid control ), Kappado- kia, and the city of Iasos in Asia Minor. The Hellenistic Empires 35 archiai , toparchiai and oikonomiai ). 64 The second major difference was that the Se- leukids tended to settle their soldiers in cities rather than the country side. Higher military personnel and civil beneficiaries who received large plots of land as gift estates attached their holdings to the land that belonged to the cities. 65 This seems to have led to special administrative districts ( hyparchiai or oikonomiai ) attested in the Greek poleis of Asia Minor and the former Lydian capital Sardis. These adminis- trative districts, much like the toparchies in Egypt, seem to have been in charge of the revenue from the cities and their hinterland. In second-century inscriptions from Ai Khanum, too, several oikonomoi are documented as delivering tax money to the treasury , suggesting that though the city was not the capital of the Bactrian king- dom, it was the center of several districts. It is best to conclude that while adminis- trative institutions and some principles of collecting revenue were comparable throughout the Hellenistic world, administrative geographies and hierarchies could vary depending on the established structures of the fiscal unit. In the second centu- ry , local variation increased even further as a result of weakening central authority and the reemergence of local authority structures. 66 VI Settlement and Citie s The Hellenistic period was a time of urban growth and development. 67 Alexander is well known for the large number of cities he founded in the course of his cam- paigns. Many were attributed to him later, and the early successors, most notably Seleukos and Antiochos, followed his practice. Urban foundations not only served to settle and resettle veteran soldiers, civil immigrants, indigenous populations, and personnel in charge of military control and tax collection, but also created an impe- rial landscape marked by settlement and urbanization. Several hundred new cities emerged in the first century of the Hellenistic period, forming part of an urban de- velopment that reached from the Mediterranean coasts through Iran to central Asia and northern India. Some of these cities are still visible, such as Alexandria in Egypt, Samarkand in modern Uzbekistan, and Ai Khanum in Afghanistan. Parts of the material culture of these cities were recognizably Greek, such as gymnasia , theater buildings, and libraries that became part of the representational quarters of royal cities. We also find portrait sculpture, votive offerings, vase paint- ings, and pottery of typically Greek iconography and manufacture. For a long time,  Capdetrey 2007 , 257 – 265 with some alternative nomenclature for satrapal subdivisions.  Aperghis 2004, 99 ‒ 107 (land grants) and 148 – 153 (revenue from cities).  Well shown by Thonemann 2013 for the Attalid kingdom after 188  , and Manning 2003 for Egypt.  Billows 2003; Kosmin 2014, 183 – 221; Mairs and Fischer-Bovet forthcoming; Cohen 2006, 2013 for a survey . 36 Sitta von Reden archaeologists have latched on to the Greek remains, emphasizing the Hellenizing influence of Greek urbanization in Asia and Egypt. Yet with increasing sensitivity to transcultural practices, larger amounts of non-Greek artifacts, as well as objects of mixed styles, are being identified in Hellenistic urban sites, suggesting various types of convergence, interaction, and segregation. Ai Khanum is a particularly in- structive place for analyzing Greek and local social and cultural interaction, includ- ing common use of temples and sanctuaries by worshippers belonging to different religious communities. 68 Alexandria in Egypt, a truly multicultural space where people of various origins met and settled, also offers opportunities for studying growing degrees of cultural interaction, as the number of artifacts and buildings of Egyptianizing styles increased in the course of time. 69 On closer inspection, a very heterogeneous picture of Hellenistic urbanization emerges. First, very few of the sites were actually new foundations. Some big cities, such as the royal residences of the Seleukids, do seem to have been new founda- tions (e.g., Seleukeia- Tigris, Seleukeia-Pieria, and Antiocheia-Orontes). Yet Alexan- dria in Egypt incorporated a local village, Rhakotis, which continued to be remem- bered in the mythographies of the city . 70 Moreover, many cities were refounded to accommodate the cult for a new dynastic founder and to receive a dynastic name. 71 Thus, Susa became Seleukeia-Eulaios, Uruk became Antiocheia-Ishtar, Gaza be- came another Seleukeia, and Jerusalem only just escaped the fate of becoming an- other Antiocheia. 72 Foun dation s also di ffer ed in siz e and org aniz ation. Scholar s have at temp ted to clas sify Hel lenist ic fou ndat ions, dis tingu ishi ng betwe en large prima ry , medium- size d second ary , and small terti ary urb an cente rs, as wel l as betwe en villa ges from larg e to small . 73 Arch aeolo gical s urve ys of se lecte d Helle nist ic sett lemen ts sug gest that t here were in deed cer tain ca tegori es of urban fo unda tions. T he sumpt uous roy al capi tals of S eleuke ia-P ieria , Antioch eia- Oront es, Laod ikeia , Apamei a in no rth Syri a and Pt olemai s Hermio u (Uppe r Egypt ) w ere all in th e range of 205 to 2 80 ha in size, whil e second ary urb an sit es such as D ura-E uropo s on the Eu phrate s, Phi ladelp hia in the Fay um, and Ber enik e Troglod ytike on th e Red Sea, me asure d betwee n 50 and 70 ha. As t he evid ence f or urban areas is gen erall y too sca nty and f raugh t with a r- chae ologic al probl ems, i t would be unwise to speak of a co here nt str ategy of Helle- nist ic city f oundat ions . Yet it i s still p ossi ble to ar gue tha t set tlemen t sizes w ere  Hoo forthcoming, with further literature.  Goddio 1998; Ashton 2004.  von Reden and Strootman forthcoming.  Sherwin-White and Kuhrt 1993, 161. Seleukos I is said to have founded 9 Seleukeias, 16 Anti- ocheias, 5 Laodikeias, 3 Apameias and 1 Stratonikeia, see Kosmin 2014, 183 – 184 and map 3, below . The number of known Ptolemaic foundations include 24 Arsinoes, 19 Ptolemaeis, 13 Berenikes, 5 Philoteras, 5 Kleopatreis, and 4 Euergeteis, see Mueller 2006, 14.  One of the reasons for the Maccabean revolt, see Billows 2003, 198.  Grainger 1990, 91 – 99; Mueller 2006, 89 – 106. The Hellenistic Empires 37 plan ned in re lation t o their f unct ion and l ocatio n, th eir impo rtance in wid er regi onal urba n networ ks, and t heir role in contro lling agrarian la nd, com munic ation l ines , and lo cal popu lation s. 74 Some Hellenistic cities developed from mid-sized urban sites to gigantic cities – largely by cramming more people into the urban space and developing new hinter- lands for the supply of urban demand. There is abundant evidence that the Fayum in Egypt, for example, supplied Alexandria with food and other agricultural prod- ucts. The Seleukid and Ptolemaic capitals developed into urban centers far bigger than any of the biggest towns in the Classical period. Athens, Syracuse, and Miletus once the largest poleis in the classical period had about 100,000 to 125,000 free inhabitants. Antiocheia-Orontes and Seleukeia- Tigris, in contrast, reached popula- tions on the order of 300,000, while the population of Alexandria, the biggest town in the Mediterranean before Rome, numbered an estimated half million inhabitants by the mid-third century  . Many secondary cities and local metropoleis still reached population numbers of 50 – 100,000 people, while Berenike Troglodytike at the Red Sea or Ai Khanum are estimated to have had populations on the order of 20,000 – 30,000. The vast majority of new settlements, however, was in the category of large to small villages numbering about 5,000 inhabitants. 75 There were also different types of political organization. Our evidence unfortu- nately is too limited to construct a typology of Hellenistic cities, but we know that many communities were organized, or organized themselves as poleis with citizen bodies usually drawn from the Greek ethno-class and a set of well-functioning civic institutions: councils, magistracies, and popular assemblies. Yet not all Hellenistic foundations were Greek poleis with citizen bodies and civic institutions. Out of the 49 cities known to have been founded by Alexander and the early Ptolemies in Egypt, just two cities were organized as poleis : Alexandria and Ptolemais Hermiou, the second Ptolemaic administrative center founded by Ptolemy I for the administra- tion of Upper Egypt. The Seleukids founded many more cities, the majority of which are believed to have been organized as poleis . Other cities, such as Jerusalem, grad- ually developed Greek civic structures in the course of the Hellenistic period, a trend that has been termed ‘ poliadization ’ or ‘ polification ’ by modern scholars. 76 The process of poliadization meant not just the development of civic institutions, but also the adoption of a more comprehensive civic model in terms of architecture, socio-political behavior, education, and urbanization. Archaeological research has brought to light the spatial politics of Hellenistic settlement and urbanization. It is obvious that Alexandria and the Seleukid Tetra- polis of capitals on the Mediterranean coasts marked the orientation of Hellenistic  Mueller 2006, 85 – 184 for city sizes and discussion. Alexandria quickly outgrew the size of any of the other capitals, e.g., Monson 2012, 40 ‒ 41.  Kosmin 2014, 189 .  Clancier 2017 (for Babylonia); Thonemann 2013 (for the cities of Asia Minor). 38 Sitta von Reden Map 3: City foundations under Seleukos I and Antiochos I and II (after K osmin 2014, map 8). © Peter P alm. rulers toward the Mediterranean. The choice of Seleukeia- Tigris as Seleukid capital instead of using the major Persian residence in Persepolis was also a sign of the western gravity of Seleukid imperial politics. 77 It was part of a new Seleukid geogra- phy for which the development of the Persian Gulf, its riverine connections into northern Iran and Syria, as well as the agrarian development of the Tigris valley , played a crucial economic role. Yet smaller settlements add to the picture. West of the Euphrates, 87 Seleukid settlements are attested and just 35 in the larger expanse between the Euphrates and Sogdiana. All but a dozen of these eastern foundations, moreover, were located in the lower middle Tigris and Gulf region, areas that were marginal during the Achaemenid period, but became centers of strategic economic development under the early Seleukids. A similar combination of urban, strategic, and economic development can be observed in northern Syria. The development of northern Syria was a direct result of the Seleukid loss of southern Syria to the Ptolemies and the fragmentation of the Levantine coast between the late fourth to the early second century  . The Ptolemies, too, pursued an urbanization policy , if at a smaller scale. 78 One focus of new settlement was the Fayum Oasis in Lower Egypt, just 45 miles away from the former capital of Memphis. Ptolemy II tripled the area of cultivation by draining Lake Moeris (modern Birket Qarun). The Fayum not only became a center  Kosmin 2014, 188 – 195 for this and the following.  Müller 2006; Fischer-Bovet in Mairs and Fischer-Bovet forthcoming. The Hellenistic Empires 39 of Greek settlement, but also a center of agrarian development and innovation. An- other focus of economic development and settlement was the Eastern Desert and the Red Sea coast. On the northwestern stretch of this coast, two important harbor towns, Myos Hormos and Berenike Troglodytike, were founded and linked by cara- van routes to the Nile. Substantial forts with large capacities for grain storage and water supply have been excavated along these routes, demonstrating substantial Ptolemaic interest in this area. 79 At the northern tip of the Red Sea, Ptolemy II also reopened a canal that linked the Bitter Lakes via Heroonpolis (Pithom/Tjeku) with the eastern branch of the Nile delta (see ch. 8 C, map 2, this volume). Where the canal connected with the sea, a city called Arsinoe was founded, and the opening of the canal was celebrated in grand style by the priesthood of Pithom/Tjeku. 80 Above all, the urban and infrastructural development of the Red Sea coast and its linkage with the Nile served as the means to transport a supply of war elephants from Nubia and the East African coast to Alexandria. But in time, it acquired further economic functions and laid the foundation for the thriving trade between Egypt, the Arabian Peninsula, and India. The largest number of new settlements were founded in Egypt, but Ptolemaic foundations spread throughout the provinces of Cyrene, Cyprus, and Asia Minor. Small settlements and garrisoned fortresses suggest that they served as nodal points for the control of local taxation and administration, as well as for the control of networks of communication in areas of imperial and economic interest. 81 VII Language and Institutions Greek language, but also Greek political terminology in the Hellenistic kingdoms, created a sense of cultural homogeneity that was greater in appearance than in practice. As far as language was concerned, written documents increasingly used a common Greek dialect ( koine ) that replaced numerous local dialects of Greek (i.e., Doric, Aeolic, Ionic etc.) as well as local administrative languages. The Hellenistic koine was a slightly modified version of the Attic dialect, which shows the cultural role of Athens in the formation of Hellenistic culture. 82 It is difficult to estimate how far the use of Greek spread down the social scale (many Egyptian priests learnt Greek and fixed religious texts and Egyptian literature in Greek writing), but Greek language offered access to the administration and to Greek legal practice, which was favored by many indigenous inhabitants. 83  Sidebotham and Gates-Foster 2019 .  Mueller 2006, Appendix II; Tuplin 1991 for the previous Achaemenid canal that had silted up.  Mueller 2006, 41 – 84; Kosmin 2014, 183 – 221.  Billows 2003. Continuing Athenian influence can be seen in the Athenian weight standard of most Hellenistic coinages, Athenian legal traditions in Hellenistic law , and the emulation of Athens as educational center by many Hellenistic capitals, especially Alexandria.  Clarysse and Thompson 2009 , 36 – 89 . 46 Sitta von Reden ries of the Romans and their allies in Syria, Asia Minor, and Greece. The small king- dom of the Attalid dynasty , in control of the territory around the city of Pergamon since 283/2  , gained a large territorial expanse stretching across the rich and densely populated river valleys of northwestern Asia Minor, Phrygia, and the south- ern Pontic Coast. 106 Within one generation, Rome had risen to be an aggressive military factor, protector state, and arbitrator within the imperial Hellenistic space. A third war between Rome and Macedonia ended with a crushing defeat of the Macedonians at Pydna (168  ). This led to the fragmentation of the kingdom into four independent but separate regions. Large numbers of slaves from Illyria, Epiros, central Greece, and the Peloponnese were deported to Rome. Ptolemies and Seleu- kids changed the nature of the control over their remaining territories. 107 Relation- ships with local elites and the distribution of revenue were renegotiated, as the famous decree on the Rosetta Stone suggests. 108 In most areas, we see the reemer- gence of independent local social structures and the devolution of tributary and administrative tasks to local elites. 109 Internal secessions ensued, and were success- fully repressed in the case of the Theban revolt in Upper Egypt (204 – 185  ), but not so in the case of the Maccabean revolt in Judaea (167 – 160  ). The years from ca. 160 to 30  are characterized by extended dynastic con- flicts, which weakened the cohesion of the Hellenistic Empires in both their cores and the peripheries. Concomitantly , there was further growth of Roman power, which also mobilized increasing resources and ambition within Rome itself. 110 Fur- ther expansion of the Parthians into Media, Elymais, and the important region of Babylonia threatened the Seleukids from the east. 111 The dynamic expansion of the Pontic king Mithridates VI (120 – 63  ) into Bithynia, Kappadokia, Galatia, Pa- phlagonia, Asia Minor, and the Greek mainland not only reduced the Seleukid sphere of influence in northwestern Asia, but also triggered a final phase of Roman imperial expansion into the Hellenistic world under Sulla, Lucullus, and Pompey . By 129  , Seleukid possessions were reduced to the territory of northern Syria before becoming a Roman province in 63  . In 148  Macedonia had been made the first Roman province in the Hellenis- tic imperial space, followed by the Peloponnese, central Greece, and Thessaly , which became the province of Achaia in 146  . The aggressive expansion, and increasing popularity of Mithridates VI, a rather typical Hellenistic king who, though descending from the Achaemenid royal family , presented himself as a typi-  Thonemann 2013, 2 – 3.  Manning 2003, 43 – 54.  Derow and Bagnall 2004, 165; see also n. 50 above.  Thonemann 2013 for the Attalid kingdom. In Egypt, the most striking evidence for the reemer- gence of local autonomy and power is the Ptolemaic local bronze coinage, which reverted to Egyp- tian standards of weight and value, Faucher and Lorber 2010.  Weaverdyck, ch. 7 , this volume.  Derow 2003 for the following. The Hellenistic Empires 47 cal Greek benefactor and defender of Greek interests against the Roman barbarians, led to the confrontation between Mithridates and three Roman legions under Sulla in Greece. Mithridates ’ s retreat and the Roman advance into Asia during the second and t hird Mit hrida tic W ars (83/ 2  and 7 4 – 6 3  ) in volv ed almo st all par ts of the Hel lenis tic worl d and ca used ma ssi ve human a nd fina ncia l losse s. It eve ntua lly end - ed wi th the r eorga niz atio n of Asi a un der Pom pey in 6 3   . Bith ynia , Po ntos, K iliki a, and S yria bec ame Ro man pr ovinc es, wh ile tre atise s of fr iends hip an d submi ssio n wer e concl uded wi th th e kings o f Arme nia, K appa dokia , Kol chis, a nd Gala tia . The Ptolemies, too, gradually lost their peripheral territories. In 94  , Rome had seized Kyrenaika according to the will of Ptolemy Apion, and Cyprus in 58  . Soon after, however, thanks to her remarkable alliances with Julius Caesar and Marc Antony , Kleopatra VII was able to reestablish the empire, if at the cost of Egyptian independence. After the final defeat of Kleopatra and Antony at Actium in 31  , Egypt became a Roman province in 30  . X Decline and T ransf ormation This is not the place to discuss the reasons for the loss of the Hellenistic Empires to the Romans. Polybios and subsequent historiography put much emphasis on the dynastic and moral decline of the Hellenistic courts. The luxury and debauchery of the kings bewildered Roman armies, magistrates, visitors, and envoys, known for their austerity . Some modern scholars suspect that the Ptolemies turned mad in the course of their repeated incest. 112 Yet, strong rulers such as Ptolemy VIII and Kleopa- tra VII argue against this suspicion. More convincing is the argument that Roman expansion in the Eastern Mediter- ranean was an extraneous shock. 113 The military machine Rome had been able to develop in the course of the third century was simply more powerful than that of the Hellenistic armies, and became increasingly so the more the Romans were able to tap Hellenistic financial resources. 114 Access to manpower, as well as agrarian and monetary revenue that paid for armies, fleets, and campaigns were crucial to imperial power in the Hellenistic world. Some extraordinarily wealthy regions, such as Babylonia, Media, and Egypt (including Cyprus and Cyrene), had been indispen- sable for imperial success. Transportation routes and nodal cities that connected  From the marriage between Ptolemy II and his full sister Arsinoe II, brother-sister marriages were frequent among the Ptolemies, probably in order to act against the problem of unclear lines of succession. This marriage policy , however, was not successful, as the dynastic conflicts sur- rounding the reigns of Ptolemy VIII and Kleopatra II suggest, Hölbl (1993) 2001, 194 – 203, 222 – 256.  Fischer-Bovet 2014.  Weaverdyck, ch. 7 , this volume. The capture of foreign treasuries was indispensable for further conquest, de Callataÿ 1989; Taylor 2014. 48 Sitta von Reden these regions were similarly important. This explains the strategic importance of Syria-Phoenicia and its coastal harbor towns, of the Euphrates and the Tigris as riverine routes, the Aegean islands connecting Northern Greece and the Propontic coast with the Western Asian coast and Egypt, and the Pontos region that linked the Black Sea with Kilikia. Above all, the city leagues of central Greece, the Pelopon- nese, and Macedonia may have functioned as symbolic centers for Hellenistic rulers and their followers who drew identity and cultural orientation from them. The loss of these regions to the Romans and the Arsakids respectively cut the courts off from their lifelines. Finally , there was the problem of imperial expansion turned inward. As was mentioned several times in this chapter, all Hellenistic rulers, including new con- tenders in Asia Minor, Pontos, and Armenia, were spurred by the ideological drive for universal empire. For this goal, alliances were created and negotiated, probably always to the advantage of the highest bidder. Those potential allies who promised to make available most revenue, troops, and transportation routes were won over by means of benefactions, promises of friendship, and protection. In this fiercely competitive sphere, no single Hellenistic ruler was able to develop stable alliances in the long-term (Egypt, Cyrene, and Cyprus perhaps being an exception). Dynastic conflicts and changing marriages added to the instability of the Hellenistic Empires. Eventually , the new powers from the West (Rome) and East (Parthia) seem to have struck the best deals in this ever-changing imperial space. Yet approaching t he imperial tr ansformation of th e Hellenistic Empires in terms of decl ine is a mislea ding perspective i n the context o f world history . Under Hellenistic influenc e, structures develope d that outlived the rule of the Hellenis tic kings. Greek continue d to function as common l anguage in the eastern R oman provinces. Helle nistic elites, althoug h gradually adopting Ro man habits, contin- ued to fill the mos t important administra tive positions und er the Roman Empire. Greek and Greek-s tyle coinages remained the domin ant media of payment and e x- change in Egypt an d in many regions of Asia that lay b eyond the Roman fis cal- military regime b ased on Roman coinage. T he greatest Hellenisti c capitals, Alexan- dria and Antioche ia-Orontes, conti nued to be the uncon tested metropoleis o f the Eastern Mediterr anean. The dev elopment of produc tive economic regi ons such as the Fayum, the Eastern Desert, n orth Syria, and the Tigr is valley provided the fo un- dations for the econom ic prosperity and v ibrant trade of subsequent c enturies. Rul- er cults and political generosi ty ( euergetism ) remain ed important e lements of polit- ical identity- an d community-building i n the Eastern Medit erranean under Roman rule. It is thus with some justif ication that scholars sugge st that the Hellenistic World did not decline cultu rally before the Arabian conques ts of Asia and Egypt in the seventh cen tury  . The Hellenistic Empires 49 Ref erences Ager, S. L. 2003. “ An uneasy balance: From the death of Seleukos to the battle of Raphia. ” In Erskine (2003), 35 – 50. Allsen, T . T . 2011. “ Premodern empires. ” In J. H. Bentley (ed.), The Oxford handbook of world history , 361 – 378. Oxford: O xford University Press. Aperghis, G. G. 2004. The Seleukid royal economy: The finances and financial administration of the Seleukid Empire . C ambridge: Cambridge University Press. Ashton, S.-A. 2004. “ Ptolemaic Alexandria and Egyptian tradition. ” In A. Hirst and M. S. Silk (eds.), Alexandria: Real and imagined , 15 – 40 . Aldershot: Ashgate. Badian, E. 1965. “ The administration of the empire. ” Greece and Rome 12.2, 166 – 182. Bagnall, R. S. 1976. The administration of the Ptolemaic possessions outside E gypt . Leiden: Brill. Barkey, K. 2008. E mpire of difference: The Ottomans in comparative perspective . C ambridge: Cambridge University Press. Bernard, P . 2012. “ Aï Khanum: A Greek colony in post-Alexandrian C entral Asia, or how to be Greek in an Oriental milieu. ” In J. Aruz and E. V altz Fino (eds.), Afghanistan: F orging civilizations along the Silk Road, 42 – 53. New Y ork, NY: Metropolitan Museum of Art. Billows, R. 2003. “ Cities. ” In Erskine (2003), 196 ‒ 215. Bowman, A. K. 1996. Egypt after the Pharaohs, 332 BC – AD 642: From Alexander to the Arab conquest . 2nd ed. Berkeley, CA: University of California Press. Briant, P . 1988. “ Ethno-classe dominante et populations soumises dans l ’ empire achéménide: Le cas de l ’ Égypte. ” In A. Kuhrt and H. Sancisi-W eerdenburg (eds.), Achaemenid history . Vol. 3, 137 – 173. Leiden: Nederlands Instituut voor het Nabije Oosten. – . 2002. From C yrus to Alexander: A history of the Persian Empire . W inona Lake, IN: Eisenbrauns. – . 2017a. Kings, countries, peoples: Selected studies on the Achaemenid E mpire . A. Kuhrt (trans.). Stuttgart: Franz Steiner. – . (1994) 2017b. “ Greco-Hellenistic sources, Persian and Macedonian institutions: C ontinuities, changes and ‘ bricolages. ’” In Briant (2017a) 498 – 525. Originally published as “ Institutions perses et institutions macédoniennes: C ontinuités, changements et bricolages. ” – . (2003) 2017c. “ When kings write history: The Persian period in Ptolemaic inscriptions. ” In Briant (2017a), 247 – 258. Originally published as “ Quand les rois écrivent l ’ histoire: La domination achéménide vue à travers les inscriptions officielles lagides. ” Bringmann, K., and H. von Steuben, eds. 1995. Schenkungen hellenistischer Herrscher an griechische Städte und Heiligtümer . V ol. 1, Zeugnisse und Kommentare . Berlin: Akademie- V erlag. Bugh, G. R., ed. 2006. The Cambridge companion to the Hellenistic world . C ambridge: C ambridge University Press. Capdetrey, L. 2007 . Le pouvoir séleucide: T erritoire, administration, finances d ’ un royaume hellénistique, 312 – 129 avant J.-C . Rennes: Presses universitaires de R ennes. Chaniotis, A. 2003. “ The divinity of Hellenistic rulers. ” In Erskine (2003), 431 – 445. – . 2007 . “ Mythos und Religion. ” In W eber (2007), 139 – 157 . Clancier, P ., and G. Gorre. Forthcoming. “ The integration of indigenous elites and the development of poleis in the Ptolemaic and Seleucid Empires. ” In C. Fischer-Bovet and S. von Reden (eds.), Comparing the Ptolemaic and S eleukid Empires . C ambridge: C ambridge University Press. Clancier, P . 2017 . “ The polis of Babylon: A historical approach. ” In B. Chrubasik and D . King (eds.), Hellenism and the local communities of the Eastern Mediterranean: 400  – 205  , 53 – 82. Oxford: O xford University Press. Clarysse, W ., and D . J. Thompson. 2009. Counting the people in Hellenistic Egypt . 2nd ed. V ol. 2, Historical studies . Cambridge: C ambridge University Press. 50 Sitta von Reden C ohen, G. M. 2006. The Hellenistic settlements in S yria, the Red Sea Basin, and North A frica . Berkeley, CA: University of California Press. – . 2013. The Hellenistic settlements in the East from Armenia and Mesopotamia to Bactria and India . Berkeley, CA: University of California Press. Davies, J. 2013. “ C orridors, cleruchies, commodities and coins: The pre-history of the Athenian Empire. ” In A. Slawisch (ed.), Handels- und Finanzgebaren in der Ägäis im 5. Jh. v. C hr. / T rade and finance in the 5th c. BC Aegean world , 43 – 66. Istanbul and Vienna: E ge Y ay ı nlar ı and Phoibos. de Callataÿ, F . 1989. “ Les trésors achéménides et les monnayages d ’ A lexandre: Espèces immobilisées et espèces circulantes? ” Revue des Études Anciennes 91, 259 – 276. – . 2005. “ L ’ instauration par Ptolémée Ier Sôter d ’ une économie monétaire fermée. ” In F . Duyrat and O . Picard (eds.), L ’ exception égyptienne? Production et échanges monétaires en Égypte hellénistique et romaine: Actes du colloque d ’ Alexandrie, 13 – 15 avril 2002 , 117 – 134. C airo: Institut Français d ’ Archéologie Orientale. – . 2011. “ Quantifying monetary production in Greco-Roman times: A general frame. ” In F . de Callataÿ (ed.), Quantifying monetary supplies in Greco-Roman times , 7 – 29. B ari: Edipuglia. Derow, P . 2003. “ The arrival of Rome: From the Illyrian W ars to the Fall of Macedon. ” In Erskine (2003), 51 – 70. Derow, P ., and R. S. Bagnall. 2004. The Hellenistic period: Historical sources in translation . 2nd ed. Malden, MA: Blackwell. Erskine, A., ed. 2003. A companion to the Hellenistic world . Malden, MA: Blackwell. Faucher, T ., and C. Lorber. 2010. “ Bronze coinage of Ptolemaic Egypt in the second century BC. ” American Journal of Numismatics 22, 35 – 80. Feyel, C., and L. Graslin- Thomé, eds. 201 4. L e projet politique d ’ Antiochos IV : Journées d ’ études franco-allemandes, Nancy 17 – 19 juin 2013 . Nancy and Paris: ADRA and De Boccard. Fischer-Bovet, C. 2014. Army and society in Ptolemaic Egypt . C ambridge: Cambridge University Press. – . Forthcoming. “ The Ptolemaic and Seleucid Empires. ” In P . F . Bang, C. A. Bayly, and W . Scheidel (eds.), The Oxford world history of empire , XX- XX , Oxford: Oxford University Press. Gehrke, H.-J. 2008. Geschichte des Hellenismus . 4th ed. Munich: O ldenbourg. Goddio, F . 1998. Alexandria: The submerged royal quarters . London: P eriplus. Grainger, J. D . 1990. The cities of S eleukid S yria . Oxford: C larendon. Hauben, H., and A. Meeus, eds. 2014. The Age of the Successors and the creation of the Hellenistic kingdoms (323 – 276 BC). Leuven: Peeters. Hölbl, G. (1993) 2001. A history of the Ptolemaic Empire . London: R outledge. Originally published as Geschichte des Ptolemäerreiches . Hoo, M. Forthcoming. E urasian localisms: T owards a translocal approach to Hellenism and inbetweenness in central Eurasia, c. 300  ‒ 1  . Stuttgart: Franz Steiner. Howgego, C. 1995. Ancient history from coins . London: Routledge. Jacobs, B. 1994. Die Satrapienverwaltung im Perserreich zur Zeit Darius ’ III . Wiesbaden: R eichert. – . 2011. “ Achaemenid satrapies. ” Encyclopaedia Iranica, online edition. http:/ / www.iranicaonline.org/articles/achaemenid-satrapies. Jones, C. P . 2010. New heroes in antiquity: From Achilles to Antinoos . C ambridge, MA: Harvard University Press. Koenen, L. 1993. “ The Ptolemaic king as a religious figure. ” In A. Bulloch, E. S. Gruen, A . A. Long, and A. Stewart (eds.), Images and ideologies: Self-definition in the Hellenistic world , 25 – 115. Berkeley, CA: University of California Press. Kosmin, P . 2014. The land of the elephant kings: Space, territory, and ideology in the Seleucid Empire . C ambridge, MA: Harvard University Press. Kuhrt, A., and S. M. Sherwin-White. 1991. “ Aspects of Seleucid royal ideology: The cylinder of Antiochus I from Borsippa. ” Journal of Hellenic Studies 111, 71 – 86. The Hellenistic Empires 5 1 Kulke, H., and D . Rothermund. 2018. Geschichte Indiens: V on der Induskultur bis heute . 3rd ed. Munich: C. H. Beck. Lanciers, E. 1993. “ Die Opfer im hellenistischen Herrscherkult und ihre Rezeption bei der einheimischen Bevölkerung der hellenistischen Reiche. ” In J. Quaegebeur (ed.), Ritual and sacrifice in the ancient Near East: Proceedings of the international conference organized by the Katholieke Universiteit L euven from the 17th to the 20th of April 1991 , 203 – 223. Leuven: Peeters. – . 2018. “ The emergence of Ptolemaic honorific court titles. ” Ancient Society 48, 49 – 82. Le Rider, G. 1986. “ Les A lexandres d ’ argent en Asie Mineure et dans l ’ Orient séleucide au IIIe siècle av. J.-C. (c. 275 – c. 225): Remarques sur le système monétaire des Seleucides et des Ptolémées. ” Journal des Savants 1/3, 3 – 57 . Le Rider, G., and F . de Callataÿ. 2006. L es Séleucides et les Ptolémées: L ’ héritage monétaire et financier d ’ Alexandre le Grand . Monaco: Éditions du R ocher. Lewis, N. (1986) 2001. Greeks in Ptolemaic Egypt: Case studies in the social history of the Hellenistic world . Oakville, CT : American Society of Papyrologists. Lloyd, A. B. 2000. “ The Late Period (664 – 332 BC). ” In I. Shaw (ed.), The Oxford history of ancient Egypt , 364 – 387 . Oxford: O xford University Press. Lorber, C. 2012. “ The coinage of the Ptolemies. ” In W . E. Metcalf (ed.), The Oxford handbook of Greek and Roman coinage , 211 – 234. Oxford: O xford University Press. Lowry, S. T . 1987 . The archaeology of economic ideas: The classical Greek tradition . Durham: Duke University Press. Ma, J. 2000. Antiochos III and the cities of western Asia M inor . Oxford: O xford University Press. – . 2011. “ C ourt, king and power in Antigonid Macedonia. ” In R. J. Lane F ox (ed.), Brill ’ s companion to ancient Macedon: Studies in the archaeology and history of M acedon, 650 BC – 300 AD , 521 – 543. Leiden: Brill. – . 2013. “ Hellenistic Empires. ” In P . F . Bang and W . Scheidel (eds.), The Oxford handbook of the state in the ancient Near East and Mediterranean , 324 – 357 . Oxford: Oxford University Press. Mairs, R. 2014. The Hellenistic Far East: Archaeology, language, and identity in Greek Central Asia . Berkeley, CA: University of California Press. Mairs, R., and C. Fischer-Bovet. Forthcoming. “ R eassessing Hellenistic settlement policies: The Seleukid Far East and Ptolemaic Red Sea B asin, Egypt and Nubia. ” In C. Fischer-Bovet and S. von Reden (eds.), Comparing the Ptolemaic and S eleucid Empires . C ambridge: Cambridge University Press. Manning, J. G. 2003. Land and power in Ptolemaic E gypt: The structure of land tenure 332 – 30  . Cambridge: C ambridge University Press. – . 2010. The last Pharaohs: E gypt under the Ptolemies, 305 – 30 BC . Princeton, NJ: Princeton University Press. Meadows, A. 2012. “ P tolemaic possessions outside Egypt. ” In R. S. Bagnall, K. Brodersen, C. B. Champion, A. Erskine, and S. R. Huebner (eds.), The encyclopedia of ancient history . V ol. 10, 5625 – 5629. Malden, MA: Wiley-Blackwell. Mitchell, L. 2013. The heroic rulers of archaic and classical Greece . London: Bloomsbury. Mittag, P . F . 2006. Antiochos IV . Epiphanes: Eine politische Biographie . Berlin: Akademie- V erlag. Monson, A. 2015. “ Hellenistic empires. ” In A. Monson and W . Scheidel (eds.), Fiscal regimes and the political economy of premodern states , 169 – 207 . Cambridge: C ambridge University Press. Mooren, L. 1977 . La hiérarchie de cour ptolémaïque: Contribution à l ’ étude des institutions et des classes dirigeantes à l ’ époque hellénistique . Leuven: Katholieke Universiteit L euven. Moyer, I. S. 2011. Egypt and the limits of Hellenism . C ambridge: C ambridge University Press. Mueller, K. 2006. Settlements of the Ptolemies: C ity foundations and new settlement in the Hellenistic world . Leuven: Peeters. Ogden, D . 1999. Polygamy, prostitutes and death: The Hellenistic dynasties . L ondon: Duckworth. 52 Sitta von Reden Pfeiffer, S. 2008. Herrscher- und Dynastiekulte im Ptolemäerreich: S ystematik und Einordnung der Kultformen . Munich: C. H. Beck. Pfeiffer, S., and H. Klinkott. Forthcoming. “ Legitimizing the foreign king in the Ptolemaic and Seleucid Empires: The role of local elites and priests. ” In C. Fischer-Bovet and S. von Reden (eds.), Comparing the Ptolemaic and Seleucid E mpires . C ambridge: Cambridge University Press. Rice, E. 1983. The grand procession of Ptolemy Philadelphus . O xford: Oxford University Press. Sherwin-White, S., and A. Kuhrt. 1993. From Samarkhand to Sardis: A new approach to the Seleucid Empire . Berkeley, CA: University of C alifornia Press. Sidebotham, S. E., and J. Gates-Foster. 2019. The archaeological survey of the desert roads between Berenike and the Nile Valley: Expeditions by the U niversity of Michigan and the University of Delaware to the Eastern Desert of E gypt, 1987 – 2015 . Boston, MA: American Schools of Oriental Research. Stephens, S. A. 2003. Seeing double: Intercultural poetics in Ptolemaic Alexandria . Berkeley, CA: University of California Press. Strootman, R. 2014a. “ Hellenistic imperialism and the ideal of world unity. ” In C. Rapp and H. A. Drake (eds.), The city in the classical and post-classical world: Changing contexts of power and identity , 38 ‒ 61. Cambridge: C ambridge University Press. – . 2014b. “‘ Men to whose rapacity neither sea nor mountain sets a limit ’ : The aims of the Diadochs. ” In H. Hauben and A. Meeus (eds.), The Age of the Successors and the creation of the Hellenistic kingdoms (323 – 276 BC) , 307 ‒ 322. Leuven: Peeters. Strootman, R., and M. J. V ersluys, eds. 2017 . Persianism in antiquity . Stuttgart: Franz Steiner. T aylor, M. J. 2014. “ Sacred plunder and the Seleucid Near East. ” Greece and Rome 61, 222 – 241. Thompson, D . J. 1984. “ Agriculture. ” In F . W . Walbank, A. E. Astin, M. W . Frederiksen, and R. M. Ogilvie (eds.), The Cambridge ancient history. V ol. 7 .1, 363 ‒ 370. C ambridge: Cambridge University Press. Thonemann, P . 2013. “ The Attalid state, 188 – 133 BC. ” In P . Thonemann (ed.), Attalid Asia M inor: Money, international relations, and the state , 1 ‒ 47 . Oxford: O xford University Press. – . 2015. The Hellenistic world: Using coins as sources . C ambridge: C ambridge University Press. T uplin, C. 1991. “ Darius ’ Suez Canal and P ersian imperialism. ” In A. Kuhrt and H. Sancisi- Weerdenburg (eds.), Achaemenid history. V ol. 6, 237 – 283. Leiden: Nederlands Instituut voor het Nabije Oosten. – . 2009. “ The Seleucids and their Achaemenid predecessors: A Persian inheritance? ” In S. M. R. Darbandi and A. Zournatzi (eds.), Ancient Greece and ancient Iran: Cross-cultural encounters: 1st international conference, Athens, 11 – 13 November 2006 , 109 – 136. Athens: National Hellenic Research F oundation. von Reden, S. 2007 . Money in Ptolemaic E gypt: From the Macedonian conquest to the end of the third century BC . Cambridge: C ambridge University Press. – . 2019. “ Monetization of the countryside. ” In K. V andorpe (ed.), A companion to Greco-Roman and Late Antique Egypt , 217 ‒ 232. M alden, MA: Wiley-Blackwell. von Reden, S., and R. Strootman. F orthcoming. “ Imperial metropoleis and foundation myths: Ptolemaic and Seleucid capitals compared. ” In C. Fischer-Bovet and S. von Reden (eds.), Comparing the Ptolemaic and Seleucid E mpires . C ambridge: Cambridge University Press. Walbank, F . W . 1981. The Hellenistic world . London: Fontana. Weber, G., ed. 2007 . Kulturgeschichte des Hellenismus: Von A lexander dem Großen bis Kleopatra . Stuttgart: Klett-C otta. Lauren Morris 2 C entra l Asian Empires I Introduction: C entral Asia as an Imperia l Spac e I.  From Imperial F rontier to C enter: 300  – 300  During the mid to late first millennium  , Central Asia underwent enormous polit- ical and cultural change. Through the military campaigns of the Achaemenid kings Cyrus II and Darius I, Central Asia came to form the empire ’ s northeastern frontier by the late sixth century  . 1 As with the rest of the Achaemenid Empire, this space was subdivided into dahyava (countries) – preexisting cultural units refigured as satrapal administrative units – from which taxes, military personnel, and laborers could be levied. Bactria was an especially resource-rich and significant country , whose satraps were often drawn from the Achaemenid royal family , but it was also unruly and prone to revolt. 2 Here, the Achaemenid administration further developed the irrigation systems that had existed since the Bronze Age, 3 and introduced the use of Aramaic as a chancellery language; the use of Elamite has been additionally attested in Arachosia. 4 Between 329 and 327  , Alexander the Great campaigned in Central Asia for two difficult years in the pursuit of Bessos (the satrap of Bactria who had murdered the king Darius III), all the while adapting himself to Persian modes of rule and largely preserving administrative structures he found in place. 5 One important cul- tural ramification of Alexander ’ s conquests was the introduction of demobilized Greek and Macedonian soldiers into the population through the foundation of new settlements with a strong military component. 6 Garrisons were established within existing cities or as new outposts, existing cities were refounded, and new founda- tions were also established, such as Alexandria Eschate on the Jaxartes (mod. Syr  On Achaemenid Central Asia, see generally Briant 1984. Pre-Achaemenid forms of state control or development in this space are unclear, especially since Sancisi-Weerdenburg 1988 cast doubt on the existence of the Median ‘ Empire. ’  Wu 2017 ; Henkelman 2018.  Gardin 1998, 185; Mairs 2014, 38 – 39 .  See discussion and bibliography in Morris, ch. 9 , this volume.  On Alexander in Bactria, see Holt 1988.  On prior deportations of Greek communities to Bactria during the Achaemenid period, see Marti- nez-Sève 2012, 372. On the foundations and settlement policies of Alexander in Central Asia, consult the discussions in Fraser 1996; Fischer-Bovet and Mairs forthcoming. Note: Thanks are due to Olivier Bordeaux, Omar Coloru, and Gunnar Dumke for their comments on earlier versions of this chapter. Open Access. © 2020 Lauren Morris, published by De Gruyter. This work is licensed under the Creative C ommons Attribution-NonCommercial-NoDerivatives 4.0 License. https:/ /doi.org/10.15 15/97831106077 41-004 54 Lauren Morris Map 1: C entral Asia and northern India at the time of the Greek Kingdoms. © Peter P alm. C entral Asian Empires 55 Darya) River. 7 Although two settler revolts followed in 325 and 323  , 8 Graeco- Macedonian occupation likely continued to some degree. Following Alexander ’ s death in 323  and the first wars of his successors, Seleukos I led a campaign in ca. 305  into Central Asia. An unsuccessful military venture across the Hindu Kush against the Mauryans resulted in a treaty: Arachosia, the Paropamisadai, Gandh ā ra, and northwest India were ceded to Chandragupta Maurya along with a Seleukid princess, in exchange for five hundred war ele- phants. 9 Seleukos appointed his son Antiochos I as co-regent in ca. 294  with a particular concern for Bactria, due in part to the region ’ s strategic significance but also Antiochos ’ s own family ties; his mother, Apama, was the daughter of the Sog- dian noble Spitamenes, who had been a prominent antagonist during Alexander ’ s campaigns. Seleukid rule in Central Asia further culturally and economically inte- grated this space into the wider Hellenistic world. In Bactria and Sogdiana, the Greek language came into wider use, garrisons and larger settlements were founded and refounded (such as Antiocheia in Margiana), new settlers were introduced, and monetization began through the establishment of royal mints striking gold, silver, and copper coinage to the Attic standard used throughout the Hellenistic East. 10 Then, in the mid-third century  , a new geopolitical phenomenon occurred, which shaped the following centuries in this space. Bactria was transformed from an eastern imperial frontier to become the locus of two new successive Central Asian ‘ empires ’ : that of the Greek Kingdoms (map 1), and the Kushan Empire (map 2). 11 The rule of the Greek Kingdoms (ca. 250  – 10  ) is usually regarded to begin with the secession of Diodotos I, the former Seleukid satrap of Bactria. He and sub- sequent kings ruling in Bactria are conventionally referred to as ‘ Graeco-Bactrian. ’ After the southern conquests of Demetrios I (ca. 200 – 180  ), the subsequent Greek kings thought to have ruled in Arachosia, the Paropamisadai, Gandh ā ra, and the Punjab are now usually referred to as ‘ Indo-Greek. ’ 12 So much about these kingdoms remains unclear. Strabo, drawing on the now-lost P arthika of Apollodoros of Artemi- ta (ca. 50  ), provided the most informative account surviving today . He described an independent Greek polity which had emerged from the agricultural prosperity of Bactria to become a far-reaching and multiethnic political entity: The Greeks who caused Bactria to revolt grew so powerful on account of the fertility of the country that they became masters, not only of Ariana, but also of India, as Apollodorus of  The site of which remains archaeologically unconfirmed, see Holt 1988, 54 – 59 .  See further in Holt 1988, 70 – 87 .  For this episode, see Karttunen 1997 , 260 – 263.  On Seleucid Central Asia, see Holt 1999 , 21 – 29; Capdetrey 2007 , 79 – 81; Martinez-Sève 2012, 375 – 376. For more on the Seleukid Empire, see von Reden, ch. 1, this volume.  There are many orthographic variants for ‘ Kushan ’ in scholarship, and preferences differ ac- cording to disciplinary , historical, and methodological positionality . ‘ Kushan ’ is in widest use, but ‘ Ku ṣ ā ṇ a ’ or ‘ Ku šā n ’ are also commonly seen.  For more on these terms, see below . 62 Lauren Morris and sizes, while nominal sovereignty could be asserted over peoples in marginal areas (i.e., desert, steppe, and mountains) within the boundaries of the state. 51 Envi- ronment and climate may also shape power structures in other ways, for example in the development of seasonal capitals between which the court would travel. His- torically , Kapisa and Kabul (located in the ancient Paropamisadai) have served as the location of a summer highland capital for multiple powers who also had winter lowland bases to the east at Jalalabad, Peshawar, or Taxila. 52 I.  Environment, C limate, and Natural R esourc es The environment of Central Asia is very diverse, encompassing oasis, desert, moun- tain, and steppe landscapes. Its climatic and ecological zones are accordingly var- ied, but the lowlands in the most part have a semiarid steppe climate with low rainfall occurring mostly in winter and spring. Although the region ’ s climate ap- pears to have been relatively stable for the last five thousand years, climate varia- tions have been documented through proxy records, including a humid period be- tween ca. 0 – 410  . 53 Sedentary settlement was clustered around piedmonts and river plains composed of fertile loess, and sustained by runoff from the high moun- tains of the region. These river plains supported denser flora, particularly tugai- dense riparian forests thriving in arid regions in addition to agricultural cultivation, which flourished since at least the Bronze Age through the installation of artificial irrigation systems. These systems continued to be developed and extended through- out the Hellenistic and Kushan periods. 54 On these irrigated plains, or where dry farming was possible, the main food resources cultivated appear to have been grains such as wheat, barley , and millet, i.e., both summer and winter crops. The agricultural output of Bactria was famously outstanding, a point emphasized in multiple Greek and Latin texts. 55 Meat and secondary animal products could be obtained through hunting and pastoralism practiced along a spectrum of mobility . 56 Comparably , mountainous regions, such as those of the Paropamisadai and Aracho- sia encompassed a great diversity of ecotopes created by rapid altitude changes. Lowland valleys, like Laghman and Jalalabad at less than 1,000 m above sea level, are relatively warm year-round and in modern times produce wet rice and citrus crops. However, the highlands (mod. Kabul at 1,800 m, Ghazni at 2,200 m, and Gar-  Barfield 2010, 67 – 68.  Barfield 2010, 53.  See Gentelle 1989 , 61 – 79; Yang et al. 2009 .  The available data is uneven. A synthesis of data from the Chalcolithic to the Iron Age is to be found in Francfort and Lecomte 2002. The ample data for east Bactria are presented in Gentelle 1989 , 81 – 106.  See further in Morris, ch. 9 , this volume.  See, for example, the analysis of faunal remains at the Achaemenid period site of Kyzyltepa in north Bactria, indicating that the local economy included a market-oriented pastoral component, in Wu et al. 2015. C entral Asian Empires 63 dez at 2,300 m) have cold winters which have traditionally supported the cultivation of wheat, barley , grapes, and fruit and nut trees. In a parallel manner to the ‘ vertical archipeligo ’ model of the Andes, the close proximity of such diverse ecotopes was likely attractive to the states wishing to exploit them. 57 Central Asia is host to rich metal deposits including silver, gold, tin, copper, and lead, and to precious and semiprecious stones, including lapis lazuli and gar- net. Although direct evidence for the location of mines and their operation during antiquity is lacking, 58 they were likely economically significant and possibly con- trolled by imperial monopoly . Somewhat better documented is the case of lapis lazu- li, which was found in Badakhshan and had been exploited and distributed through wide networks of exchange since the third millennium  . The most important sources were the Kokcha valley and Sar-i Sang, although archaeological research has not been undertaken at the ancient mining works. 59 However, at Ai Khanum, a Graeco-Bactrian city which lay downstream at the Kokcha ’ s junction with the Panj, a bag-shaped deposit of roughly 75 kg of unworked lapis was found abandoned in one of the rooms of the treasury , 60 suggesting some kind of state involvement with the stone ’ s exploitation. No archaeometallurgical research has yet proven the precise sources of gold, silver, and copper used in the production of Graeco-Bactrian, Indo-Greek and Ku- shan coinage. However, gold deposits are located in Badakhshan and the upper Indus. Potential sources of silver include the galena deposits in Panjshir valley and Farenjal in the Ghorband valley , both in the central Hindu Kush. The major copper deposit of Mes Aynak in the modern Logar province was certainly exploited from the late Kushan period, although earlier exploitation is also expected. The Zar- kashan-Anguri formation, located in Arachosia, may have been a major source of copper, iron, gold, and tin. 61 II The Greek Kingdoms of C entral Asia The Greek Kingdoms of Central Asia existed from ca. 250  until 10  (map 1), emerging on the frontier of Seleukid rule in Bactria, with the last kings ruling in the eastern Punjab. These kings – basileis like their Hellenistic counterparts – were perhaps 45 in number, known mostly from coins. Only eight are mentioned in trans- mitted texts. 62 It has become conventional to describe these kings as either ‘ Graeco-  Barfield 2010, 52.  A project dedicated to ancient mining and resource use in Afghanistan has been established by the Deutsches Archäologisches Institut , see Thomalsky et al. 2013.  Bernard 1978.  Rapin 1992, 50.  For further information on all deposits, see Thomalsky et al. 2013.  Bordeaux 2018, 29 , n. 28. 64 Lauren Morris Bactrian ’ or ‘ Indo-Greek, ’ designations which are partly geographical, chronologi- cal, and numismatic. The convention is most justified from a numismatic perspec- tive. Graeco-Bactrian coinage developed out of Seleukid coinage and was struck to the Attic weight standard current in the Hellenistic east ( tetradrachm of 16.80 g), while Indo-Greek coinage was struck to a lower standard ( tetradrachm of 9 .80 g). 63 The design of Indo-Greek coinage also differs. The names and titles of kings are usually given in G ā ndh ā r ī on the reverse of coins, sometimes Indic deities or sym- bols are depicted, and copper coins were often struck from square flans. The precise implications of these numismatic changes for the political and cultural orientation of the Greek Kingdoms remains an open question. In fact, some later Indo-Greek kings who certainly did not politically control Bactria also minted coinage of a Grae- co-Bactrian style and weight; explanatory hypotheses suggest that these coins may have been intended for commercial exchange with Bactria, or as used as tribute payments for antagonistic neighbors. 64 Indeed, from the available evidence thus far, it appears that the Greek King- doms were modelled foremost on the political and administrative structures current in other Hellenistic monarchies, especially that of their Seleukid predecessors. 65 It is accordingly important to be aware of the possibility of political complexity that is not explicitly visible in our limited sources. Although the reconstruction of the sequence and territorial association of these kings seems exceedingly fragmented and disconnected, the possibility of coregencies and even alliances between kings must be taken into account, as well as inter- and intra-dynastic conflict (fig. 1). This is made plain by the first lines of the Asangorna parchment, an unprovenanced tax receipt from Bactria. This features a dating formula according to multiple kings: “ In the reign of God Antimachos and Eumenes and Antimachos … year 4, month of Olöus, in Asangorna. ” 66 This date appears to be given in the Greek era discussed above, possibly inaugurated to commemorate an alliance made between Antima- chos I and Apollodotos I, meaning this document would have been drawn up in 171  . 67 Indeed, recent analysis with infrared photography has proposed to recon- struct the lacuna to read “ [and] Apollodotus ” although this reading has yet to be adequately debated and accepted in scholarship. 68 Numismatic evidence indicates that Antimachos I ruled in Bactria, while Antimachos II ruled in the Indo-Greek realms (as did Apollodotos I). Eumenes is otherwise unknown. Essentially , while the terms Graeco-Bactrian and Indo-Greek are retained here for lucidity , they are obviously problematic vis-à-vis their ethno-cultural baggage, as well as for obscur- ing the political and administrative organization of the period.  Bopearachchi 1991, 13.  Bopearachchi 1990.  For which see von Reden, ch. 1, this volume.  Rea, Senior, and Hollis 1994; see also Bernard and Rapin 1994.  Rapin 2010.  Jakobsson and Glenn 2018. C entral Asian Empires 65 Fig. 1: Kings and putative regional extent of the Greek Kingdoms (after Bopearachchi 1991, tab. 5; Coloru 2009, tab. 2; Bordeaux 2018). 66 Lauren Morris II.  Politic al History In the mid-third century  , Diodotos I, the Seleukid satrap of Bactria, reportedly revolted and established an independent kingdom. 69 The precise date, between ca. 250 – 235  , and nature of the Bactrian secession remains a subject of dispute. 70 Diodotos I was succeeded by a homonymous son, who was reported by Justin to have made an alliance with Arsakes, the founder of the Arsakid dynasty . This alli- ance apparently ended conflict between Parthia and Bactria and helped the other former Seleukid satrapy to likewise revolt. 71 The Diodotids appear to have been overthrown by Euthydemos I (ca. 230 – 190  ), who Polybios described as being a native of Magnesia. 72 The Hellenistic historian provided further details about a certain historical episode of conflict be- tween the Seleukids and Graeco-Bactrians. In his grand eastern campaigns, the Se- leukid king Antiochos III defeated Euthydemos in a battle near the Arius River (mod. Harirud ) in 208  , and Euthydemos retreated to his capital at Bactra, where he was besieged for two years. 73 During the siege, Euthydemos justified his actions to a Seleukid messenger as action taken against the original rebels, underlining his faculty as the last line of defense against approaching hordes of nomads. The siege concluded with a peace agreement, ratified in person by Euthydemos ’ s son Deme- trios I. This recognized Euthydemos ’ s legitimacy , giving also a daughter of Antio- chos III to Demetrios in marriage, while Antiochos III received Euthydemos ’ s ele- phants and grain rations for his troops. 74 Demetrios (ca. 190 – 180  ) enacted the first major southward expansion of Graeco-Bactrian territory . Likely in imitation of Alexander the conqueror, he was represented on his coins as a bust wearing the scalp of an elephant. 75 His conquests included at least part of Arachosia, and possi- bly coincided with the foundation of a new city , Demetrias. 76  Just. Epit. 41. 4. 3 – 5.  Holt ’ s interpretation based on Diodotid coinage is most generally accepted. Holt concluded the secession was a gradual process beginning around 250  , with Diodotos I first minting coins with his own portrait in Antiochos II ’ s name, followed by complete independence in the reign of Diodo- tus II (Holt 1999 , 87 – 125). Jakobsson proposed alternatively that the coins minted in the name of Antiochos were issued by a hitherto unknown Graeco-Bactrian king named Antiochos, a son of Diodotos I, who ruled after Diodotos II. This would imply Diodotos I ’ s secession was sudden, ruling between ca. 255 – 250  (Jakobsson 2010).  Just. Epit. 41.4. See also the discussion in Fabian, ch. 6, this volume.  Polybios (Polyb.) 11. 34. On Antiochos III ’ s eastern campaign, see von Reden, ch. 1, this volume.  Polyb. 10. 49 .  Polyb. 11. 34. 8 – 10.  See von Reden, ch. 1, fig. 1, this volume. The elephant scalp iconography was widely adopted by other kings.  Mentioned only in the Parthian-era Greek itinerary text Stathmoi Parthikoi 19 , of Isidoros of Charax. C entral Asian Empires 67 Demetrios was succeeded by Euthydemos II (ca. 180 – 177  ). 77 Euthydemos II ’ s coinage has many similarities with certain issues of two contemporary Indo-Greek kings, Pantaleon (ca. 180 – 175  ) and Agathokles (ca. 180 – 174  ), who are thought to have ruled in Arachosia and the Paropamisadai. Agathokles also minted Graeco-Bactrian weight and style coins. Some political arrangement between the three kings has been inferred. 78 Pantaleon and Agathokles perhaps expanded their control into Gandh ā ra, issuing the first Indo-Greek rectangular copper coins, with bilingual legends and depictions of Brahmanical deities rather than royal por- traits. 79 Euthydemus II ’ s reign seems to have been brought to a premature close by Anti- machos I Theos (ca. 177 – 171  ). Antimachos and Agathokles minted special series of commemorative or ‘ pedigree ’ coins in silver, depicting portraits of prior rulers. This has been interpreted to indicate their political rivalry . 80 As discussed above, it is possible that an alliance was made between Antimachos I and the Indo-Greek king Apollodotus I (ca. 174 – 160  ). An apparent co-regency between Antimachos I and (his sons?) Eumenes and Antimachos II is documented in 171  . During this time, a certian Demetrios II (ca. 174 – 171  ) may have ruled in east Bactria. 81 Then, an usurper, Eukratides I (ca. 171 – 145  ), wrested power in Bactria. The Indo-Greek king Menander I (ca. 165 – 130  ) perhaps emerged shortly after in the Paropamisadai, and it is possible that during this early period, Antimachos II ruled in Arachosia subordinately to him. 82 Eukratides ’ s early reign appears to have been marked by violent conflict, as Justin mentioned his wars with Sogdians, Drangians, and Indians. 83 Yet, his campaign into the Indic borderlands, probably in the 160s  , seems to have been successful. Its material outcome appears to be visible in the finds at the treasury of Ai Khanum, with luxury objects representing the capture of booty or even extraction of tribute. 84 A Hellenistic phase at Marakanda-Afrasiab in Sogdiana is thought to be associated with a brief reconquest by Eukratides, al- though the evidence is not conclusive. 85 The immense prosperity of Eukratides ’ s reign is indicated by his prolific coinage and the enactment of a new building pro- gram at Ai Khanum, but it was short lived. Justin reported that, on the king ’ s return from India, he was killed by his son, who drove his chariot through his father ’ s blood and ordered the body to be cast out unburied. 86  Either the younger son of Euthydemus I (Narain 1957 , 23; Bopearachchi 1991, 55) or the elder son of Demetrios I (Coloru 2009 , 195).  Bopearachchi 1991, 56.  Bopearachchi 1991, 57 ; Coloru 2009 , 203 – 206.  Bopearachchi 1991, 61; Coloru 2009 , 200 – 202.  Coloru 2009 , 208.  Bordeaux 2018, 140 – 141.  Just. Epit . 41. 6.  Rapin 1992, 281 – 287 .  Lyonnet 2012, 167 .  Just. Epit . 41. 6. 68 Lauren Morris As a result of Eukratides ’ s southeast expansion, conflict with Menander seems to have broken out. Menander probably also enacted an eastward program of con- quest, as well as eventually reconquering the territory lost to Eukratides. Later Bud- dhist tradition places the capital of Menander at S ā gala (probably mod. Sialkot ) in the Punjab. 87 Of all the Indo-Greeks, it is supposed that Menander ’ s conquests were most extensive, if their precise limits remain uncertain. Later Indic literary tradition, foremost the Y ugapur ā ṇ a , seems to reflect a memory of a Yavana (here, Indo-Greek ) invasion across the Gangetic plain. According to this text, the Indo-Greeks invaded S ā keta and P ā ṭ aliputra in an alliance with the Pañcalas and M ā thuras. Having over- thrown P ā ṭ aliputra, leaving all desolate and in disorder, they returned to their own realm as conflict had broken out there. 88 Equally difficult to validate is the claim presented in Strabo that the Greeks of Bactria took possession of Patalene, the king- dom of Saraostos, and Sigerdis, which perhaps also reflected the state of expansion under Menander. 89 The subsequent political history of the Greek Kingdoms is more poorly attested. In Bactria, Eukratides ’ s campaigns might have spread power and resources too thin- ly . Internal pressures, such dynastic conflict and possibly a local uprising, in combi- nation with external ones, such as the Arsakid conquests of the western provinces and nomadic migrations into Bactria, caused the royal abandonment of Ai Khanum in ca. 145  , marking the end of Greek control over the important region of east Bactria. 90 Plato and Heliokles I (ca. 143 – 130  ) appear to have been the last Greek rulers of west Bactria. The first nomadic group that migrated into Bactria during the second century  were perhaps the Saka/Sai, displaced by the arrival of the Y ue- zhi (Tochari/Tocharoi?), although the archaeological identification of these groups remains problematic. 91 After Menander ’ s probably violent death in ca. 130  , Greek power in the Paropamisadai, Arachosia, Gandh ā ra, and the Punjab seems to have become more fragmented. Bopearachchi ’ s numismatic study still provides the basis for the chro- nology and location of the kings, but the picture continues to be reassessed through  Milindapañha 1.2, a paracanonical Buddhist text that relates a fictitious dialog between king Menander and a Buddhist sage. Its first book was likely composed between 100  – 200  . For further discussion, see Morris, ch. 9 , this volume.  Yugapur ā ṇ a , 47 – 48; 56 – 57 . See a parallel in Patañjali ’ s commentary on P ā nini ’ s grammar, pro- viding similar information in the form of a grammatical example ( Mah ā bh ā ysa 3.2.11). The Y ugapu- r ā ṇ a is a Sanskrit text of uncertain age, concerned with the decline of the dharma , which ‘ prophe- sizes ’ the Y avana invasion in the Kali Y uga (the age of strife). This should be compared to a wider tradition about foreign invasions into India ’ s northern plains within Buddhist prophecies of de- cline, for which see Nattier 1991.  Strabo 11. 11. 1, discussed in Coloru 2009 , 241.  For historiographical discussion on the treatment of the fall of Greek Bactria in modern scholar- ship, see Mairs 2014, 146 – 176. For the palatial abandonment of Ai Khanum and its later occupation, see Martinez-Sève 2018.  See Abdullaev 2007 . C entral Asian Empires 69 ongoing research. Coloru has observed that Indo-Greek rule at this point appears to have been divided into two large blocs governed by dynasties often in mutual con- flict. Hypothetically , the western bloc might have comprised the Paropamisadai, Arachosia, and western Gandh ā ra. The eastern bloc perhaps included eastern Gan- dh ā ra to the region surrounding S ā gala, with the Sutlej River marking a frontier with Ś unga territory . 92 Zoilos and Diomedes are thought to have ruled in Arachosia and the Parop- misadai after ca. 130  , while joint issues of Strato I and Agathokleia (perhaps Strato ’ s wife) have been placed in eastern Gandh ā ra and the Punjab, followed by sole issues of Strato. 93 Afterward, Antialkidas and Lysias perhaps ruled jointly in the Paropamisadai and Arachosia. 94 A unique document of diplomatic relations from this period is the Heliodoros pillar, an inscribed stone column topped with a sculpture of Garuda (the mount of Vishnu) erected in ca. 113  in Vidisha (Madhya Pradesh ). The two Prakrit inscriptions written in the Br ā hm ī script describe the pil- lar ’ s establishment by Heliodoros of Taxila, a devotee to the god Vasudeva and the Yona (Greek ) ambassador from mah ā r ā ja Amtalikita (i.e., Antialkidas) sent to Bh ā gabhadra, who was either a local ruler or member of the Ś unga dynasty . 95 Around the beginning of the first century  , Philoxenos might have briefly reunified the western and eastern blocs, followed by the rule of several coeval kings. Hermaios (ca. 90 – 70  ) was the last Indo-Greek king to rule in the Paropamisadai. His coinage was posthumously imitated in this region until the conquest of the Ku- shan king Kujula Kadphises. The authority responsible for striking these imitations remains unclear; perhaps it was the Y uezhi. 96 Between ca. 90  and 10  , Arachosia, Gandh ā ra, and the Punjab developed as spaces of intensified political competition and diversity . Explication of the abso- lute chronology , political organization, and familial relations and alliances between rulers of this period remains a work in progress. 97 Indo-Greek kings ruled over a declining space in the east. Maues – the first of the so-called Indo-Scythian (Saka) kings – appears to have taken control of Taxila, with his dynasty ruling over Ara- chosia and Gandh ā ra thereafter. 98 Apollodotos II perhaps regained control over part of Gandh ā ra to the eastern Punjab, a space then divided by Hippostratos and Dio- nysios, while the Indo-Scythian king Azes followed Hippostratos at Taxila. Indo- Greek rule ended with the reigns of Zoilos II, Apollophanes, Strato II, and Strato III  Coloru 2009 , 245.  Coloru 2009 , 246.  Coloru 2009 , 247 .  Discussed further in Mairs 2014, 117 – 133.  Bopearachchi and Rahman 1995, 37 – 38.  For one recent exposition, see Errington and Curtis 2007 , 57 – 66.  A familial relationship between Maues and the Indo-Greek Artemidoros has been postulated from the interpretation of a G ā ndh ā r ī legend of a rare copper series of Artemidoros, but for a critical discussion thereof see Bopearachchi 2008, 28 – 34. 70 Lauren Morris in the eastern Punjab. At the turn of the first century  , another dynasty initiated by the king Gondophares emerged in Arachosia and Gandh ā ra. This dynasty is con- ventionally referred to as the Indo-Parthian dynasty , due to the similarity between their own and Arsakid coins, although their political affiliation remains unclear. The Indo-Parthians ruled much of Gandh ā ra and Arachosia until the Kushan inva- sion of Kujula Kadphises. During this period, other local rulers and dynasties were entangled with regional political affairs and patronage of Buddhism: the Indo- Scythian K ṣ atrapas; the Apracar ā jas, kings of the Apraca royal house in Bajaur who had repeated connections with Indo-Scythian rulers; and the neighboring O ḍ i dy- nasty in Swat. 99 II.  Kingship The logic and challenges of Graeco-Bactrian and Indo-Greek kingship appear to have been similar to those of Hellenistic monarchies, so only characteristics specific to the Central Asian milieu are noted here. 100 In a political context of apparently constant violence and turmoil, the Graeco-Bactrian kings were compelled to com- municate their legitimacy to a competitive peer group – the small Graeco-Macedoni- an settler class – as well as the Bactrian aristocracy who had allied with the con- querors. 101 Strategies of communication included royal displays of resources, such as Eukratides I ’ s building program at Ai Khanum, 102 and propagandistic displays of a king ’ s military achievements, most clearly seen on coinage. For example, the ideological justification of rule over spear-won land ( doriktetos chora ) 103 is reflected in some of Eukratides I ’ s coinage, as on types where he is depicted in a helmet with cuirass (instead of the usual mantle), or on the new type he initatied depicting the spear-throwing king from behind. 104 This type was repeated by subsequent kings, in addition to Eukratides ’ s epithet megas ( ‘ the great ’ ) and his choice of the Dioscuri as reverse deities. These references to Eukratides ’ s military campaigns communicat- ed a distinctively martial royal identity which was evidently attractive. The choice to depict certain deities on the reverses of coins were perhaps in- tended to communicate dynastic legitimacy , such as the Diodotid use of Zeus with thunderbolt, or wider familial associations, such as Euthydemos II ’ s depiction of Apollo on his nickel and copper coinage in reference to the Seleukids. 105 Likewise,  For more on the Apracar ā jas, see Falk 1998, and for the kings of O ḍ i, see von Hinüber 2003.  Further discussion on Hellenistic kingship is found in von Reden, ch. 1, this volume.  See further in Bernard 1981.  Martinez-Sève 2015.  For the significance of spear-won land, see von Reden, ch. 1, this volume.  Bopearachchi 1991, Eukratides I series 8.  Bopearachchi 1991, 56. For a different perspective, see Coloru 2009 , 195. C entral Asian Empires 71 the commemorative issues of Antimachos I and Agathokles may be interpreted as attempts to convey legitimacy , used by both kings to connect themselves with past sovereigns of Hellenistic Central Asia. 106 Gaining the acceptance of wider subject groups was an important task that in- volved new cultural and linguistic challenges, and to this end, the ‘ religious policy ’ of the Greek kings of Central Asia appear to have been similar to that of their Seleu- kid contemporaries. They depicted or referenced popular local deities on their coins, used sanctuaries of local and syncretic deities as platforms for showcasing royal power, and most probably developed royal cults. 107 On Indo-Greek coinage, this manifested in the use of bilingual legends and the portrayal of Indic symbols and deities; most vividly , certain issues of Agathokles depicted the Brahmanical deities Balar ā ma-Sa ṃ kar ṣ ana and V ā sudeva-K ṛṣṇ a. Of course, such strategies are not evi- dence for any ruler ’ s adherence to local religions. Whether Menander converted to Buddhism remains a matter of debate, but for political purposes he might have supported the Buddhist community in some way; the use of the symbol of the wheel on some of his copper coins may be read as an attempt to communicate with his Brahmanical and Buddhist subjects. 108 II.  Settlement, Fortific ations, and Urbanism Whi le Gree k and Lati n lit erary s ourc es men tion th e names o f nume rous f ounda tio ns of th e Hellen istic p eri od in Cen tral Asi a, th e admit tedl y uneve n arch aeolo gic al evi- den ce indi cate s that i t was mo re com mon for p ree xisti ng cit ies a nd set tleme nts t o be de velo ped an d rena med d uring this per iod, rather than fo r new f ound ation s to be es tabli shed . 109 Und er Ale xande r, the Se leuk ids, and t he earl y Grae co-Ba ctr ians, new f ortif icat ions an d garri sons we re pro babl y added to al l majo r urban ce nter s, inc ludin g Mara kand a-Af rasia b and B actra , and sma lle r fortr esse s were c reat ed, lik e Kur ganzo l in nort her n Bactri a, and K ampy rtep a, whi ch was loc ated a t a stra tegi c cro ssing p oint on th e Oxu s. The la tter be cam e more sig nifi cant i n the Gra eco- Bac tri- an p eriod , inco rpor atin g a res ide ntial a rea. 110 A netw ork of bor der fo rtif icat ions acr oss the Hissar Ra nge w as als o deve loped possib ly as ea rly as t he S eleuk id per iod and m ainta ined und er th e Graeco -Bac tria ns. Th is inclu ded th e Iron Ga te wall ne ar Der bent, a nd Uzun dar a, a for tres s loca ted on t op of the S uzi tag co ntro llin g a secon d  Antimachos I depicted portraits and legends naming Diodotos I and Euthydemos I, while Aga- thokles depicted Alexander the Great, Antiochos II, Diodotos I, Diodotos II, Euthydemos I, Deme- trios I, and Pantaleon.  Martinez-Sève 2010.  Coloru 2009 , 242 – 244.  Leriche 2007 , 138. Consult also Cohen 2013; Mairs and Fischer-Bovet forthcoming.  Martinez-Sève 2012, 376. On Kurganzol, see Sverchkov 2014. On Graeco-Bactrian Kampyrtepa, see Bolelov 2018, 11 – 18. 78 Lauren Morris kings ruling in west Bactria. These were followed by the enigmatic silver coinage of ‘ Heraios ’ (misread from an apparent title in the Greek legend turannountos eiaou korranou ). These depict a distinctive bust of a moustachioed and diademed ruler on the obverse, and on the reverse the same ruler in Central Asian-Iranian dress on horseback, with a bow hanging from the saddle (ch. 9 , fig. 4). This coinage seems to have emanated from the realm thought to have been ruled by the Kushan yabgu , and the issuing authority is thought to have been a pre-imperial Kushan yabgu , or perhaps the first known Kushan king, Kujula Kadphises, who came to power in ca. 50  . 141 At the end of the pre-imperial period, the Arsakid king Vardanes I reported- ly sought refuge in the plains of the Bactrians during a power struggle with his brother Gotarzes II, before regaining authority and leading a successful campaign against Gotarzes ’ s Dahae (a nomadic group) army . 142 This event may help to inter- pret the painted clay sculptures of Khalchaian, which is a Y uezhi temple or ceremo- nial pavilion of contested date (perhaps mid-first century  ). Several figures have similar physiognomy and dress to that on the ‘ Heraios ’ coins, and the sculptural program, which includes a mounted battle scene, appears to celebrate a victory over a Saka group. 143 The central sculptural panel includes an individual clearly identifiable as an Arsakid or Parthian noble, which Grenet has hypothesized may constitute a portrait of Vardarnes I, although an earlier date and different historical circumstances remain possible. 144 The beginnings of the Kushan Empire remain shrouded in mystery , best known from their pithy retelling in Hou H anshu ( Documents of the Later Han ), compiled in the fifth century  : More than a hundred years later, the prince [ xihou / yabgu ] of Guishuang, named Qiujiuque [Kujula Kadphises], attached and exterminated the four other xihou . He established himself as king, and his dynasty was called that of Guishuang. He invaded Anxi and took the Gaofu region. He also defeated the whole of the kingdoms of Puda, and Jibin. Qiujiuque was more than eighty when he died. 145 Thus, Kujula seized power among the Y uezhi, and captured at least Kapisa, Gandh ā - ra, and Taxila; the toponyms given in this account are difficult to precisely locate. 146 The Hou H anshu also provides information about the entanglement of the Kushans in the military and diplomatic activities of Ban Chao (Han protector-general of the Western Regions), which also sheds light on the Kushans ’ own diplomatic relations during this period. Reportedly , Ban Chao attacked Yarkand in 84  . Then, the king  See discussions in Falk 2015, § 57 ; Cribb 2018a, 11 – 14.  Tacitus Annales (Tac. A nn .) 9 . 8 – 10.  Bernard 1987 .  Grenet 2000; Olbrycht 2015, 347 – 349 .  Hou H anshu 88.2921, trans. after Hill 2015, 1: § 13.  See Hill 2015, 2: appendices D, F , G. For a numismatic perspective on Kujula ’ s conquests, see Jongeward et al. 2015, 24. On the identification of Central Asian toponyms and ethnonyms in Chi- nese texts, see Morris, ch. 9 , this volume. C entral Asian Empires 79 AN S 1 9 44 . 1 00 . 59 4 28 K i p un a d h a ( c a . 3 3 5 – 3 5 0 ) AN S 1 9 73 . 5 6. 2 20 K u j ul a K a d p h i s e s ( c a . 5 0 – 9 0 ) AN S 1 9 44 . 1 00 . 59 4 69 V i m a T a k t u ( c a . 9 0 – 1 1 3 ) AN S 1 9 44 . 1 00 . 30 1 62 V i m a K a d p h i s e s ( c a . 1 1 3 – 1 2 7 ) AN S 1 9 44 . 1 00 . 30 7 43 K a n is h k a I ( c a . 1 2 7 – 1 5 1 ) AN S 1 9 86 . 1 49 . 12 H u v is h k a ( c a . 1 5 1 – 1 9 0 ) AN S 1 9 44 . 1 00 . 63 8 27 V a s ud e v a I ( c a . 1 9 0 – 2 3 0 ) AN S 1 9 67 . 1 54 . 11 K a n is h k a I I ( c a . 2 3 0 – 2 4 7 ) AN S 2 0 12 . 3 1. 3 V a s is h k a ( c a . 2 4 7 – 2 6 7 ) AN S 1 9 44 . 1 00 . 59 4 45 K a n is h k a I I I ( c a . 2 6 7 – 2 7 0 ) AN S 1 9 44 . 1 00 . 63 8 31 V a s ud e v a I I ( c a . 2 6 7 – 3 0 0 ) AN S 2 0 12 . 2 .1 M a h i ( c a . 3 0 0 – 3 0 5 ) AN S 1 9 44 . 1 00 . 48 1 07 Ṣ a k a ( c a . 3 0 5 – 3 3 5 ) Fig. 2: The Kushan kings. All photographs © American Numismatic Society. of Kashgar revolted, receiving military help from the Kangju 康 居 (another nomadic group controlling Chach and much of Sogdiana). The Kushans and the Kangju were newly linked by a recent marriage, so Ban Chao sent gifts of silk stuffs to Kujula and persuaded him to convince the Kangju to discontinue their support for the king of Kashgar. 147 Whe n Ban Cha o defea ted Y arkan d, the K usha ns se nt an env oy with tri bute of p rec ious st ones, antelo pes, an d lio ns; ho weve r, the ir re quest f or a Han pri ncess i n marria ge wa s reject ed, p rompt ing ani mos ity towa rd the H an ther eaft er. 148  Hou H anshu 47 .1579 – 1580.  Hou H anshu 3.158; Hou H anshu 47 .1580. 80 Lauren Morris In Ku jula ’ s fin al year s, he a lso ap pears t o ha ve init iate d the ne w ‘ So ter Meg as ’ cop - per c oinag e – with the l ege nd basi leus b asi leuo n [sic ] sot er mega s ( ‘ k ing of kin gs, gr ea t savior ’ ) – which was struck to a reduced Attic standard. The majority of this coinage however appears to have been minted by Kujula ’ s son and succes sor, Vima Taktu. 149 Vima Taktu continued attempts to expand Kushan rule through conquest. Chi- nese historical sources mention a failed military enterprise into the Tarim Basin in 90  . Vima Taktu reportedly led a force of 70,000 to attack Ban Chao west of Ku- cha. Short on provisions, Vima sent envoys with gold, silver, pearls, and jade to the king of Kucha to request aid, but Ban Chao laid an ambush to kill the envoys. Vima then sent an envoy to Ban Chao in a gesture of peace, and thereafter the Kushans reportedly paid tribute to the Han every year. 150 Vima ’ s exploits in the southeast were more successful. He conquered Tianzhu 天 竺 (northwest India up to Mathura) and installed a general in charge of its administration, which reportedly made the Kushans extremely rich. 151 Vima appears to have consolidated his control in the Mathura region by establishing a devakula (house of the gods) at M ā ṭ . This poorly documented structure was furnished with over life-sized sculptures of the Kushan kings (some accompanied with inscriptions) and also deities, although the precise religious beliefs and practices involved remain unclear. 152 Vima Taktu also estab- lished a trilingual inscription on a boulder on a ridge of Mount Qarabayu, perhaps intended as a proclamation of power or boundary marker. The languages used were Bactrian (the native spoken idiom of the country written in a modified Greek alpha- bet ), G ā ndh ā r ī , and the so-called ‘ unknown language ’ in the undeciphered ‘ un- known script. ’ 153 Vima Taktu ’ s successor, Vima Kadphises, is best known for introducing a regu- lar gold coinage, struck to a new unit (ca. 8.0 g) conventionally referred to as the dinar . 154 On these coins, Vima Kadphises established distinctive and innovative de- signs reflecting Kushan royal identity . 155 A single rock inscription at Khalatse (near to Leh, see map 2 ) dated to 112  bears his name. 156 Between the years of 114 – 120  , i.e., during Vima Kadphises ’ s reign, the Kushans also reportedly have be- come entangled with dynastic problems of Kashgar. The king o f Kas hgar ’ s mate rnal unc le, Che npan , was exi led to the Ku shan s, who are s aid to h ave beco me ver y fond  Cribb 2014; Jongeward et al. 2015, 5.  Hou H anji 13; Hou H anshu 47 .1580. For the interpretation of Vima Taktu ’ s name in these texts, see Falk 2015, § 74 – 75.  Hou H anshu 88.2921.  For a summary of the excavations, see Rosenfield 1967 , 140 – 142. For the inscription on the base of the statue of Vima Taktu, Falk 2015, § 88.  Fussman 1974, 2 – 50; Sims-Williams 2012, 76 – 77 . See further in Morris, ch. 9 , this volume.  See further in Morris, ch. 16, this volume.  Jongeward et al. 2015, 7 , 53.  Edition in Falk 2015, § 91. C entral Asian Empires 81 of h im. Whe n the ki ng o f Kashg ar di ed wit hout an hei r an d a you ng rela tiv e was ins talle d on th e thro ne, Kus han so ldie rs es corte d Che npan ba ck to K ashg ar whe re he wa s made ki ng. 157 Vima Kadphises was succeed ed by his son, Kanishka I, who is re latively well known to history through memories and legends in later Buddhist texts relating his conquests, supernatural powers, wickedness, and his (historically dubious) conversion to and personal patronage of Buddhis m. 158 Thankfully , two contempo- rary Bactrian inscription s provide more concrete insight into Kanish ka ’ s accession and Indian conquests. The R abatak foundation inscription was discove red by chance at an unexcavated bagolango (image temple) in 1993, and is packed with historical information that revolut ionized Kushan studies. 159 This inscription men- tions Kanishka ’ s inauguration of a new era (year 1 = 127  ) and notes that the king was busy pacifying “ all Ind ia ” for the first six years of his r eign. 160 Here “ all India ” is conceived as the famous u rban centers of the Gangetic plain that were captured, including S ā keta, Kau śā mb ī (Kaushambi), P ā ṭ al iputra (Palibothra), until Ś ri-Cam- p ā . 161 The inscription on the silver dish of Nukunzuk records that the king returned from India to Bactria- Tokharistan in ca. 137  . 162 The historicity of Kanishka ’ s other military campaigns and extent of power, transmitted in later texts, is difficult to pin down. The seventh-century  Chinese Buddhist pilgrim Xuanzang reported that he had learned that Kanishka had en- larged his territory into the Tarim Basin. 163 A fifth-century  Chinese Buddhist text described an attack from the king of Anxi (the Arsakids), from which Kanishka emerged victorious. 164 The same text gives a moralizing and mythologized account of the king ’ s end, in which his ministers – sick of Kanishka ’ s insatiable desire for conquest – smothered him to death after he fell ill. 165 Yet Kanishka was not just a conqueror. The Rabatak inscription is an unequivo- cal representation of Kanishka ’ s interest in new material and symbolic expressions  Hou H anshu 88.2927 .  See discussion in Rosenfield 1967 , 28 – 39 . A range of evidence, especially epigraphic, disputes the traditional association of the Kushans with Buddhism and its expansion. As Fussman (2015, 153) stated plainly , “ The Surkh Kotal and Rabatak inscriptions prove beyond doubt that Kani ṣ ka was no Buddhist, and, judging by their coinage, no Kushan Emperor was ever a Buddhist. ”  The first publication is Sims-Williams and Cribb 1996. Refer now to the third edition and trans- lation in Sims-Williams 2004.  Rabatak, lines 2, 19 .  Rabatak, lines 5 – 6.  Silver dish of Nukunzuk, line 4; edition and trans. in Sims-Williams 2015, 257 .  Xuanzang Da T ang Xiyu ji , Taish ō Tripi ṭ aka, vol. 51, no. 2087 , 1.0873c23; trans. Zürcher 1968, 377 .  Fu F azang Yinyuan zhuan , Taish ō Tripi ṭ aka, vol. 50, no. 2058, 5.316b16 – 18; trans. Zürcher 1968, 386.  Fu F azang Yinyuan zhuan , Taish ō Tripi ṭ aka, vol. 50, no. 2058, 5.317a4 – 18; trans. Zürcher 1968, 387 . 82 Lauren Morris of his royal and dynastic authority . This is not only seen in his inauguration of a new regnal era, but in the central part of the Rabatak inscription, which records Kanishka ’ s order for a karalrang (a high officer, discussed below) to establish a temple to a pantheon of gods (a number of which being Zoroastrian in origin), to be fitted with the images of these gods, and the images of Kanishka and past kings. The foundation act recalls Vima Taktu ’ s establishment of the devakula at M ā ṭ , and that of the bagolango of Surkh Kotal. 166 Kanishka ’ s reign also marks the abandon- ment of Greek as an imperial official language and its replacement by the Bactrian language. This is indicated in the Rabatak inscription, which states that Kanishka “ issued a Greek edict (and ) then he put it into Aryan, ” 167 and likewise Kanishka ’ s coin legends began to be inscribed in Bactrian exclusively . Thus, Kanishka ’ s reign appears to see a conscious revival of an Iranian vocabulary of power. 168 The dimensions of the Kushan Empire appear to have crystallized in the reign of king Huvishka, yet some political problems encountered during his reign may be inferred. For example, the weight standard of his coinage dropped, 169 and the main inscription at Surkh Kotal refers to Nukunzuk the karalrang ’ s installation of a well at the sanctuary , an act framed as a pious response to an enemy attack. 170 Similarly little is known about the reign of Huvishka ’ s successor, Vasudeva I, but some indi- cations of diplomatic activities during his rule are available. Most significantly , Va- sudeva appears to have sent envoys with tribute to the Chinese capital in 230  , 171 possibly seeking assistance in anticipation of the Sasanian foe. Nonetheless, a crushing blow was struck to the empire with the loss of Bactria- Tokharistan to Sasanian invaders under Ardashir I in ca. 230  , followed by the installation of the so-called Kushano-Sasanians (ca. 230 – 365  ) whose kings called themselves Kushanshah . The Kushano-Sasanians were possibly a semiautonomous cadet branch of the Sasanian royal house, perhaps named ‘ Kushan ’ for ruling over an area of former Kushan control. 172 While these kings bore names from the Sasani- an world – Ardashir, Peroz, Hormizd, Varahran – their coinage mingled Sasanian and Kushan designs. The process and chronology of the Sasanian/Kushano-Sasani- an advance across the former Kushan realm still requires further clarification. It would appear that Bactria- Tokharistan was invaded by the Sasanian king Ardashir I in ca. 230  , establishing a homonymous Kushano-Sasanian Ardashir at Bactra. Expansion into Kapisa might have begun around ca. 260  under the Kushano-  See the discussion of the Kushan pantheon in Grenet 2015.  Rabatak, lines 3 – 4; trans. Sims-Williams 2004. “ Aryan ” here does seem to simply designate the Bactrian language, but was a politically , ideologically , and ethnically charged term used in a parallel manner in Darius I ’ s Bisitun inscription (Panaino 2015, 96 – 101).  As proposed already in Fussman 1977 .  Jongeward et al. 2015, 89 – 90.  Surkh Kotal 4 M, trans. Sims-Williams 2012, 78 – 79 .  Sanguo zhi 3.97 ; trans. Zürcher 1968, 371. See also the sources discussed in Falk 2015 § 112 – 116.  Rezakhani 2017 , 72 – 73. C entral Asian Empires 83 Sasanian king Peroz I. 173 The trilingual inscription of Shapur I (240 – 270  ) at Ka ’ ba-i Zartosht (Naqsh-i Rustam, Fars, Iran) described the Kushanshahr as reach- ing “ as far as Peshawar, ” 174 and Gandh ā ra appears to have remained a contended space between the Kushano-Sasanians and Kushans for almost a century thereafter. The Sasanian rock relief of Rag-i Bibi in northern Afghanistan (ca. 25 km southeast of Surkh Kotal ), which depicts a figure identified as Shapur I hunting rhinoceros, attests to a direct Sasanian presence in this territory . 175 The historical record pertaining to the Kushan kings after Vasudeva is poor and contested. Indeed, we only know that the last three Kushan kings were named Mahi, Ṣ aka, and Kipunadha, due to the Br ā hm ī inscriptions on their coins. A plausible reconstruction suggests that the Kushan realm gradually constricted in the wake of the expansion of burgeoning Kushano-Sasanian and Gupta power. The presence of both the Kushano-Sasanians and Kushans in Gandh ā ra and the northern Punjabwas finally wiped out through the expansion of the Kidirate Huns from Bactria- Tokharis- tan. 176 III.  Kingship The logic of Kushan kingship has not been comprehensively treated in scholarship, but some general remarks may be offered. The Kushan monarchy appears to have been absolute and hereditary . The origins and legitimacy of the dynasty ’ s kingship were regarded to derive from the Iranian goddess Nana, the head of the Kushan pantheon. More light on the conferral of kingship is shed by an unprovenanced painting on cotton of an investiture scene. The king in the center, perhaps a depic- tion of Huvishka, is depicted here accompanied by two figures to his left – perhaps representing military and civil estates – while the figures to his right appear to be Zoroastrian priests, preparing to solemnize a service to the sacred fire. A putto flies above, presenting an untied diadem. Although the central part of the scene is miss- ing, it has been tentatively reconstructed to depict the investiture of the crown prince, to whom the king hands over a bow and quiver. 177 The Kushans drew on multiple vocabularies of power in their expressions of kingship, stemming from Iranian, Greek, and Indic worlds. In particular, the Central Asian-Iranian tradition was central to articulations of royal imagery . Kushan royal costume – the caftan, felt boots, and a cap – recalled the traditional dress of the region, and the Kushans also drew heavily on Arsakid royal iconography , as the  According to the reconstruction proposed in Jongeward et al. 2015, 197 – 198.  ShKZ I:24; trans. Huyse 1999 .  Grenet et al. 2007 .  See the extended discussion in Cribb and Bracey forthcoming.  Marshak and Grenet 2006; Grenet 2015, 225 – 226. 84 Lauren Morris predominant imperial power in the Iranian world of that time. 178 On the other hand, some elements of Kushan royal imagery appear to recall the nomadic heritage of the dynasty . 179 From late in the reign of Kujula Kadphises, the Kushans referred to themselves as ‘ King of Kings ’ in the languages of their empire: in Greek basileus basileon , Bactrian shaonano shao , as well as the G ā ndh ā r ī maharaya rayatiraya and Sanskrit mah ā r ā ja r ā j ā tir ā ja ( ‘ great king, King of Kings ’ ). This title of Achaemenid origin had been previously introduced into the numismatic and epigraphic reper- toire of Gandh ā ra by the Indo-Scythian king Maues, and its use by the Kushans was perhaps influenced by contemporary Arsakid usage. A variety of epithets were also attached to the Kushan kings, perhaps the most contested being the Bactrian bagopouro , equivalent to the G ā ndh ā r ī and Sanskrit devaputra ( ‘ son of god ’ ). Scholars disagree as to whether this epithet actually indi- cated the king ’ s divinization. 180 Opinions likewise diverge as to the interpretation of the bagolango / devakula sanctuaries. Although frequently described as dynastic temples, they were fitted with both images of the gods of the Kushan pantheon and the Kushan kings. Whether the kings were in fact worshipped within a dynastic cult remains an open question. 181 The Kushans were tolerant of the many religions cur- rent in their empire, which solidified the legitimacy of their rule among their subject populations. This approach is most visible in the occasional depiction of Indic and Greek deities on their coins, and donations made by officials and lay followers in Buddhist contexts seeking to confer merit on the king. 182 A crucial piece of the puzzle – the nature of the socio-political structures which bound the king to his inner circle and upper officials (see below) – remains obscure. The inscription of the silver dish of Nukunzuk may hold a clue. Here, Nukunzuk began by describing himself as a marego ( ‘ servant, slave ’ ) to Vima Kadphises in his former position as a (lower) official, an amboukao . He then framed his subsequent promotion under Kanishka I in similar terms of servitude: “ he established me (as) equal(?) with (his) father ’ s and with (his) grandfather ’ s servants [ maregano ], with the foremost (people). ” 183 This calls to mind the high-ranking men described as servants (O. Pers. bandaka ) of Darius I, mentioned in the Bisitun inscription, who were chosen by the king as generals against those who rebelled against him. 184 These noblemen were apparently ‘ bound ’ to the king, a relationship perhaps sealed  Sinisi 2017 .  Grenet 2012, 12 – 17 .  See V erardi 1983; Panaino 2009 .  Compare most recently Shenkar 2017 with Grenet 2015, 209 – 210.  For religion under the Kushans, see Fussman 2015; Grenet 2015. Skinner 2017 approaches don- ative spheres, especially Buddhist, as central to imperial stability .  Silver dish of Nukunzuk, line 2; trans. Sims-Williams 2015, 257 .  Throughout DBp; edition and trans. in Schmitt 1991. I thank Frantz Grenet for drawing this to my attention. C entral Asian Empires 85 with an oath. 185 The information provided about Darius ’ s bandaka indicates that they were drawn variously from (at least ) noble families and satraps , and were not always Persians. Yet, multiple sources for the comparative Kushan case may be hypothesized. For example, the institution of the comitatus (as it was known in Latin sources) – a loyal war band of friends sworn to defend their lord to the death – has been analyzed as a crucial component of the ‘ Central Eurasian Culture Com- plex, ’ key to the success of nomads in warfare, and a central and enduring structure in their rule. 186 III.  Settlement and Urbanism Two principle problems pervade research on Kushan-era settlement and urbanism: first, the relevant archaeological documentation is of uneven coverage; and second, the extent of imperial control is unclear and difficult to determine archaeologically . Indeed, widespread material cultural phenomena in northern India of the early his- toric period – such as baked bricks, red stamped ware, and terracotta figurines – have been traditionally characterised as ‘ Ku ṣ ā ṇ a, ’ making it difficult to assess any specific characteristics of Kushan urbanism. 187 Kushan imperial control may be more safely postulated from the presence of inscriptions mentioning Kushan kings and substantial site or surface finds of copper coins. Thus, for example, from Vima Taktu ’ s Dasht-i Nawur inscription and the copper Kushan coins in the collection of Kandahar Museum, 188 it may be inferred that Arachosia – a politically contested space between the fourth century  to the mid-first century  – came under some form of Kushan control or influence. Yet, this mountainous and sparsely settled region remains largely archaeological terra incognita. The most intensive research on settlement patterns and urban archaeology of the Y uezhi-Kushan period has been conducted across disparate micro-regions of the Kushan Empire ’ s heartland, Bactria- Tokharistan. 189 In north Bactria, the number and coverage of sites increased, and former military colonies and fortresses became regional hubs, such as Kampyrtepa, or even urban centers, like Dal ’ verzintepe and Termez. Another urban center, although only partly archaeologically documented, is believed to have been established also at Shakhrinau, in addition to a possible regional hub at Garav Kala (Iavan). The number of sites in east Bactria seem to have receded, but a new urban center in central Bactria was created, the unexcavated Qala-i Zal. Bactra was refortified, and the apparent former military colony  Eilers and Herrenschmidt 1989 .  Beckwith 2009 , 12 – 25.  Pons 2016; Prasad 1973.  MacDowall and Ibrahim 1978, 68.  For the following, see Leriche 2007 . For the main sites of Bactria- Tokharistan ’ s Kushan period, see also Staviskij 1986, 261 − 279 . 86 Lauren Morris Dil ’ berdzhin was developed into an urban site. The expansion of Buddhism during this period also affected the archaeological landscape, seen in the construction of new Buddhist monasteries and monuments around old urban centers such as Bactra and Begram, and the new center of Termez. 190 III.  Military Expansion by military conquest was the foundation of the Kushan imperial project, so much that it came to the attention of outside commentators; the author of the Periplus Maris Erythraei (first century  ) knew the ‘ Bactrians ’ only as a “ very war- like people. ” 191 However, there is limited concrete evidence for the shape and opera- tion of the Kushan army . 192 The Khalchaian sculptures (discussed above) depict the Y uezhi vividly as light horsemen armed with bows against an opponent using armored horsemen. The Ku- shans probably drew on their nomadic heritage through an intensive use of light cavalry units. The use of elephant units may be inferred from the iconography of Huvishka ’ s coinage, which sometimes depicts the ruler holding an elephant goad or riding an elephant. Nikonorov has postulated the existence of a professional or- der of ‘ horsemen ’ forming a military class of light- and heavy-armed mounted troops in the royal army , led by a high dignitary entitled ‘ chief of cavalry . ’ 193 The size of the Kushan army is also not clear from our sources. The Hanshu recorded that the Kushans possessed 100,000 men who could bear arms, 194 and the force led by Vima Taktu against Ban Chao comprised either 70,000 horsemen or 70,000 soldiers, according to two different Chinese historical chronicles. 195 These sources give rough and probably stylized figures, and therefore hardly constitute reliable evidence for the manpower that the Kushans could mobilize for any given campaign. III.  Administration Although direct evidence remains limited, we may speculate that the Kushan Em- pire was supported by a complex administrative system. Epigraphic evidence attests to offices held by the intermediaries between the king and his subjects, if the nature of their roles and the hierarchy remains unclear. Bactrian-language inscriptions mention one of the highest offices, the title of karalrang ( ‘ lord of the marches, ’ ‘ mar-  See generally Fussman 2015.  P ME 47 .  The only extended treatment is Nikonorov 1997 , 56 – 68.  Nikonorov 1997 , 56.  Hanshu 96A.3890, repeated in H ou Hanshu 88.2920.  Hou H anji 13; Hou H anshu 47 .1580. C entral Asian Empires 87 grave ’ ). 196 This is equivalent to the later attested Middle Persian kanarang , a title given to the Sasanian northeastern frontier province warden. 197 A similar function may be inferred for the Kushan context, although the epigraphic evidence only doc- uments karalrangs establishing or enriching bagolangos . The Surkh Kotal inscrip- tion clarifies the high status of this position, as Nukunzuk the karalrang is also described as “ the lord ’ s favorite, who is most dear to the king … the second-in- command(?). ” 198 The fact that two individuals, Pyash and Shafar, are associated in the Rabatak inscription with the title karalrang indicates that the position might have existed for multiple frontier provinces. The function of other titles mentioned in the Bactrian inscriptions, such as amboukao and hasht- walg , are not yet clear. However, in the epigraphic record, we witness most likely the same Nukunzuk rise from an amboukao under the reign of Vima Kadphises, to hasht- walg under Kanish- ka, and finally reaching karalrang in the reign of Huvishka. 199 This indicates a hier- archy of high offices, as well as mobility between them. Other documented offices look more familiar to us. The titles of k ṣ atrapa ( ‘ sa- trap ’ ) and mah ā k ṣ atrapa ( ‘ great satrap ’ ), which are clearly of Iranian origin, are attested already in Gandh ā ra in the mid-first century  . 200 Yet, when attached to the names of various Indo-Scythians from Jalalabad to Ujjain in the following centu- ries, these titles appear to variously indicate nobility , regional governance, or indep- dendent rule. 201 Nonetheless, the use of the title in clear contexts of Kushan rule may indicate a governing office. A unique legend on a ‘ bull and camel ’ coin issued by Vima Taktu (usually attributed to Kashmir), appears to name a descendant of the king as mah ā k ṣ atrapa . 202 Likewise, a G ā ndh ā r ī -inscribed reliquary casket from Manikyala stupa (east of Taxila) refers to a k ṣ atrapa of Kapisa, the son of another k ṣ atrapa ; 203 thus perhaps this office could be hereditary . With the advent of the Kushan invasion of India, professional titles of Iranian origin begin to appear in inscriptions in Epigraphical Hybrid Sanskrit. 204 There are numerous examples, such as the titles ham ā rak ā ra and gañja-h ā m ā rakara, which denote an account-keeper of the treasury . Another term, khar ā salerapati , has been explained as “ chief of the army chiefs. ” 205 As documentary texts from the Achaemenid and Kushano-Sasanian periods in Bactria have recently materialized, it is probably just a matter of time before similar  For example, in Rabatak, lines 7 – 8, 15 – 16; Surkh Kotal 4 M, lines 20 – 25.  The etymology first proposed in Henning 1965, 77 – 79 .  Surkh Kotal 4 M, lines 6 – 9; trans. Sims-Williams 2012, 79 .  Silver dish of Nukunzuk, lines 1 – 2; Rabatak, lines 16 – 17 ; Surkh Kotal 4 M, line 7 .  On the etymology of k ṣ atrapa , see Salomon 1974, 12 – 14.  Falk 2010, 74, 78.  Falk 2010, 78, Fig. 3.  CKI 150 in Baums and Glass, 2002 – .  See Falk 2010, 78.  Falk 2010, 78; Falk 2015, 121 – 122. Mamta Dwivedi 3 E arly Historic South Asia I Geography, C limate, and the Human Landsc ape The g eogra phic al bor ders of p re-n atio n-st ates or p oli ties ar e dif ficul t to cha rt out cle arly . Yet it is mor e oft en than n ot that g eog raphy s hape s poli tica l and cult ural ent ities . Sout h Asi a form s one su ch ent ity , havin g phys iog raphi cal mar ker s tha t de- fin e the spa ce both ge ogr aphic ally a nd cul tura lly . 1 Thi s space is b ounde d by the Bal ochis tan hi ghla nds to the w est, the S wat v alley i n the nort hwe st, the Hi mala yas in th e north , the mer idio nal mou nta in chai n of Ind ochi na in the ea st, a nd the pe nin- sul ar regi on in the so uth . The pen insul ar so uth is sur roun ded by t he Arab ian Se a in the w est, t he Ba y of Be ngal in the ea st, an d the I ndia n Oce an in th e sou th. A n exte n- siv e coast line of 1 1,1 04 km in total s urro unds t he Indi an su bcont inen t (map 1). 2 Three physiographic divisions mark the mainland of the subcontinent: the Him- alayas, the Indo-Gangetic Plains, and the peninsula. 3 The Himalayas are young tec- tonic mountains with various glacial snow formations. Melting glacial deposits form three perennial river systems, the Indus, the Ganga, and the Brahmaputra, with various distributaries and tributaries. These river systems inundate the northern plains and form deposits of fertile alluvial soil. The Indus River system flows into the Arabian Sea, while the Ganga and the Brahmaputra flow into the Bay of Bengal. The alluvial plains are bordered by the Aravali and Vindhya Ranges in the south. The Aravali plateaus are the oldest physiographical formations, and it is here that the Deccan plateau and peninsular India begin. Pe nin su la r In di a is ch ar ac te riz ed b y sm all er z on es, s uc h as th e la va tr ap t op ogr a- ph y wit h bl ac k soi l in th e we st ern a nd u pp er De cc an , 4 gra ni te a re as in t he ea st er n re gio n, an d re d soi l in th e so uth er n pen in su la. 5 The re ar e al so po ck et s of all uv ia l soi l zo nes i n the N ar ma da, T apt i, M ah an ad i, Go da var i, K av er i, an d Kri sh na Ri ve r val le ys . 6  South Asia comprises the modern states of India, Pakistan, Nepal, Bhutan, Bangladesh, and Sri Lanka. It is located between 5° and 37° N. The areal expanse of the subcontinent is more than 4.2 million sq km (Spate and Learmonth [1954] 2007 ; Xue and Yanai 2005, 115).  The total coastline area includes the coastline of the modern nations of India (7 ,500 km, includ- ing the islands), Pakistan (1,365 km), Sri Lanka (1,585 km), and Bangladesh (654 km), (Snead 2010b, 1059; 2010a, 1078; Nayak and Hanamgond 2010, 1065; Swan 2010, 1072).  Spate and Learmonth (1954) 2017 , 6 – 7 ; Tandon et al. 2014, 3.  The black soil, also called the regur type, is rich in ferromanganesian and aluminum compounds because of the volcanic lava content (Randhawa 1980, 8).  This region consists of the oldest rock constitution containing prevalent crystalline schists and ferromanganesian minerals (Randhawa 1980, 8).  Spate and Learmonth (1954) 2017 , 16. Note: I would like to thank Professor H. P . Ray for her suggestions and comments on this chapter. Open Access. © 2020 Mamta Dwivedi, published by De Gruyter. This work is licensed under the Creative C ommons Attribution-NonCommercial-NoDerivatives 4.0 License. https:/ /doi.org/10.15 15/97831106077 41-005 96 Mamta Dwivedi Map 1: Major cities and routes in early historic South Asia (after C handra 1977; Neelis 2011). © Peter P alm. Early Historic South Asia 97 Of fsh or e, t he So ut h As ia n re gio n in cl ud es th e fo ll owi ng m ai n isl an ds : Lak sh ad we ep of f the w es t co as t, A nd ama n- Ni co bar I sl an ds in t he B ay o f Be ng al, 7 an d Sri L an ka i n th e Ind ian O ce an . The S ri La nk an h int er la nd co ns is ts of al lu vi al t ra ct s, la goo ns , an d co ast al r eg io ns . 8 The climate of the subcontinent varies considerably . The subcontinent has six major ecological zones: rainy tropical, humid subtropical, tropical savanna, moun- tain, arid desert, and steppe grassland. 9 The extreme north has a temperate climate, while the northeastern regions, the central part, and the south are influenced by the tropical monsoon. The southernmost part of the mainland and Sri Lanka have an equatorial climate. Seasonal changes in temperature and rainfall throughout the subcontinent depend on the monsoon. 10 In most regions, 80 percent of the rainfall is the result of the southwestern monsoon, lasting for four months from June to September. The northernmost regions, by contrast, experience precipitation during the winter, while parts of the southeastern coast, the islands in the Bay of Bengal, and northeastern Sri Lanka have rainfall during the retreating monsoon from Sep- tember to January . 11 The diversity of the topography , soil type, climate, and rainfall are one explana- tion for the diversity of vegetation in South Asia. 12 The other is the human impact on the landscape. Apart from archaeobotanical finds, early historical literary texts mention a great variety of agricultural crops and methods of cultivation. 13 Arguably , the particularly long tradition of local domestication combined with the dispersion and adoption of grains via land and sea routes since the third and second millennia  account for the greater variety of grain crops in early historic South Asia than in any other world region. 14 Double-cropping in the form of winter and summer crops, and multicropping (growing more than one crop in a single season) were common practices in the subcontinent as early as 3000  . 15 Rice, millet, mustard, sesame, cotton, hemp, and some pulses formed the main crops of the summer culti- vation cycle, irrigated by the monsoon rain from July to August. 16 Winter cropping included wheat, barley , pulses (horse gram, mung bean, pea, chickpea, grass pea), flax, and safflower. Winter crops were watered by the winter rainfall in the north- west and by residual soil moisture or river irrigation in other regions.  Nayak and Hanamgond 2010, 1066.  Swan 2010, 1073.  McColl 2014, 453.  Singhvi and Krishnan 2014.  Dash 2005, 509; Xue and Yanai 2005, 115; Randhawa 1980, 21.  The types of vegetation include the temperate Himalayan type, the tropical thorn forest, the dry deciduous forest, the tropical evergreen rain forest, and the mangrove and beach forest. For a de- tailed account of the division of vegetation types, see Randhawa 1980, 25 – 43.  Randhawa 1980; Raychaudhuri and Roy 1993; Srinivasan 2016.  Murphy and Fuller 2017 , 6; Fuller et al. 2011.  Murphy and Fuller 2017 , 8; Petrie and Bates 2017 , 83 – 84, 89 .  Petrie and Bates 2017 , 89; Murphy and Fuller 2017 . 98 Mamta Dwivedi II P olitic al F ormations in Early South Asia Henige suggests that the study of political chronology can be placed into a typologi- cal continuum. At one end are those incontrovertibly documented societies about which there is no dispute over timing and sequence. As we move back in time or to areas of restricted literacy and documentation, problems of chronology become more numerous and refractory . In these cases, the available evidence makes it virtu- ally impossible to be certain about the timing and sequence of even the known events. 17 The chronology of early India fits the latter end of this continuum. Since it was the British who first wrote histories of the subcontinent in the eighteenth and nine- teenth centuries, all early political chronologies of Indian history are based on the Christian calendar and relate to Hellenistic events. For the colonial rulers, the only authentic sources of history were Greek historiographies. The Pur ā ṇ as , in contrast, were condemned as being mythological. 18 Thus, the date for the reign of Candra- gupta (Chandragupta) Maurya (ca. 320  ) is derived from references to him in the accounts of the immediate successors of Alexander of Macedonia in the Indian satrapies (ca. 325  onward ). This was followed by dates of five Hellenistic rulers mentioned in Rock Edicts (RE) II and XIII of A ś oka (ca. 268 – 231  ). 19 Also, the dates commonly found on Indic inscriptions of the Samvat and Ś aka eras are attrib- uted to 56  and 78  based on the chronology of the Indo-Scythian rulers. 20 Given that the dating of Buddhist and Hindu texts is also highly controversial, his- torical chronologies and sequences are often only relative in nature and must be treated with utmost caution. The period between 300  and 300  is marked by a variety of political formations: (1) so-called empires; (2) satrapies and independent monarchies; and (3) janapadas (coin-issuing communities in the Indo-Gangetic divide). 21 The period begins with the emergence of the Maurya dynasty (ca. 320 – 185  ), generally re- garded as the first empire of South Asia, followed by a period of fragmentation and foreign domination. It ends with the rise of the Guptas (320 – 550  ), another indigenous dynasty with its center in the northern alluvial plains. Scholars various- ly consider the post-Mauryan period as a phase of invasions, confrontations, inter- actions, innovations, and urban development prompted by external influences from  Henige 1986, 58.  Ray and Potts 2007 .  Antiochos II Theos (261 – 246  ), Ptolemy II of Egypt (285 – 247  ), Antigonos II Gonatas of Macedonia (278 – 239  ), Magas of Cyrene (300 – 258  ), and Alexander II of Epiros (272 – 258  ).  Bhandare 2006, 69 .  The janapadas here are used in the meaning Shrimali suggests. He finds janapada has connota- tions of both monarchical and non-monarchical forms of political organization, and implies both the communal and territorial aspect of a polity (Shrimali 1985, 3 – 4). Early Historic South Asia 99 the northwest into the northern part of the subcontinent. 22 Western influences re- sulted in the emergence of satrapies and monarchies usually referred to as Indo- Greek, Indo-Parthian, Indo-Scythian or Ś aka, and Ku ṣ ā ṇ a. 23 Among the local poli- ties in the post-Mauryan phase, the most prominent were the local ga ṇ a-samghas in the Indo- Yamuna divide; the dynastic rule of Ś ungas in the north, S ā tav ā hanas and Kalinga in the Deccan and Odisha; and the kingdoms of Cola, Cera, and P ā ṇḍ yas in the peninsular south. The historiography of the period under consideration is influenced by two tradi- tions, one emphasizing imperial unity , the other local autonomy . 24 The first focuses on imperial state formation explained by surplus production and the concomitant emergence of social hierarchy and political organization. This approach explains history in terms of the rise and fall of empires, the emergence and fall of cities and cultures, and the dynamics of political vacuums created and filled. It originated in early nineteenth-century scholarship, which searched for empires as markers of ear- ly civilization. 25 By the middle of the twentieth century , the tradition transformed into concepts of state formation and urbanization and was strongly influenced by Marxist historiographic models introduced into Indian historiography by D. D. Ko- sambi. 26 Thus in the Indian context, the study of empire is strongly associated with concepts of state formation and unifying processes. An empire is understood as a well-developed, centralized state that is territorially expansive. Within the imperial approach, the subcontinent is constructed as one historical unit in which different areas played their parts. The northwest is described as a region ever riddled by incessant invasions and imperial endeavors. 27 The northern alluvial plains are regarded as the nucleus of civilization and state formation, while the Deccan and the southern peninsula played catch-up. 28 The state of the Mauryan dynasty , with its center at Magadha (modern Bihar), came to be considered the first and archetypical Indian empire. This characterization of the Mauryan dynasty owes much to the idea of the Mauryan state as a unifier of South Asia under one ruler,  Basham (1954) 1986; Thapar 2003; Chakravarti 2016; U. Singh 2008.  See Morris, ch. 2, this volume.  Chakravarti 2012, 14.  The late eighteenth- and early nineteenth-century Indologists like William Jones and James Mill were familiar with the king named Candragupta of Mauryas from references in ancient Greek sour- ces as a contemporary of the successors of Alexander of Macedonia. Also, A ś oka was known from the Sri Lankan Pali chronicles of the fourth and sixth centuries  . With the successful decipher- ment of Kharo ṣṭ h ī and Br ā hm ī scripts by James Prinsep in the 1830s, the discovery of A ś oka as an Indian emperor was made. The deciphering of the scripts also enabled the identification of the Ku ṣ ā ṇ as as a ruling dynasty of early India (Prinsep 1838; 1858; Dwivedi 2015, 208 – 210).  Kosambi 1956. Gurukkal (2008) describes the tradition as influenced by the “ Kosambi effect. ”  Prakash 1964; 1971.  Seneviratne 1981; Champakalakshmi 1996; Thapar 2003; Basu Majumdar 2017 . 100 Mamta Dwivedi no doubt inspired by the colonial idea of ruling and governing India as a single imperial unit. 29 Since the nineteenth century , the identification of the Mauryas as an empire, particularly under A ś oka, has been based on identifying the criteria for imperial status and arguing that the Mauryas fulfilled them. An empire is defined as a uni- form territory divided into centers and peripheries; it is marked by the introduction of a uniform script and state-issued coinage; there is evidence for royal patronage of art leading to representative monuments; and it has a state ideology . 30 To demon- strate Mauryan imperial homogeneity , scholars have pointed to the ubiquity of northern black polished ware (NBPW) throughout the subcontinent, which in turn was labled Mauryan pottery . They saw the rock edicts as marking the territorial expanse of A ś oka ’ s sphere of influence (see ch. 10.A, map 1), and his policy of dhamma (Buddhist ethical teachings) as the unifying ideology . 31 In this approach, A ś oka stands at the core of the Mauryan dynasty as the ideal emperor. The decipher- ing of Br ā hm ī and Kharo ṣṭ h ī scripts by James Prinsep in the early 1830s enabled the attribution of the titles dev ā nampiya and piyadasi to A ś oka. 32 Since then, A ś oka has remained an important historical figure, both to be studied and admired. 33 Any other political formation was understood in terms of being part of the proc- ess of incipient state formation. Where the influence of the Mauryan imperial state formation process could not be identified, it was suggested that these polities re- mained uninfluenced by Magadha. 34 The imperial model also fostered ideas of ur- banization and so-called secondary state formation in imperial vicinities, and the establishment of long-distance trade routes and religious contacts as part of the imperializing process. Secondary state formation and secondary urbanization were the explanatory models for the rise of complex state-like polities and cities in areas adjacent to the imperial centers in the middle Ganga valley as well as in Sri Lanka. 35 Finally , the spread of the imperial state was associated with the processes of what may be called Sanskritization and Brahmanization. Complex imperial state forma- tion in this model was influenced by orthodox Brahmanical ideas vis-à-vis hetero- dox republican ideas that were prevalent in areas farther away from the Ganga val- ley . 36  Chattopadhyaya 2015, 3 – 4.  Thapar 2003; 2006; (1961) 2013. For criticism, see Morrison 1995; 1997 ; M. Smith 2005; Ray 1986; 2008.  Thapar 2003; (1961) 2013; Allchin 1995; U. Singh 2008; Chakravarti 2016.  Prinsep 1838. In Sanskrit dev ā n ā mpriya means ‘ beloved of the gods ’ and priyadar ś in , ‘ he who regards everyone with affection. ’  For an account of scholarly interest in A ś oka and the Mauryan dynasty , see Lahiri 2015.  Thapar 2003, 158.  The suggested regions of impact are Bengal, Odisha, the Deccan, and the southern peninsula, as well as Sri Lanka, Thapar 2003, 211; Basu Majumdar 2017 ; Chakravarti 2017 , 333 – 338.  The influence of Brahmanical ideals on the emergence of complex state structures is ascribed to the complex and hierarchical ‘ Brahmanical ’ social institutions, which contrast with simpler egali- Early Historic South Asia 101 The second method for studying this region calls for examining socio-cultural aspects of South Asia beyond the question of state formation and political expan- sion. These approaches explain early historical India in terms of autonomous spaces and consider regional social variations instead of uniform processes of state forma- tion. 37 They use fluid concepts of historical orbits, 38 and describe the political space “ as networks of resource acquisition in which territories and boundaries are porous, permeable, flexible, and selectively defended. ” 39 The identification of certain geo- graphical areas as cradles of civilization and starting points of political development is rejected in favor of approaching regions as more than either perennial nuclear regions or backward and tribal parts of the subcontinent. The search for imperial structures and centralization is largely abandoned in favor of studying areas and political formations in relation to their specific social, economic, and political con- texts. 40 The new perspective also focuses on power nodes and social functionaries that used to be overlooked, such as merchants, small landowners, and religious agents. 41 This method denounces the compartmentalized view of political geograph- ies and advocates for the possibility of a more continuous ecological and cultural divide between the northern and southern polities of the subcontinent. II.  Politic al Sc enarios in South Asia II.  .  The Mauryas and the North At the time of Alexander ’ s invasion of the subcontinent, the Nandas ruling at Maga- dha were a formidable political power with a grand standing army . It was in 321/ 20  that the Nandas were usurped by Candragupta Maurya (known as Sandro- kottos to the Greek historians) who founded the Maurya dynasty . 42 Control over the northwestern region of the subcontinent under the Mauryas is credited to Candra- gupta and is based on the record of the gift of Arachosia, Gedrosia, and Paropami- tarian political models found in regions west of the Yamuna and identified as more ‘ heterodox ’ (Thapar 1978). In the Deccan and southern regions, the polities are considered tribal until Brahman- ical ideas were spread through the process of secondary state formation and the propagation of V edic- Ś astric-Pur ā n  ic ideas (Sahu 2001). This idea does not find any support in the archaeological evidence. The archaeological sites of Mathura, Kaushambi, Varanasi, and others exhibit a mixed archaeological assemblage with common and shared iconographies.  Chattopadhyaya 2003a.  Chakrabarti 2010a.  M. Smith 2005, 835.  Lahiri 2015, 172.  M. Smith 2005, 836; Ray 2008, 11.  The date for Mauryan accession depends on the two debated dates of the Buddha ’ s parinirv ā ṇ a (death ), 486 and 483  . 102 Mamta Dwivedi sadai (Kandahar, south Balochistan, and Kabul ) that Seleukos Nikator made in ex- change for 500 elephants. 43 It is also possible that Candragupta Maurya had the chance of retaining some regions controlled by the Nanda rulers, possibly with the exception of Kalinga. 44 Candragupta Maurya ’ s grandson, A ś oka (r. 268 – 232  ) is regarded as the first ruler to have united the subcontinent under one rule, after which he adopted a pacifist policy changing bherigho ṣ a to dhammagho ṣ a – from announcements of war to propagation of dhamma (Buddhist teaching) – as declared in RE IV . A ś oka is also known to have added the Kalinga region to the Mauryan sphere of influence after a battle recorded in RE XIII, which expresses remorse for the losses suffered in war. 45 The Maur ya dynasty en ded with the ass assinati on of the last king B ṛ ha drath a by his mi nister Pu ṣ yamitr a Ś unga , who estab lished a n ew dyn asty cente red in Mag a- dha. T he legend of his access ion, how ever, is deb atabl e as it is dra wn fro m a roman- tic play of th e fourth cent ury  . 46 The pos t-Maury an period is identif ied in th e impe- rial app roach as a p hase of politi cal uphe aval when ki ngdom s that sprang up were in const ant confli ct. The Ś ung as are said to ha ve campa igned again st their so uthern neigh bors in the Decc an, against th e Hellen istic Gree ks in the northw est, and again st the Kaling a in the southe ast. 47 The po litical a ctiviti es in the post- Mauryan perio d, moreov er, are c onsidere d to be th e resul t of poli ties emer ging to f ill in t he vacuu m created by the decline o f the emp ire. The nor thwest and west exper ienced the pres ence of In do-Gree k kingdom s and Sel eukid s atrapie s . 48 After the Ma uryas, only the Ku ṣ ā ṇ as in the first centur y  were able to cl aim a large pa rt of the no rth under th eir rule, as men tioned in the Ra batak ins criptio n. 49 The northwestern region was occupied by the janapadas and ga ṇ a- samghas (political conglomerates) who are known by their coin issues as Yaudheyas, Ā rjun-  Chakravarti 1986, 49 .  In the west, at Junagarh (Gujarat ), Candragupta Maurya is also noted to have commissioned the construction of Sudar ś ana Lake, which was then repaired at the time of A ś oka and then again in 150  by Rudrad ā man, a Ś aka ruler. In the east, in modern Odisha, a post-Mauryan eulogistic rock inscription ascribed to Kh ā ravela refers to a water tank constructed by one of the Nanda rulers. It is possible that Kalinga was controlled by the Nandas but was perhaps lost by early Mauryas and was conquered by A ś oka (Jayaswal and Banerji 1929). The extent of the Kalinga region in the early historic period is not clear. The region might have included areas of the present states of Odisha, Chhatisgarh, Telangana, and Andhra Pradesh.  However, the war with Kalinga referred to in RE XIII is not mentioned in the major rock edicts found in Odisha (at the Dhauli and Jagauda sites), a part of the traditional Kalinga region itself. Guruge (1994, 54 – 55) questions the historical reliability of the content of RE XIII as it has as many as eight versions found in different regions, however none near Odisha.  The M ā lavik ā - Agnimitram is a Sanskrit play in the k ā vya tradition authored by K ā l ī d ā sa sta- tioned at the court of the Guptas.  Thapar 2003, 210.  For discussion of Indo-Greek, Indo-Parthian, and Ku ṣ ā ṇ a rule, see Morris, ch. 2, this volume.  For the Rabatak inscription see, Sims-Williams and Cribb 1995. See also, Morris, ch. 2, this vol- ume. Early Historic South Asia 103 ā yanas, Trigartas, Ku ṇ indas, Ś ibis, and some other names. 50 They are often regard- ed as tribal kingdoms filling the vacuum created by the end of the Mauryan rule. Yet one finds references to these ga ṇ a-samghas in the pre-Mauryan grammatical work by P ā ṇ ini, and they are also mentioned in the Alexander historiography . 51 In the Indian grammatical tradition, these groups are identified as ś astropaj ī vi-sa ṁ gha and ā yudhaj ī v ī (a group living by the profession of arms or warfare). 52 They are also mentioned in the Mah ā bh ā rata and the fourth-century  inscriptions of Samudra- gupta. Coin issues suggest that the three communities of Yaudheyas, Ā rjun ā yanas, and Ku ṇ indas formed an alliance to eventually defeat the Ku ṣ ā ṇ as in the northern region. They issued coins or seals commemorating their victory with the legend yaudheya-ga ṇ asya- jaya (victory of the Yaudheya alliance). 53 Similarly , an inscrip- tion ascribed to the dynasty of the V ā k ā ṭ aka (third to fourth centuries  ), suggests that the imperial N ā gas of K ā ntipuri (identified as modern Mirzapur district of Uttar Pradesh ) ousted Ku ṣ ā ṇ as from the Ā ry ā varta region. 54 II.  .  Politic s in the Decc an Scholars who write the early history of South Asia give relatively less space to the history of the peninsula because there is less evidence for a well-developed territori- al state system. Many scholars suggest that the political formations in the Deccan and the south were secondary state formations, adopting administrative institutions under the influence of the Mauryan metropolitan state ’ s administrative structures. 55 An example of a region supposedly experiencing secondary state formation is the central India region (present Chhattisgarh ), a land corridor of primarily forested areas communicating between the southern and northern regions. While travelers frequented it over centuries, its particular landscape formed a territorial zone favor- ing a particular kind of settlement. Inscriptions from the Sitabenga and Jogimara caves in this region, dating to the late third or early second centuries  , refer to military and administrative offices that might have shared features with offices at- tested in the territorial polities ( mah ā janapadas ) of the Ganga valley during the pre- Mauryan period. 56 Yet the idea of a southward influence over land is problematic when one looks at the archaeological evidence, sculptural and architectural re- mains, and the history of maritime activities. 57  Handa 2007 ; Shrimali 1985; Gupta 1996; Agrawal 1953, 457 ; Allan 1936.  E.g., Arrian Indica 5.  Handa 2007 , 149 .  Altekar and Majumdar (1946) 1986, 23 – 33; S. Ghosh 2012, 49 – 51.  Jayaswal 1933, 5. However, this has been contested by Altekar and Majumdar (1946) 1986, 25 – 28.  Seneviratne 1981; Thapar 2003, 60, 211; Chattopadhyaya 2003b; Basu Majumdar 2017 .  Basu Majumdar 2017 , 123. For the inscription and translation see H. Sastri 1925 – 1926, 152 – 156.  Maloney 1970; Ray 1986; 2003; 2008; Morrison 1995; 1997 ; Abraham 2003. 110 Mamta Dwivedi However, emphasis on divinity of the king is also found in the Indic literary tradi- tion. One of the earliest references comes from the M anusm ṛ ti ( MS ), 105 one of the earliest dharma śā stras dated between the second and third centuries  . 106 In the Sangam literature, we find a reference to the king as the center and embodiment of administration, encapsulated in the titles ko , mannan , vendan , and iraivan . K o is also suggested to have meant god or ‘ god incarnate. ’ 107 The idea of royal paternalism is also attested in various sources. 108 R ā jadharma ( ‘ duties of the king ’ ) are found in early Hindu political and economic treatises, the KA and the MS . The KA refers to the king ’ s happiness ( sukham ), but his welfare ( hita ) lies in the happiness and welfare of the people. 109 Like a father, the king helps initial settlers with various exemptions and grants favors. 110 A reckless king, by contrast, can lose his position as a result of his people ’ s anger and revolt ( janapada- kopa ). 111 Buddhism also has the concept of political society as a family presided over by a morally elevated father figure. 112 The Buddhist universal monarch, the cakkavatti ( cakravarti ), is considered the beloved of the subjects. 113 The statement of the A ś okan edict at Dhauli that “ all men are my children ” goes in the same direc- tion, though in the imperial historiographical tradition this statement has been mis- taken as some kind of paternal despotism. 114 Standing armies are frequently mentioned in the sources, and the military cam- paigns in northern India are well known. The KA gives a detailed account of meth- ods and tactics of warfare and siege, various ranks and their duties, training, army organization, salaries, camping, transport of armies, and the duties of the army commander. 115 The army traditionally was fourfold, comprised of foot soldiers, horsemen, chariots, and elephant forces. Apart from the standing army , the king could deploy additional “ hereditary troops, hired troops, corporate troops, troops supplied by the ally , troops supplied by the enemy , and tribal troops. ” 116 The pur-  Manusm ṛ ti ( MS ) 7 . 4. The king is further equated with the gods Fire, Wind, Sun, Moon, Yama, Kubera, and Indra ( MS 7 . 7). Apart from association with Kubera, the deity of wealth, the king is also elevated to a higher socio-ritual position, and in his “ benevolence lies Padm ā , the goddess of prosperity ” ( MS 7 . 11). A verse very similar to that in MS 7 . 4 is also cited in MS 5. 96, where the king is mentioned as the embodiment of the ‘ eight guardians ’ ( as  t  ā n ā m lokap ā la ) and thus is con- sidered pure at all times (Jayaswal 1924, 55).  Olivelle 2005, 25.  Subramanian 1966, 40.  Bandopadhyaya 1927 , 64. See also Thapar (1961) 2013, 121.  KA 1. 19 . 34.  KA 2. 1. 17 – 18.  KA 1. 13. 20.  Gokhale (1966, 21) in reference to Majjhima Nik ā ya 3. 176.  Gokhale 1966, 21.  Thapar (1961) 2013, 121.  KA 2. 33. 6 – 10.  KA 9 . 2. 1, trans. Olivelle 2013, 352. Early Historic South Asia 111 pose and conditions for the recruitment of these groups are also explained. 117 The arms were to be returned to the royal armory , the horses and elephants to the royal stables. 118 The elephants had three mounted archers and a mahout. Greeks showed a great interest in the use of elephants in the army , and there are references to the importance of elephants in the indigenous military . 119 The KA refers to officers of horses, elephants, chariots, and infantry ( a ś v ā dhyak ṣ a , hastyadhyak ṣ a , rath ā - dhyak ṣ a , and pattyadhyak ṣ a ). 120 The Buddhist text D ī gha Nik ā ya refers to the wheel- treasure, the elephant-treasure, and the horse-treasure as parts of the seven jewels of an empire. 121 It is believed that it was because of this form of military machine that it was possible for the Mauryas to subdue most of the northern region. 122 Even in various works of Greek Indography , the military strength of Indian rul- ers is praised. Plutarch mentions that Candragupta Maurya subdued India with an army of 80,000 horses, 200,000 foot soldiers, 8,000 chariots, and 6,000 ele- phants. 123 Pliny mentions walled towns of the eastern Deccan, and that the rulers of the powerful tribe Andarae (probably referring to the Ā ndhras/S ā tav ā hanas) 124 maintained an army of 100,000 infantry , 2,000 cavalry , and 1,000 elephants. 125 In the southern context, a similar picture has been suggested, with a fourfold army – ox-drawn chariots, elephants, cavalry , and infantry . 126 War drums were symbols of sovereignty and an important part of warfare. Each ruler and chieftain had a war drum among his insignia. 127 The symbolism of war drums is clearly attested in a reference to the change in A ś oka ’ s policy toward con- quest and expansion. His expansionist policy is referred to as bherigho ṣ a , literally meaning the ‘ sound of the war drums, ’ i.e., declaration of war. 128 In the Sangam texts, there are also references to capturing the war drums of different groups as a sign of control over them.  KA 9 . 2. 13 – 20.  Trautmann 2009 , 233.  Trautmann 2009 .  KA 2. 30. 1 – 2. 33. 11.  Ghoshal (1959) 1995, 77 .  Trautmann 2009 , 233.  Plutarch Life of Alexander 62. 3; Majumdar 1960, 192 – 193, 198; see also Chakravarti 1986, 48.  Aiyangar 1941, 46.  Plin. HN 6. 22. 67 ; see von Reden, ch. 10.B, this volume.  N. K. A. Sastri (1955) 1995, 133.  U. Singh 2008, 385.  In RE IV , A ś oka claims the replacement of bherigho ṣ a with dhammagho ṣ a ( ‘ call of dhamma ’ ). It is understood as a change in policy that is from aggressive conquest to dhamma . The term dham- ma in this context is understood by some scholars as the spread of Buddhism, while others under- stand it as moral governance, righteousness, and social ethics. See V . A. Smith (1901) 1920, 29 – 31; Raychaudhuri (1923) 1972, 170 – 178; Basham 1986, 56 – 58; Thapar 2003, 200 – 204. 112 Mamta Dwivedi II.  Regions, Networks, and C onnectivity The Sanskrit textual sources refer to the northern region as the Ā ry ā varta , the abode of the noble, expanding from the Himalayas in the north to Vindhyas in the south. 129 Within this the Ganga- Yamuna region is considered the M adhyade ś a ( ‘ Mid- dle Country ’ ), expanding toward the river Sarasvat ī in the west. 130 Within the period of our concern, epigraphic sources also refer to the subcontinent as Jambudv ī pa (literally ‘ the island of rose apple ’ ) 131 and Bh ā ratavar ṣ a ( ‘ land of the descendants of King Bharat ’ ). 132 The region south from the Vindhya Ranges to the river Krishna was known as the Dr ā vi ḍ ade ś a , which in the Graeco-Roman Periplus Maris Erythraei ( P ME ) and Ptolemy ’ s Geographia is frequently referred to as Dachinabades and Limyrike ( Dimirike ), the hinterland of Barygaza (Bharuch ). 133 The southernmost re- gion of present Tamil Nadu and Kerala are identified as the Tamilakam, which is a linguistic rather than a geopolitical entity . 134 Literary sources do not only identify geographical regions, but also networks and corridors. Such corridors of travel and communication appear in texts from the sixth to fifth centuries  onward. 135 Of interest here are the Uttar ā patha ( ‘ north- ward route ’ ) and the Dak ṣ i ṇ ā patha ( ‘ southward route ’ ). The term Uttar ā patha is found in one of the earliest surviving Sanskrit grammatical texts, P ā ṇ ini ’ s A ṣṭ ā dhy ā - y ī , dated between the fifth and fourth centuries  . 136 The route certainly facilitat- ed long-distance journeys through networks of roads toward northerly regions with multiple functions and associations depending on geographical contexts. 137 The networking routes are said to have filled important roles for administrative purpos- es. They were hubs, semiautonomous regional centers, and clusters of cultic activi-  Here the Sanskrit/traditional textual sources include a wide array of normative texts (the śā s- tras , pertaining to general livelihood, morality , legality , and administrative matters), grammatical texts, and the itih ā sa-k ā vya tradition commonly considered related to the epic tradition.  MS 2. 21 – 22. The river Sarasvat ī has not been located, and it is mostly understood as part of a mythical geography . See also, Chattopadhyaya 2015, 23.  RE I from Maski refers to the subcontinent as J ambudvipa . Sircar points out that this could have been a reference to the A ś okan Empire, which in Dhauli RE V is called prithvi ( ‘ the entire earth ’ ). This may be an exaggeration, however. The concept of J ambudv ī pa is also seen in Buddhist Visuddhimagga , Vinaya texts, and the Pur ā ṇ as , which identify Jambudv ī pa as the region between the Himalayas and the southern sea (Law 1955, 8 – 9; Sircar [1979] 2000, 62, 66 – 68, 84; Thapar [1961] 2013, 26).  Jayaswal and Banerji 1929; Bhattacharyya 2009 , 21 – 22.  P ME 51; Ptol. Geog . 7 . 1. 8; 1. 7 . 6; 7 . 1. 85; Casson 1989 , 213; Chakravarti 2016, 187 .  Abraham 2003, 207 – 208, 212; 2008, 53.  Chandra 1953, 45 – 68.  For the commodities procured along the northern route, A uttarapathika , see Agrawal 1953, 244.  Neelis 2013, 14. Early Historic South Asia 113 ties. 138 It is also suggested that the network was used by both regional and foreign political units for expansion into northern India. 139 This interpretation is based on an understanding of the routes as part of an imperial communication network, de- scribed by Megasthenes as royal roads with pillars to mark distances, extending from Susa in Iran to Palibothra (Pataliputra, modern Patna in Bihar) under the Mauryas. 140 The two Aramaic edicts of A ś oka from Laghman mention the term KRPTY ( karapathi ), which is considered to mean ‘ royal roads. ’ 141 The KA explains “ t he east and the west trad e routes ” further by talking about “ the route to the H imavat ” (possibly a ref erence to the Uttar ā patha ) an d the r out e to th e so uth ( th e Dak ṣ i ṇ ā pa tha ). 142 The reference is made within a deb ate about which trade r oute is better. The KA emphas izes the benefits of trade to the south. The author expres ses a clear preference f or land routes, and here in particu lar wheel tracks rath er than footpaths. 143 Among differen t kinds of water pat hs, he prefers coastal and ri verine routes over the o pen sea. The influence of topograp hy on the routes is most pr ominent in peninsu lar India. Because of th e varied terrain and smaller s tretches of alluvi al soil, the ag ricultural tracts were relatively sma ller in comparison to t hose of the norther n plains. Yet t he uneven distribu tion of the rich mineral reso urces in the penins ular region necess itated an internal e xchange system. The K A explicitly refer s to the profitable trad e in minerals in the D eccan. 144 Sanskrit Buddhis t texts divide India into thr ee parts, Madhyade ś a , Uttar ā path a , and Dak ṣ i ṇ ā patha , whe re the two latter terms are not see n as routes but as re- gions. 145 The interaction of different ecological zones stands out clearly in the Sangam literature. The texts are strongly aware of the geographical and ecological divisions of the southern Indian region. Central to this perception is the notion of tinai . 146 The five tinais are first mentioned in T olkappiyam , the earliest extant Tamil text, where they form zones characterized by particular landscapes and occupations: the Kurinji (areas with hilly terrain), Mullai (pastures and woodlands), P alai (arid stretches), Neidal (littorals), and Marudam (agrarian tracts). Tinais also had their  Thapar 2003, 196; Fussman 1987 – 1988, 66 – 68; Neelis 2013, 14.  Neelis 2013, 14.  Megasthenes ad Strabo 15. 1. 11; 15. 1. 50; see also Neelis 2013; von Reden, ch. 10.B, this vol- ume.  Chakravarti 2017 , 309 . For the inscriptions, see Mukherjee 1984.  KA 7 . 12. 22 – 26. The route to the Himavat, or the H aimavatapatha , in Kautiliya ’ s Artha śā stra ( KA ) has been identified as the route from Balk to Taxila based on the items that are said to have been traded in the region, such as horses, woolen cloths, hides, and furs (Chandra 1977 , 5, 78, 79). For the Skt. text and translation of the KA, see Kangle (1960 – 65) 2014, vols. 1 and 2 respectively  KA 7 . 12. 13 – 26.  KA 7 . 12. 13 – 26.  Law 1955, 14. Based on the PME , De Romanis (2012) also discusses the possible evolution of the term Dak ṣ i ṇ ā patha from a hodonym (name of a road ) to a choronym (name of a region).  For discussion on Sangam literature, see Dwivedi, ch. 10.A, this volume. 114 Mamta Dwivedi own heroes and deities, and thus formed some kind of identity group. 147 Yet the precise meaning of the term tinai can only be conjectured, as the semantic field of the term ranges from ‘ space, ’ ‘ land, ’ and ‘ abode, ’ to ‘ genre, ’ ‘ genealogy , ’ and ‘ situation. ’ 148 Thus, some scholars take them as ecological zones that harbored dif- ferentiated economic activities. The littoral zone was associated with fishery and pearl diving, while agricultural groups that were always under the threat of attack by raiders from the P alai areas occupied the fertile M arudam . In this approach, the fertile agricultural tracts of Marudam created a prerequisite for state-like political formations, as particular ecological pockets had access to urban centers and the inland capitals of chieftains. 149 Other scholars regard them as just mythopoeic cate- gories, or semiotic tools for organizing a lyrical landscape with no background in a real world. 150 However, regardless of whether the tinais are considered as real eco- logical zones in southern India or as mere mythopoetic concepts, Sangam poetry provides us with an insight into how early writers conceptualized and represented regions, occupations, and polities in ecological terms. The knowledge of geography or place names of the southern region in northern Sanskrit texts does not indicate a gradually increasing awareness of southern geog- raphy in the post-Mauryan period. Rather, the Indo-Sri Lankan coast was already well known in pre-Mauryan times. 151 If we believe the Roman geographer Strabo, Sri Lanka (Taprobane in Greek ) was known to Alexander ’ s pilot Onesikritos when the Macedonian army was stationed near the river Indus. 152 It is also remarkable that early sources refer to Sri Lanka as Tamrapar ṇ ī . 153 The Tamrapar ṇ ī River (now Thamirabarani) near the coast of the southern peninsula may represent a close con- nection between the riverine and maritime routes to Sri Lanka, connecting coastal ports with inland riverine ports. 154 The early historical donative records, found throughout the subcontinent, are also an interesting source for understanding social, economic, and ecological con- nectivity . Donative texts on stone were not the result of any “ royal decree but con- structed through the generosity of the common man, by a process of collective do-  Gurukkal 1993, 7 .  Champakalakshmi 1996; Chakravarti 2016; Devadevan 2006, 200.  Champakalakshmi 1996, 28 – 32; Gurukkal 1995.  Devadevan 2006; Selby 2008.  Maloney 1970, 606; Ray 2003; 2008; Abraham 2003; 2008.  Strabo 15. 1. 14 – 15; Maloney 1970, 606. Sri Lanka was also known to Arrian (second century  ). On the knowledge of Sri Lankan natural phenomena by Arrian, see Ray 2003, 168 – 172.  A ś okan RE II and XIII, as discussed above, and the KA 2. 11. 1 – 2.  The similarity of archaeological assemblage even in the megalithic phase between South Indi- an and Sri Lankan sites, such as Adichanallur and Pomparippu, has often been suggested (Kennedy 1974, 24). Further, there have been studies about the cosmopolitan nature of Anuradhapura from the fourth century onward. It is situated in the North-Central province of the island and transformed from an Iron Age village into a metropolis due to extensive development of irrigation facilities (Coningham, Manuel, and Shoebridge 2017). Early Historic South Asia 115 nation that is attested by the masses. ” 155 The donations or gifts, referred to as d ā na , include anything from single railing pillars, cross bars, sculptures, and stone to images of the Buddha, Bodhisattvas, Jain T ī rthankaras, and Hindu deities. The do- nors often identified themselves with their and their ancestors ’ names, references to their home city or region, and their occupations. It is interesting that most of the donors were monks, nuns, and laypersons. 156 Donative records from Mathura most- ly refer to the occupations of the donors, such as perfume merchants, courtesans, blacksmiths, goldsmiths, ivory workers, performers, rich merchants, travelling mer- chants, and others. 157 Studies of the donative inscriptions from the Sanchi st ū pa have been used to identify kinship relations. The references to kinship ties of the donors along with references to their occupation and places of origin have been used to understand the complexity of identities as a characteristic of urban socie- ties. Individuals occupied many roles, vis-à-vis their religious, economic, political, and regional affiliations. 158 Similarly , donative inscriptions from areas near Anura- dhapura have been used to understand the limited role of kings in patronage of religious structures, water body management, and administration. Like in other parts of the subcontinent, Anuradhapura also demonstrates the importance of so- cial dynamics as opposed to state-driven developments. 159 II.  .  Foreigners in Loc al T exts References to outsiders are commonly found in Indic texts and are often identified with the term yavana . References to rulers like Antiochos of Syria, Ptolemy of Egypt, Antigonos of Macedonia, Megas of Cyrene, and Alexander of Epiros as yona- l ā ja in RE II and XIII have been mentioned above. In RE V , A ś oka includes yavanas among his subjects at the northwestern border. 160 RE XIII refers to kingdoms situated along his borders and includes the country of the yavanas . 161 The presence of yavanas in the northwest is also associated with the presence of the bilingual A ś okan edict at Kandahar in Greek and Aramaic. 162 Another piece of epigraphic evidence comes from Kh ā ravela ’ s Hathigumpha inscription dated to the second century  . Here, the king is eulogized for having defeated and pushed back a certain yavana-r ā ja  Dehejia 1992, 35.  U. Singh 1996.  V . L. Singh 2005; Bhattacharya 2008, 495 – 500. For the inscriptions from Mathura, see, Lüders 1912.  Basant 2009; U. Singh 1996.  Coningham 1995.  Sircar (1979) 2000, 15; 22 – 23; 31 – 33.  Ray 1988, 312.  Sircar (1979) 2000, 113. 116 Mamta Dwivedi out of R ā jag ṛ ha to Mathura. 163 There are, furthermore, records of donations made by yavanas in central and western India, for example, the records of a gift at Sanchi b ya yona living in Setapatha, of gifts at Karle by yavanas living in Dhenuk ā kata, by the V ī tasamgata from the unidentified place Umehanakata, and at Junnar by a yavana from Gata. 164 While most of the epigraphic records are post-A ś okan, references to yavanas are also found in pre-A ś okan Sanskrit grammatical works. 165 From around 500  onward, yavanas appear to have been included in the category of people who spoke a mlecchas bh ā ṣ ā . 166 The areas where these people lived were designated as mleccha de ś a and included frontier zones such as the countries where the yavanas and k ā m- bojas lived. Subsequently , rich mythologies emerged, giving the yavanas a local northern Indian origin. 167 In Tamil texts, references to Graeco-Roman outsiders or strangers are attested rather late. The term yavanar seems to have been triply imported. It is the Tamilized form of yavana deriving from the Sanskritized version of ‘ Ionian, ’ in turn entering Sanskrit/Prakrit through the Old Persian term Y auna denoting Ionian Greeks, who were conquered by Cyrus in the sixth century  . 168 Originally denoting Greeks or Romans, the word came to refer to any unfamiliar being, foreigner, or stranger. In the context of Tamil literature it occurs for the first time in the P uran ā n ū ru (ca. 350  ) where the yavanars are described as wine merchants coming in boats. It is interesting that while Greek and Roman trade is attested in earlier texts, the term yavanar enters Tamil literature much later, though with no conspicuous description. It comes “ very much in passing and almost offhand, as if the ‘ Greeks ’ were a part of the ordinary daily existence. ” 169 II.  .  Inland Networks Indic connections to the north and beyond the Himalayas are clear in the material and architectural remains. This is often studied in connection to the Gandh ā ran material culture and spread of Buddhism. The expansion of Indic traditions beyond the Indus is associated with the legendary account of colonization of Khotan by Kun ā la, a son of A ś oka. By the fourth century  Khotan had become a center for  Jayaswal and Banerji 1929 .  Ray 1988, 314 – 315.  Karttunen 2015, 42.  The term mleccha refers to ‘ non- V edic, ’ ‘ barbaric, ’ ‘ non- ā rya , ’ an outcast, or a foreigner. Mlec- cha bh ā ṣ ā means the language of the mlecchas .  Ray 1988, 321 – 322.  Ray 1988, 312; Selby 2008, 82.  Selby 2008, 83. Early Historic South Asia 117 the diffusion of northwestern Prakrit. 170 The accounts of a Chinese Buddhist monk, Faxian (fourth to fifth century  ) refers to the Gomat ī vih ā ra of Khotan, famous for housing 3,000 monks, where Sanskrit manuscripts of medical knowledge and other types of texts were reproduced and translated. 171 The Kharo ṣṭ h ī documents from Niya also refer to a monetary denomination of masa (Skt. m ā ṣ aka ) and other San- skritized names in the context of Buddhist vih ā ras that point to a prolonged connec- tion with Indic traditions. 172 Routes through the Swat valley are considered the most common and conven- tional connections between the Indian subcontinent and Central Asia. This region is considered as a corridor toward the Indus and then into the mainland for contact and invasions. Other than this corridor, scholars point to another overland route along the Karakorum highway connecting Kashmir with Kashgar, including the sites of Gilgit, Chilas, Skardu, etc., based on the rock engravings and epigraphic material. 173 This route is identified as the Jibin (Kashmir) route in the Chinese sour- ces of the Han period. It was suggested this route would be a shorter way to directly connect Kashgar to the subcontinent in comparison to the route from Kashgar to Kabul and then to the Peshawar- Taxila region. 174 Material remains in archaeological contexts have also been used as evidence for interregional connections within mainland India. Raw materials as well as finished products of agate, steatite, carnelian, ivory , gold, copper, and iron have been used to chart inland communication between northern and southern regions. 175 Archaeo- logical assemblages of various megalithic sites contain material remains that have been used to identify connectivity between inland and coastal areas, so much so that recent studies have questioned the concept of ‘ hinterland ’ and ‘ foreland. ’ 176 The distribution of Buddhist sites and their contextualization in their respective geo- graphic settings has also allowed the identification of different passes and arterial routes (map 1). 177  Sastri and Srinivasachari 1970, 229 .  Sastri and Srinivasachari 1970, 229 .  Hansen 2017 , 83 – 93.  Inscriptions in Kharo ṣṭ h ī , Br ā hm ī , Chinese and Sogdian have been reported along with rock engravings of stupa worship and horses brought for trade (Jettmar 1989; Neelis 2000; Chakravarti 2017 , 311 – 314).  Chakravarti 2017 , 313. For a discussion on problems of identification of such toponyms as mod- ern regions, see Morris, ch. 9 , sec. II.3, this volume.  Lahiri (1992) has illustrated the distribution of various raw materials in different parts of the subcontinent. She explains the possible distributive networks from the Bronze age up to 200  . Chakrabarti and Lahiri (1996) charted out the distribution of copper sources as well as their find spots in the subcontinent, which allows one to identify channels of communication.  Bauer 2016.  Ray 1994a; 1994b; Chakrabarti 2005; 2010b; Neelis 2011. 118 Mamta Dwivedi II.  .  Maritime Networks South Asia has a long history of maritime connections to both western and eastern regions of the Indian Ocean. 178 The transmission of flora and fauna from the African continent and Southeast Asia dates back to the second and first millennia  . 179 Maritime networks involving coastal centers in Gujarat (western India) date back to the third millennium  , while Tamil maritime connections can be traced to the first millennium  . A third region of maritime activity was that of the Bay of Ben- gal, although it contrasts with the other two as having more extensive inland and riverine connections. 180 Here, the location of coastal ports may not have responded to maritime trade, but rather to the location of resources in the hinterland. 181 The connection of regional coastal routes with the hinterland via river is confirmed by the distribution of rouletted ware (RW) in Bangladesh and West Bengal. 182 In the Indian context, trade and exchange are often part of the ritual and social roles of communities and not the consequence of settled agricultural life associated with surplus production and state formation. 183 Trading activity , including seafaring in the Indian Ocean and Arabian Sea, must be dissociated from the question of state formation and the great empires as prime movers of trade. 184 If one considers that the vast majority of produce shipped between and through Indian ports consisted of local subsistence products rather than luxuries, the role of small-scale entrepre- neurs in multitude comes into view . 185 The vast majority of pottery found in coastal sites along the Indian littoral is now identified as being of local origin, although many varieties of coarse ware are still difficult to place precisely . 186 South Asia ’ s role in the maritime networks of the western Indian Ocean is usual- ly studied in connection with trade in the Arabian Sea, Persian Gulf, and the Red Sea. 187 The Periplus Maris Erythraei (mid-first century  ), with its extensive record of travel and navigation through ancient ports, is one of the most commonly used  Among the large amount of literature devoted to Indian Ocean trade, see the most recent volumes by De Romanis and Maiuro (2015); Mathew (2017); Cobb 2018.  Fuller et al. 2011.  Ray 2003, 20.  Most of the early western sources refer to the already existing ports. There is hardly any indica- tion that the activities of the traders either established or enhanced the trading stations (Ray 2003, 23 – 24).  Schenk 2006; Tomber 2008, 45 – 46.  For discussion see, A. Ghosh 1973; Ray 1994a.  For a discussion on internal and external impetuses for the development of trade in India and urbanization and secondary state formation, see Chakravarti 2017 , 333 – 338.  Ray 2003 82 – 125; Fuller et al. 2011. The non-luxury items may include food stuffs, inexpensive textiles, spices, medicines, and ritual commodities (Seland 2014, 386).  On the misattribution of conical amphorae and rouletted ware in particular, see Tomber 2008, 44 – 45. See Dwivedi, ch. 10.A, this volume for the history of rouletted ware.  Mathew 2017 ; De Romanis and Maiuro 2015; Seland 2014, 368. Early Historic South Asia 119 sources for the study of maritime activity in this region. 188 As a result, port sites on the western coast of the mainland and in Sri Lanka have been studied through the lens of Roman trade with India, most prominently at the sites of Arikamedu and Pattanam (alleged Muziris). More recently , Roman and Indian traders have been considered as just two players active along the long-established routes of trade and exchange from East Africa to Sri Lanka via Arabia, southern Iran, and the Indian subcontinent. 189 Recent archaeology focuses more carefully on the identification and analysis of South Asian artifacts at sites across the Arabian Sea. One example is the discovery of Indian pottery dating back to centuries long before the Common Era from sites in southern Arabia (Khor Rori), the Arabian Gulf (Mleiha), and the Red Sea (Berenike and Myos Hormos). 190 Finds of grain from Mleiha and ed-Durr in Umm al Qaiwain, early Indian coins of Ujjain and and the S ā tav ā hanas, and inscrip- tions in Tamil-Br ā hm ī script are also important markers of connectivity . 191 Indian pottery found in the port towns on the Red Sea coast are now considered as relating to Indian communities settling there, rather than as representing Roman imports of ceramics or their contents. 192 This is particularly true for rouletted and other fine wares that are represented in small numbers (ca. 20 vessels) in Myos Hormos and Berenike. 193 Southeast Asian connections with South Asia based on the similarity of religio- political concepts has been a topic of discussion since the colonial period. The ‘ Indi- anization ’ or ‘ Indicization ’ of Southeast Asia, including the spread of Buddhism, was considered a form of colonization undertaken by the South Asian empires and states, especially during the first millennium  . 194 However, archaeological, epi- graphical, and socio-anthropological research has pointed to maritime connections with Southeast Asia as early as the second millennium  . Furthermore, shared seafaring technology , suggesting that the development of boat-building techniques in India were influenced by Southeast Asian practices, indicates transfers of knowl- edge. 195 Another type of shared knowledge was that of metallurgy . Bimetallic arti- facts of bronze and iron from sites in South Asia, east Java, and Vietnam date back to the first millennium  , and are still evident in the beginning of the Common Era. Glass, pottery , and carnelian beads also appear as common remains linking South and Southeast Asia over long periods of time. 196  Ray 2011; Seland 2014.  Fitzpatrick 2011, 30.  Reddy 2016, 55 – 68.  Reddy 2016, 71 – 72; Haerinck 1998, 293 – 296.  Tomber 2008, 74; Thomas 2012, 180.  Tomber 2008,74.  For an overview of the argument and criticisms, see Mabbet 1977 ; M. Smith 1999; Winzeler 1981, 459 – 466; Ray 1996, 422 – 431.  Fuller et al. 2011, 551 – 553.  Ray 2003, 120 – 123. 126 Mamta Dwivedi – . 2011. “ Writings on the maritime history of ancient India. ” In S. Bhattacharya (ed.), Approaches to history: Essays in Indian historiograph y , 27 – 54. Delhi: Primus. Ray, H. P ., and S. Mishra. 2018. “ Introduction. ” In Bibliography on Sailing to S uvarnabhumi . AIC -RIS. http:/ /ris.org.in/aic/bibliography. Ray, H. P ., and D . T . Potts, eds. 2007 . Memory as history: The legacy of Alexander in Asia . New Delhi: Aryan Books International. Ray, H. P ., and C. M. Sinopoli, eds. 2004. Archaeology as history in early S outh Asia. New Delhi: Aryan Books International. Raychaudhuri, H. (1923) 1972. Political history of ancient India: From the accession of P arikshit to the extinction of the Gupta dynasty . 7th ed. Calcutta: University of C alcutta. Raychaudhuri, S. P ., and M. Roy. 1993. Agriculture in ancient India: A report . New Delhi: Indian C ouncil of Agricultural Research. Reddy, A. L. 2016. “ Archaeology of Indo-Gulf relations in the early historic period: The ceramic evidence. ” In H. P . Ray (ed.), Bridging the Gulf: Maritime cultural heritage of the western Indian Ocean , 53 – 78. New Delhi: Manohar. Sahu, B. P . 2001. “ Brahmanical ideology, regional identities and the construction of early India. ” Social Scientist 29.7/8 , 3 – 18. Sastri, H. 1925 – 1926. “ Brahmi inscription on a wooden pillar from Kirari. ” Epigraphia Indica 18, 152 – 157 . Sastri, N. K. A. (1955) 1995. A history of South India: From prehistoric times to the fall of Vijayanagara . 4th ed. Delhi: O xford University Press. Sastri, N. K. A., and G. Srinivasachari. 1970. Advanced history of India . Bombay: Allied Publishers. Schenk, H. 2006. “ The dating and historical value of rouletted ware. ” Zeitschrift für Archäologie außereuropäische Kulturen 1, 59 – 90. Seland, E. H. 2014. “ Archaeology of trade in the western Indian Ocean, 300 BC – AD 700. ” Journal of Archaeological Research 22.2, 367 – 402. Selby, M. A. 2008. “ Representations of the foreign in classical T amil literature. ” In G. R. Parker and C. M. Sinopoli (eds.), Ancient India in its wider world , 79 – 90. Ann Arbor, MI: University of Michigan. Seneviratne, S. 1981. “ Kalinga and Andhra: The process of secondary state formation in early India. ” In H. J. M. Claessen and P . Skalnik (eds.), The study of the state , 317 – 338. The Hague: Mouton. Shrimali, K. M. 1985. History of the Pañc ā la . V ol. 1, A study . New Delhi: Munshiram Manoharlal. Sims-W illiams, N., and J. Cribb. 1995. “ A new B actrian inscription of Kanishka the Great. ” S ilk Road Art and Archaeology 4, 76 – 142. Singh, U. 1996. “ Sanchi: The history of the patronage of an ancient Buddhist establishment. ” Indian Economic and Social History Review 33.1, 1 – 35. – . 2008. A history of ancient and early medieval India: From the stone age to the 12th century . Delhi: Pearson Longman. Singh, V . L. 2005. Mathur ā : The settlement pattern and cultural profile of an early historical city . New Delhi: Sundeep Prakashan. Singhvi, A. K., and R. Krishnan. 2014. “ Past and the present climate of India. ” In V . S. Kale (ed.), Landscapes and landforms of India , 15 – 23. New Y ork, NY: Springer. Sircar, D . C. 1965. Select inscriptions bearing on Indian history and civilization . V ol. 1, From the 6th century BC to the 6th century AD. 2nd ed. Delhi: Motilal B anarsidass. – . (1979) 2000. A ś okan studies . C alcutta: Indian Museum. Smith, M. 1999. “‘ Indianization ’ from the Indian point of view: T rade and cultural contacts with Southeast Asia in the early first millennium  . ” Journal of the Economic and Social History of the Orient 42.1, 1 – 26. Early Historic South Asia 127 – . 2005. “ Networks, territories, and the cartography of ancient states. ” Annals of the Association of American Geographers 95.4, 832 – 849. Smith, V . A . (1901) 1920. Asoka: The Buddhist emperor of India . 3rd ed. Oxford: C larendon. Snead, R. E. 2010a. “ Bangladesh. ” In E. C. F . Bird (ed.), Encyclopedia of the world ’ s coastal landforms . V ol. 1, 1077 – 1080. New Y ork, NY: Springer. – . 2010b. “ Pakistan. ” In E. C. F . Bird (ed.), Encyclopedia of the world ’ s coastal landforms . V ol. 1, 1059 – 1064. New Y ork, NY: Springer. Spate, O . H. K., and A. T . A . Learmonth. (1954) 2017 . India and Pakistan: A general and regional geography . London: Routledge. Srinivasan, T . M. 2016. “ Agricultural practices as gleaned from the T amil literature of the Sangam age. ” Indian Journal of History of Science 51.2, 167 – 189. Subbarayalu, Y . 2014. “ Post- Satavahana and post- Sangam polities. ” In N. Karashima (ed.), A concise history of South India: Issues and interpretations , 55 – 63. New Delhi: O xford University Press. Subramanian, N. 1966. Sangam polity: The administration and social life of the Sangam T amils . Bombay: Asia Publishing House. Swan, B. 2010. “ Sri Lanka. ” In E. C. F . Bird (ed.), Encyclopedia of the world ’ s coastal landforms . V ol. 1, 1071 – 1075. New Y ork, NY: Springer. T andon, S. K., P . P . Chakraborty, V . Singh, and V . S. Kale. 2014. “ Geographical and tectonic framework of India: Providing context to geomorphologic development. ” In Landscapes and landforms of India , 3 – 14. New Y ork, NY: Springer. Thapar, R. 1978. “ The scope and significance of regional history. ” In R. Thapar (ed.), Ancient Indian social history: Some interpretation , 361 – 376. New Delhi: Orient Longman. – . 2003. The Penguin history of Early India: From the origins to AD 1300 . London: Penguin. – . 2006. “ The Mauryan Empire in early India. ” Historical Research 79.205, 287 – 30 5. – . (1961) 2013. A ś oka and the decline of the Mauryas . 3rd ed. London: O xford University Press. Thomas, R. 2012. “ Port communities and Erythraean sea trade. ” British M useum Studies in Ancient Egypt and S udan 18, 169 – 199. T omber, R. 2008. Indo-Roman trade: From pots to pepper . London: Duckworth. T rautmann, T . R. 2009. “ Elephants and the Mauryas. ” In The clash of chronologies: Ancient India in the modern world , 229 – 254. Delhi: Y oda Press. Winzeler, R. L. 1981. “ The study of the Southeast Asian state. ” In H. J. M. C laessen and P . Skalnik (eds.), The study of the state , 455 – 467 . The Hague: Mouton. Xue, Y ., and M. Y anai. 2005. “ Asia, climate of south. ” In J. E. Oliver (ed.) Encyclopedia of world climatology , 115 – 120. New Y ork, NY: Springer. Kathrin L eese-Messing 4 The Qin and Han Empires I Introduction The early Chinese imperial period comprises the reigns of the short-lived Qin 秦 dynasty (221 – 207  ) and the subsequent, long-lasting Han 漢 dynasty (206  – 220  ). The early Chinese empire ranks among the largest empires of its time. It was comparable to the Roman Empire in terms of its peak population (up to 60 mil- lion) 1 and territorial size (up to 6.5 million sq. km). 2 One distinguishing feature of its spatial dimensions is that following a short period of rapid expansion, the empire under Qin rule already comprised large parts of what would be the peak of the Han Empire ’ s territorial dimensions, which the latter then kept 3 (and, during certain phases, further extended ) until the early third century  . The area that came to be the capital area of both the Qin and Former Han Em- pires is traditionally referred to as Guanzhong 關 中 , the ‘ area within the passes ’ (map 1). With the fertile lower Wei 渭 River valley at its core, the area is ringed by hills and mountains and was therefore accessible only through a number of passes. From here, several dynasties would unify their empires, and it served as the capital area for 12 dynasties from the Zhou 周 (ca. 1045 – 770  ) up to the Tang 唐 (618 – 907  ) period. A traveler leaving the Guanzhong plains toward the east, where the Yellow Riv- er (Huanghe 黃 河 ) cuts through the mountains at the Hangu 函 谷 Pass, reaches the westernmost edges of the North China Plain. 4 With the Yellow River at its center, it constitutes China ’ s largest alluvial plain. With its fertile loess soil, but only a moder- ate amount of precipitation and a relatively short growing season, the North China Plain is particularly suitable for growing drought-resistant and quickly maturing crops like millet and soy beans. The area was home to many prehistoric cultures  Bielenstein 1947 ; Nishijima 1986, 595 – 596.  Turchin 2009 , 202; Taagepera 1979 , 128. These figures also include, for instance, the territory of the Tarim Basin (comprising about 1 million sq. km), in which Han control definitely was of a much more limited scale than in the empire ’ s more central regions. Furthermore, due to its dry climate, only a small proportion of the vast Tarim Basin was populated.  The Han initially controlled a smaller territory than the Qin Empire at its peak. They then recon- quered the lost territory and eventually went beyond their predecessor ’ s ambitions. See section IV .4 below .  The Later Han dynasty would eventually move its capital to the city of Luoyang 洛 陽 situated in this region (in the west of modern Henan Province). Note: I would like to thank Armin Selbitschka and T sang Wing Ma for valuable comments on an earlier draft of this chapter. Open Access. © 2020 Kathrin Leese-Messing, published by De Gruyter. This work is licensed under the Creative C ommons Attribution-NonCommercial-NoDerivatives 4.0 License. https:/ /doi.org/10.15 15/97831106077 41-006 130 Kathrin Leese-Messing Map 1: The Han Empire and its regional setting. Extent of the empire ’ s regions, commanderies, and kingdoms outlined in orange; Western R egions protectorate outlined in green. © Peter P alm. The Qin and Han Empires 131 and to the rulers of the Shang 商 (seventeenth to eleventh centuries  ), known as China ’ s first historical dynasty . The North China Plain is therefore often credited with being the cradle of Chinese civilization, although archaeological findings in other areas have put this myth of a single cultural origin into question. 5 Toward the north and northwest, the Yan 燕 and Taihang 太 行 Mountains sepa- rate the North China Plain from areas characterized by steppe, desert, and forest landscapes. During the phase of major expansion in the decades around 100  , the Han Empire expanded its territory northwards to these areas of limited agricul- tural potential. In the northeast, the empire extended as far as the northern parts of the Korean peninsula. In the north and northwest, the Han got hold of the Ordos, an area enclosed by a large rectangular bend of the Yellow River (the Ordos Loop), and of the so-called Hexi 河 西 corridor. The latter is a depression with a width of less than 80 km and a length of over 960 km, with the Qilian 祁 連 Mountains and the Altun Shan to its south, and the Gobi Desert to its north. 6 The conquest of the Hexi corridor further provided direct access to the Tarim Basin (in modern Xinjiang 新 疆 Uyghur Autonomous Region), an endorheic basin dominated by the Taklama- kan Desert at its center. Intermittently , the oasis states surrounding the desert areas encountered varying degrees of Han control. On the southern edges of the North China Plain, the Qinling 秦 嶺 Mountains (in the west ) and the Huai 淮 River (in the east ) demarcate the separation between the areas of the Yellow River and the Yangtze (Changjiang 長 江 ) River Basins. The latter basin was accessible from the Guanzhong capital area via its southeastern gate, the Wu 武 Pass. The Yangtze River Basin enjoys favorable agricultural conditions, with a climate characterized by abundant rainfall and moderate temperatures that enable a long growing season. Paddy field rice, in combination with beans, vegetables, and fruits, constituted the most important agricultural products of this area. During ear- ly imperial times, the land south of the Yangtze River was a region that northern people preferably avoided even though centuries later it evolved into the political, economic, and cultural center of China. The vast majority of the early empires ’ pop- ulation lived north of the Yangtze River. The area to its south was often seen as a place of exile, characterized by excessive heat and wetness, disease, and poisonous air. Even though militarily occupied, particularly the subtropical and mountainous area south of the Nanling 南 嶺 range including the Pearl River (Zhujiang 珠 江 ) Basin remained largely outside the cultural sphere of the Chinese heartland. To the southwest of the ‘ area within the passes ’ lies an area with considerable geo-strategical importance for the formation of the Chinese empires: the fertile Si- chuan 四 川 Basin. Surrounded by mountains on all sides, it initially constituted a rather remote area from the outward perspective of the so-called Chinese heartland, even though it had been the home of a highly developed Bronze Age civilization  V on Falkenhausen 1995.  On the geography of the Hexi corridor, see Chang 2007b, 3 – 4. 132 Kathrin Leese-Messing contemporaneous to the Shang. 7 During the fourth century  , it was made acces- sible from the Guanzhong area via a network of newly opened mountain roads and evolved into a breadbasket for the forming empire. From a bird ’ s-eye view , the larger core of the early Chinese empires (including the Guanzhong area, the North China Plain, the Yangtze River Basin, and the Sich- uan Basin), is a relatively segregated part of the world. It is surrounded by the high mountains bordering the Qinghai- Tibetan Plateau in the west, by steppe and deserts in the north and northwest, the Pacific Ocean in the east, and a subtropical, moun- tainous zone to the south. Although these buffer zones limited potential contacts to other centers of civilization in Inner Asia, the Near East, South Asia, and Europe to a certain degree, even in prehistoric times the area that is now central China was never isolated from the rest of the world. Long before the Han Empire ’ s ‘ penetration to the Western Regions ’ 通 西 域 8 (via the Hexi corridor into the Tarim Basin and beyond ) and the alleged beginning of Silk Road trade, archaeological evidence testi- fies to long-range intercultural exchange with far-reaching historical impacts: Wheat was introduced to northern China from Western Asia during the mid-fifth millennium  , the westward export of Asian millets from China reached Greece by the third millennium  , and via the Eurasian steppe, the chariot found its way from Western Asia to the North China Plain, where it was adapted by the Shang at around 1200  . 9 With the expansion of the early Chinese empires, contacts and exchange intensified in all directions. The history of the Qin and Han Empires is inapprehensible without taking into account the historical foundations that these empires were built upon. Section II will therefore introduce certain political, institutional, and cultural developments that characterized the pre-imperial polities that had coexisted in the Yellow River and Yangtze River Basins prior to 221  . Thereafter, section III will provide a large- ly chronological account of the period comprising the rise of the state of Qin, the unification, and the short rule of the Qin Empire, before offering a brief chronologi- cal overview of the long-lasting Han period. Finally , section IV will focus on several characteristic institutions of the Qin and Han Empires.  On the Sanxingdui 三 星 堆 culture in Sichuan, see Bagley 2001.  This expression is frequently used in Ban Gu ’ s 班 固 (32 – 92  ) Documents of the Han ( Hanshu 漢 書 ) with regard to Emperor Wu 武 (r. 141 – 87  ) of Han and his famous envoy Zhang Qian 張 騫 (195 – 114  ).  Lightfoot, Liu, and Jonse 2013; Bestel et al. 2014; Barbieri-Low 2000. The Qin and Han Empires 133 II From a Dec entralized Network to a B ureauc ratic Empire II.  Before Empire: P olitica l Fragmentation versus C ultural C onvergence during the E astern Z hou Period (770 – 25 5  ) By 260  , the large core area of the later Chinese empires consisted of seven major (and some minor) independent states, whose power and territorial extensions were partly comparable to each other (map 2). Only 40 years later, the whole area was united under the rule of the first Chinese emperor. This fast political unification calls for an explanation that can only be provided by a look at some historical devel- opments characterizing the pre-imperial period. The centuries preceding the political unification of the Chinese heartland are known as the period of the Eastern Zhou 東 周 (770 – 255  ). As received literature would make us believe, the Zhou had provided for political stability and moral in- tegrity in their extensive realm before being forced by a neighboring nomadic peo- ple to leave their old power base in the Wei River valley in the eighth century  . The larger Zhou realm was a territorially decentralized network of individual region- al domains, based on a hierarchy of personal and kinship relations that centered on the person of the king. 10 After the involuntary eastward move of their capital to Wangcheng 王 城 (near later Luoyang) in 771  , the Zhou rulers found themselves progressively losing their authority over the other regional lords. 11 The era of the Eastern Zhou up to the Qin unification is traditionally characterized as a phase of gradual deterioration of the political-ritual order of old. But in fact, the Eastern Zhou period witnessed an unprecedented coalescence of the region nominally still under Zhou rule. 12 It is arguably with reference to this period that the designation ‘ China ’ (as a cultural unit ) can be used as a meaningful term in the first place. 13 During the Eastern Zhou period, which is traditionally divided into the Spring and Autumn (Chunqiu 春 秋 , 722 – 481  ) and the Warring States (Zhanguo 戰 國 , 475 – 221  ) periods, this trend was reinforced. Territorial expansion of the individ- ual polities resulted in increasing military and cultural contacts between them. Mili-  Khayutina 2010.  On the Western Zhou and their gradual loss of power, see Li 2006.  The roots of this cultural convergence can be traced to the ‘ ritual revolution ’ during the ninth century  . As archaeological evidence has brought to light, the city-states had standardized their systems of aristocratic display during this period, e.g., by unitizing the number of particular vessels admitted as burial objects according to rank. Rawson 1999; von Falkenhausen 1999 .  The name ‘ China ’ (or any of its linguistic variants) did not exist at that time. Nor was it ever used by any imperial Chinese polity as an autonym. The exonym is most commonly believed to have been derived from ‘ Qin, ’ the name of the first imperial dynasty . For one of the alternative suggestions, see Wade 2009 . Chinese polities employed their dynastic names or generic terms such as Zhongguo 中 國 , the ‘ Central State(s), ’ as self-references. 134 Kathrin Leese-Messing Map 2: The Warring States in 260  and the subsequent extension of the Qin Empire. © Peter P alm. tary confrontations led to drastic changes in the political landscape. While far more than one hundred smaller city-states coexisted under formal Zhou sovereignty in the early Chunqiu period, the number had already decreased to around 20 by the early Warring States era. By 260  , only seven major states were left – albeit much larger and territorially bounded. Amidst the enormous bloodshed caused by these states ’ incessant warfare, philosophical notions of universal, transcending forces and the idea of political unity were born. 14 One of these states was the state of Qin 秦 , situated at the western margins of the former Zhou ecumene, which finally brought about the first political unification of this vast area under imperial rule.  Pines 2000. The Qin and Han Empires 135 II.  T oward Empire: General Developments of the W arring States Most contemporaries would not have regarded the Qin victory as predictable even only a few decades before the actual unification took place. 15 The unification was the result of a complex interlocking of certain economic and demographic develop- ments, which served as an important historical foundation of other drastic changes that affected all of the Warring States. From the fourth century  onward, the emergence of the iron industry made field work much more effective, encouraged the use of draft animals, the construc- tion of large-scale irrigation systems, and extensive deforestation for the reclama- tion of new farmland. The ensuing increase in agricultural production enhanced trading activities and economic prosperity . An immense growth of population pre- cipitated both interior colonization and exterior expansion of the states. 16 Demographic growth had manifold implications on Warring States societies, for example with regard to military organization. A development that had started dur- ing the Spring and Autumn period with a gradual extension of military service from the nobility and its dependents to the entire population 17 was now brought to its completion by the new disposability of mass armies. Small formations led by aristo- cratic warriors, whose ‘ ritual violence ’ had characterized the wars up to the Spring and Autumn period, 18 were now replaced by infantry and cavalry forces comprising hundreds of thousands of peasant levies. They were fighting with technically im- proved swords and halberds partly made of iron, and with the newly invented and highly effective crossbow . 19 These trends precipitated a new mode of warfare char- acterized by large-scale campaigns, long-term sieges, extreme rationalization, and devastating numbers of casualties. It has been argued that the degree of militariza- tion and severity of warfare in Warring States China were higher than anywhere else in the ancient world. 20 These military changes also heavily affected late Warring States societies in general. Military merit became a primary source of social advancement, which meant new potential for social mobility for the lower levels of society while simulta- neously disintegrating the privileges of the old aristocracy . 21 Experience with the  On the changing perception of Qin during the Warring States period, see Pines et al. 2014, 5 – 6; Pines 2013a.  For the development of ironwork in pre-imperial China, see Wagner 1993; 2008, 83 – 170; von Falkenhausen 2006, 409 – 410. For the enhancement in agricultural production, see Zhou 2010, 192 – 204, 276 – 304; Li 2009; Bray 1984, 130 – 240. For demographic growth, see Ge 2005, 291 – 300; von Falkenhausen 2006, 244 – 288.  Lewis 2007 , 30.  Lewis 1990, 15 – 52.  On the crossbow , see Yates 1994, 120 – 184; Major and Cook 2017 , 151.  Kiser and Cai 2003, 522 – 526.  An intriguing demonstration of this social mobility by military merit can be seen in an excavat- ed letter sent from the front by two Qin soldiers shortly before the unification. The soldiers ask if 238 Lara Fabian – . 2016c. “ V ologases I, Pakoros II and Artabanos III: C oins and Parthian history. ” Iranica Antiqua 51, 215 – 233. Pappalardo, E. 2010. Nisa P artica: I rhyta ellenistici . Florence: Le lettere. Perevalov, S. M. 2006. “ Аланский набег 136 г . н . э . в страны Закавказья : проблемные вопросы ” [The Alan raid in 136  into the countries of the T ranscaucasus: Problematic issues]. In Античная цивилизация и варвары , 318 – 355. Москва : Наука . Pilipko, V . N. 1989. “ К вопросу о локализации Парфавнисы ” [On the question of the localization of Parthian Nisa]. Известия Академии наук Туркменской ССР . Серия общественных наук , no. 2, 17 – 25. – . 2001. Старая Ниса : Основные итоги археологического изучения в советский период [ Old Nisa: The principal results of archaeological research in the Soviet period ]. Москва : Наука . – . 2008. “ The central ensemble of the fortress Mihrdatkirt: Layout and chronology. ” Parthica 10 , 35 – 51. Posch, W . 1998. “ C hinesische Quellen zu den Parthern. ” In W iesehöfer (1998), 355 – 364. Potts, D . T . 1999. The archaeology of E lam: Formation and transformation of an ancient Iranian state . Cambridge: C ambridge University Press. – . 2002. “ Five episodes in the history of Elymais, 145 – 124 BC: New data from the astronomical diaries. ” In P . Huyse (ed.), Iran: Questions et connaissances: Actes du IV e congrès européen des études iraniennes oraganisé par la Societas Iranologica Europaea, P aris, 6 – 10 Septembre 1999 . V ol. 1, 349 – 362. Paris: Association pour l ’ avancement des études iraniennes. – . 2007 [2012]. “ Cataphractus and Kam ā nd ā r : Some thoughts on the dynamic evolution of heavy cavalry and mounted archers in Iran and C entral Asia. ” Bulletin of the Asia Institute 21, 149 – 158. Pugachenkova, G. A. 1971. Скульптура Халчаяна [ The sculpture of Khalchaiana ]. Москва : Искусство . Rajak, T . 1998. “ The Parthians in Josephus. ” In W iesehöfer (1998), 309 – 324. Rapp, S. H. 2009. “ The Iranian heritage of Georgia: Breathing new life into the pre-Bagratid historiographical tradition. ” Iranica Antiqua 44, 645 – 692. Rezakhani, K. 2010. “ The road that never was: The Silk R oad and trans-Eurasian exchange. ” Comparative Studies of S outh Asia, Africa and the M iddle East 30.3, 420 – 433. Rose, C. B. 2005. “ The P arthians in Augustan Rome. ” American Journal of Archaeology 109.1, 21 – 75. Rostovtzeff, M. I. 1935. Dura and the problem of Parthian art . New Haven, C T : Y ale University Press. Russel, J. R. 1987 . Zoroastrianism in Armenia . C ambridge, MA: Harvard University Press. Sauer, E. W ., H. O. R ekavandi, T . Wilkinson, and J. Nokandeh. 2012. “ Innovation at P ersia ’ s frontiers: Sasanian campaign bases and defensive barriers. ” In L. V agalinski and N. Sharankov (eds.), Limes XXII: Proceedings of the 22nd International Congress of Roman Frontier Studies, Ruse, Bulgaria, September 2012 , 327 – 331. Sofia: National Archaeological Institute with Museum. Schlude, J. M., and A. J. Overman. 2017 . “ Herod the Great: A Near Eastern case study in Roman- Parthian politics. ” In J. M. Schlude and B. R eubin (eds.), Arsacids, Romans and local elites: Cross-cultural interactions of the Parthian E mpire , 93 – 110. O xford: Oxbow. Schlumberger, D . 1970. L ’ Orient hellénisé: L ’ art grec et ses héritiers dans l ’ Asie non méditerranéenne . Paris: Albin Michel. Schuol, M. 2000. Die Charakene: Ein mesopotamisches K önigreich in hellenistisch-parthischer Zeit . Stuttgart: Franz Steiner. – . 2017 . “ Isidor von Charax und die literarische G attung der stathmoi . ” In Wiesehöfer and Müller (2017), 71 – 86. The Arsakid Empire 239 Seland, E. H. 2011. “ The Persian Gulf or the R ed Sea? T wo axes in ancient Indian Ocean trade, where to go and why. ” World Archaeology 43.3, 398 – 409. – . 2014. “ The organisation of the Palmyrene caravan trade. ” Ancient West and East 13, 197 – 211. Sellwood, D . 1980. An introduction to the coinage of P arthia . London: Spink. Shayegan, M. R. 2011. Arsacids and Sasanians: Political ideology in post-Hellenistic and Late Antique Persia . C ambridge: C ambridge University Press. – . 2017 . “ Persianism: Or Achaemenid remembrances in the Iranian and Iranicate world(s) of antiquity. ” In R. Strootman and M. J. V ersluys (eds.), Persianism in Antiquity , 401 – 455. Stuttgart: Franz Steiner. Simonetta, A. M. 2001. “ A proposed revision of the attributions of the Parthian coins struck during the so-called ‘ Dark Age ’ and its historical significance. ” East and W est 51.1 – 2, 69 – 108. – . 2009. “ The coinage of the so-called Parthian ‘ Dark Age ’ revisited. ” E lectrum 15, 141 – 194. Sinisi, F . 2014. “ Sources for the history of art of the Parthian period: Arsacid coinage as evidence for continuity of imperial art in Iran. ” Parthica 16, 9 – 60. Strootman, R. 2017 . “ Imperial Persianism: Seleukids, Arsakids and Fratarak ā . ” In R. Strootman and M. J. V ersluys (eds.), Persianism in antiquity , 177 – 200. Stuttgart: Franz Steiner. Strugnell, E. 2008. “ Thea Musa, R oman Queen of Parthia. ” Iranica Antiqua 43, 275 – 298 . V an der Spek, R. J. 1998. “ C uneiform documents on Parthian history: The R ahimesu archive. ” In Wiesehöfer (1998), 205 – 258 . V ardanyan, R. 2000. “ Zur Deutung des K önigsprofils auf parthischen Münzen. ” Archäologische Mitteilungen aus Iran und T uran 32, 253 – 259. – . 2001. “ T endenze culturali e ideologiche nell ’ impero partico riflesse dalla monetazione. ” Parthica 3, 25 – 132. Widengren, G. 1983. “ Sources of P arthian and Sasanian history. ” In E. Y arshater (ed.), The Cambridge history of Iran . V ol. 3.2, 1261 – 1283. Cambridge: Cambridge University Press. Wiesehöfer, J. 1996. “’ King of Kings ’ and ‘ Philhellên ’ : Kingship in Arsacid Iran. ” In P . Bilde (ed.), Aspects of Hellenistic kingship , 55 – 66. Aarhus: Aarhus University Press. – . 1999. Das frühe Persien: Geschichte eines antiken Weltreichs . M unich: C. H. Beck. – . 2007 . “ Fars under Seleucid and Parthian rule. ” In V . S. Curtis and S. Stewart (eds.), Age of the Parthians , 37 – 49. London: I. B. T auris. – . 2011. “ Frataraka rule in Seleucid Persis: A new appraisal. ” In A. Erskine and L. Llewellyn-Jones (eds.), Creating a Hellenistic world , 107 – 122. Swansea: C lassical Press of W ales. – . 2016. “ Greeks, Iranians and Chinese on the Silk R oad. ” In S. N. C. Lieu and G. B. Mikkelsen (eds.), Between Rome and China: History, religions and material culture of the S ilk Road , 1 – 17 . T urnhout: Brepols. Wiesehöfer, J., ed. 1998. Das P artherreich und seine Zeugnisse / The Arsacid Empire: S ources and documentation . Stuttgart: Franz Steiner. Wiesehöfer, J., and S. Müller, eds. 2017 . Parthika: Greek and Roman authors ’ views of the Arsacid Empire / Griechisch-römische Bilder des Arsakidenreiches . W iesbaden: Harrassowitz. Wissemann, M. 1982. Die Parther in der augusteischen Dichtung . F rankfurt: Peter Lang. Wolski, J. 1946. “ Arsaces II. ” E os 41, 1 56 – 165. – . 1965. “ Le rôle et l ’ importance des mercenaires dans l ’ état Parthe. ” Iranica Antiqua 5, 103 – 11 5. – . 1981. “ L ’ aristocratie foncière et l ’ organisation de l ’ armée parthe. ” Klio 63, 105 – 112. – . 1993. L ’ empire des Arsacides . Leuven: Peeters. Zanous, H. P ., and J. Y ang. 2018. “ Arsacid cities in the Hanshu and Houhanshu . ” Iran and the Caucasus 22.2, 123 – 138. Eli J. S. W eaverdyck 7 The Roman Empire I Introduction The Roman Empire was one of the largest and longest lasting of all the empires in the ancient world. 1 At its height, it controlled the entire coast of the Mediterranean and vast continental hinterlands, including most of western Europe and Great Brit- ain, the Balkans, all of Asia Minor, the Near East as far as the Euphrates (and be- yond, briefly), and northern Africa as far south as the Sahara. The Mediterranean, known to the Romans as mare nostrum ( ‘ our sea ’ ), formed the core. Th e Me dit er ra ne an b as in i s c har ac te ri zed by e xt re me v ar ia bi lit y ac ro ss b ot h sp ac e an d tim e. Ge olo gi ca ll y , the a re a is a l ar ge s ubd uc ti on z one b et we en t he A fri ca n a nd Eu rop ea n te ct on ic pl at es . Th is no t on ly pr od uce s vo lc ani c an d se is mi c act iv it y , it als o me ans t ha t the m os t com mon ly e nc ou nte re d bed ro ck i s upl ift ed l im est on e, wh ic h is ea sil y er od ed by w at er . Mu ch o f the c oa st li ne is m ou nt ai nou s wi th d eep r iv er v all ey s. Th is ru gg ed t opo gr ap hy me an s th at ev en b roa dl y si mil ar c li ma ti c con di ti on s can p ro- du ce dr as ti ca ll y dis si mi la r mic ro cl im ate s wi th in ve ry s ho rt di st an ce s. In a dd it ion , st ron g in te ra nn ual var iab il it y in pr ec ip it ati on m ea ns t ha t loc al f oo d sh or tag es w er e an en de mi c fe at ur e of Me di te rra ne an a gr ic ul tur e. I n co mbi na ti on , thi s te mp ora l an d sp ati al v ar iab il it y me an t th at r isk -b uf fe rin g me ch an ism s i ncl ud in g div er si fi cat io n, st ora ge , an d di st rib ut io n of go od s pla ye d an im po rt an t rol e in an ci en t Med it er ra nea n su rvi va l st ra te gie s. C on ne cti vi ty h as al wa ys c ha rac te ri ze d the M ed it er ra nea n. 2 While geography encouraged mobility , the empire accelerated that tendency , inducing the transfer of people, goods, and ideas on a scale never seen before. 3 This mobility , combined with increased demand and the efforts of the imperial govern- ment to mobilize specific products, led to the rise of broad regional specializations, particularly in staple foods and precious metals. 4 The results of this increased con-  It has also been the subject of more scholarship than any other empire treated in this volume. The modern study of the Roman Empire is generally thought to have begun in the eighteenth centu- ry with works by the French political philosopher Montesquieu (1734) and the English historian Edward Gibbon (1776 – 1789). For an overview of the study of the Roman Empire in the twentieth century , see Potter 2006.  Horden and Purcell 2000, particularly 175 – 230. The environmental history of the ancient Mediter- ranean is entering new territory as more and better data and models become available. See Harris 2013; McCormick et al. 2012 for introductions.  For the mobility of people in the Roman empire, see the essays in de Ligt and Tacoma 2016. Recent overviews of the mobility of goods include Wilson and Bowman 2018; Morley 2007 .  For example, Sicily , Egypt, and North Africa exported grain on a large scale (Erdkamp 2005, 206 – 257 ; Kessler and Temin 2007). Olive oil for export was produced in North Africa and the Iberian Peninsula (Marzano 2013; Mattingly 1988). Gold and Silver mines were concentrated in the Iberian Peninsula and in the Balkan and Carpathian Mountains (Wilson 2007). Open Access. © 2020 Eli J. S. Weaverdyck, published by De Gruyter. This work is licensed under the Creative C ommons Attribution-NonCommercial-NoDerivatives 4.0 License. https:/ /doi.org/10.15 15/97831106077 41-009 242 Eli J. S. Weaverdyck Map 1: The Mediterranean world prior to Roman conquest. © P eter Palm. The Roman Empire 243 sumption and connectivity are most visible today in the ruins of ancient cities, where certain forms of monumental architecture, built in Roman concrete, are re- peated again and again across Europe, northern Africa, and western Asia. 5 These far-flung remains are the result of a long history of imperial development. Roman history can be divided into several major epochs, only two of which fall within the chronological limits of this handbook: the Republic and the Principate. The city of Rome was ruled according to an oligarchic system of government from the late sixth to the late first century  (the Republican period ), at which point it became a monarchy (the Principate). Within these epochs a period of social restruc- turing in the fourth century  inaugurated the Middle Republic, an era of stability that lasted until the late second century  , at which point further social upheaval took hold and lasted for about a century . This period is known as the Late Republic. In the last third of the first century  , one man, Augustus (63  – 14  ), success- fully consolidated supreme power in his own person and passed it on to his chosen heir, establishing the rule of Roman emperors. Traditionally , the first three centuries of the Common Era are known as the Principate. The word emphasizes the emper- or ’ s status as merely the preeminent citizen, first among equals, princeps . 6 After a long crisis of government in the third century  , the empire was refashioned along more authoritarian lines. The period from 300  to 300  , then, roughly coin- cides with the Middle Republic, the Late Republic, and the Principate. Rome acquired its massive empire over centuries of more or less continuous warfare. After a brief discussion of the reasons why Romans were so often willing to go to war, their long-term military success, and the consequences of conquest for the Roman state in section two, section three describes the structures through which the empire was ruled and how these changed over time, particularly between the Republic and the Principate. It begins with the military apparatus that provided the means of physical coercion, then describes the political structures that administered the empire, and finally the infrastructure, both physical and institutional, that sup- ported imperial governance. In addition to the strategies and goals of the central state, the participation of other agents and the impacts of empire are highlighted throughout. The fourth section explores two aspects of society in the ancient Medi- terranean that were affected by the Roman Empire and that contributed to its inte- gration. First, it describes long-distance social networks, particularly those based on patronage, those based on Italian origin, and those based on familial ties. Then it discusses the identities through which people understood themselves and related to each other within an imperial context.  See Woolf 1998, 112 – 126 for Roman Gaul; Woolf 2012, 288 – 300 more generally .  The term ‘ principate ’ is found in ancient literature. The Roman historian Tacitus makes a point of saying that principatus and libertas (freedom) were irreconcilable until the rise of the good em- peror Nerva (Tacitus Agricola [Tac. Agr. ] 3. 1). 244 Eli J. S. Weaverdyck II Achieving Supremacy: C ity, State, and Empire Roman power was established and, in large part, sustained through military su- premacy . 7 During the Middle Republic, when the Romans defeated the other hege- monic powers in the Mediterranean, they were in the habit of sending out armies every year. At the end of the Late Republic – when powerful generals conquered vast swaths of territory in Europe and the Near East, and when the entire Mediterra- nean was convulsed by Roman civil wars – the portion of Italian young men serving under arms reached as high as 25 percent. 8 Under the Principate, the Roman ap- proach to war changed, and the pace of expansion slowed drastically . Nevertheless, the army remained the single largest item of state expenditure, and some emperors still launched campaigns of conquest. The fact that Rome was a city-state when it acquired its empire sets it apart from the other polities discussed in this volume. Rome ’ s oligarchic internal socio- political structure incentivized persistent military aggression. An expansive ap- proach to citizenship and a tolerance for the partial continuation of others ’ civic structures allowed for the political integration of new territory and provided access to vast reserves of manpower for further wars. Rome ’ s civic ideology , centered on the flexible idea of the res publica , provided the thread of continuity that allowed the empire to survive a traumatic period of civil war and a shift to a monarchic form of government. Roman expansion can be explained in large part as a consequence of the partic- ular nature of aristocratic competition within the Roman Republic. 9 Since the mid- dle of the fourth century  , the ruling class in Rome was defined primarily by election to office, an honor bestowed by the Roman people on those who had served the state that provided further opportunities to serve and accumulate glory . 10 The most important form of service was military achievement. Distinction in battle was a valuable asset for candidates and those who held the highest offices achieved further glory by successfully commanding Rome ’ s armies. This encouraged Roman  The sources of Roman power have intrigued scholars since the days of Machiavelli in the six- teenth century , and the Roman example has been influential in broader sociological works (e.g., Doyle 1986; M. Mann 1986). Recent overviews of Roman power by Roman historians include Harris 2016; Morley 2010; Woolf 2012, which build on numerous narrower, more detailed investigations. See, for example, influential discussions on the cultivation of loyalty to Rome in the provinces (Ando 2000); the role of honor in binding elites to the imperial system (Lendon 1997); and the position of the emperor (Millar 1977).  Brunt 1987 , 509 – 512.  Modern debate on Roman imperialism is focused on the extent to which Rome ’ s military aggres- siveness during the Republic was unusual and driven by internal factors (Harris 1979) or typical of contemporary Mediterranean city-states in the context of inter-state anarchy (Eckstein 2006). For overviews, see Morley 2010; Hoyos 2013.  For the origins of this ethos, see Hölkeskamp 1993; Raaflaub 1996; cf. Eckstein 2006, 229 – 237 . The Roman Empire 245 commanders to be aggressive during their time in office. Even in defeat, a com- mander could win glory by living up to the aristocratic ideal of virtus : manly , stead- fast courage. 11 This drove individual generals and the aristocracy as a whole to per- sist in fighting after losses that might have been war ending for other polities. Beyond glory , war brought substantial economic benefits that extended beyond the political aristocracy and ensured the support of other segments of society for Roman war making. The most immediate benefit was booty . 12 After a successful campaign, the spoils of war were paraded through the streets of Rome in a ‘ tri- umph, ’ a grand procession celebrating the victorious general, 13 who distributed the booty in varying proportions between the soldiers, the treasury , and often construc- tion projects (especially monumental temples) that celebrated the victory . Among the booty were large numbers of slaves. Selling prisoners of war into slavery was common practice in antiquity , and Rome ’ s nearly constant warfare and the huge scale of some campaigns brought massive quantities of slaves into Italy . 14 Although natural increase and trade would replace warfare as the primary source of slaves, the influx during the Mid-Republic facilitated a new type of market-oriented agricul- ture practiced especially by elites. 15 The land acquired was another benefit of con- quest. The Romans would often confiscate a portion of their enemies ’ land, some of which was auctioned off, some rented out, and some distributed to colonists. 16 Rome established colonies throughout Italy in the Mid-Republic and, in the Late Republic and Early Principate, beyond. 17 The extraction of resources continued after conquest. In the third and second centuries  , the Romans imposed large indem- nities – fixed sums to be paid over a specific period of time – on their defeated rivals. 18 In the second century  , it became more common to impose taxes, which were paid in perpetuity . 19 Captured mines also produced a great deal of revenue. 20 The exploitation of mines, collection of taxes, and renting of public land were all managed through public contracts auctioned off to groups of individuals called pub- licani , so even public revenues benefited private people. 21  McDonnell 2006; Rosenstein 1990.  Kay 2014, 29 – 35.  See Beard 2007 for the Roman triumph.  For quantitative estimates of slavery in Republican Italy , see Scheidel 2005, with further litera- ture on the sources of slaves.  Hopkins 1978 provides the reference point for modern debate about land, slavery , and the eco- nomic impact of Roman conquest on Italy . For a more recent account, see Kay 2014, 133 – 188.  Roselaar 2010, 31 – 54.  The classic account of Roman colonization is Salmon 1969 . For more recent perspectives, see the essays in Stek and Pelgrom 2014.  Kay 2014, 37 – 42.  Tan 2015.  Kay 2014, 43 – 58.  Badian 1972 remains fundamental on publicani . See also Kay 2014, 49 – 54 (mining); Tan 2017 , 40 – 67 (tax collection). 246 Eli J. S. Weaverdyck Outside of Rome, the complex network of relationships, often antagonistic, be- tween polities of different types and sizes that stretched across the Mediterranean basin enabled, encouraged, and directed Roman war making. 22 In the third century  , the Mediterranean contained a plethora of political actors of various configura- tions, each with a complex set of relationships and most in some form of subordina- tion to another (map 1). 23 The ubiquity of substantial power differentials created a lively market for protection, with smaller polities seeking the aid of larger powers against other larger powers. Pleas for assistance by allies were the proximate causes of many of Rome ’ s wars of conquest. It was hazardous to Rome ’ s standing to refuse such a request: if an ally were overrun, the value of Roman friendship was dimin- ished and other allies would be more likely to join a rival hegemon. The political and economic benefits of warfare and the politically complex Med- iterranean context help to explain Roman aggression, but they do not explain Ro- man success. A habit of persistence in the face of defeat was important, but this was only possible due to the overwhelming manpower advantage that Rome held over its rivals. 24 The soldiers in Roman armies came not just from the city of Rome but from a network of municipia , colonies, and allies across Italy . In the process of conquering Italy , some defeated cities were absorbed into the Roman state and called municipia . Although the physical infrastructure of these cities remained, and they continued to govern themselves in internal matters, they had no independent foreign relations and their free populations were made Roman citizens. This was unique in the ancient Mediterranean, where access to citizenship was normally closely guarded. Also unique to Rome was the practice of bestowing citizenship on freed slaves (freedmen/women). As a result, Rome had more citizens than any other city-state. 25 They served in the legions, units of 4,000 to 6,000 men. 26 Those ene- mies not incorporated became allies ( socii ), maintaining an independent political identity but subject to Roman commands and expected to provide troops on request. They served in distinct auxiliary units alongside the legions. Most of the colonies that Rome established in Italy were independent allies (albeit with an especially  Gruen (1986) first argued that Roman expansion in the East was driven by involvement in re- gional politics. Eckstein (2006) developed the idea using modern International Relations theory . Morley (2010) and Woolf (2012) integrate internal and external factors in their accounts of Roman expansion.  V on Reden (ch. 1, this volume) describes the most powerful polities in the Eastern Mediterra- nean and provides further literature.  Brunt 1987 is the standard work on Italian demography and Roman manpower in the Republic. More recently , Rosenstein (2004) has provided a model for the demographic sustainability of mili- tary recruitment. For an overview of the debate on Italian demography in this period and further literature, see Lo Cascio 2013.  Sherwin-White 1973 is fundamental for Roman citizenship. Brunt 1987 provides quantitative estimates. See below , sec. IV .2.4 for more on the meaning of Roman citizenship.  For the (variable) size of the legion, see Brunt 1987 , 671 – 676, 687 – 693. The Roman Empire 247 close relationship to their metropolis ), but some were inhabited by citizens. This arrangement not only expanded Roman manpower, it also integrated distant poli- ties into the Roman Empire without straining the bureaucratic capacities of a city- state. Rome ’ s allies outside of Italy were also instrumental in its success. 27 They often fought for Rome, both alongside Roman armies on the battlefield and more inde- pendently in a broader campaign. Their diplomatic connections in areas far from Italy could prove invaluable in facilitating troop movements and in mediating new relationships. They also provided invaluable logistical support. They furnished not only supplies and equipment, but transportation, ports, and secure supply bases far from Italy . 28 This allowed Roman armies to overcome the limitations on cam- paigning imposed by ancient technology . Allies were often former enemies, those wishing to avoid Roman enmity , or those hoping to benefit from Roman power. Further, each new alliance brought with it a new network of relations stretching into new areas, new potential allies and enemies. Thus, Roman conquest both en- abled and invited further conquest. This dynamic drove Roman state formation. I use the term ‘ state ’ loosely to refer to the imagined community of people who identified or were identified as Romans and the institutions that existed, in theory , to serve the interests of that community . The phrase that Latin speakers used to describe this entity most often was res publi- ca , 29 a term that encompasses the affairs that concern the Roman people and de- mand communal action and the entity that emerges from the interactions of Romans acting as Romans. The flexibility of the concept combined with its ideological power helps explain its continued relevance in ancient political discourse over the centu- ries, and the durability of the res publica as a symbol helps explain the ability of the Roman Empire to survive in recognizable form through changes in government, ruling dynasty , and even state religion. The process of conquest strengthened both the state ’ s ideological power and its institutional capacity . 30 Military service was the primary contribution of both citi- zens and allies to the Roman state. Outside the city of Rome, it was the main way in which citizens experienced and reproduced the state and their membership in it. For the Italian allies, conscription and campaigning – which involved not only obeying Roman commanders, but the repeated construction of camps that articulat- ed social power relations in the spatial layout of barracks and command centers – were the core expressions of Roman domination. 31 The extension of Roman war  See P . J. Burton 2011, 172 – 205 for the obligations of Rome ’ s allies.  Roth 1999 , 169 – 177 (supply lines), 227 – 230 (provision of grain).  Recently , Moatti ’ s work on the meaning of this phrase and what it reveals about the changing nature of the Roman state has been especially influential (2018).  This paragraph is based on Eich and Eich 2005.  Polybios (Polyb.) 6. 26 – 42. 254 Eli J. S. Weaverdyck Balkan Mountains and the Danube, but the Tetrarchic and Constantinian emperors invested in fortifications and campaigned north of the river, eventually reestablish- ing Roman dominance. The rural economy recovered, with new settlements spring- ing up in agriculturally productive areas. In the late fourth century , the frontier collapsed again and the area faced a mass migration that the empire was powerless to stop. Eventually , the immigrants forged a treaty with the emperor and a sem- blance of Roman authority was reestablished, but there was no longer a line of forts along the Danube protecting the countryside. Instead, the army fortified strategic population centers and supply depots. Not only did the number of settlements in the countryside plummet, people now lived in areas with easy access to a naturally defensible refuge, leaving the most productive areas empty . 62 The spatial layout of the frontier varied greatly by region and period (map 2). In Europe, army bases mostly lined the Rhine and Danube Rivers, which eased sup- ply and communications between them while simultaneously creating a surveil- lance network that could monitor populations and control movement across the frontier. 63 It also provided the staging ground for expeditions beyond the frontier that enforced and extended Roman supremacy . So, for example, the initial construc- tion of the chain of forts on the lower Rhine in the mid-first century  has been linked to the conquest of Britain. 64 In the East, the geography of the frontier was more varied and dynamic. Egypt can be dealt with briefly , as its garrison was relatively stable in size and distribution. This province, crucially important for the grain supply of the capital, contained first two legions and then only one stationed near Alexandria, far from the edge of Ro- man territory . This megalopolis had to be firmly controlled to ensure traffic flowed smoothly between the Mediterranean and the Nile. From the late first century  , there were also approximately 5,000 to 7 ,000 auxiliary soldiers. Some of these were concentrated with the legions near Alexandria, while others were stationed in the region of the First Cataract, protecting Egypt from incursions from the south and preventing uprisings in the Thebaid region. Many soldiers were scattered in smaller outposts, particularly in the Eastern Desert, guarding roads and controlling imperial quarries. A detachment of legionaries from Egypt was even stationed on the Farasan Islands at the mouth of the Red Sea, though their precise function – to suppress piracy and smuggling, to maintain Roman supremacy and control of the Red Sea, or both – is not entirely clear. 65 The distribution of Roman forces in western Asia was less stable, shifting to address two main phenomena: the strength of the Arsakid Empire in the north and  For the effects of the Gothic invasion on the rural economy of the central Danubian Plain, see Poulter 2004; 2013. I identified the widespread shift in settlement location preferences and argued that it could be attributed to lack of security in Weaverdyck 2016.  Dobson 2009; Karavas 2005; C. S. Sommer 2009 .  Polak 2009; van Dinter 2013.  Haensch 2012. For the function of the garrison, see Cobb 2018, 118 – 120 with literature. The Roman Empire 255 the rebelliousness of local populations, especially the Jews, in the south. 66 In con- trast to Europe, the states that controlled large parts of this area were more central- ized, and the rulers of the westernmost states had maintained good relations with Rome since Pompey ’ s eastern campaigns in the 60s  at the latest. As a conse- quence, the initial occupation force was smaller and more concentrated than in the West. Throughout the first and second centuries, their kingdoms were annexed and relations with more distant kings established, so the external borders of Roman provinces also changed more frequently . Also in contrast to the West, units in the East were more often stationed in cities, so military occupation is much harder to track archaeologically . We know that for most of the first century  , the province of Syria contained four legions, of which one was stationed at Kyrrhos and one at Raphanaia. In the 50s and 60s  , a major war erupted over the appointment of the king of Armenia – a perennial bone of contention between Rome and the Arsakids – leading to a major reorganiza- tion under V espasian (r. 69 – 79  ): kingdoms were annexed to enlarge the provinces of Syria and Kappadokia, the road network was expanded, fortifications built, and four legions were stationed on or near the Euphrates at Melitene and Satala in Kap- padokia, and at Samosata and Zeugma in Syria. 67 Syria contained a third legion still at Raphanaia, and Judaea, in response to the Jewish revolt of 66 – 73  , contained one in Jerusalem. V espasian also strengthened the Roman presence in the Black Sea, establishing a fleet and stationing troops in cities on the east coast. This well-developed military infrastructure, further elaborated by V espasian ’ s successors, allowed Trajan (r. 98 – 117  ) to launch a major invasion of Mesopotamia from 114 to 117  , although his territorial gains in Mesopotamia and Armenia could not be held. Trajan did, however, annex the Nabataean Kingdom, the last independ- ent kingdom west of the Euphrates, which became the province of Arabia in 106  . One legion was transferred from Egypt to Arabia ’ s capital, Bostra, and a major road, the Via Nova Traiana , was built to connect this city to the Red Sea. 68 Following Trajan ’ s Parthian campaigns, legions were shuffled, but the bases remained mostly the same. After the Bar Kokhba revolt in the 130s  , a second legion was added to Judaea, now called Syria Palaestina, at Caparcotna. The next Parthian war came in the 160s  and resulted in the occupation of territory across the Euphrates and an extension of Roman control down the river to at least Dura Europos. Another war in the 190s further expanded Roman territory in Mesopotamia and led to the estab- lishment of new provinces across the Euphrates. 69 Two legions were stationed here  Isaac 1990 underpins most recent scholarship on the eastern frontier. Wheeler 2007 is a useful summary that takes a large-scale, strategic perspective. M. Sommer 2005 and Edwell 2008 provide detailed accounts of the Upper and Middle Euphrates frontier. For the Arabian frontier, see Kennedy 2004.  Dart 2016; Edwell 2008, 18 – 20.  Speidel 2019 .  For the provincialization of northern Mesopotamia, see Speidel 2009 , 181 – 210. 256 Eli J. S. Weaverdyck from the 190s, one in Singara near the Tigris, the other possibly in Nisibis, though that is not certain. Although many details elude us, it seems that the Roman army in the East was intended to control major population centers and important nodes in transportation networks, including crossing points on the Euphrates and oases in the Egyptian and Arabian deserts. 70 Controlling routes could check major invasions, but the history of Roman aggression suggests a generally offensive rather than defensive stance: control of major river crossings allowed the Romans to threaten, attack, and domi- nate their neighbors. The old idea that the Arabian frontier was meant to check nomadic incursions has been well refuted. 71 Rather, control and taxation of the lu- crative trade from Arabia, the Indian Ocean, and the East might have been a moti- vating factor. 72 Compared to the great riverine frontiers of Europe, Rome ’ s eastern frontier is much less linear and much more focused on the internal population, not only in Judaea but in Egypt as well. III.  .  Supplying the Army Relative to the population of the empire, which is usually estimated to have con- tained between 60 and 70 million people in the mid-second century  , 73 the size of the army under the Principate was smaller than that of the Republic, numbering probably between 300,000 and 400,000 soldiers, but they were now permanently on duty and many were stationed in areas that had hitherto been sparsely populat- ed. The state ’ s efforts to ensure the supply of these troops with food and equipment had profound and far-reaching economic effects. 74 The army in northern Europe, in particular, is seen as a transformative force, not only spurring greater agricultural production but introducing new plants, breeds of animals, and technologies affect- ing both agriculture and communications. 75 Furthermore, some important Mediter- ranean staples – particularly olives – do not grow in northern Europe and had to be imported. Supplying these armies, therefore, was a major undertaking that had long-term and far-reaching effects. Military supplies were acquired and transported to the army through a complex and variable network of exchanges and interactions. 76 The major point of debate  Kennedy 1996; Parker 2006.  Isaac 1990, 68 – 77 .  Speidel 2016.  Hanson 2016, 71 with literature.  MacMullen 1963; Wierschowski 1984 are foundational for the economic impact of the army . See also the contributions in Erdkamp 2002 and Blois and Lo Cascio 2007 . For military supply , see Roth 1999; Mitthof 2001.  Stoll 2016 summarizes many of these themes.  I focus here on the supply of troops during peace time, setting aside questions of supply during campaigns. The Roman Empire 257 has been whether military supplies were acquired through taxation and the extrac- tion of in-kind rents from imperial estates, purchase at a fixed price, or free market purchase and which commodities were obtained in which way . 77 The problem is that the state acquired goods, in particular grain, in a number of ways, and precise origins are rarely stated in the documentary evidence that survives. So, for example, an ostracon may affirm that a certain person delivered “ public wheat ” to a soldier, but for the purposes of the people named, who grew the wheat and whether or not they were paid for it is immaterial. 78 A line from Pliny the Younger ’ s speech in praise of the emperor Trajan shows how blurry the line between purchase and requi- sition could be: “ The fiscus bought whatever it seemed to buy . ” 79 The implicit com- parison is revealing. If requisition was masked as purchase, then purchase was seen as normal. In the same breath, Pliny also notes how extraordinary Trajan was for buying goods at a negotiated price. Documentary evidence, sparse and unevenly distributed as it is, reveals that there was no single, universal system of acquisition. 80 The precise mechanisms of supply at work will have depended heavily on contingent circumstances shaped by local ecologies, agricultural practices, social structure, and histories of surplus extraction as well as military need. Supply systems in Egypt and Britain, for exam- ple, must have been very different and must have changed over time. A receipt preserved on papyrus shows that, in Egypt, the barley needs of each unit were re- ported up through the military hierarchy to the praefectus of Egypt, who then dis- tributed the burden of supplying those needs down through the civilian hierarchy until the two met in the form of a low-ranking officer collecting grain from village elders. Although the transaction is described in monetary terms, we hear of no mon- ey changing hands. 81 An earlier papyrus describes barley as having come from the public account of a village and makes no mention of purchase. 82 In Britain, the wooden writing tablets from the fort of Vindolanda suggest that the needs of the military were met through a more complex and irregular system that included entre- preneurs operating within a free market context. 83 One letter, addressed to an offi- cer, discusses various business arrangements, including the purchase of a large quantity of grain that has been interpreted as military supply . 84 The writer has put down a deposit and requires the balance of the price or he risks losing the grain,  Summaries can be found in Kehne 2007 ; Rathbone 2007 .  O .Petr. 245 with Adams 1995.  Pliny Panegyricus 29 . 5.  For the use of documentary evidence, see von Reden, ch. 8.C, this volume.  P Amh . 107 with Adams 1999 , 120 – 121. Note, however, that in Adams ’ s translation, the prefect orders the grain “ to be brought up, ” but the Greek verb συνωνηθ ῆ ναι indicates purchase and is better translated as “ to be bought up, ” as in Campbell 1994, no. 235.  SB XIV 12169 with Mitthof 2001, 38 – 39 .  Evers 2011; Whittaker 2002.  T .Vindol. II , 343. 258 Eli J. S. Weaverdyck the deposit, and his pride. If the writer is engaged in supplying the army , it is nota- ble that this connection does not guarantee access to the grain he is buying. The evidence from Egypt, which had a long history prior to Roman conquest of paying taxes in grain, might be giving a false impression about the prevalence of state- owned grain elsewhere in the empire. Beyond the local economy of the frontier zones, the soldiers ’ demand for Medi- terranean products led to the modification of long-distance exchange networks and the creation of new ones that had not previously existed. Mediterranean traders had been active in the areas that would become frontier zones long before the influx of soldiers, but the establishment of the military cordon created new demand and, more importantly , made such trade into a structural support of imperial power. The trade in olive oil from Baetica in southern Spain is especially visible because it was transported in a distinctive amphora known as Dressel 20. 85 While most types of pottery decline in frequency with distance from the place of production, Dressel 20 does not. It is found in very large numbers on the Rhine and in Britain, suggesting directed distribution. There is no evidence that the state ever collected taxes in oil, but there is widespread evidence for the existence of military conductores , people who arranged or entered into contracts for the supply of goods to different units. 86 Contracts and trade relationships that, once established, were jealously guarded by powerful actors could account for these distributions just as well as state-run redistribution. 87 Transporting these supplies could be a lucrative business. Not only were the shippers ( navicularii ) paid for their services, the goods that they transported for the government were not taxed. The temptation to clandestinely mix other goods in with the tax-free portion of the cargo was irresistible, as we learn from a second- century edict: On th e sub jec t of p rop erty w hic h go vern ors d ire ct to be br ough t to t hem f or th eir u se, t he d eif ied Had ria n wrot e to gove rno rs sa yin g that wh en a prov inc ial go ver nor or a le gio nary c omma nd er or a pr ocu rato r of suc h a per son di spa tch es so meon e to mak e a pur cha se, he s hou ld ind ica te thi s in a memo ran dum s igne d in h is own ha nd an d sh oul d send thi s mem ora ndum to a ta x far mer s o tha t any thi ng tha t is br ou ght i n exc ess of wha t was o rde red c an be s ubje ct to t ax . 88 Similar edicts demonstrate the persistent difficulty of distinguishing between tax- free and taxable goods. Even the honest navicularius , however, benefited from gov- ernment contracts beyond the value of the contract itself. In the ancient world, where information on market conditions was difficult to come by , the government  The debate about the mechanisms by which Baetican olive oil reached the northern frontier is conveniently summarized in Tchernia 2016, 255 – 264.  Whittaker 1994, 108 – 113.  Tchernia 2016, 97 – 114, 255 – 264.  Digesta 39 . 4. 4. 1, trans. Watson. The Roman Empire 259 contract ensured a profitable voyage, allowing the merchant to speculate on extra cargo with little risk. The stability and size of the military market, coupled with its ability to operate over long distances and the privileges granted to its suppliers, encouraged the development of a class of business people that was particularly self- aware in western Europe. 89 This group of people had social contacts throughout the empire and business expertise that allowed them to undertake substantial commer- cial operations both for the army and for nonmilitary markets as well. In this way , the frontier might have contributed to economic integration and development be- yond its own supply needs. III.  Politic al Structure s The authority of Rome rested on more than the coercive threat posed by the military . Indeed, the empire consisted of a great number of administrative units tied together in a relatively shallow hierarchy through diverse vertical relationships with the im- perial center. Compared to the Chinese empires, the Roman state delegated a great deal of responsibility onto relatively autonomous but firmly subordinate political units. These came in a variety of forms. Strabo, writing a geography of the entire Roman world under Augustus, captures this diversity: Of all of these lands under the Romans, some are ruled by kings, others they themselves hold calling them provinces and sending governors and tax collectors. There are some free cities, some having come as friends from the beginning, and others they themselves set free as an honor. There are also some dynasts and commanders and priests under them: these live ac- cording to their ancestral laws. 90 In addition to the lands that the Romans “ hold themselves, ” there are free cities, kingdoms, and a variety of other places that live by their own laws. In what sense were these places “ under the Romans ” ? The form of Roman dominance differed from case to case, but all acknowledged Roman supremacy , and all were expected to obey Roman commands. The degree of Roman interference in the lives of their subjects also varied, but a general diachronic trend toward greater interference, especially in the Principate, can be observed. III.  .  C lient Kings In Strabo ’ s conception of the empire, monarchies and provinces are parallel forms of Roman control over large territories. The monarchs who ruled these lands are  For the negotiatores and navicularii as a social class, see V erboven 2007 . Broekaert 2013 provides a prosopography of these people. For the economic benefits of state contracts, see Tchernia 2016, 103 – 111.  Strabo 17 . 3. 24. 260 Eli J. S. Weaverdyck often called ‘ client kings ’ (queens were rarer) by modern scholars to emphasize their subordination to Rome, but in antiquity the relationship was expressed in terms of friendship and alliance. 91 This diplomatic language allowed for a great deal of flexi- bility and diversity . The precise relationship between a king and Rome, as with oth- er allies, varied according to local circumstances. Many were Roman citizens and had grown up in Rome. Herodes and his successors, who ruled Judaea, had no dynastic claim and were entirely dependent on Roman backing for their position, while others had much more powerful, local bases of support. Mithridates VI of Pontos on the southeast coast of the Black Sea and Jugurtha of Numidia in North Africa were both powerful enough in their own right to wage large-scale wars against Rome in the late second and early first centuries  . 92 For the most part, though, Rome ’ s royal allies were loyal and their kingdoms were seen as part of the Roman Empire. 93 Indeed, territories could be transferred between province and kingdom relatively easily , though it was more common for royal land to become provincial than vice versa. 94 At the same time, it would be a mistake to regard these kingdoms as simply provinces by another name. There is no unambiguous evidence of kings paying taxes to the Roman treasury , though they might pay indemnities for fixed periods of time, and they expended a great deal of money , ultimately extracted from their subjects, cultivating the favor of powerful Romans both in the Republic and in the Principate. 95 The Roman state also provided money to some kings. These gifts could ensure loyalty , strengthen the king ’ s position within his kingdom, and allow the king to advance Roman interests, but they could also be protection money , pay- ments to avoid war. 96 The exchange of money was an ambiguous but normal form of gift exchange that established and maintained social relationships. 97 For the kings, friendship with Rome provided powerful support against their enemies both inside and outside their kingdom, and enhanced their prestige and influence, especially when they enjoyed close personal relations with powerful Romans. 98 For the Ro- mans, these relationships provided stability in areas that were difficult to control otherwise – kings were often asked to control bandits – and protection against more distant, hostile groups, but kings could also act as intermediaries between Rome and these groups. 99 While each king and kingdom was unique, with its own back-  Braund 1984 is foundational. Braund 2015; Kemp 2018 are more recent.  See von Reden, ch. 1, this volume.  Braund 2015, 150 – 154.  Braund 1984, 84 – 85.  Braund 1984, 58 – 66.  Kemp 2018, 97 – 100 sees a shift from the former to the latter in the late second century  .  Braund 1984, 62 – 63.  Braund 1984, 82 – 83.  Braund 1984, 91 – 103; 2015. The Roman Empire 261 ground, goals, and strategies, they all played a role in supporting (usually) Roman power, and this role was distinct from that played by the provinces. III.  .  Province s in the Repub lic The English term ‘ province ’ denotes a territorial administrative section of a state or empire. Under the Principate, Roman provinciae conform well to this definition, but this was not always the case. Tracing the changing meanings of the Latin word provincia through time can illuminate the changing nature of the Roman Empire from one based on the military might and the authority of Rome and Roman repre- sentatives vis à vis other polities to one based on Roman possession and administra- tion of territories around the Mediterranean and beyond. In the third and early second centuries, a provincia was a task or responsibility assigned to a magistrate. 100 Thus, the treasury was the provincia of the urban quaes- tor . While most provinciae in this period were named for a geographic area, this does not imply that those areas had been or were going to be annexed. 101 Rather, these geographic provinciae defined the area in which a magistrate was expected to operate and in which his authority superseded that of other, equal magistrates. The borders of a provincia were often vague, but sometimes clear geographical limits were defined, as when the Senate ordered the praetors responsible for the provinciae of Nearer and Farther Spain to fix the boundary between the two in 197  to clarify their areas of competence. 102 This gradually changed as conquered territories were assigned year after year, and the number of magistrates increased to oversee them. For Cicero, writing in the mid-first century  , provincia had a broader range of meanings, including not only a task but also a territory that is in some way owned by the Roman people and continued to exist in the absence of a magistrate as well as the community of people living in that territory . Contemporary authors like Cae- sar also used the term to refer to a set of legal and administrative norms according to which the territory was governed. The sense in which the provincia were owned by the Romans is more difficult to tease out because the provincia included numerous subject communities, each with their own relationship to Rome. The provincia of Asia in western Anatolia illus- trates the complexity (map 2). 103 In 134  , the last king of Pergamon died without an heir and left his kingdom to Rome. 104 After defeating a claimant to the Perga- mene throne, the commander and a board of ten commissioners established the  The following discussion is based on Richardson 2008. See also Lintott 1993, 22 – 32.  The provincia of Macedonia is a good example, for which see Kallet-Marx 1995, 12 – 42.  Richardson 1986, 77 – 78.  Kallet-Marx 1995, 98 – 123.  For the reasons behind this bequest, see Sherwin-White 1984, 80 – 92. 262 Eli J. S. Weaverdyck relationships that Rome would have with the various parts of the old kingdom. Many cities were declared free, and large areas were handed over to Rome ’ s allies, but some communities, particularly those that had supported the pretender, were taxed, and formerly royal lands became Roman public land to be sold or rented. Magistrates were sent every year to see to this new provincia , but they did not re- strict themselves to the land that was now owned by the Romans nor to the non- free cities. Indeed, during peacetime, the magistrates seem to have operated prima- rily out of Pergamon and Ephesos, two free cities. 105 The shift in the meaning of provincia from a task to a territory owned and ad- ministered by the Roman government was exemplified and confirmed by Pompey when, after his military campaigns in the East, he unilaterally created the provincia of Syria and appointed one of his subordinates to govern it because he could find no local candidate able to rule it. The Senate had never before allocated Syria as a provincia , so this was a significant usurpation of its functions, but what is more, the provincia was established explicitly as a territory that would be governed by a Roman magistrate not as a war to be fought or a task to be completed. Never before had a provincia been created after the war had been won, but this would soon be- come the standard practice. 106 By the time Augustus took control of the Roman state in the late first century , this territorial administrative meaning was dominant. In 27  , the first official attempt to define Augustus ’ s position within the Roman state divided the Roman Empire into provinciae that belonged to the People of Rome and whose governors (as we can now call them) were selected according to traditional practice, and pro- vinciae that belonged to Augustus, who chose representatives ( legati Augusti pro praetore ) to govern them in his stead. 107 These are clearly territorial entities existing separate from the magistrates who ruled them, but they also reveal a conception of the provincia as the standard administrative subdivision of the empire. Only now is it appropriate to translate the word as “ province. ” Over time, the provinces would be transferred between senatorial and imperial control, expanded, and subdivided. The provinces had become units of Roman territory . The emergence of the idea that provinces were territorial sections of Rome ’ s empire had implications for the way they were governed, at least ideally . 108 The rise of Augustus to supreme power brought the reality closer to that ideal. Roman politi- cal thought maintained a theory of beneficent imperialism. 109 Cicero expressed the  Kallet-Marx 1995, 115.  Richardson 2008, 106 – 116.  Strabo 17 . 3. 25.  For provincial administration, see Ando 2006; Bowman 1996; Eck 2000a; Lintott 1993; Richard- son 1992.  For an introduction to Roman political theory of absolute rule beginning with Cicero in the Late Republic, see Noreña 2009 . For the theory of beneficent imperialism in the Republic, see Braund 1998; Brunt 1990, 288 – 323. The Roman Empire 263 idea often. In a letter to his brother, who was governing Asia at the time, he says, “ In my opinion all who govern others are bound to regard as the object of all their actions the greatest happiness of the governed. ” 110 As long as provinciae were wars to be won, the inhabitants of the area could be treated as enemies or inconsequen- tial, unless they had formed an alliance with Rome. When they became provinces, part of the Roman Empire, the inhabitants were assumed to be allies and thus ought to be treated well. The utilitarian rationale was that contented subjects were less likely to revolt. 111 The actual conduct of Republican governors and their subordinates rarely lived up to this ideal. 112 Starting in the mid-second century  , laws were established that allowed for the prosecution of governors by individuals seeking to recover mon- ey that had been improperly extorted from them. These laws were amended, re- vised, and multiplied, but there is little evidence, aside from a few spectacular cas- es, that they had the effect of ensuring good governance. The plaintiffs had to travel all the way to Rome and secure the services of an aristocratic patron (see below) to represent them in court against another powerful aristocrat. Even then the accused was often acquitted. 113 Cicero ’ s letter to Quintus, mentioned above, illustrates several aspects of Roman governance and the abuses that Rome ’ s subjects normally suffered. 114 Cicero dwells at length on Quintus ’ s exercise of self-control in such a wealthy province: What can be imagined so striking or so desirable as … that the inhabitants are not being ruined by your progresses, drained by your expenses, agitated by your approach? That there is the liveliest joy , public and private, wheresoever you come, the city regarding you as a protector and not a tyrant, the private house as a guest and not a plunderer? 115 The role of the governor was still based on that of a military commander who oper- ated without any check on his power. As a governor traveled throughout his prov- ince with his entourage, he was entitled to requisition supplies, lodging, and mon- ey , and most took full advantage of this privilege. Hosting Roman officials and billeting soldiers could be a heavy burden, and cities would sometimes pay large sums of money to the governor to avoid it. 116 In addition, cities would vote to give large sums of money as honorific gifts not only to the governor, but to the People of Rome. Cicero boasts of outlawing the voting of money for public spectacles in  Cicero Epistulae ad Quintum fratrem (Cic. QFr. ) 1. 1. 24, trans. Shuckburgh.  See, for example, Livy 8. 13. 16; Sallust Bellum Iugurthinum 102. 6.  For the ways Republican Roman officials enriched themselves in the provinces, see Tan 2017 , 68 – 90.  Betts and Marshall 2013; Brunt 1990, 53 – 95.  I focus here on Cicero ’ s praise of his brother ’ s exceptionally good government of Asia because it reveals what was considered normal more clearly than his condemnation of V erres ’ s bad govern- ment of Sicily in In V errem (Cic. V err. ) 1 and 2.  Cic. QFr. 1. 1. 8 – 9 , trans. Shuckburgh.  Cicero, Epistulae ad Atticum 5. 2. 1. 270 Eli J. S. Weaverdyck Greece sent numerous embassies to the Senate to settle disputes among themselves. Under the Principate the same sorts of conflicts were settled either by the emperor or the governor. In addition, cities could ask governors to step in to enforce the obligations of individuals, such as the payment of debts or fees by officials. 147 Indi- viduals, in turn, could turn to the governor or the emperor to obtain immunity from office holding and the financial burdens associated with it. 148 Governors were also called upon to ensure the cities ’ rights to public lands that had fallen into private hands. Essentially , when a city ’ s authority was not sufficient to deal with a problem, it called upon the central government to step in. An important area in which imperial officials intervened against the wishes of cities was in the management of civic finances. 149 Under the Principate, cities were generally forbidden from collecting direct taxes and had to raise money through indirect taxation, public monopolies, the leasing of public land and buildings, en- dowments, and fees on magistrates. They were also forced to obtain imperial per- mission to levy new taxes or to increase the size of their council (and thus the fees that were paid by councilors). Governors had the right to audit a city ’ s accounts, unless the city enjoyed the privilege of immunity from such oversight. Sometimes a city ’ s finances were so troubled that the governor or emperor appointed a special magistrate with extraordinary powers to correct them. Occasionally , such an official was appointed to an entire province. This allowed the Roman state to bypass the privileges granted to certain communities without destroying the entire system. These extraordinary officials represent the most intensive expression of central pow- er in municipal affairs, but they were only occasionally appointed and held office for relatively brief periods. 150 More pervasively , the central state imposed control over cities ’ largest expenses: public buildings. These buildings were an important part of the intercity competition described above. 151 Governors were frequently in- volved in building projects in various capacities, including organizing their financ- ing and maintenance. By the early third century at the latest, it was illegal for a city to build a large, public building without the permission of the emperor, and it was the governor ’ s duty to inspect existing buildings, see that dilapidated ones were repaired, and ensure the completion of any unfinished constructions. 152 Roman governors had the discretionary power to interfere deeply in municipal administration, but in practice their ability to do so was constrained by time and the personnel at their disposal. The paucity of officials sent out by the imperial government led Peter Garnsey and Richard Saller to describe Roman administration  E.g., Pliny Epistulae (Plin. Ep .) 10. 17 , 23.  E.g., Aelius Aristides Orationes (Aristid. Or. ) 50; Eck 2000a, 277 .  G. P . Burton 2004.  G. P . Burton 2004, 336 – 341.  Gleason 2006.  Dig. 1. 16. 7 . [Document text truncated for crawler view.]