Indian Ocean Trade in the First Millennium c.e.: Taking the Romans out of Indo-Roman?

Abstract

The study of the Western Indian Ocean in the first millennium is a dynamic and exciting field, in which scholarship, especially from within the Indian Ocean region itself, is expanding rapidly. It is experiencing a period of major, but not necessarily disruptive, change, to its core questions, terminology and periodisation. This article offers an overview of the study of Roman trade with the Western Indian Ocean (sometimes termed ‘Indo-Roman studies’) from the early 2000s to the present. It examines key developments in the field, including the changing scope of analysis in terms of period, region and evidence; the impact in the field of an increasingly global focus and efforts to decolonise a subject historically deeply rooted in colonial processes; and specifically the effort to provincialise or decentre Rome in historical narratives. It then suggests directions in which the field appears to be developing and makes tentative suggestions for future work.

I INTRODUCTION

The study of commercial interactions across the Western Indian Ocean in the early first millennium c.e. is a field with a long pedigree.¹ Roman texts mentioning trade with ‘India’ had been known in Europe throughout the Middle Ages, but closer European interaction with South Asia from the fifteenth century onwards provoked more intensive scholarship. The identification of Roman coins in South Asia, from at least the sixteenth century, apparently offered material proof that claims in texts such as Strabo's Geography and Pliny's Natural History were neither mere moralising hyperbole nor fantasy.² From the nineteenth century, when the first widely available edition was followed by numerous translations, the first-century c.e. Greek text known as the Periplous of the Erythreian Sea gave further impetus to a subject which was increasingly known as ‘Indo-Roman’ studies.³ This remarkable anonymous work describes what goods could be traded at harbours between the Red Sea and peninsular South Asia, alongside occasional navigational notes and other observations.

The first synthetic studies dedicated to the topic emerged in the late nineteenth and early twentieth centuries, bringing together what was by then a sizeable body of narrative and material evidence. The wider context for these studies was European imperial control of territories around the world and the global maritime routes connecting them. Since those early publications, much work has been done to find new evidence, refine our understanding of existing evidence, broaden the scope of the field (and therefore what may be considered evidence) and produce new syntheses. These new syntheses respond both to new materials and to changing social and political realities, from decolonisation to globalisation.

An excellent overview of the historiography of Indo-Roman studies up to the early 2000s was provided by Roberta Tomber in her 2009 book, Indo-Roman Trade: From Pots to Pepper. Tomber's work drew extensively on her unparalleled archaeological knowledge of the field and remains the clearest summary of current knowledge about movement in the Western Indian Ocean in the first six centuries of the first millennium c.e. In 2008, Grant Parker's The Making of Roman India did a similarly effective job of presenting the ancient and modern historiography pertaining to Roman ideas about India. The aim of this article is to summarise some of the distinctive features of ‘Indo-Roman studies’ from the early 2000s onwards. These have been two and a half decades of very active scholarship, such that comprehensiveness is neither achievable nor attempted. The terms, scope and practice of the field have been substantially redefined, though it remains recognisably indebted to its earlier history. One aspect of this redefinition has been an expanded field of view: Roman connections with the Indian Ocean and Roman understandings of the Indian Ocean are no longer the only or, often, the most important focus of study. Nevertheless, for this article I have aimed to retain a perspective on the subject likely to be most relevant to specialists in the Roman Empire.

To summarise so much scholarship in a clear and usable form, I have arranged this survey under three themes, which I believe are central to the concerns and dynamic of the field: interdisciplinarity and the quest for new evidence; the global turn and decolonisation; and the move to ‘provincialise’ Rome. I then turn to what I believe are the most important directions of travel in current research. Inevitably, such a survey must be personal, so I begin with a brief biography, as it necessarily informs my views, priorities and identification of key changes.

My background was originally in history, with a focus on the medieval Eastern Roman (Byzantine) Empire. This was followed by the study of archaeology, especially in the Aegean. My route into Indian Ocean studies was via numismatics, with the study of the Late Roman/Byzantine coins found in peninsular South Asia.⁴ My subsequent work has examined Mediterranean perspectives on the Indian Ocean, the movement and reception of coins within the Indian Ocean, and especially in peninsular South Asia, and the development of state structures in the region.⁵ While I have endeavoured here to write a helpful review for readers with a general interest in the development of Indo-Roman studies, and one that is fair to the subject as a whole, it is inescapable, in a field that welcomes so many branches of scholarship and such an international body of scholars, that my survey will be different from that which another might have written. Perhaps most obviously, the importance of periodisation, scholarship from South Asia and numismatics loom large in my work and find a prominent place in the pages that follow.

Before moving to discuss key developments in the field, it is worth saying that the study of ancient Indian Ocean trade has always been, and remains, a multi-lingual field, with substantial and significant scholarship in various languages, and especially in some areas.⁶ Nevertheless, this survey focuses mainly on English-language material, both for reasons of space but also because English is currently the primary language of publication by scholars working within the Indian Ocean region. The expansion of scholarship produced within the Indian Ocean region, and by scholars from those regions (unlike earlier phases, which were substantially led by colonial operatives from Europe), has been a major and welcome change in the field. Inclusion of such work is privileged wherever practical.

II NEW EVIDENCE, NEW TERMS

As it has been from its inception, the study of Indian Ocean trade in antiquity remains strongly interdisciplinary. Early scholars usually had education in Latin and Greek, and the interpretation of texts in those languages, especially the Periplous, remains a productive source of research.⁷ Nevertheless, as already noted, the combination of texts and coin evidence was crucial to the early development of the field. When, from the mid-twentieth century onwards, the archaeological findings of sites in South Asia (including Sri Lanka), along the Red Sea coast and further south in East Africa and in southern Arabia came to light, these too were rapidly incorporated into discussion, offering insight into inter-regional trade networks and daily life in port and coastal sites, and indications of the sharing of technology as well as goods.⁸ Most studies today make use of a range of evidence types, even if the selection is shaped by each scholar's starting expertise.

Nonetheless, though interdisciplinarity is baked into the field, new sources have significantly changed not just the scope of Indian Ocean studies, but also some of its core debates, approaches and terminology.⁹ The largest and fastest growing pool of new evidence for ancient Indian Ocean trade is undoubtedly coming from archaeology. Not only do new sites continue to be identified, but long-term excavation at various places has improved techniques and created changing and more nuanced narratives of life. In addition, archaeology is also providing the most significant evidence for deepening our understanding of the wider economies within which long-distance trade operated, as discussed further below. In all areas of the Indian Ocean, archaeology now offers independent narratives that do not simply corroborate textual sources, but exist alongside, challenge or refine them.¹⁰

Recent changes in the archaeological understanding of this topic have included the re-identification and re-dating of ceramic types from South Asia. This has revealed trade networks within the region, emphasising the shorter-distance and denser local connections of ports that were traditionally viewed primarily in terms of their long-distance links, especially to the Mediterranean.¹¹ In addition, examination of food use, particularly the large-scale consumption of rice, at the Red Sea site of Berenike suggests the probable existence of a resident, or seasonally resident, South Asian population at the port.¹² This and other archaeological clues indicate that a wider range of people actively participated in long-distance trade than comes across in texts mostly written from a Roman perspective.¹³

In 2009 the discovery on the island of Socotra, just south of Arabia, of hundreds of late antique graffiti, mainly by South Asian sailors from the region of modern Gujarat, probably en route to Arabia, underscored the mobility of varied groups across the ancient Indian Ocean.¹⁴ The Socotra finds also added to the wider contribution made by the documentation, decipherment and interpretation of inscriptional evidence all around the Western Indian Ocean. Such research, especially using the complex but rich epigraphy of southern Arabia, has highlighted the role of maritime trade in enabling local elites to consolidate and expand their power in the area.¹⁵ This has mirrored similar discussion of the importance (or insignificance) of overseas trade to state formation in southern peninsular South Asia, based on land-grant and temple inscriptions.¹⁶ Inscriptions and graffiti, once thought unpromisingly brief or lacunose, are now recognised as a vital body of evidence. In particular, they enable comparison of the political and social structures of areas outside the Roman Empire with the much better known landscape of the Roman Empire and can help to fix sites, events and individuals in chronological relation to one another.

Analysis of biological material at various sites, apart from food remains, has yielded information about the wood used to make and repair ships, exposed techniques of sail-making and boat repair, and illustrated a Western Indian Ocean craft sphere in which local practices and products were shared and combined in ways that increasingly obviate earlier ideas of ‘Roman boats’ (which were, for example, hypothesised in earlier scholarship, on the basis mainly of textual evidence, to have been uniquely able to exploit monsoon winds because of being sturdier than Arabian or South Asian craft).¹⁷ There is still some discussion of the possible logistical preference on the part of Roman merchants, especially in the second century, for exceptionally large ships, but this can no longer be seen in terms of technological determinism or Mediterranean superiority.¹⁸

There remain deep imbalances in the spread and density of archaeological evidence across the Western Indian Ocean, owing to a number of factors, including geopolitical disruptions, local heritage policies and structural biases within the field. Diagnostically Roman objects found in the Indian Ocean region but outside the empire are still easier to identify than non-Roman objects in the Roman Empire, due to the Roman Empire's creation of a very distinctive material cultural signature but also to overwhelmingly greater scholarly attention given to Roman material over generations compared with perhaps any other Afro-Eurasian political system. Nevertheless, considerable effort is gradually redressing this balance. A small but growing number of ostraca in the Red Sea with South Asian graffiti on them contradicts traditional stereotypes of dynamic, mobile Romans versus passive local populations.¹⁹ A silver coin of a South Asian Western Kshatrapa ruler has also been excavated at Berenike.²⁰

As these developments suggest, the geographical scope of the category ‘Indo-Roman’ has widened considerably. Roman texts, most especially the Periplous, make it clear that both East African and Arabian coastal communities were deeply enmeshed in Indian Ocean trade.²¹ Nevertheless, it has been archaeological excavations at coastal sites in East Africa and in Arabia that have expanded the provable range of communications, especially southwards, into modern Tanzania.²² Meroë, too, has emerged as linked to an Indian Ocean network that reached much further inland than earlier scholarship often guessed.²³ Analysis of the genetic structure and movement of rice has shown deep historical ties between South Asia and Africa. This exchange predated any significant Roman presence in the Western Indian Ocean but seems to have intensified in the first century c.e., likely as a consequence of growing Roman demand for Indian Ocean products.²⁴ Though requiring caution when applied over such long distances in time, ethnographic work with fishing, sailing and shipbuilding communities has also come to play an increasingly important role in Indian Ocean studies.²⁵

Shipwreck archaeology has not yet proved as promising in the Western Indian Ocean as in either the Mediterranean or, for slightly later centuries, the Eastern Indian Ocean,²⁶ but the discovery and excavation of the Godavaya wreck off the coast of Sri Lanka in 2008 has shown the potential for such material to come to light in future and, in this case, has illuminated the shorter-distance exchanges which characterised Sri Lankan and peninsular South Asian trade.²⁷ The wreck, dated to around the first century c.e., carried glass ingots, ceramics and quern stones. Its location is significant, lying off the coastal site of Godavaya, where a second-century c.e. inscription granted trade revenues to a local monastery and where a port and village have also been located.²⁸ The discovery speaks to a recurring question: how and when did political authorities involve themselves in long-distance maritime trade?

Alongside other material culture, numismatics continues to be an important sub-field within Indian Ocean studies, with frequent new finds of Roman coins or imitations.²⁹ How, or whether, to integrate finds coming to light via the open market in antiquities is a growing methodological challenge. There is also now far more awareness of non-Roman coins circulating in the Indian Ocean region as these types become better documented and are more actively looked for in western contexts.³⁰ In addition, imitation and reuse have been recognised and are now studied as creative and proactive responses to cultural interaction, rather than as acts of dependence or ‘influence’, complicating simpler narratives of Roman cultural dominance. Close analysis of how, when and where Roman coins in South Asia were imitated, made into jewellery, or incorporated into thought worlds has shifted perspective from the movement of Roman goods to their meaning and use in foreign contexts.³¹

Numismatic data have also been fundamental to pinning down the date of the South and Central Asian ruler Kanishka.³² The reign of the Kushan emperor was the basis for an era used extensively in dating clauses in South and Central Asian inscriptions. Some of these in turn referred to people and events related to Indian Ocean trade, but until the early 2000s no agreement had been reached, despite decades of debate, about the correlation of the Kanishka Era with modern calendars. The resolution of the start of the Kanishka Era to 127 c.e. has given greater certainty to inscriptional, textual and archaeological evidence and enabled the aligning of chronologies between the Mediterranean and large parts of the Indian Ocean littoral.

Papyrology has become yet another new frontier in debates about maritime trade. Efforts to catalogue many of the world's disparate papyri collections are ongoing and the http://papyri.info online catalogue now enables researchers to bring together fragments, compare translations and view transcriptions even of material kept and published in very different places. The particular grammar, vocabulary and palaeography of papyrological sources is consequently now far better understood than decades ago. At the centre of these papyrological debates has been an enigmatic document often referred to as ‘the Muziris papyrus’ (PVindob. G40822), which details a loan arrangement for the movement of goods from the Red Sea to Alexandria and taxes payable on the cargo of a ship called the Hermapollon, which seems to have carried pepper, malabathron, Gangetic nard, ivory tusks, ivory fragments and tortoiseshell.³³ The papyrus has been dated to the mid-second century c.e. and numerous scholars have refined its interpretation, culminating in 2020 in one of the most important single studies on Roman trade with South Asia, The Indo-Roman Pepper Trade and the Muziris Papyrus by Federico de Romanis.

It is not a survey of Indian Ocean trade per se but a focused study on the Muziris papyrus itself; but such a description undersells a work which ranges very broadly, from the practice of trimming the tusks of captive elephants, to the best conditions for pepper farming, to how Greek speakers talking about sea journeys might have used the idea of ‘to’ and ‘from’ to refer to journeys out of or into the Mediterranean.³⁴ De Romanis also underscores the value of papyrological sources which do not necessarily have any connection to the Indian Ocean. They are used throughout the book to contextualise Indian Ocean material by showing how specific phrases or words might be used in non-literary settings or how widely products circulated within the Roman Empire.³⁵ The picture that de Romanis paints of second-century c.e. Roman trade with South Asia is minutely researched and moves various debates forward significantly, including about the structure of taxation along the Red Sea coast and at the borders of the Roman Province of Egypt. De Romanis's work has been critiqued for its emphasis on Roman agency, at the expense of earlier periods or other places, and its comparatively limited engagement with (South Asian) archaeology, in contrast to various kinds of written source material.³⁶ The treatment of some medieval sources, such as the Peutinger Table, is also markedly less detailed or critical than that of the Roman sources which are at the heart of de Romanis's argument.³⁷ Nevertheless, The Indo-Roman Pepper Trade and the Muziris Papyrus is without doubt a landmark in the field.

All of these developments in the range and treatment of sources have enriched the study of what was until recently termed Indo-Roman trade, so that the label no longer seems appropriate and is therefore passing away with little of the controversy or fanfare that has attended calls for terminological change in some other fields.³⁸ Most importantly, it is not simply a change of terminology, but an increasing recognition that what we are trying to understand simply cannot be reduced to trade between ‘India’ and the Roman Empire.³⁹ In this respect, the field has been shaped by trends in the world around us. The implications of a more global and inclusive framing of Indian Ocean trade are not simply an issue of identifying more and different sources of evidence. This shift has entailed, and continues to demand, other changes of practice and perspective.

III POSTCOLONIALISM AND THE GLOBAL TURN

One major change that has resulted from a more global perspective on Indian Ocean studies has been increasing interest in the Eastern Indian Ocean. The Western Indian Ocean demarcated the functional limits of Roman knowledge and experience, and therefore the effective limit of classical sources, but was not hermetically sealed from contact with or changes in the Eastern Indian Ocean.⁴⁰ While direct journeys from the Roman Empire to Southeast Asia were likely rare or non-existent, goods from the Mediterranean have been found, especially in modern Thailand but also Indonesia.⁴¹ Roman coins and their imitations demonstrate the potential value of these rare foreign objects. A lamp from the late Roman Empire has also been tracked backwards from its discovery in Southeast Asia. Drawing on Chinese-language sources for trade and travel in the first-millennium Eastern Indian Ocean, Lin Ying and Xuefei Han illuminate the importance of routes across the Bay of Bengal for connecting the Mediterranean, via South Asia, with regions further east.⁴² The work of Himanshu Prabha Ray pioneered a growing focus on Buddhism as a driver of connectivity and movement eastwards from South Asia, including by sea, and on activity in the Bay of Bengal.⁴³ While the framing of these links as proto-globalisation seems to me to go beyond the very ephemeral traces we have, they show the potential of Roman artefacts to travel far beyond their point of origin and to engage audiences even potentially at many removes.

In some ways a more difficult expansion in perspective has been chronological. Traditionally, ‘Indo-Roman’ studies formed the ‘ancient’ component of a tripartite division of Indian Ocean studies, in which ‘medieval’ has often been described in terms of the dominance of Islam or Arabs and ‘modern’ was equated with the dominance of European powers. The gaps between the most active centuries of each of these apparently quite distinct and separate eras could be quite large, so that their historiographies developed in relative isolation. In the case of Indo-Roman studies, this meant a focus on the first two centuries c.e., then until recently a comparative lack of interest in the following six or seven centuries until the conventional beginning of the ‘medieval’ period. This would then, generally, be studied by other scholars, with different linguistic, historical and archaeological specialisms in the period c. 900–1300.

There are good reasons for the development of this siloed chronology, rooted in ‘hotspots’ of evidence relating to moments of particularly intense Indian Ocean traffic. There are also good reasons for looking across the established divides. Activity in the Western Indian Ocean is extensively shaped by long-term structural features, such as the monsoon winds, which blow annually across the sea space from west to east and back again.⁴⁴ Comparisons across periods are therefore often both obvious and helpful.⁴⁵ There are, however, dangers to a longue durée perspective. Diachronic studies sometimes promise extensive coverage while in fact reifying the overriding importance of a few highly visible moments. This is particularly a problem in edited volumes with chronologically broad remits which are, in reality, a series of studies centred solidly on the traditional chronological ‘hotspots’. Other works, with a more methodological focus on the long term, can flatten out difference, assuming that evidence can be read uniformly across centuries. As my own work argues, exploring the ‘in between’ centuries in the traditional periodisation has great potential to explain the whys and hows of change. Conscious comparison between specific periods of heightened activity also has great potential to open up new lines of inquiry or resolve questions posed by lacunose evidence.⁴⁶

Archaeology and changing perspectives, for example, on Late Antiquity as a period in its own right, are both changing this landscape.⁴⁷ In 1926, Charlesworth dismissed the notion of significant activity after the reign of Constantine I (306–337 c.e.).⁴⁸ It does seem to be the case that there was a marked drop in activity during the third century c.e., in contrast to an exceptional intensity in the first and second centuries.⁴⁹ However, it is now evident that trade activity increased again between the fourth and sixth or seventh centuries, albeit not to former levels, and that this later phase of trade should be considered alongside the first- and second-century boom in Indian Ocean trade.⁵⁰ This shift in perspective emerged from the excavation of sites along the Red Sea coast, which frequently showed clear revival from the fourth century. It is also reflected in coin finds and scrutiny of texts from the fourth century onwards, including early Christian writing and the expansion of Christian communities into the Indian Ocean. Nevertheless, all of this was only possible because of the much wider change in the historiography of the Roman Empire, framing Late Antiquity as a distinct period and arguing for a smoother and longer transformation of the Mediterranean than earlier accounts had proposed. On one hand, this has opened up valuable new avenues for research. On the other, it shows the continued link between Roman and Indian Ocean historiography.

Even if some of the impetus for the expansion of Indian Ocean studies remains connected with developments in Roman studies, however, the result is inevitably to diminish the dominance of Rome in the narrative. Recently, there have also been efforts to push analysis further backwards and to see what kinds of long-distance trade might have provided the foundation for later structures, further broadening the field.⁵¹ This has included Troy Wilkinson's work on Ptolemaic activity in the Western Indian Ocean, building on references to elephant hunting in East Africa and trade routes associated with it,⁵² and Ephraim Lytle's publication of a new Hellenistic text addressing the Red Sea,⁵³ but it also extends back to Bronze Age exchanges in the Indian Ocean, for example in the recent and groundbreaking publication of Sureshkumar Muthukumaran.⁵⁴

A further challenge of the broadening scope of Indian Ocean studies concerns the relationship between geography and culture. One problem with the concept of Indo-Roman studies, and a reason for its diminishing currency, is that it pairs mismatched categories. While the Roman element refers to a distinct political unit, the Indian element was defined by how the Romans themselves understood a blurry, Eastern ‘other’ place. India, as a concept, emerged in the Hellenistic world and had no ancient analogy in the South Asian subcontinent, which was politically plural and religiously and culturally diverse.⁵⁵ Efforts to address this have been rooted in the closely linked ideas of post-colonialism and decolonisation.⁵⁶ Both terms occur widely in current discussions of ancient Indian Ocean studies, or at least their influence can be detected. They can mean a range of things, including recognition of the involvement of varied participant groups in Western Indian Ocean networks,⁵⁷ questioning of narratives giving primacy, hierarchical precedence or initiative to Romans in Indian Ocean activity,⁵⁸ and identification of activity and underlying structures which connected people across the Western Indian Ocean independent of any involvement with the Roman Empire.⁵⁹ There are different dynamics to this act of reexamination within the work of each individual scholar and across the field as a whole. For some, it has an explicitly political dimension. For many, it has opened up space for recognition of topics which earlier struggled for attention.

While they can all be seen to be driving ancient Indian Ocean studies in similar broad directions, works in the vein of post-colonial, decolonising and global scholarship can have very distinct aims. For some scholars discussed below, and predominantly based in Europe and North America, the issue is one of ‘provincialising Rome’. For other scholars, often though not exclusively working in Indian Ocean states, there is an emphasis not on provincialising Rome but on locating historical connections that were either distinct from subsequent European colonisation or opposed to its dynamic. Alternatively, works identify the ways in which modern colonial imperatives and knowledge structures shaped the interpretation of the more distant past.⁶⁰

Perhaps one of the most important works in this vein has been Rajan Gurukkal's (2016) Rethinking Classical Indo-Roman Trade: Political Economy of Eastern Mediterranean Exchange Relations. In a new reading of the Periplous of the Erythreian Sea, rather than seeking evidence for an Indian Ocean world in antiquity that simply worked without Roman involvement, Gurukkal proposes a framework rooted in analysis of more recent colonial structures. This adopts and subverts a trend visible in early Indo-Roman studies, in which it was common to model Roman trade with South Asia on a nineteenth-century interpretation of British trade with India. The result then was to project a broadly peaceful and mutually beneficial relationship shaped by the greater will, dynamism and capability of the British/Roman actors.⁶¹ Gurukkal follows the practice of drawing out similarities between modern and ancient imperial contact between South Asia and Britain and the Roman Empire, respectively, but looks instead for similarities rooted in much more recent interpretations of empire. As such, he identifies a fundamentally extractive imperial force involved in an exploitative relationship with a South Asia, or at least a peninsular South Asia, lacking the social structures to resist or engage to its own benefit. He rejects the idea of ‘Indo-Roman trade’, not because the label gives too much prominence to Rome, but because he says that there was neither a sufficiently capable ‘Indo’ element to justify such an equal pairing, nor even much that constituted trade, as opposed to naked exploitation. This study stands out not just for its desire to suggest a new framework, which is badly needed, even if criticisms of that model can be made,⁶² but also in its eschewing of a pacifist and cosmopolitan vision of the Indian Ocean past, which still tends to dominate in the field at large.

IV PROVINCIALISING ROME?

The study of Roman trade in the ancient Indian Ocean still often borrows much of its conceptual framework from the imperial and then western globalised economic models within which earlier Indo-Roman studies developed. It remains not uncommon for works to cite balance of trade or supply and demand as trans-temporally applicable concepts, or to assume state management of trade, by analogy with the role and importance of external trade to modern states. In this respect, studies at times accept one side of the now more than century-long debate about the nature of the the Roman economy and the role of commerce in the empire, without recognition of its contested nature. Alternatively, paradigms for understanding Indian Ocean trade can seem isolated from those discussions.⁶³ Mental frameworks from colonial contexts also remain visible in some of the core questions of Indian Ocean studies. The idea that there were ‘here’ places and ‘there’ places in antiquity, and that the job of Indian Ocean studies is to understand how these places were linked by arrows, marking the flows of distinct ‘here’ and ‘there’ goods, is deeply embedded in the field.⁶⁴ ‘Here’ is often defined, implicitly or explicitly, as the Roman Empire, associated as it came to be, ideologically and spiritually, with European empires, and especially the British Empire.⁶⁵ ‘There’ was traditionally defined in terms which Said so effectively exposed as an ontologically constructed and distinct ‘other’, as noted above in terms of the use of ‘Indo’ as a category label.⁶⁶ The colonial overtones of this understanding are increasingly recognised, but can only be addressed by a long and slow process of filling out the history of all of the ‘there’ places which lay around the shores of the Western Indian Ocean and their many interactions with one another.

One alternative is the idea of ‘provincialising’ (or marginalising, or decentring) Rome. This approach is developed explicitly in the works of some scholars. In other works, various approaches to decentring Rome may be more implicit, but still guide the questions asked and the conclusions reached. Some describe and analyse an Indian Ocean in which the Roman Empire was not such an important player.⁶⁷ Others present an Indian Ocean in which the Roman Empire was not present at all in many interactions.⁶⁸ Still others offer scenarios in which the Roman Empire held a subordinate or dependent position.⁶⁹ The aim in these cases is to disrupt the idea of the ‘here’ and a tendency, still also visible in the field, to interpret change and dynamism in terms of Roman influence and contact.⁷⁰

Attempts to quantify Roman trade with the Indian Ocean have been particularly important in this specific area of debate. Some strands of Indian Ocean studies have argued that Rome was not only not the major player in, but was even dependent upon, Indian Ocean trade for its expansion and survival. In its most extreme expressions, this impulse has taken the form of spinning the sparse remarks of Strabo and Pliny, most famously Pliny's claim that ‘in no year does India absorb less than fifty million sesterces of our empire's wealth, sending back merchandise to be sold with us at a hundred times its prime cost’,⁷¹ into elaborate mathematical calculations, concluding that trade with India amounted to anywhere between 10 per cent (perhaps plausible) and a third (entirely implausible) of the Roman empire's total revenue.⁷²

It is the position of this author ‒ not one that is shared unanimously in the field, as the foregoing makes clear ‒ that this particular direction within the study of ancient Indian Ocean studies is incompatible with the critical fact that almost all of the economies around the Indian Ocean in antiquity, and certainly the Roman economy, were agrarian at their base.⁷³ Within this economic framework, in a world without post-industrial production, transportation and communication technology, there was simply no possibility that long-distance trade, i.e. trade outside the territories controlled by these large states, in mostly lightweight, luxury products, could have underpinned a state economy to any significant degree.⁷⁴

In my view it is therefore untenable to decentre Rome by making it dependent on Indian Ocean trade, but such attempts point to a bigger problem: analysing the Roman Empire as dependent on Indian Ocean trade is still making Indian Ocean studies about the Roman Empire. The challenge of marginalising the Roman Empire is how easily it can reconstitute Rome as the lens through which Indian Ocean interactions are understood. The Roman Empire undoubtedly was a major source of demand, production and symbolic power in the world around it, including in the Indian Ocean, and attempting to situate it as economically dependent on those external interactions is not an antidote to earlier tendencies to see it as dominating them completely. This brings me, accordingly, to what I believe are the most important future directions in ancient Indian Ocean studies.

V FUTURE DIRECTIONS

Most obviously, and already a feature of the field, Indian Ocean studies is no longer the study of Roman interactions alone. The centrality of South Asia in Indian Ocean networks is now recognised, but so too is the importance of the East African coast, not just in terms of its littoral rim, but also of the internal social and economic dynamics which linked coastal and inland societies.⁷⁵ In addition to continued archaeological excavation of sites all around the Western Indian Ocean, more critical editions and translations of texts from outside the Roman Empire will also help to deepen and nuance this more balanced picture. One example is the vast Pali commentarial literature from Sri Lanka, which was compiled over centuries and includes details which, like the papyrological evidence from the Mediterranean, reveal everyday life, vocabulary and problems which are often not the concern of more literary texts.⁷⁶

Finally, further numismatic work depends on the continued effort of cataloguing, dating and pinpointing geographically the huge outputs of coinage from areas outside the Roman Empire and the complex and different ways in which they could be embedded into local and long-distance social systems.⁷⁷ A better understanding of coinage economies around the Indian Ocean will also be critical to identifying flows of trade and places where coinage can, and cannot, stand as a proxy for overseas commerce, and where it illuminates other cultural and social phenomena, such as the prestige use of foreign coins as objects of power in religious and royal practice.⁷⁸

Decentring Rome must (and should) inevitably mean altering the amount of attention that it takes up in Indian Ocean studies. Nevertheless, agents of the empire, themselves linguistically, culturally and ethnically diverse, were, especially in the first and second centuries c.e., significant participants in Indian Ocean networks.⁷⁹ The urban and moneyed wealth of the empire was also a major impetus to the expansion and densification of trade routes. As such, the Roman Empire cannot, will not and should not be too aggressively minimised in the future of Indian Ocean studies. Continued work on the internal configuration of the Roman economy, as well as ongoing consideration of how the Roman Empire handled its own diversity, will help to ensure that parallels between modern colonial structures and the Roman Empire are put in their historical place, rather than operating as ongoing interpretative paradigms for antiquity.⁸⁰ Moreover, recognising the importance of the Roman Empire in Indian Ocean studies, alongside a more balanced volume of scholarship on other regions, will hopefully enable it to take up a proportionate position in our understanding of historical Indian Ocean dynamics.

The legal, social, cultural and economic role of resident and visiting foreigners in the Roman Empire is another area of study which has clear relevance for Indian Ocean studies. Recent work, for example, has complicated the ideas of both ‘foreigner’ and ‘insider’ and ‘internal’ and ‘external’ mobility. Examining the massive internal movement within the Roman Empire and how this shaped concepts like diplomacy can contribute directly to breaking down legacy impressions of people as being in some sense proxies for monolithic political blocks (e.g. Roman sailors or Indian travellers).⁸¹

Each of these new directions, well under way as I write, has obvious benefits, but poses a challenge, too, if ancient Indian Ocean studies is to retain coherence and the sense of community that has made it an especially welcoming and friendly field of study, with the connected scholarly benefits of generous sharing of ideas and data and valuable mentorship for younger practitioners. This is a golden moment to discuss terminology. The term Indo-Roman is dropping quietly out of use, or perhaps taking on a narrower meaning, but Indian Ocean history remains substantially labelled by Mediterranean-centric terms — ancient, medieval and modern — that are themselves becoming increasingly elastic. Likewise, regions around the ocean and even the ocean itself remain labelled by terms which overwhelmingly derive from western perspectives.⁸² We may wish collectively to continue using some or all of these labels, and all terms are labels of convenience, which will inevitably be contingent and flexible. Still, having explicit conversations about why, how and what we mean by words we all use will be useful going forward.

It seems unlikely, for example, that the term ‘Indian Ocean’ is going anywhere soon, but that does not mean we can ignore what is concealed and foregrounded by that choice, especially the diminution of the role of Africa in the history of human mobility. Meanwhile, there are calls from various quarters to avoid using names closely associated with modern nation states, when discussing ancient (and, indeed, any pre-modern) global history. Most prominently, there have been calls for ‘China’ to be avoided as a label referring to the distant past, as doing so endorses modern narratives, promoted by the Communist Party of China, of ‘China’ as a timeless and geographically fixed entity, which in turn has implications for claims made by the state over contested territory. It can be replaced with labels for specific political formations, such as the Han or Tang [dynasties/empires], or by speaking of specific parts of continental East Asia or Eurasia.⁸³

Beyond terminology, another area of development continues to be chronology. The field of ancient Western Indian Ocean studies must continue to grapple with the relationship between the ‘ancient’ Indian Ocean and the coming of Islam in the seventh century. At present, the Islamic conquests function as a hard barrier within textual studies and an extremely porous one in broadly archaeological studies. This is explicable in terms of the evidence available: in the textual tradition, the Islamic conquests are presented by contemporary sources as a break, marked also by a shift in the languages of most Mediterranean sources dealing with Indian Ocean interactions, from Greek and Latin to Arabic and Persian.⁸⁴ By contrast, archaeologically it is not possible to date many ceramic types, and therefore whole sites, clearly to before or after the Islamic conquest, which may reflect very real continuities in economic, social and cultural life. Above all, this difference between literary and archaeological evidence reflects the perspective shift inherent to these sources, from ‘the Islamic conquests’ as a large-scale political and military phenomenon recorded in writing, to ‘the Islamic conquest’ of a particular community, region or village, as attested in material culture. The latter might take centuries and range from violent and dramatic to slow and incremental, not just from place to place but between different social processes within each location.⁸⁵

In a strongly cross-disciplinary field, however, and one which does not always acknowledge its own emotional preference for particular narratives (continuity of connection, cosmopolitanism, peaceful hybridity), this inevitable difference between pictures which can be drawn from material and textual sources can enable the cherry-picking of data to suit particular outcomes. Alternatively, it can simply cause researchers to back away from the complexities of the emergence of Islam in the Indian Ocean region, by ending studies of the ‘ancient’ phase in the sixth century and beginning studies of the ‘medieval’ in the ninth or tenth. Some studies have sought to cross this chronological divide, often at the same time as breaking down the division between eastern and western halves of the Indian Ocean, but more work in this direction seems both likely and desirable.⁸⁶

Finally, the study of ancient Indian Ocean exchange cannot be separated from the study of other long-distance trading networks. Some of the contours of its historiography parallel those in the field of trans-Saharan trade, which has likewise tended to be seen as having an ancient (Roman) followed by a medieval (Islamic) phase, and which was traditionally seen through the lens of Roman priorities and activity, but now shows similar shifts towards a more global point of view.⁸⁷ Most influential in the development of Indian Ocean studies, however, is the study of what is often termed the Silk Road(s).

Silk Road studies are enjoying something of a Renaissance at present, not least because of the impetus given to them by Chinese state support, related to the One Belt, One Road initiative. This has resulted in a plethora of conferences, exhibitions and research projects and, beyond direct support, generated a context in which there is a clear need to contextualise and historicise long-distance connections between large East Asian states and the lands to their west.⁸⁸ Silk Road studies also provide a framework within which to re-think some of the same challenges of decolonisation and post-colonialism which affect Indian Ocean studies. Responding to this surge in popularity and shared questions and methods, the concept of the Maritime Silk Road (regularly abbreviated, especially in East Asian scholarship, to MSR) has become popular as a way of framing Indian Ocean studies and connecting the study of landward and maritime exchange.⁸⁹

This development does make it easier to compare evidence, share frameworks and apply methods between the study of maritime and terrestrial trade routes. However, the concept of the Silk Road(s) remains subject to critiques, which have existed since the term was invented: that it overstates, simplifies and above all reifies relationships that were in reality much more ephemeral, contingent and locally embedded than the metaphor of a trans-continental street map might suggest.⁹⁰ In the case of the Maritime Silk Road, the metaphor can seem even more strained, involving as it did neither roads nor probably very much silk, in comparison to other goods. The idea of the Maritime Silk Road is a fashionable means to raise the profile of Indian Ocean studies, a move which should not be dismissed in an environment in which support for humanities research must be fought for competitively, but it remains to be seen whether it has lasting analytical merit.

VI CONCLUSION

The study of the ancient Western Indian Ocean is a vibrant, international and interdisciplinary field that is coming to terms with global changes in the world in which our scholarship takes place. As we see the present differently, the past inevitably looks different, too. It is no longer as obvious as it may have been a few generations ago that the Roman Empire was dominant in Indian Ocean exchange networks, but that does not make it easy to identify what its role was. Nor does this change of perspective invalidate the work of scholars over more than 150 years, upon whose findings our current conclusions depend. If now is the time to rethink some of their assumptions, it must be with humility and the certainty that many of our own will one day seem obviously erroneous.

In a growing field, in which terminology and scope are all up for grabs, it will also take work in the coming years to ensure that the field remains both coherent and collegiate. At present, it is a conspicuously friendly environment and one that has historically been welcoming to diverse perspectives. For that to continue, we need to foreground conversations about what we mean and how we say it, to agree to compromises and to continue to seek the best in interdisciplinarity, while also appreciating the immense value of specialist studies, including by amateur scholars, collectors and local enthusiasts. It is no longer possible for us all to have read everything or to expect that of one another, but that is a challenge for every field. It is a challenge that Indian Ocean studies, with its growing network of journals, conferences and online resources, is well placed to address proactively through the continuing work that so often goes unrecognised in institutional incentives, such as reviewing publications, peer reviewing and cataloguing, editing and translating material.⁹¹

The continued potential of the subject is significant not, I believe, because trans-Indian Ocean trade was central to major political structures — pursuing that theme is to subordinate the field to a narrow paradigm for what is important. Instead, its richness lies in the capacity to mould, model and maintain conversations across nations, regions, different religious and linguistic communities and across disciplinary groups in the present. It lies, too in the importance of Indian Ocean networks for showing how large-scale political and social systems are constituted out of local-scale, smaller networks and communities, which, at their edges, may both define and defy the power structures that claim them. The Indian Ocean is a liminal space that has never been under any single political regime, and this makes it a powerful place from which to take a new perspective on the things around its edges. Uncovering human experiences that often took place far from the centres of political power and literary output can offer new insights into those centres, but is also worthwhile in itself for foregrounding ways of life that were distinctive, real and fully alive in their own terms.