The publication in 1974 of the first volume of Immanuel Wallersteins The Modern WorldËSystem¹ was a milestone in the historiography of world history. For the first time the whole complex history of the world since the sixteenth century CE had been presented as part of single integrated process in which social and political structures were correlated with regional economic roles. Wallersteins by now familiar classification of world regions into core, semi-periphery, and periphery provided an elegant framework for analyzing the socio-economic history of the various regions in terms of their function in the World System. Wallersteins scheme has been criticized on many grounds including putative Eurocentrism and overemphasis on the significance of economic factors in history.²  Probably, no aspect of his work has drawn more criticism, however, than his insistence that the world system was uniquely characteristic of modern history because its appearance was dependent on the emergence of modern capitalism.

      Wallerstein did not deny the existence of interregional trade prior to the sixteenth century CE, but he insisted that it was primarily a small-scale trade in luxuries that operated within an essentially political framework provided by what he called world empires. Wallersteins denial of the applicability of World System analysis to antiquity and the middle ages has seemed unconvincing to many scholars, and there have been numerous attempts to identify earlier World Systems. The most far-reaching of such attempts has been that of Andre Gunder Frank,³ who has argued in numerous works that the current World System did not originate in the sixteenth century CE but actually has existed for over five thousand years, beginning with the economic integration of Egypt and Mesopotamia in the early fourth millennium BCE, and expanding outward continuously from that original core until it reached its present world-wide extent. 

      Gunder Franks expansive reconstruction of the history of the Modern World System has found few followers.  More common have been attempts—primarily by archaeologists‹to identify smaller and more localized pre-modern øWorld SystemsÓ in which various regions formed functionally integrated trading systems organized around a regional economic core.⁴ All such attempts to extend strict World System analysis into pre-modern times, whether general such as that of Gunder Frank or more limited archaeologically defined World Systems, have foundered, however, on their inability to demonstrate that ancient or medieval states were able to project sufficient power‹political, military, or economic‹to reconfigure neighboring regions as semi-peripheries and peripheries as required by conventional World System theory.

      Janet Abu-Lughod in Before European Hegemony proposed a less ambitious but potentially more widely applicable model. According to Abu-Lughod, a Medieval World System that was made possible by the Mongol Conquests of the thirteenth century CE preceded Wallersteins modern world system. Unlike the Modern World System, however, the Medieval World System was not functionally integrated around a central core, but segmentary, consisting of a series of regionally defined trading subsystems—each with its own core and periphery—that interlocked to create a single world trading system extending from Europe to China.  In this system the Indian Ocean and its two great gulfs‹the Red Sea and the Persian‹formed a key link, connecting the Mediterranean basin, Near East, and East Africa to the Indian subcontinent, itself the bridge between the Indian Ocean circuit and Southeast Asia and China.⁵

THE ANCIENT INDIAN OCEAN TRADE

      Although Abu-Lughod confined her analysis to the high Middle Ages, the economic pattern described by her was not unique. Similar economic configurations also can be identified in antiquity. Specifically, an almost identically structured ancient world trading system in which the Indian Ocean also served as a key link between east and west emerged in the first century BCE and lasted until the third century CE. Study of the Ancient World System as a whole has hardly begun.  Synthetic works are few,⁶ while most scholarship has been philological in character, concentrating on identifying the places and products mentioned in the sources, relating written and archaeological evidence, and establishing a reliable chronology for the various phases of the Ancient World System in general and the Indian Ocean trade in particular. Moreover, as is true in the case of the Medieval World System, the primary focus of scholarship concerning the Indian Ocean trading system in antiquity has been its connection with historical developments in the Indian subcontinent.⁷ The purpose of this paper is modest: to outline the principal changes in the Indian Ocean trading system in the late first century BCE and the mid-first century CE and to examine the impact of those changes on the two principal states in ancient Northeast Africa, Kush and Axum. 

      The difficulties in such an undertaking are formidable. The predominantly philological character of ancient Indian Ocean studies is a reflex of the inadequacies of the extant sources when compared with what is available to historians dealing with the Modern World System or even the Medieval World System. Although written sources exist in a wide variety of languages including Greek, Latin, Sanskrit, Geez, ancient south Arabic, Tamil, Sanskrit, and even Chinese, these sources are limited in quantity and quality. The vast majority are brief references or allusions without clear chronological or socio-economic context that occur in a wide variety of literary and technical works dealing with other subjects. Only one short work‹the so-called Periplus of the Erythraean Sea written by an unknown but clearly knowledgeable author‹describes the Indian Ocean trading system as whole, and it has required over a century of scholarship just to convincingly demonstrate that this text is to be dated to the mid-first century CE.⁸

      Also lacking are detailed business records and correspondence that would permit studies of changes over time in the mechanics, content, and volume of the trade, although the recent publication of a papyrus discovered in Egypt detailing a consignment of goods from India in the early second century CE raises the hope of the discovery of further examples of this type of evidence.⁹ These deficiencies in the written sources are only partially offset by the steadily growing body of archaeological evidence, which allows us to track the origin and distribution of goods but not how they reached their various destinations or why.¹⁰ As a result, studies of the Indian Ocean trade and its historical significance are inevitably schematic with numerous gaps in our knowledge and understanding and greater clarity concerning the general outlines of the trades history and structure than its details, and that is no less true of this paper.

      As already mentioned, thanks to the fortunate survival of the Periplus of the Erythraean Sea we have a comprehensive overview of the Indian Ocean trading system at one critical point in its history, namely, in the mid-first century CE, when the effects of the wide spread use of the southeast monsoon by ships bound for India were making themselves generally felt for the first time. Archaeological evidence and cuneiform sources, however, indicate that the origins of the Indian Ocean trading system lay much further back in history. References to the presence of Indian language interpreters and resident Indians in southern Mesopotamia suggest that a well organized coastal trade between the Indus Valley civilization and Mesopotamia via the coasts of the Indian Ocean and Persian Gulf already existed in the mid-third millennium BCE.¹¹ 

      While the collapse of the Indus Valley civilization in the mid-second millennium BCE temporarily interrupted this trade, the connection was not only reestablished in the early first millennium BCE but extended westward to include already existing trading systems in southern Arabia and the Red Sea basin, thereby making it possible for the first time for Indian goods to reach Egypt and the eastern Mediterranean basin entirely by sea as well as by the millennia old river and caravan routes running through Mesopotamia and Syria.¹²  Explicit evidence for the scale and contents of the early Indian Ocean trade is lacking. The emphasis on the rarity of Indian goods in the extant sources suggests, however, that it was still relatively small in scale.  References to the presence in western Asia of various Indian agricultural products including rice and humped cattle, domesticated chickens and peafowl, as well as to Indian textiles, and the existence of loan words in Greek and other Mediterranean languages for Indian spices such as pepper and ginger indicate that many of the products typical of the fully developed late Hellenistic and early Roman period trade were already being traded westward.¹³

      Although the geographical framework of the Indian Ocean trading system established in the early and middle first millennium BCE remained essentially unchanged for the rest of antiquity, the Periplus of the Erythraean Sea and other sources including Pliny the Elders Natural History and Ptolemys Geography reveal major changes in its scale, organization, and conduct during the late Hellenistic and early Roman imperial periods. Those changes were three fold. First, direct sea contact between India and the Mediterranean became more common as a result of the increasing use of the southeast monsoon by merchantmen based in Roman Egypt. Second, the explosion in Roman demand for south Indian products including textiles, pearls and especially pepper, which rapidly became a staple of Mediterranean cuisine,¹⁴ increased the scale and value of the trade to a level clearly beyond that of a limited trade in øluxuries,Ó particularly with the use of large ships, each one of which could carry more than a hundred times the bulk of the consignment of three and a half tons of ivory, textiles and spices detailed on the papyrus document mentioned above.¹⁵ Third and finally, Roman demand for south Indian goods meant that the focus of the trade shifted during the first century CE from its traditional center in northwest India to southwest Indian ports such as Muziris, modern Cranganore in southern Malabar. 

       The organization and development of the direct sea trade to India is outlined in an important passage of the Natural History of the first century CE encyclopedist Pliny the Elder,¹⁶ which makes it clear that the underlying reason for the radical transformation of the Indian Ocean trade at this time was the desire by Roman merchants to increase profits by gaining direct access to the sources of the desired products. As was true in the Middle Ages, the ancient Indian Ocean trading system traditionally had consisted of a series of interlocking regional trading networks, so that the same person rarely traversed the whole system. Goods bound for the west or east, instead, passed through the hands of numerous intermediaries with attendant rises in cost as they changed hands at each stage. There was, therefore, a strong incentive for peoples at either extreme of the system to reduce costs and increase profits by eliminating the various intermediate stages and the ømiddlemenÓ who dominated them.

      The initiative in establishing direct contact between the Mediterranean and India appears to have come from the Indian side in the late second century BCE.   The Ptolemaic rulers of Egypt responded in kind, dispatching at least two official trading missions to India while establishing a Greek presence on the island of Socotra just outside the southern entrance to the Red Sea alongside the already existing Indian and Arabian settlements on the island and charging a special official with general oversight over commerce in the Red Sea and the Indian Ocean.¹⁷  The Ptolemies Roman successors continued these initiatives with such dramatic results that by the beginning of the Common Era the Greek geographer Strabo could observe that øin earlier times not twenty vessels dared to traverse the Red Sea so as to emerge beyond the straits; but now great fleets are sent as far as India€whence the most valuable cargoes are brought back to Egypt and thence exported again to other places.Ó¹⁸ 

      The ramifications of these developments for countries involved in the Indian Ocean trade were far-reaching.  One result was the establishment of relatively regular diplomatic relations between Rome and India with the first Indian embassy to Rome being attested as early as 20 BCE, barely a decade after the Roman conquest of Egypt.¹⁹ Unfortunately, the sources are silent about the purposes of these embassies, but the Periplus of the Erythraean Sea suggests that facilitating trade between Roman Egypt and India may have been among them since it reveals the existence in the first century CE of a sophisticated commercial infrastructure intended to accommodate such trade including a system of designated ports of trade together with harbor pilots, legal and military protection for foreign traders,²⁰ and the presence of a small western trade diaspora in a number of Indian ports. Contemporary archaeological and textual evidence documents the substantial impact made by the increased availability of Indian on Roman imperial material culture.²¹

       Equally important, the sources also clearly indicate that there was a close correlation between state formation and political relations between states in western and especially southern India and shifts in the focus of the Indian Ocean trade, encouraging, according to Kenneth Hall:²²

The focus of scholarship on the eastern segments of the Indian Ocean trade is understandable since, as was true of the Medieval World System, India was the primary goal of the trade and the sources, not surprisingly, are fullest for that aspect of it.  It has had, however, the unfortunate effect of obscuring the fact that expansion of the Indian Ocean trade also had similarly important impacts on the development of states at the other end of the system in northeast Africa.²³

ANCIENT NORTHEAST AFRICA AND THE INDIAN OCEAN TRADE

      Like its eastern counterpart, the western segment of the Indian Ocean trading system was already millennia old and complex in structure, when the Periplus of the Erythraean Sea reveals it as it functioned in the mid-first century CE. Also like its eastern counterpart, the western segment of the Indian Ocean trading system was not a single unified system but was itself composed of three relatively independent regional trading networks: a southern network comprising approximately the area of present day Somalia and the coast of East Africa at least as far south as Zanzibar that was dominated by the kingdom of Himyar in ancient southern Arabia; another dominated by the kingdom of Axum located in the highlands of modern Eritrea and focused on the southwestern coasts of the Red Sea, a region that was known in antiquity as the Incense Bearing Country; and, and, finally, an inland system centered on the Nile that was dominated by the Kingdom of Kush in the Central Sudan.  What was common to all three of these networks was their function as a periphery supplying primarily African products‹gold, ivory, incense, hard woods such as ebony, exotic animals and animal skins, and, of course, slaves‹to an economic core formed by Egypt and its neighbors in the eastern Mediterranean basin.

      While all three components of the western segment of the Indian Ocean trading system were exposed to the stresses resulting from the rapid transformation of the Indian Ocean trade in the late first century BCE and the first century CE, it was the Nile valley portion of the system, which was the oldest and for most of its history the most important of them, that proved to be most vulnerable to those stresses. The Nile Valley has been described for good reason as a øCorridor to Africa.Ó²⁴  It forms an over a thousand mile long narrow oasis running through the great desert that extends across the whole of North Africa from the Atlantic Ocean to the Red Sea, thereby providing the most convenient route by which various sub-Saharan African products could be transported to Egypt whence they could be distributed throughout the Mediterranean basin.  Throughout antiquity and beyond the rulers of Egypt dreamed of gaining direct access to these sources by securing control of the whole of the Nile corridor.

      The formidable glacis of lower Nubia with its harsh deserts and numerous stretches of rapids, the so-called cataracts of the Nile, however, hindered achievement of that goal. These obstacles were not, of course, insurmountable. The kings of the New Kingdom in the second millennium BCE and the Twenty-Fifth Dynasty in the First Millennium BCE demonstrated that they could be overcome with sufficient effort, but most Egyptian dynasties were unwilling to expend the required effort and resources, preferring instead to achieve their goals by less costly means. They maintained strong defensive positions in Lower Nubia and freely used military force to intimidate or even subdue potentially hostile states further south in Nubia, while relying on diplomacy and trade to assure the steady flow of African products north to Egypt.  Geography, therefore, created the opportunity for the emergence in the second and first millenniums BCE of a series of states between the third and sixth cataracts of the Nile‹the region the Egyptians referred to as Kush‹that prospered by organizing the collection and transport to Egypt of African products.²⁵

      Despite superficial differences in the details of their organization, these states shared certain common characteristics. Specifically, they all were quasi-confederate in structure, consisting of a union of local chiefdoms united by their allegiance to a paramount chief or king, who was believed to be the divinely appointed ruler and protector of Kush. Ideology was reinforced by the kings ability to attract the loyalty of followers by various means including the Egyptianization of elite culture and especially the kings control of  øprestige-goods exchange, the wealth of which materialized a ruling ideology and acted as a political currency to finance the leadership.Ó²⁶ In the case of Kush that meant goods imported from Egypt, which Kushite kings used to enhance the unique status of themselves and their supporters through redistribution, a process that is reflected archaeologically in the fact that such imported prestige-goods are found only in Kushite royal and elite residential and burial sites.

      It was the significance of such prestige-goods in the political economy of Kush that determined how trade was conducted in the Nile corridor. Although a limited non-official trade in African goods existed, the role of private merchants was limited. Instead, trade between Kush and Egypt was organized primarily as a royal monopoly in which Kushite kings provided goods directly to their Egyptian counterparts or their representatives. The system is known primarily from Egyptian sources, which refer to the goods involved as inw, øofficial giftsÓ, in the Pharaonic period and dora, øgiftsÓ, or xenia, øfriendship giftsÓ, in the Hellenistic Period and interpret their arrival in Egypt as recognition by Kush of Egyptian suzerainty in return for which they were rewarded with what Egyptian texts call  øthe breath of lifeÓ.²⁷ Archaeological evidence from Kush and textual evidence from the Near East, where the system also functioned, leave no doubt that the Kushites received not only the øbreath of lifeÓ but also substantial counter-gifts which provided them with the prestige goods they redistributed to their supporters. As Timothy Earle has pointed out, however, societies that rely on such  øinternational prestige-gift exchangeÓ are øvulnerable to forces beyond the influence of local action,Ó²⁹ and Kush was no exception. Unlike the situation in India, those forces did not involve changes in the source of goods—the various goods involved in the western trade network were readily available throughout northeast Africa‹but in the emergence of the Red Sea as the primary trade route in the region as a result of the great expansion of direct sea-borne commerce between Roman Egypt and India in the late first century BCE and the first century CE.

      Egyptian attempts to acquire an alternate source of African and south Arabian goods by establishing contact via the Red Sea with the region they called  øPuntÓ, essentially the coastal regions of the Sudan and Eritrea,³⁰ are attested as early as the third millennium BCE, but the costs and difficulties of maintaining a naval presence on the Red Sea combined with the convenience of the Nile corridor limited the frequency and extent of such trade, which remained largely confined to episodic incidents of øprestige-giftÓ exchange, until the Hellenistic period, when the construction of a canal linking the Nile and to the Red Sea³¹ and the creation of a series of an Elephant hunting stations on the Egyptian and Sudanese coasts of the Red fostered the development of a small private trade with these regions along side the long established official øprestige-giftÓ exchange system.³² The late Hellenistic and early Roman period expansion of the Indian Ocean trade rapidly transformed this almost three millennia old system by greatly increasing the number and size of private ships traversing the Red Sea and using its ports, where African goods could be acquired for shipment either north to Roman Egypt or south and east to ports further along in the broader Indian Ocean trading system.^32a Geography dictated that it would not be the largely landlocked kingdom of  Kush but that of Axum in the highlands of modern Ethiopia that benefited from these changes.

      Evidence for early history of the kingdom of Axum is scant, but it indicates that, unlike Kush, Axum was of relatively recent origin.³³ Although archaeological evidence suggests that its roots lie in the early first millennium BCE, when South Arabian colonists founded a series of small kingdoms in the territory of modern Eritrea and Tigray in Ethiopia, the kingdom of Axum began its rise to prominence only in the late Hellenistic period, when the ruler of one of these kingdoms, that of  the Habasha (=Abyssinians), astutely chose the site of the future city of Axum as his capital. Located on the Ethiopian plateau with ready access to the upper Nile Valley and its hinterlands on the west and to the Red Sea on the east through the port of Adulis near modern Massawa, Axum was ideally situated to profit from the new conditions in the Red Sea created by the expansion of the Indian Ocean trade, and the kings of Axum took advantage of this opportunity, creating as was being done in contemporary India, the essential infrastructure to accommodate the increase in shipping by establishing in agreement with Rome Adulis as a designated port of trade and encouraging the development of a trade diaspora in their kingdom.³⁴

      As a result of these developments, Adulis had become by the mid-first century CE the principal center for the export of goods from the coastal regions of the Red Sea and its hinterlands. Indeed, the Periplus of the Erythraean Sea reveals that by that time ivory collected in Sennar in Kushite territory was being exported through Adulis instead of being brought to Meroe for transportation to Egypt through the Nile corridor.³⁵ Their second and third century CE successors continued these policies, extending their rule over most of the hinterlands of the southern Red Sea basin and building and maintaining a caravan route to Egypt that bypassed the Nile corridor entirely,³⁶ thereby confirming Axums rule as the principal supplier of African goods to Roman Egypt in particular and the Mediterranean basin in general and winning recognition as one of the four great kingdoms in the world known to the peoples of the Near East and Mediterranean basin.³⁷   

      It took time for the effects of these developments on Kush to become evident. Indeed, the late first century BCE and the first half of the first century CE was in many ways the climax of Kushite history. Peaceful relations between Rome and Kush resulted in unprecedented prosperity. The evidence is primarily archaeological and includes: widespread building and repair of temples, palaces, and other elite structures, expansion of agricultural settlement and installations, and, of course,  numerous objects imported from Roman Egypt that have been discovered on Meroitic sites.³⁸ Contained in this latter substantial but little studied body of evidence is a wide variety of small but high quality metal, glass, and ceramic objects including eating utensils; items related to personal adornment such as rings, jewelry, beads, and mirrors; and household furnishings and decorative objects.  In addition, the existence of classical motifs and images in Roman period Meroitic art and architecture, Greek auloi and statues of aulos and lyre players,³⁹ and a column drum inscribed with the Greek alphabet may imply the presence also in Kush either of individuals skilled in their production and use or appropriate models. Kush even played a minor role in the Indian Ocean trade through a narrow corridor leading from the vicinity of the fourth cataract of the Nile to the coast of the Red Sea near modern Suakin.⁴⁰ 

      By the end of second century and early third CE, however, conditions had changed dramatically. The emergence of Axum as Romes primary source of African goods gradually reduced diplomatic contact between Kush and Roman Egypt as Roman policy toward the Upper Nile valley increasingly focused on the defense of the southern frontier of Egypt. Consequently, relations with the peoples of Lower Nubia took precedence of those with Kush, thereby drying up the prestige goods trade on which the Kushites monarchys wealth was based. Again the evidence is primarily archaeological. At Meroe, the Kushite capital, significant new temple construction virtually ceased, the size of royal pyramids shrank to a fraction of that of their first century CE predecessors, and, most importantly, evidence for imported Roman luxury goods in royal burials almost totally disappears. Increasing impoverishment of the Kushite central government was paralleled by the appearance for the first time of rich elite residences and cemeteries in or near important regional administrative centers in Lower Nubia together with epigraphical evidence pointing to the transformation of important administrative and military offices in the region into hereditary fiefdoms,⁴¹ thereby, weakening the Kushite monarchys control over its peripheral territories and exposing it  to attack and ultimately conquest by neighboring peoples and states such as the Noba and Blemmyes and especially Axum. As usual, the sources to trace the process in detail are lacking.  When the evidence allows us again to observe conditions in the region in Late Antiquity, however, the kingdom of Kush has disappeared, being replaced by a series of warring successor states, and the Nile corridor has ceased to be a significant artery for the transmission of African goods to Egypt; while Axum has become Romes principal ally in Africa and the main intermediary between India and the Mediterranean basin in a reorganized and smaller but still vigorous Indian Ocean trade.  

CONCLUSION

      This paper is only a preliminary sketch of a long and complex historical process. Much of importance has had to be omitted, partly for lack of space such as the impact of changes in the political economy of Roman Egypt on trade relations with Kush and Axum or the role of trade diasporas in the spread of Christianity and, later, Islam in the Red Sea basin; and partly for lack of evidence such as the impact of the Indian Ocean trade on cultures located on the east coast of Africa.⁴² Hopefully, however, it has also demonstrated the error of treating of developments in ancient northeast Africa in isolation from the rest of the history of events in the Indian Ocean region as a whole.