Navigation – Plan du site

AccueilThématiques06Connexions de basse intensité et ...India in Africa: Trade goods and ...

Connexions de basse intensité et symboliques partagées

India in Africa: Trade goods and connections of the late first millennium

L’Inde en Afrique : marchandises du commerce et connexions à la fin du premier millénaire
Jason D. Hawkes et Stephanie Wynne-Jones


Les approches archéologiques pour étudier les connexions de l’océan Indien ont tendance à se concentrer sur les objets « étrangers » qui apparaissent dans différents contextes. En Afrique de l’Est, ces objets sont trouvés sur des sites situés sur ce qui devint la côte swahili, et ils montrent que ces sites étaient liés aux réseaux de l’océan Indien dès le viie siècle ap. J.C. Leur faible quantité fait que ces premières connexions sont difficiles à voir ; elles deviennent bien plus faciles à identifier à partir du xie siècle, lorsqu’un plus grand nombre de biens est connu. Cependant, notre capacité à retracer ces premières connexions n’est pas uniquement affectée par la quantité de traces à notre disposition. La manière dont les archéologues ont abordé ce sujet a été dominée par des paradigmes historiques qui mettent l’accent sur le golfe Persique et l’entremise des marchands arabes et qui considèrent les connexions dans l’océan Indien principalement en termes de commerce. Cet article examine la manière dont cette surévaluation du rôle du commerce entraine une façon de penser particulière dans les études archéologiques. Il discute le fait que, dans ce cadre, les connexions entre l’Inde et l’Afrique sont mal prises en compte. Cela amène alors à reconsidérer les traces de ces connexions dans l’archéologie du viie au xe siècle dans ces deux régions. S’appuyant sur la capacité de l’archéologie à s'interroger sur la signification et l’utilisation des objets matériels, cet article explore la manière dont la variété des catégories d’artefacts évoque un certain nombre de types différents de connexions entre les populations d’Inde et d’Afrique à travers l’océan Indien.

Haut de page

Notes de l’auteur

This work was completed during a period of research funded by the Entrepot Project, Danish Research Council Sapere Aude fund. We are grateful to the fund and to the PI, Dr Søren Sindbaek, for their support.

Texte intégral


  • 1 M.C. Horton, 1996.
  • 2 M.C. Horton, T.R. Blurton, 1988.
  • 3 Ibid., p. 19.
  • 4 Ibid., p. 20.
  • 5  L. Fouché, 1937.
  • 6  D. Miller et al., 2000.
  • 7  A. Oddy, 1995, p. 186.

1Excavations at the Swahili stonetown of Shanga, on the northern Kenya coast,1 recovered one of the more enigmatic and suggestive artefacts known from the eastern African coast during the pre-colonial trading period. The bronze figure of a lion, dated to c. AD 1100, is unique among finds from the coast of eastern Africa, yet is typical of a number of similar figurines found in India—specifically the Deccan Plateau—known to have been used in Hindu rituals.2 One explanation for the presence of this figurine in an archaeological context in East Africa is that it was brought by an Indian merchant or traveller, or imported by a member of the increasingly rich local elite; Shanga was deeply connected with Indian Ocean routes of commerce and interaction dominated by Islamic traders. Yet, the Indian technologies employed in the manufacture of the Shanga lion tell only half the story. Despite its South Asian style, the figurine appears to depict an African—not an Indian—lion, with a wild unkempt mane running down its back. Indian examples have the neatly trimmed collar of an Asian lion. This means that its maker was familiar with both this style of Indian sculpture and with African lions themselves, and it suggests the possibility that it was made by an Indian craftsman present in Africa. The metal content of the lion further supports this latter suggestion. This is somewhat different from comparable statuettes found in Indian contexts.3 The closest parallels to the lion’s specific alloy are with contemporary Chinese coins, leading to the suggestion that it may have been cast from melted coins, as part of a wider practice of metal recycling in the Indian Ocean world.4 The single find of the Shanga lion therefore opens up a world of possibilities, to the movement of people, materials, ideas, styles, and religions. Further glimpses of such movements are evident in other isolated finds, such as the 13th century gold rhinoceros figurine from Mapungubwe in southern Africa.5 Made from wood and covered in locally sourced gold sheet,6 the figure of the rhinoceros itself has only one horn and may thus have been based on an Indian model.7 Beyond these individual objects, however, these movements are poorly accounted for by archaeologies, which tend to assume trade as the explanation for foreign objects in far-flung contexts.

  • 8  M.C. Horton, 2004.
  • 9  See also E. Lambourn, 1999.

2Horton is also the only archaeologist to have considered the ways that archaeology might reveal a wider web of relationships between India and Africa than that suggested by trade goods alone.8 Drawing on the evidence of the Shanga lion, he suggests the possibility of the movement of artisanal communities around the Indian Ocean and hypothesizes a two-way exchange between India and Africa.9 At the sites of Shanga and Tumbatu, Horton points to some of the more sizeable collections of Indian goods known on the coast. These are mainly ceramics, and he deals largely with the period from the 11th century onwards, when the largest numbers of goods are known. This study offers an important challenge to some of the ways that Indian Ocean connections are understood, which have focused on trade relationships. Here, we build on these suggestions, exploring known assemblages from the period before c. AD 1000 on the East African coast and placing these into the context of what is known from the Indian subcontinent at this time. As Horton points out, these traces are much less tangible than those after AD 1000, when the beginnings of a recognizable diaspora might be seen. Rather than seeing this as solely a problem of recognition, we use this scarcity as a provocation, challenging us as archaeologists to explore other means of connection than simple trade relationships. In fact, the rarity of pre-AD 1000 goods suggests that trade relationships between India and Africa were not significant in economic terms at this time, and we suggest some other explanations for those few objects found. Like Horton, we point to the movement of technologies and knowledge and think through some of the mechanisms for this dispersal.

Archaeology and history

3The settlements of eastern Africa’s coastline are known mainly through archaeology, although written histories add detail for specific places and times. From the 7th century onwards, archaeology has demonstrated the existence of a series of densely settled villages on the littoral and offshore islands (Figure 1).

Figure 1: The coast of eastern Africa, showing 7th–10th century sites mentioned in text

Figure 1: The coast of eastern Africa, showing 7th–10th century sites mentioned in text

Illustration: Stephanie Wynne-Jones

  • 10  Cf. E. Wolf, 1982; J. Blaut, 1993; P. O’Brien, 2006.

4These were linked into oceanic networks from the start, particularly with the Persian Gulf: exotic goods such as turquoise-glazed ceramics are scattered through the assemblages of sites of this period. As such, we have two ways of understanding the coastal sites on what would become known as the Swahili coast. First, there is a rich archaeological record, suggestive not just of exchange relationships with far-flung contacts, but also of the daily life in which these relationships were managed. Second, archaeologies of the coast position the inhabitants as part of a well-attested network of trade and exchange which united regions of the Indian Ocean rim from earliest times, with contacts becoming denser and better documented from the Islamic period onwards. Despite this wealth of archaeological evidence, the explanatory framework that dominates the ways we explore this process and envisage these connections originated in, and is still largely defined by, historical scholarship. For various historiographical reasons surrounding the development of “world histories” in European scholarship,10 historical literature tends to chart cycles of “globalization” that preceded the arrival of Europeans and considers Indian Ocean connections primarily in terms of trade.

5Here, we suggest that exploring connections through archaeology on the Swahili coast might offer an important addition to the understandings afforded through historical approaches to the study of the Indian Ocean. First, we review the ways that this commercial emphasis creates a particular way of thinking about past connections and discuss the fact that this accounts poorly for connections between India and Africa. The second part of the paper then moves on to think about traces of these connections in the archaeology of the 7th to 10th centuries in both of these regions. Drawing on the strengths of archaeology in thinking through the meaning and use of material objects, we explore the ways in which a variety of artefact categories evoke a number of different types of connections between the people of India and Africa across the Indian Ocean.

Indian Ocean economies

  • 11  I. Wallerstein, 1974.
  • 12  J. Abu-Lughod, 1989.
  • 13  A. Wink, 1990.
  • 14  J. Abu-Lughod, 1989, p. 266.

6In many cases, historical traditions of the pre-modern Indian Ocean have been positioned as a challenge to models of European hegemony, identifying earlier cycles of “globalization”, particularly those that accompanied Islamic trade. Reacting to Wallerstein’s world-systems thinking,11 these understandably focus on and emphasize economic concerns. For example, Janet Abu-Lughod set out explicitly to write an economic history of an eastern “world-system” preceding the European dominance of the 16th century.12 As such, she outlines trade patterns and cycles connecting regions of the Indian Ocean that peak in the 13th century AD. A similar approach is taken by Wink’s treatment of the “Indo-Islamic” world, although this work places greater emphasis on events and developments on the Indian coast and a shift in chronology to consider periods from the 7th century rise of Islam.13 For both scholars, Arab navigators are seen as “keeping lines to the east open” after the decline of Roman trade14—with Persian Sassanids and later Islamized Arabs the primary agents in shipping—as well as establishing communities or “colonies” at ports in India, notably in Gujarat.

  • 15 Ibid., p. 272.

7These portrayals of interaction are based around large-scale, commodified commerce, connecting “world cities” in a network that rivals that of the modern world-system. Transactions themselves are envisaged as mediated by currency exchange, itself to a large degree under state control. The “system” is that world of state-sponsored commercial exchange, often regulated by treaty, in which only interactions on a wide gauge would show up. As part of the endeavour to counter Western hegemonic models, this is entirely appropriate. Yet, such an approach almost certainly under-represents non-commercial forms of interaction, privileging the areas with a strong textual tradition of traveller’s tales. It is then unsurprising that East Africa does not feature in this analysis, although it is at times referred to in the context of Arab and Indian trade. For example, in AD 1500 when Barbosa recorded Gujarati ships at Malindi and Mombasa, the “presence of Gujaratis in East Africa was neither unusual nor new”.15

  • 16  P. Beaujard, 2005.
  • 17 E.g. K. Ekholm, J. Friedman, 1982; F. Frank, B.K. Gills, 1993.
  • 18  Also P. Beaujard, 2012, 2013.

8Beaujard also positions himself in opposition to Wallerstein’s concept of a world-system,16 although he builds the work of other systems theorists into his account of the pre-modern Indian Ocean.17 He too identifies the emergence of world-systems pre-dating the dominance of European economies in the 16th century and sees the Indian Ocean as a critical space within which this emerged. In doing so, Beaujard broadens previous configurations of Indian Ocean spheres of interaction to include the East African coast in the formulation of an “Afro-Eurasian” world-system.18

  • 19  Drawing on E. Morin, 1990.
  • 20  P. Beaujard, 2005, 2012.
  • 21  P. Beaujard, 2005, p. 241.

9More than simply mapping exchange networks and revising the economic history of these developments, Beaujard attempts an entire re-working of the way we think about world-systems to encompass non-economic aspects and complex relations.19 Although limited by a necessary reliance on secondary sources, this allows Beaujard to accommodate the existence of multiple cores in a single temporal scheme—and, importantly, to incorporate both different mechanisms of interaction (political and religious as well as purely economic) and the wide variety of relationships that these interactions gave rise to. Accordingly, Beaujard charts these multiple factors through examination of the interplay between the different cores that existed in the Indian Ocean (such as the Persian Gulf and the multi-centred core of India) and their peripheries (such as East Africa).20 Cities are recognized as the main nodal points of core–periphery relations, being at once centres of political power and the forum within which commercial routes and religious networks intersected with each other. In doing so, he identifies four distinct phases (“pulsations”) in the emergence and development of the Afro-Eurasian world-system: the emergence of international maritime connections during the third millennium BC; the crystallization of these connections into a world-system between the 1st and 6th century AD; the reconfiguration of west Asian and Indian cores between the 6th and 10th century; and from the 15th or 16th century to the industrial revolution in the mid-18th century.21

An “Indian Ocean” identity

  • 22  K.N. Chaudhuri, 1990.

10A parallel historical literature has sought to conceptualize the Indian Ocean as a cultural space. This can be seen as part of a post-colonial challenge to Orientalist discourse, in which the “East” (particularly India) has been defined as complex, powerful, and distinctive. The classic example of this tradition is found in the work of Kirti Chaudhuri, where “Asia before Europe” has been described, characterized, and worked into complex trading systems that pre-date modern world-systems in this region.22 This work sees the Indian Ocean primarily as an Asian sea and draws large-scale patterns about social identities and relationships to landscape. Chaudhuri’s work serves as an important counterpoint to studies that locate agency in the Persian Gulf and with Islamic merchants. Like Abu-Lughod, he identifies civilizations or interaction networks stretching east, with the Islamic “system” being only one of four important Indian Ocean civilizations. Chaudhuri also adopts a Braudelian chronology, exploring the rhythms of time and history that have united and disunited the region.

  • 23  M. Pearson, 2003. See also K. McPherson, 1993; H.P. Ray, E.A. Alpers, 2007.
  • 24  As opposed to attempting to define an ‘Indian Ocean culture’, cf. K.N. Chaudhuri, 1990.
  • 25  M. Pearson, 2003, p. 5.

11Annales history also informs Pearson’s analysis of the Indian Ocean, a nuanced account of interaction between regions as part of an attempt to write a history of the sea, rather than of the various countries that surround it.23 Pearson analyses the myriad ways that the ocean facilitated commerce, yet he also attempts to explore other types of interaction, notably the movement of ideas and religion. His account of the period AD 700–1500 is dominated by Muslim traders, sailors, and proselytizers. Yet, Pearson does not restrict his analysis to movement from an Islamic heartland in the Arabian Peninsula or Persian Gulf. Instead, he recognizes a rich web of travellers, including Islamic converts from East Africa or Gujarat, moving around the networks established by the demands of trade and contributing to a “littoral society” that is somewhat distinct from that of the hinterland at their backs. Pearson conceives of this society in interesting ways, seeking to explore the ways that peoples of the littoral might be distinctive.24 In doing so, littoral peoples emerge as cosmopolitan, accustomed to a transient way of life and sensitive to the rhythms of seasonal arrivals and departures. While recognizing diaspora communities, this approach to merchants and travellers seeks to include a much larger range of people—those who travel and return, who are linked by blood or religion, and who transact across and along the ocean’s shores as part of their daily existence. Thereby, he seeks to explore both “material and mental frameworks” that create an Indian Ocean world.25

  • 26  Following M. Pearson, 1998.
  • 27  P. Beaujard, 2013.
  • 28  See also A. Sheriff, 2010.

12By attending to the people of the littoral, Pearson is able to build local models for the conduct of trade which refine the larger-scale dynamics discussed. This is particularly well developed for the Swahili coast,26 a region where merchants are seen as active in assigning and determining value as well as in negotiating exchange.27 Attention at the micro-scale also follows from the observation that most connections [in the Indian Ocean] were rather minor, with long-distance trade occurring amid a web of smaller-scale and shorter-range transactions and interactions around the ocean rim.28 This coasting trade is more rarely considered as part of Indian Ocean connectivity and complicates models that are based only on “big picture” modelling of economic transaction. Yet, to a great extent, the forms of interaction noted are still defined by economic concerns.

Micro-histories and things

  • 29  R. Chakravarti, 2004.

13An interesting addition to this literature on macro-scale patterns has been the range of histories that seek to explore micro-scale processes and perspectives within these early world-systems. In large part, this shift in scale is created by the sources that are drawn upon, particularly in thinking about adding Indian perspectives to these narratives. For instance, among the primary sources for the “early medieval” period in India (c. 6th to 13th century AD) are a series of copperplate grants that record landholdings and agreements. In doing so, they provide insight into the ethnic and religious complexity of coastal communities, urban organization, and the role of guilds, governors, and merchants in political affairs and juridical practice.29

  • 30  E. Lambourn, 2014b.
  • 31  E.g. E. Lambourn, 2003a, 2014a.
  • 32  E. Lambourn, 2014b.
  • 33  E. Lambourn, 1999, 2003a, 2003b.
  • 34  F.B. Flood, 2009.
  • 35 Ibid., p. 11.

14The work of Elizabeth Lambourn, in particular, seeks to supplement these “small” histories through attention to a range of sources,30 including the personal inventories and stories of Jewish merchants in India preserved in the Geniza archive, historical linguistics, and a close attention to the qualities of objects that speak of relationships beyond commercial exchange.31 Even as she struggles with how to make this series of micro-histories add up to a “big picture” like that of Arab-dominated commerce, Lambourn evokes a world of connection and exchange that transcends the purely economic.32 Her work deals extensively with the issue of a diaspora among Indian craftsman, using the record of stone carving from sites across the Indian Ocean as a guide to explore the movement of people and, further, the shared aesthetics created through this movement.33 In this, her findings can be viewed in the same vein as those of Finbarr Flood, who re-imagines medieval encounters between India and the Arab world by engaging with a series of objects.34 His work acknowledges multiple routes for the movement of objects and, further, the movement of ideas about objects, encompassing aspects of style, form, and craftsmanship. Objects, for Flood, represent networks of affinity, bounded not by “religious, ethnic, or linguistic identity, but by possession, consumption and display”.35 As such, he focuses on “things” to reconceptualize medieval relationships between the Indian subcontinent and the Islamic heartlands, arguing for dynamic models of encounter and exchange that move beyond “civilizational histories”. He acknowledges that all cultures are already transcultured and instead explores the ways in which the material record can illustrate this.

Points of departure

  • 36  E. Lambourn, 2015 (forthcoming).
  • 37  S.D.S. Jayasuriya, 2003.

15These fine-grained histories that centre on objects rather than records of trade can evoke a richer world of connection. Flood suggests attention be given to looting, gifting, and trade as mechanisms of circulation. He also explores the issues of diaspora as a two-way movement between India and the Persian Gulf. Lambourn advocates attention to religious networks, to the movement of craft workers, and to cultures of collecting and connoisseurship.36 The obvious additions to this list would be pilgrimage, particularly among the Muslim communities that occupy the Indian Ocean rim, and processes of enslavement.37 These processes, to an extent elided by macro-histories that focus on trade and commerce, all combine to add to a picture of a rich web of connections across the Indian Ocean.

  • 38  J. Hawkes, 2014.

16Archaeology has only rarely contributed to this picture. In part, this is due to certain limitations inherent in the comparability and “quality” of the available archaeological evidence. The identification, archaeologically, of such connections and interactions and the processes that caused them requires a thorough knowledge of the material from multiple regions. However, the insularity of scholarship within different regions often means that this is not achieved. The effective consideration of evidence across regions—especially those as distant and diverse as around the Indian Ocean—also requires datasets that are comparable. Yet, different trajectories of archaeological research in different countries have resulted in widely divergent methods of excavation, interpretive models, and modes of reporting research findings. These, in turn, have lead to different scales of archaeological knowledge and understanding, not just of the artefacts themselves and the sites in which they were found, but also of the wider cultural contexts in which they existed. Investigations in East Africa, for example, frequently involve the excavation of small test-pits that account for relatively small areas of wider sites. While undoubtedly providing a great deal of valuable evidence, the results of these works cannot always be incorporated with those of large-scale “horizontal” excavations that characterize the investigation of many urban sites on the Persian Gulf. Similarly, the interpretation of settlements in India is plagued by problems surrounding their chronologies.38 This limits the extent to which they can be compared with evidence from specific periods in other locations.

17Such differences in the “quality” of archaeological evidence relative to one region and another certainly pose significant difficulties in the identification of connections across the Indian Ocean. Yet, these limitations are not insurmountable, and they do not fully explain the extent to which archaeology has so far contributed to our understanding of this richer web of connections. Arguably, this has more to do with the dominance of existing historical paradigms in archaeological research on the Indian Ocean. The histories of archaeological research and the relationships between archaeology and history are, of course, different in each country, and it is not the intention to review these here. Notwithstanding these differences, when archaeologists have sought to approach the study of the Indian Ocean, they have tended to do so with recourse to the explanatory frameworks defined in history. With their emphasis on macro-histories and economics (as described above), the movement of objects has normally been understood somewhat simplistically as indicative of trade connections. At the same time, topics that are peripheral to this main thrust of historical enquiry have rarely been explored.

18It is the contention here that a reinvigoration of an archaeological approach to the examination of objects can reveal traces of Indian Ocean connections that extend far beyond those related to trade and monetary exchange, and it has the potential to speak to these broader understandings and to offer a different perspective from the “big picture” afforded by economic histories. Archaeology as a discipline is, by its very nature, concerned with the qualities of things and can thus afford a similar type of analysis to that achieved by Flood or Lambourn. It must be said that archaeologists are rarely confronted with the type of unique dataset encapsulated in the Shanga lion. Rather, we explore a palimpsest of daily activities, revealing a different scale and temporality of connections. In addition, archaeology is able to focus on earlier periods. Most of the studies mentioned above explore the second millennium AD, at which time a series of connections across the Indian Ocean are attested through documentary sources. The world of the first millennium is much hazier, and yet there are glimpses of connections that might be explored; it is here that we wish to focus for the rest of this paper. The Swahili coast of eastern Africa was in contact, either directly or indirectly, with the Indian subcontinent from the 7th century AD, and we suggest that there are a number of classes of evidence that can illuminate this relationship and produce a richer understanding of oceanic connections at this time.

The eastern African coast

19Africa’s eastern coastline, where the continent meets the Indian Ocean, was named Swahili only in comparatively recent times by Arab writers and travellers; but the civilization they refer to has long been established there. Today, the shoreline and offshore islands are dotted with the remains of coral-built ports of trade with largely Muslim populations that were, and are, home to a merchant elite. The grandest monuments—mosques, tombs, and palaces—date to the golden age of Swahili trade in the 13th and 14th centuries AD. Yet, this stretch of coast had long been incorporated into Indian Ocean networks of trade and was widely known as an exporter of luxuries to the Arabian Peninsula and the Persian Gulf, and onward to markets in India and China.

  • 39 al-Mas’ūdī, 1861-1877.
  • 40  H.N. Chittick, 1984.

20In the 10th century, Al-Mas’udi reports that the people of East Africa were exporting ambergris and resins, leopard skins, tortoise-shell, and ivory, which were highly prized in the workshops of India and China.39 To this list we might add several commodities that we know of from other histories, particularly mangrove wood, gold, and slaves. Archaeologically, these exports are all difficult to recognize. Indeed, objects of a definitively African origin have yet to be identified in archaeological assemblages outside of Africa. Sometimes, however, we see hints of these exports in the archaeological record at sites in East Africa. For example, at Manda, an important trading town in northern Kenya, remains of ivory-working (a signature of production, possibly for export) were encountered during excavations.40

21The people conducting this trade lived in quite humble settlements. The buildings were all constructed of mud and thatch. Even mosques were built on a small scale using this fabric; an 8th-century mosque can be traced in the foundations of a mud and thatch structure at Shanga in northern Kenya (replaced in later centuries by a much grander coral-built mosque). This process of mosque construction and enlargement also illustrates the growing Islamic community on the coast. Although Muslims were present from the 8th century, it was only from the 11th century, after a gradual process of conversion and, perhaps, immigration, that they appear to have comprised the majority of the population.

  • 41  J.B. Fleisher, S. Wynne-Jones, 2011.

22In the 7th to 9th centuries, archaeological evidence points to the Swahili people having had a mixed fishing and farming economy. The huge majority of artefacts found were locally produced, with large quantities of ceramics of a type known as Early Tana Tradition.41 These ceramics are of a tradition shared with all coastal settlements, as well as sites far into the African hinterland (Figure 2).

Figure 2: Early Tana Tradition ceramics from Dakawa

Figure 2: Early Tana Tradition ceramics from Dakawa

Photograph: J. Fleisher.

  • 42  C. Hardy-Guilbert et al., 2004.

23Yet, it is clear that these small fisher–farmer villages were involved in Indian Ocean trade from the start. Coastal sites of the 7th century onwards have remains of imported objects among the locally produced goods. The most visible to archaeologists is the range of glazed ceramics, coming mainly from the Persian Gulf. Turquoise-glazed wares (also known as Sasanian-Islamic) are ubiquitous among cargoes of this period and are found at sites across the Indian Ocean from the 7th to 9th centuries.42 In East Africa, turquoise-glazed jars were the most common finds; they were probably containers for other goods such as date products or oil from Persia.

  • 43  F.B. Flood, 2009, p. 21.

24These trade relationships are well attested and fit within the broader-scale understandings of the Indian Ocean world, with the Persian Gulf (and particularly large urban ports such as Siraf) acting as a fulcrum for commerce in this region. East African connections with Siraf were likely the most intensive in terms of trade, and there was also an important cultural element to this connection: links with the Gulf echo through the “Shirazi” origin myths recorded along the coast into recent times. It is probable that connections with India at this time may have been mediated via ports in the Gulf and eastern Arabia, where intensive connections with Gujarat were maintained year-round, and where significant Indian diaspora communities were resident from at least the 9th century.43 There may also have been more direct connections, although these seem not to have taken the form of significant trade relationships. Instead, we can see a series of classes of object—coins, beads, and ceramics—that allow us to trace different forms of cultural interconnection, and that evoke different types of connectivity.

Shared value: Swahili coins and the Sind connection

  • 44  For instance, the 11th-century writings of al-Biruni, and the 13th-century accounts of al-Mujawir.
  • 45  C. Allibert, 1991a, 1991b; M. Tolmacheva, 1993; T. Spear, 2000.
  • 46  M. Walsh, 2010.
  • 47  D.W.I. Piggot, 1941a, 1941b.
  • 48  See M. Walsh, 2010 for a critical overview.
  • 49  H.N. Chittick, 1965; M.C. Horton, 2004.
  • 50  M.C. Horton, forthcoming, p. 40.
  • 51  M. Walsh, 2010.
  • 52  M.C. Horton, J. Middleton, 2000.

25Histories of early movements between East Africa and South Asia in the western Indian Ocean can be difficult to interpret. While later references certainly exist in the accounts of Arab geographers and travellers,44 a relative lack of written records means a reliance on oral traditions.45 It is a matter of lively debate to what extent we can rely on these traditions as fact; in general it is understood that they represent a mixture of deep memories and symbolic statements, deeply affected by the context in which they were recorded.46 Yet, a series of allusions in Swahili oral traditions contains a whisper of early connections with South Asia, most particularly with Sindh, at the eastern edges of the first millennium Islamic empire. Swahili oral histories mention the waDebuli, or waDiba, ethnonyms associated with a range of historical events and periods in different parts of the Swahili coast. These traditions are most persistent on Pemba (the waDiba) and Unguja (waDebuli) in the Zanzibar archipelago, as well as on the coast of Kenya and on Mafia Island.47 They are also mentioned at Kilwa/Songo Mnara.48 Both Chittick and Horton associate the waDebuli stories with the port of Daybul at the mouth of the Indus, based on similarities in the name and the prominence of that site from the 8th right into the 13th century.49 As such, the association might not refer to a literal origin point for immigrants so much as a “direction of cultural contact”.50 Unfortunately, it seems that connections with one particular place are difficult to sustain, due to the range of cultural practices and associations attributed to the waDebuli in different accounts. A general link with South Asia does seem to be indicated, although there is also the possibility that some Debuli may have originated from Indonesia or the Laccadive Islands. In his review of the term in East African histories, Walsh concludes that it is not possible to pick apart the elements of historical fact from myth and inaccuracy in the Debuli traditions.51 He does, however, note that the term entered the Swahili language at some point immediately after initial dispersal from a NE Bantu dialect, placing it in the last quarter of the first millennium AD. Horton is more ambitious and suggests that the waDebuli traditions represent an echo of interaction with Sindh and the port of Daybul (itself echoing his own interpretations of the more widespread “Shirazi” origin traditions relating to the Persian Gulf).52

  • 53  B.A. Collins, 1994; G.R. Smith, 2008.
  • 54  M.R. Mughal, 1992; M. Kervran, 2013.
  • 55  F.A. Khan, 1963.
  • 56  S.P. Gupta et al., 2002; S.P. Gupta et al., 2003; S.P. Gupta et al., 2004.
  • 57  D.C. Sircar, 1962.

26The port of Daybul is known mainly through the accounts of Arab and Persian writers, such as the 9th century Arab geographer Al-Muqaddasi and the 13th century Persian traveller Ibn Al Mujawir.53 The site has yet to be identified conclusively on the ground, but a growing number of scholars argue for its identification with the site of Banbhore.54 Textual accounts of Daybul and excavations at Banbhore both testify to the site having been a major coastal emporium for the city of Mansura, one of two major towns of the Amirate of Sindh, the eastern extent of the ‘Abbasid empire.55 In the 9th and 10th centuries, Sindh was particularly prominent in trade with the Islamic west, with masses of objects and goods moving into the ‘Abbasid heartland through trade and tribute. Excavations at Banbhore have revealed little in the way of material evidence for direct contact between East Africa and Sindh or support for the suspicion of the movement of people from Sindh to the East African coast. Yet, as mentioned above, we do know from wider textual sources that communities from various kingdoms in South Asia were present in cities in the Gulf, just as Jewish, Muslim, and Zoroastrian merchants from the Red Sea and the Persian Gulf travelled to India and established communities along the west coast of India. The recently excavated site at Sanjan on the Konkan coast of Gujarat,56 for instance, is known to have been founded by a community of Zoroastrian émigrés. At one time it is also recorded as having been administered by a governor named Mohammed Sugatipa, who was probably Muslim.57 Thus, we can conceive of these regions as both having been part of an “Arab common market” at this time. It is possible that this involved a number of connections that were not necessarily mediated via the Gulf and eastern Arabia.

  • 58  H.W. Brown, 1996, p. 369.
  • 59  H.P. Brown, 1996; M.C. Horton, 2004.
  • 60  A.H. Dani, 2007.
  • 61  V.A. Smith, 1889; I.E. Coins, 1998.
  • 62  N. Lowick, 1990.

27Something of the nature of these connections between East Africa and South Asia can be identified in the evidence provided by coins found in both regions. Scholars have long charted the movement of coins as evidence for the physical movement of commodities and economic exchange. Yet, as Helen Brown has pointed out, in addition to these (perhaps more obvious) examples, the coins that were produced in each region also constitute valuable forms of evidence for connections that extend beyond those accounted for by trade inventories or histories of the movement of objects.58 For instance, we see shared notions of value between these two areas, as each responds to the dirham/dinar weight standards.59 Swahili sites and Sindh both issued coins modelled on Islamic coinage after the 7th century reforms; that is, they lacked figural representation and contained only Arabic script. For Sindh, the concept of coinage was not new. Prior to the expansion of the Islamic Caliphate, the region came under the rule of various South Asian kingdoms, including those ruled by the Mauryan and Indo-Parthian dynasties during the later centuries BC60 and the Rai, Vardhan, and Chacha dynasties during the early to mid-first millennium AD, all of whom issued coins.61 Yet, new coins were minted by the Amirs of Sindh in Mansura and Multan, following the caliphal weight standard, to produce the so-called Qandhari dirhams out of silver. Copper alloy issues were also minted in great numbers. Copper coins were outside the direct fiscal control of the caliphate and so were more sensitive to regional variation in different parts of the empire.62

28On the Swahili coast, however, the tradition of coinage dates only to the Islamic period, and the concept of money itself undoubtedly had an external origin. The Arabic script that characterizes Swahili coins suggests inspiration in the Islamic world, although certain esoteric aspects of the coins point to a link with Sindh rather than with the Persian Gulf.

Figure 3: Gold coins from Kilwa Kisiwani

Figure 3: Gold coins from Kilwa Kisiwani

Photograph: Helen Brown

  • 63  J. Walker, 1939; H.P. Brown, 1993; S. Album, 1999; S. Wynne-Jones, J.B. Fleisher, 2012.
  • 64  H.W. Brown, 1996, p. 369.
  • 65  F.A. Khan, 1963, p. 45.
  • 66  M. Mitchiner, 1977, p. 131.
  • 67  H.W. Brown, 1993, p. 14.
  • 68  H.P. Brown, 1996, p. 372.

29Although the most prolific minters of coins were the second millennium sites of Kilwa Kisiwani (Figure 3), and perhaps Tumbatu on Zanzibar, the tradition of minting is attested during the 8th and 9th centuries at the site of Shanga on the northern Kenya coast. Here, early silver coins were minted in the names of Mohammed and Abd Allah. They followed a series of patterns, which were to become characteristic of later coinage along the coast.63 First, the coins seem to have been cast in a mould, indicated by a regularity of size and a much less regular weight. The coins do not follow the established weight standards for the caliphate. In fact, all 24 silver coins found at Shanga together weigh 3.82g: less than the weight of 1.5 ordinary dirhams.64 Second, the text of the inscriptions runs continuously from obverse to reverse, with a single sentence beginning on one side and finishing on the other. Finally, the coin dies are set in one of four regular positions using pegs, meaning that the text is always at one of the cardinal directions. All three of these characteristics are known in the coinage of Sind, where terracotta coin moulds were excavated at Banbhore65 and dies using pegs were also in use.66 The continuous inscription is also otherwise unique in the Islamic world, although there are some parallels with amulets—rather than coins—elsewhere.67 As Helen Brown points out: “At the very least, this must be an indication of lively contact between the two areas.”68

Shared aesthetic: Beads

  • 69  L. Dussubieux et al., 2008; P. Robertshaw et al., 2010; M. Wood et al., 2012; M. Wood, 2015.
  • 70  V. Begley, 1996; V. Begley, 2004.
  • 71  P. Francis, 2002.
  • 72  H. Morrison, 1984; M.C. Horton, 1996; M. Wood, 2002, 2012.
  • 73  M. Wood, 2015.
  • 74  P. Sinclair et al., 2012; M. Wood et al., 2012.

30Beads are another class of object suggesting relationships between India and Africa. They are also one of the key classes of trade goods in the Indian Ocean world and seem likely to have functioned as a form of currency in many regions, parallel to and in place of the coin traditions mentioned above. The Indian subcontinent was the main producer of beads, and Africa a significant consumer, so this is one of the key categories that has been mobilized in thinking through connections between the regions. “Trade wind” beads made from glass and semi-precious stones are one of the most striking features of artefact assemblages found at sites in East Africa. These beads were undoubtedly a key commodity of the pre-modern period, and the large numbers that are found not only speak of the importance of this trade to the communities that exported and imported them, but also mean that they are a good indicator of the changing volume of trade between East Africa and the regions in which they were made. Indeed, studies of Indian beads found in East Africa offer the main form of quantitative evidence for that trade.69 The most voluminous type of these trade beads are the wound and drawn “Indo-Pacific” glass beads, which appear to have originated at sites such as Arikamedu in South India70 but quickly spread throughout South and Southeast Asia.71 These point clearly to trade between multiple origin points in India and the east coast of Africa, yet only really become common after the 11th century.72 Recent research has greatly improved our understanding of the directions of this trade.73 The glass beads found at East African sites of first millennium date are much rarer and mainly of North African, Byzantine, or Middle Eastern origin; a notable category of this latter type is the Zhizo series traded into southern Africa via Chibuene.74 Yet, in addition to these glass beads, a small but significant trade in Indian semi-precious stone beads also existed. This trade in stone beads, particularly those made from carnelian, has been given less prominence in scholarship on the Indian Ocean. The evidence provided by these beads allows us to suggest a number of connections that extend beyond international trade, including shared notions of value and aesthetics.

  • 75  T. Insoll et al., 2004.
  • 76  A.J. Arkell, 1936.
  • 77  P. Francis, 2002.
  • 78  K.P. Gupta, 1972, p. 266.
  • 79  S.R. Rao, 1973, p. 102.

31The precise source of the carnelian beads found in East Africa is, of course, difficult to identify. We know, for example, that carnelian beads were sourced and manufactured in West Africa alongside imported Indian types from at least the 10th century.75 Yet, through consideration of patterns of trans- and sub-Saharan trade, and stylistic comparison on the basis of form, we can safely assume that many of the carnelian beads found in East Africa originated in India, most probably from Gujarat.76 Geologically, the region around the Narmada Valley in Gujarat is home to the richest source of carnelian. In addition, archaeological and historical sources testify to this region being the main source of carnelian in antiquity. The “Agastya Samhita” section of the Garuda Purana (dating to the early first millennium AD), and the Ratnapariksa by Buddhabatta (c. 6th century AD), both refer to the Narmada region as being rich with gemstones, including carnelian,77 with the most carnelian-rich area of the Narmada then, as now, seeming to be the area around Ratanpur. Various sites along the western Narmada, such as Telod78 and Bhagatrav,79 have been mentioned in the archaeological literature as possible sources of raw material.

  • 80  A. Juma, 2004, p. 128.
  • 81  M.C. Horton, 1996, p. 333.
  • 82  H. Morrison, 1984, p. 184.
  • 83  F. Chami, 1994, carnelian bead illustrated, p. 66.
  • 84  J.R. Walz, 2010, p. 335.

32Carnelian beads are found at sites in East Africa from at least the 7th century onwards, perhaps even earlier. The numbers are always small, which is one reason that they have not seen much discussion as a significant trade commodity, but they are consistently present at sites on the coast and into the hinterland. At most of these, precise quantities are not given. At Unguja Ukuu, for example, Juma mentions carnelians, including one etched bead, during his Period Ia (6th–9th centuries) and notes that “regular carnelian beads are present” in Period Ib (9th–10th centuries).80 Five carnelian beads are illustrated (p. 127). Likewise, Horton notes “significant numbers” of carnelian beads in Trench 6-10 at Shanga, with their numbers peaking in the 9th and 10th centuries, and again in the second millennium.81 Interestingly, he notes Cambay (modern Khambhat) in Gujarat as their likely source, but suggests that they could have been transported to East Africa as “roughouts” (which were commonly found in the excavations) and finished locally. At the site of Manda, which has the largest quantities of imports for any site during this period, a surprisingly low number of carnelian beads are reported, with only five examples, which all date to Periods I and II (c. 7th–10th centuries).82 This might perhaps relate to the intensity of connectivity between Manda and the Persian Gulf, attested in other categories of imported goods such as Sirafi storage jars. It also reflects a notably low number of stone beads in general, with only 15 specimens total, including those made from rock crystal, which were likely of local derivation. Carnelian beads also travelled to sites in the interior: Felix Chami reports stone beads (unquantified) from the 7th–9th century site of Misasa in the Tanzanian coastal hinterland, and at least one of these was of carnelian.83 Jonathan Walz has recovered eight carnelian beads in 7th–10th century layers at the site of Kwa Mgogo in the Pangani Valley.84

  • 85  L. Casson, 1989, p. 81.
  • 86  H. Wilson, 1835, p. 85-86.
  • 87  N.R. Banerjee, 1959.
  • 88  K.C. Jain, 1972.
  • 89  A.J. Arkell, 1936. For a valuable ethno-archaeological study of carnelian bead production in India (...)
  • 90  For further discussion, see P. Francis, 2002.

33Turning to consider the contexts of carnelian beads in West India, textual sources are available which tell us that during the early first millennium AD, the main centre for lapidary workshops on the Indian subcontinent was at the city of Ujjain, just north of the Narmada River in Madhya Pradesh. The Periplus of the Erythraean Sea85 and the Sanskrit Mricchakatika86 both refer to the jewels made in Ujjain. The evidence from excavations at Ujjain itself,87 as well as that from surrounding villages,88 supports this identification. These workshops fed the main market for international trade at the city port of Baruch, at the mouth of the Narmada, which has long been recognized as the main coastal port of the early first millennium. At some point in the mid to late first millennium AD, the centre of lapidary workshops appears to have moved from Ujjain to Limudra, and the main port shifted to Khambhat.89 Exactly when this shift took place and why it occurred are unclear.90 What is interesting, however, is that throughout the first millennium AD there was a clear and close spatial association between: 1) source areas, 2) production centres, and 3) ports connected to Indian Ocean. Thus, the production of carnelian beads in India seems to have been closely linked to their value as a commodity of international trade. We can therefore see their production (in South Asia) and their consumption (in East Africa) as linked processes: participation in an Indian Ocean aesthetic shaped through interaction and suggesting connectivity beyond the larger bulk cargoes of glass beads that seem to have been more commodified.

  • 91  J. Hawkes, 2015 (forthcoming).

34Equally, however, within the context of South Asia, these beads were not just items of export; they are reported at sites throughout the Indian subcontinent. Consideration of their distribution and use in India might help to identify something of their shared meaning. Sadly, due to the ways that many of these settlements in India have been excavated and reported (something that affects our understanding of urbanism during this period generally), our ability to comment on their use and distribution within India is limited. Of the 105 excavated settlements in India that have been found to date to this period, site reports exist for only 19; even when the results of excavations are published, beads have traditionally been classified under the general rubric of “minor antiquities” and thus have not been adequately reported.91

35Nevertheless, carnelian beads were found during excavations at 16 of the 19 settlements that have been published, suggesting that they were a common feature in settlement assemblages from the period. They are not confined to West India, but instead circulated throughout the subcontinent (Figure 4), suggesting that they were very much part of wider domestic networks of trade and connectivity within India.

Figure 4: Carnelian bead finds from 7th–10th century contexts across the Indian subcontinent

Figure 4: Carnelian bead finds from 7th–10th century contexts across the Indian subcontinent

Illustration: Jason Hawkes

36At all of these sites, carnelian beads were (as in East Africa) part of a much wider assemblage of beads, including those made from other types of stone, as well as from other materials such as clay, glass, and shell. Within this wider framework of bead use, carnelian and other types of stone beads do not appear to have been as common as beads made from other materials, with overall quantities of beads made from other materials predominating.

  • 92  M.N.D. Shastri, 1968, p. 207.
  • 93  Cf. L. Gopal, 1965.
  • 94  P. Francis, 2002, p. 106-107.

37However, this does not mean that they were less valued. Indian Sanskrit texts are replete with references to carnelian and other semi-precious stones and discuss them in terms that leave us in no doubt as to their perceived value. For instance, texts such as the Garuda Purana refer to “blood stones” (carnelians) as possessing “the mystic value of increasing the wealth and number of servants of the wearer”.92 It may be that they served as some sort of currency within a domestic economy and system of transaction that was monetized but is also known to have involved agricultural products, shell beads, and semi-precious stones as currency.93 Due, presumably, to their economic value, we also have a number of textual references that allude to their political importance to the kingdoms and states that ruled the areas in which they were made.94

  • 95  P. Francis, 1982.

38This paints a picture of their value in India which, when considered with reference to the value they occupied in East Africa, suggests that with the maritime trade and interaction that took place across the Indian Ocean, people in different regions developed shared tastes and aesthetics, as well as shared understandings of value. This could have meant that carnelian beads functioned almost as a currency not just within each region but also between regions of the Indian Ocean. In this connection, it is interesting to note that carnelian beads were not the most common type of stone bead in India. Far more common were beads made from agate, the quantities of which outnumber those of other beads and have led to the identification of the West Indian agate industry as having been one of the largest stone bead industries the world has ever seen.95 Given the close association between areas of raw material resourcing, centres of production, and ports of Indian Ocean trade, it is perhaps surprising that it is carnelian beads and not agate beads that are found in greater quantities in East Africa. Here, there are only one or two instances of agate beads (presumably Indian) at the sites discussed above. Due to the size and importance of the agate industry within India, it may have been that the trade of agate was more tightly controlled by political or economic elites within the subcontinent. Alternatively, this may point to different preferences, tastes, and consumer choices within East Africa.

Shared practice: Ceramics

  • 96  M.C. Horton, 1987.
  • 97 Ding Yu, pers. comm., 2014.
  • 98  S. Priestman, 2013. See also C. Hardy-Guilbert et al., 2004.
  • 99  For exception, see M.C. Horton, forthcoming.
  • 100  H.N. Chittick, 1974.

39Ceramics are, of course, ubiquitous in archaeological assemblages and also provide the most obvious indicators of interaction between different regions of the Indian Ocean. Imported ceramics have been emphasized in the archaeology of the Swahili coast, as in the Indian Ocean more generally, and used to demonstrate connections with far-flung regions, as well as charting changing directions of trade.96 Despite this, quantified approaches that attempt to reconstruct levels of trade are rare. They might point to differential patterns across the ocean and might be used to consider types of trade. For example, ongoing excavations at Mambrui on the coast of Kenya are exploring differing forms of relationship with Chinese maritime activity, as evidenced by ratios of imperial to export ceramic types.97 Likewise, a recent study of Persian Gulf ceramics from Manda (Kenya) and sites in and around the Gulf has identified a differential intensity of exchange between regions.98 Possible Indian ceramics have been less well served, partly owing to a lack of visibility and partly because of the assumptions that accompany them in African contexts.99 Chittick, for example, considered Indian jars with “local” ceramics as part of his analysis that focused on “kitchen wares” and ceramics intended for domestic use.100 Yet, little consideration was given to why Indian water jars might be part of a domestic assemblage.

  • 101  S. Wynne-Jones, in press.
  • 102  H.N. Chittick, 1974, p. 302.

40In fact, the dynamics of exchange and interaction that resulted in the record of imports in East African towns needs further consideration in general.101 The better-studied imports from the Persian Gulf and Far East were always present only in tiny quantities. The richest sites of the late first millennium—Unguja Ukuu, Tumbe, Manda, Shanga—have imported sherd ratios that never exceed 5%, and more commonly hover around 2%. Elsewhere, the early levels at Kilwa can offer only tiny amounts of imported ceramics—0.2% of the sherds recovered from 9th century levels.102

  • 103  D. Kennet, 2004.

41Local ceramics of the 7th–10th centuries are anyway poorly understood within India, having been given only cursory consideration on the subcontinent.103 Nevertheless, we are able to recognize Indian types at a number of sites on the eastern African coast between the 7th and 10th centuries. These are interesting for a number of reasons. First, the ceramic sherds that are identifiably Indian in origin point to the movement of goods from one region to another and yet are likely not objects of trade in their own right. The ceramic record, then, the archaeologist’s primary tool, may suggest that objects are moving in other ways. As the ceramics are found as part of the archaeological record of the sites, they are bound into the domestic world of the Swahili settlements, and we can perhaps begin to discern a tiny but significant Indian presence here, the ceramics having accompanied travellers, craftspeople, or co-religionists coming to the region. Second, we are able to point to a less tangible—and more tentative—connection visible via the ceramics. It is possible that some local types of ceramic during these centuries were affected by contact with or knowledge of Indian vessels, or perhaps forms of consumption shared with the Indian subcontinent. Thus, we might see again the development of a more Indian Ocean aesthetic, common to both regions and demonstrating their entanglement during these centuries.

  • 104  H.N. Chittick, 1984.
  • 105  A. Juma, 2004.
  • 106  M.C. Horton, 1996.
  • 107  C.M. Kusimba, 1993, p. 174.
  • 108  T.H. Wilson, A.L. Omar, 2010, p. 57.
  • 109  H.N. Chittick, 1974, p. 306.

42Most of the ceramics identified as Indian within East Africa assemblages are water jars, mixed into and clearly used as part of the domestic assemblages of the towns. Yet a less mundane type has also been recognized in the form of Red Polished Ware (RPW), identified in layers attributed to the 7th and 8th centuries at Manda;104 at Unguja Ukuu in layers radiocarbon dated to the 7th to 9th centuries;105 and possibly at Shanga, where quantities of an Indian “red burnished ware” (which may or may not be Indian RPW) peak later, around the mid-11th century.106 Other examples of Indian red burnished vessels are also found in layers dating to the 11th century onwards at Mtwapa,107 Pate,108 and Kilwa.109 Instances of Indian Red Polished and Burnished Wares in East Africa are thus quite rare, with only few examples having been found; these nonetheless need to be understood. In fact, their rarity is an interesting characteristic in its own right, as it suggests that these were not significant objects of trade and may instead have had special purpose or accompanied particular individuals. Although Indian ceramics have been lumped together as a more general category, RPW suggests something different, as it is known as a special type of tableware within Indian contexts.

43RPW is in fact a relatively rare type within India also, although it is well-known and easy to recognize (Figure 5).

Figure 5: Red Polished Ware from India

Figure 5: Red Polished Ware from India

Photograph: Jason Hawkes

  • 110  S.R. Rao, 1966.
  • 111  D. Kennet, 2004, 2013.
  • 112  S. Subbarao, 1953, p. 56-58.

44It has been used as a fossil-type for dating since its discovery at Amreli, Gujarat in the 1960s.110 It is traditionally dated to between the 1st and 5th or 6th centuries AD, yet these dates are based primarily on stratigraphic association with early coins at only a few archaeological contexts. The potential residuality of the coins has never been taken into account, and there have recently been calls for the chronology of RPW in India to be reassessed.111 Given the proven existence of RPW in contexts that have been radiocarbon dated to between the 7th and 9th centuries in the Persian Gulf and East Africa, there is every reason to imagine that it continued to be produced in India until at least this time. In addition, it is widely accepted that over time the manufacture of RPW in India was succeeded by a later form of Red Burnished Ware, defined as a separate ware (or series of wares) on the basis of differences in fabric and surface treatment (still fine, but slightly less so), which continued to be produced until the early second millennium.112

  • 113 Ibid.; S.R. Rao, 1966.
  • 114  S. Subbarao, 1953; N. Pinto-Orton, 1991; A. Patel, S.V. Rajesh, 2007.

45Production sites for RPW have yet to be found, and scientific source analyses have never been carried out; but based on the distribution of sites at which it is found, and the range of vessel types found at those sites, it appears to have originated in Gujarat.113 Current thinking in Indian scholarship is that after originating here, it quickly spread throughout India, either as a commodity of trade that continued to be produced in Gujarat, or as a technique of ceramic production114—scenarios that, while different, both reflect the existence of pan-Indian networks of contact and exchange that existed in the mid-first millennium. Like carnelian beads, the pattern of its distribution in India seems to have been closely tied to maritime networks: there are clusters of sites with RPW along the west coast and its immediate hinterland (Figure 6).

Figure 6: Distribution of Red Polished Ware within India

Figure 6: Distribution of Red Polished Ware within India

Illustration: Jason Hawkes

46This suggests that it was distributed at least as much via maritime connections as by land-based routes. With this in mind, and when we consider its existence in the Persian Gulf and East Africa, it may be that the appearance of RPW in, and spread from, Gujarat was in some way connected to these wider Indian Ocean maritime networks.

  • 115  For the most detailed typological survey of Red Polished Wares in India to have been published so (...)

47In India, RPW is ubiquitous in urban and religious contexts, yet is quite rare within each individual assemblage. Far more common are coarse red and grey wares, as well as other varieties of slipped wares (red and black). Precise quantities are rarely recorded in the published reports; yet, where they are, RPW accounts only for between 2.8 and 0.06% of the ceramic assemblage in any one stratigraphic layer. This, together with its fine quality and appearance, has led to RPW being thought of as a luxury ware and associated with high status use and demand by social and political elites. This is also reflected in the range of vessel types made using this ware. Most are vessels that can be categorized as “fine table wares”: small cups with conical or flared sides; shallow bowls or dishes with flared or upturned sides and a variety of rim forms; a wide variety of small pots and jars with either plain or carinated shoulders; and very distinctive high-necked vessels with either simple out-turned rims or a vertical spout with a very small opening leading to their classification as “sprinklers”.115 A very small number of larger storage jars do also occur. Yet, other than these large jars, no cooking pots or other more utilitarian vessels made from RPW have been found. Many of these vessel types appear to be replicated (in far greater numbers) in other ceramic wares—most usually other red and black slipped wares. This indicates that RPW was not necessarily reserved for any particular types of vessel in India, other than these high-necked vessels and sprinklers. Instead, its exclusivity appears to have been based on its high quality.

  • 116  J.B. Fleisher, S. Wynne-Jones, 2011.

48Over and above the existence of Indian RPW at sites in East Africa, this rarity of use and postulated “luxury” nature of RPW in India has clear parallels with some indigenous types in East Africa. In particular, locally produced ceramic assemblages of the 7th–9th centuries on the coast contain a class of red painted, burnished bowls. These are restricted to coastal settlements and to particular types of site. A recent study of ceramics of this period found that such bowls were common at Manda, Tumbe, and Unguja Ukuu (all deeply implicated in trade) but were extremely rare at hinterland sites (Figure 7).116

Figure 7: Graph of bowls vs. jars at East African sites of 7th–10th centuries AD

Figure 7: Graph of bowls vs. jars at East African sites of 7th–10th centuries AD

Illustration: Stephanie Wynne-Jones

49Most of the ceramics produced at coastal sites are akin to those from a much broader African region, with large coarse vessels linked to communal consumption predominating. The small bowls, indicative of individualized consumption, occur at only a subset of these sites, pointing perhaps to a practice of eating and a social setting that was associated with the ocean rim.

  • 117  N. Martin, 2010. It should also be noted, however, that a small number of undiagnostic imported sh (...)

50The distribution of Indian imports and of locally produced red burnished ceramics do not map directly onto one another. Notably, the site at Shanga has yielded almost no red burnished bowls, yet is a key site for recognizing Indian imports. On the other hand, Comorian sites such as Dembeni are heavily characterized by the red burnished bowls, but so far no definitive Indian ceramics have been reported.117 The African bowl forms are similar to those noted in Indian RPW assemblages, but are not seen among imports in East Africa. We do not therefore suggest that these ceramics directly mimic the RPW, but instead note that they might reflect some similar practices.

51The burnished bowls found at sites on the East African coast at this time are part of a broader category of closed bowls, with in-turned rim and often a shoulder carination, although these can also occur in grittier, less fine fabrics (Figure 8).

Figure 8: Examples of closed bowl forms from East African coastal sites

Figure 8: Examples of closed bowl forms from East African coastal sites

Illustration: Stephanie Wynne-Jones

  • 118  B. Subbarao, 1953, p. 36, fig. 15.
  • 119  V. Begley, 2004, p. 164-165.
  • 120  R.N. Mehta, 1968, p. 52, 73, 75, figs 28, 41, 43.
  • 121  S.B. Deo, M.K. Dhavalikar, 1968, p. 51-52, fig. 12.

52Although rare for East African assemblages and not reported at hinterland sites such as Dakawa or Misasa in Tanzania, these bowls do have direct affinities with types diagnostic of this timeframe on the Indian subcontinent. Direct parallels in terms of vessel form can be seen at Akota (vessel type 9);118 Arikamedu (vessel types 3.63 and 3.64);119 Nagara (vessel types 175–178, 410, and 443);120 and Paunar (vessel type 89) (Figure 9).121

Figure 9: Comparative examples of bowl forms from Indian sites

Figure 9: Comparative examples of bowl forms from Indian sites

Illustration: Jason Hawkes (after V. Begley, 2004, R.N. Mehta, 1968 and B. Subbarao, 1953).

53From the evidence provided by these direct parallels that have been found at these sites, we can also see that these vessels were distributed widely throughout South Asia, appearing as they do in west, central, and south India, and at locations both on the coast and inland. It is clear that they were part of a much more widespread and unbroken tradition of making small shallow bowls and dishes in India, which dates back to at least the early first millennium BC. Within this tradition, we see a great deal of variation of form in terms of the following: rim diameter (ranging from 10 to over 30 cm); the thickness of the body (from a few millimetres to a centimetre); the degree of curvature and carination of the shoulder; and the profile of the rim, with some varieties being less incurved, vertical, or even slightly everted—all of which can be thickened, tapering, bevelled, or blunt.

  • 122  R.C. Gaur, 1983.
  • 123  A.K. Narrain, T.N. Roy, 1976-78.
  • 124  R.N. Mehta, H.D. Sankalia, 1955.
  • 125  K.V. Soundara Rajan, 1994.
  • 126  J.M. Nanavati et al., 1971.
  • 127  M.G. Dikshit, 1955.
  • 128  H.D. Sankalia, M.G. Dikshit, 1952.

54When viewed in this light, and broadening the classificatory parameters slightly to include even just a small amount of variation in terms of the degree of curvature of the shoulder and profile of the rims, we can identify a large number of additional parallels at both these and other sites in India. These include: Atranjikhera,122 Rajghat,123 and Timbarva,124 where a number of vessels types have been found in layers dating to the first millennium BC; Kaveripattinam,125 Somnath,126 and Tripuri,127 where they occur throughout the late first millennium BC and first millennium AD; and Brahmapuri,128 where they are found in layers dating to the early second millennium AD.

55This is, of course, only a tentative suggestion. Yet, the bowls hint at a form of consumption that might, within Africa, have been associated with the Indian Ocean world and was engaged in only by coastal groups. They prefigure a move towards very different material practices in later centuries but are one of the only aspects that separate the ceramic assemblages of coastal settlements from their hinterland in the period before AD 1000. As such, they hint at a world in which certain practices of consumption were highlighted and linked to the growing interconnectivity of the Indian Ocean world. It is possible that the outlines of these forms of consumption have links to Indian practices. If this is the case, then it may be that what we are seeing in the evidence provided by these ceramics is the replication, in both places, of common understandings of form and function. This could, as Horton has suggested for later periods, suggest diaspora communities of craftsmen from India who were familiar with these bowl/dish forms and made them in local materials. Equally, it seems just as likely that we are seeing echoes of contacts that shaped an Indian Ocean form of materiality, which was enacted through daily consumption and through practices linked to food, and which began to separate the Swahili coast from its hinterland even at this early period. It is in these ephemeral ways that we might see connections between Africa and India, as the scarcity of objects of direct exchange force a rethinking of the water jars, table wares, and bowls that do exist.


56From the examination of just these four categories of artefacts, then, we can identify a number of different types of contacts between India and Africa during the late first millennium AD. The record of shared symbolism on coinage, shared aesthetics in semi-precious stone beads, and possible shared practice seen in the changing repertoire of ceramics suggest a dense yet amorphous web of changing social connections that bound communities together in various ways and at different times. During this early period, we can see only traces of this web of connections. Certainly, from the 11th century onwards, direct connectivity is easier to recognize, through a range of goods (in particular, the growing record of Indo-Pacific glass beads) that moved between regions across the Indian Ocean. It is into this denser network that the Shanga lion fits, evoking notions of diaspora and the movement of craftspeople among a cosmopolitan community of producers and consumers.

57We argue that it is precisely the elusiveness of these earlier connections that makes them interesting and important. Shared senses of value, notions of aesthetics, and social practices such as these are not recorded in the accounts of travellers or trade itineraries and do not fit into the usual models that chart the directions of trade. Nor were they always mediated by commodity trade. Recognizing these connections thus extends our conception of Indian Ocean interactions beyond the traditional understanding of trade relationships that has dominated both historical and archaeological scholarship. Furthermore, in highlighting connections between India and Africa, we can see that this network extended beyond the rigid (economically determined) core–periphery framework that is either implied by the traditional focus on the Persian Gulf and the agency of Arab merchants, or made explicit by positing India as a multi-centred core.

58That is not to say that trade was unimportant. It unquestionably was, both in terms of being the main form of maritime interactions that took place and the effects these interactions had on societal developments throughout the Indian Ocean world. Nor do we attempt to deny the economic and geopolitical importance of Arabia within this framework. The Persian Gulf in particular was undeniably a key nexus. The political and economic dynamics at play here both stimulated trade and provided a critical space in which this trade and wider interactions were mediated. Rather, we suggest that being able to identify these other, more amorphous, social connections necessarily and positively complicates this macro-scale picture of trade. At a world-systems scale, it is clear that many of the connections drawn here would not be statistically significant. Yet, they did exist. Incorporating them into our understanding enables us to move towards micro-histories of the larger region—histories that can provide a more nuanced and detailed picture of Indian Ocean interactions, involving more than trade and economic relationships, and that enable us to reconstruct something of the social milieu within which these transactional relationships were negotiated and, perhaps, facilitated.

  • 129  Cf. P. Beaujard, 2012.
  • 130  E.g. R.B. Mason, E.J. Keall, 1991; L. Dussubieux et al., 2008; R. Tomber et al., 2011; M. Wood, 20 (...)
  • 131  For instance, Nurse and Hinnebusch have highlighted a number of linguistic innovations in Proto-Sa (...)

59There would, undoubtedly, have been many other connections between communities and regions across the Indian Ocean, such as, but by no means limited to, diaspora and pilgrimages. There would also have been many other routes by which connections might be seen to have existed. Indeed, the movements of Austronesian peoples and practices, rightly emphasized by Beaujard, reflect something of the sheer geographical scale and complexity of these movements.129 Some of these connections will remain archaeologically invisible. Yet, what should by now be clear is that it is possible to both identify traces of these connections and understand more of the wider network of oceanic mobility to which different communities and regions belonged. Further work could, and should, add to this picture, finding additional traces of India in Africa during these earliest centuries. In this connection, it is possible to suggest a number of potentially useful avenues for research. Methodologically, the greater application of scientific techniques in the analysis of different artefacts would help identify the geographical source of the materials used in their production and the directions of their movements across the Indian Ocean. Mass spectrometry (LA-ICP-MS), X-ray fluorescence (XRF), and a range of petrographic analyses have all been used to good effect in the examination of ceramics, glass, and semi-precious stone.130 Isotopic analyses of worked bone and ivory would also undoubtedly prove immensely useful. Over and above the application of these techniques, a greater emphasis could also be placed on regionally (if not inter-regionally) oriented analyses in order to identify patterns of sourcing, production, and consumption on a much broader scale than is currently possible. It is clear that future work will also need to pay close attention to context, as it is the quality rather than the quantity of the data that allows exploration of these types of connection. Beyond these methodological concerns, it is also essential that archaeologists seek to move beyond the (at times) insular foci of the scholarly traditions in which they operate, and engage with the material, methods, and theories that are found and used in other countries. Doing so will not only facilitate comparison of material between regions, but also lead to interesting synergies and new interpretative models. Nor, indeed, is archaeology the only avenue open to such enquiry. Areas such as linguistics also offer a great deal of potential for future work.131 Expanding research in these ways will not only benefit understanding of events and processes on the Swahili coast, but will also lead to a more nuanced and connected account of developments that took place throughout the Indian Ocean.

Haut de page


Abu-Lughod, J., 1989, Before European hegemony: The world system A.D. 1250–1350, Oxford, Oxford University Press.

Album, S., 1999, Sylloge of Islamic coins in the Ashmolean: Arabia and East Africa, Oxford, Ashmolean Museum Press.

Allibert, C., 1991a, “Wakwak: Vegetal, mineral ou humain? Reconsidération du problème”, Études Océan Indien, 12, p. 171-189.

Allibert, C., 1991b, “La description de l’île de pines (1668): réalité historique, phénomène littéraire ou événement sociologique?”, Études Océan Indien, 13, p. 119-153.

Allibert, C., Argant, A., Argant, J., 1990, “Le site de Dembéni”, Études Océan Indien, 11, p. 63-172.

Arkell, A.J., 1936, “Cambay and the bead trade”, Antiquity, 10, p. 292-305.

Banerjee, N.R., 1959, “The technique and manufacture of stone beads in ancient Ujjain”, Journal of the Asiatic Society, 1 (2), p. 189-195.

Beaujard, P., 2005, “The Indian Ocean in Eurasian and African world-systems before the sixteenth century”, Journal of World History, 16 (4), p. 411-465.

Beaujard, P., 2012, Les mondes de l’océan Indien, 2 vols., Paris, Armand Colin.

Beaujard, P., 2013, “Systèmes-mondes anciens. Processus du domination, de co-évolution et de résistance. L’Exemple de la Côte Est-Africaine avant le xviie siècle”, Histoire Globale, 53, p. 40-62.

Begley, V., 1996, The ancient port of Arikamedu: New excavations and researches, 1989–1992, Paris, École française d’Extrême-Orient.

Begley, V., 2004, The ancient port of Arikamedu: New excavations and researches, 1989–1992, Paris, École française d’Extrême-Orient.

Blaut, J., 1993, The colonizer’s model of the world, New York, Guildford Press.

Brown, H.W., 1993, “Coins of East Africa: An introductory survey”, Yarmouk Numismatics, 5, p. 9-16.

Brown, H.W., 1996, “The coins”, in M.C. Horton (ed.), Shanga: The archaeology of a Muslim trading community on the coast of East Africa, London and Nairobi, BIEA, p. 368-377.

Casson, L., 1989, Periplus Maris Erythraei. Text with introduction, translation and commentary, Princeton, Princeton University Press.

Chakravarti, R., 2004, “An enchanting seascape: Through epigraphic lens”, Studies in History, 20 (2), p. 305-315.

Chami, F., 1994, The Tanzanian coast in the early first millennium AD: An archaeology of the iron-working, farming communities, Uppsala, Societas Archaeologica Uppsaliensis.

Chaudhuri, K.N., 1990, Asia before Europe: Economy and civilisation of the Indian Ocean from the rise of Islam to 1750, Cambridge, Cambridge University Press.

Chittick, H.N., 1965, “The ‘Shirazi’ colonization of East Africa”, Journal of African History, 6 (3), p. 275-294.

Chittick, H.N., 1974, Kilwa: An Islamic trading city on the East African coast, Nairobi and London, British Institute in Eastern Africa.

Chittick, H.N., 1984, Manda: Excavations at an island port on the Kenya Coast, Nairobi and London, British Institute in Eastern Africa.

Coins, I.E., 1998, “Pre-Muslim coins of Chach”, Silk Road Art and Archaeology: Journal of the Institute of Silk Road Studies, Kamakura, 5, p. 307.

Collins, B.A., 1994, The best divisions for knowledge of the regions: A translation of “Ahsan al-Taqasim fi Maarifat al-Aqalim”, Reading, Garnet.

Dani, A.H., 2007, History of Pakistan: Pakistan through ages, Lahore, Sang-e-Meel Publications.

Deo, S.B., Dhavalikar, M.K., 1968, Paunar excavation, 1967, Nagpur, Nagpur University.

Dikshit, M.G., 1955, Tripuri – 1952: Being the account of the excavations at Tripuri, Bhopal, Government of Madhya Pradesh.

Dussubieux, L., Kusimba, C.M., Gogte, V., Kusimba, S.B., Gratuze, B., Oka, R., 2008, “The trading of ancient glass beads: New analytical data from South Asian and East African soda-alumina glass beads”, Archaeometry, 50 (5), p. 797-821.

Ekholm, K., Friedman, J., 1982, “‘Capital’ imperialism and exploitation in ancient world-systems”, Review (Fernand Braudel Center), 6 (1), p. 87-109.

Fleisher, J.B., Wynne-Jones, S., 2011, “Ceramics and the early Swahili: Deconstructing the early Tana tradition”, African Archaeological Review, 28 (4), p. 245-278.

Flood, F.B., 2009, Objects of translation: Material culture and medieval “Hindu-Muslim” encounter, Princeton and Oxford, Princeton University Press.

Fouché, L. (ed.), 1937, Mapungubwe: Ancient Bantu civilization on the Limpopo; Reports on excavations at Mapungubwe (Northern Transvaal) from February 1933 to June 1935, Cambridge, Cambridge University Press.

Francis, P., 1982, Indian agate beads, Lake Placid, Lapis Route.

Francis, P., 2002, Asia’s maritime bead trade: 300 BC to the present, Honolulu, University of Hawai’i Press.

Frank, A.G., Gills, B.K., 1993, The world system: Five hundred years or five thousand?, London and New York, Routledge.

Gaur, R.C., 1983, Excavations at Atranjikhera: Early civilization of the Upper Ganga Basin, Delhi, Asian Humanities Press.

Gopal, L., 1965, The economic life of northern India, c. AD 700–1200, Delhi, Motilal Banarsidass.

Gupta, K.P., 1972, “Lower Narmada and its antiquities”, Journal of the Oriental Institute, Baroda, 21 (4), p. 265-274.

Gupta, S.P., Dalal, K., Dandekar, A., Aravazhi, P., Bomble, S., 2004, “On the footsteps of the Zoroastrian Parsis in India: Excavations at Sanjan on the West Coast – 2003”, Journal of Indian Ocean Archaeology, 1, p. 93-106.

Gupta, S.P., Dalal, K., Dandekar, A., Nanji, R., Mitra, R., Pandey, R., 2002, “A preliminary report on the excavations at Sanjan (2002)”, Puratattva, 32, p. 182-198.

Gupta, S.P.,Dalal, K., Dandekar, A., Nanji, R., Mitra, R., Pandey, R., 2003, “Early Medieval Indian Ocean trade: Excavations at Sanjan, India”, Circle of Inner Asian Art Newsletter, 17 p. 26-34.

Hardy-Guilbert, C., Rougeulle, A., Renel, H., Picard, C., Kervran, M., 2004, “Ports et commerce maritime islamiques. Présentation du programme APIM (Atlas des ports et itinéraires maritimes du monde musulman)”, in Actes des congrès de la Société des historiens médiévistes de l’enseignement supérieur public, 35e congrès, La Rochelle, p. 79-97.

Hawkes, J., 2014, “Mind the gap: Chronological sequences and the ‘problem’ of Early Medieval settlement in India”, Puratattva, 44, p. 208-228.

Hawkes, J., 2015 (forthcoming), “Finding the Early Medieval period in South Asian archaeology”, Asian Perspectives, 53 (1).

Horton, M.C., 1987, “The Swahili corridor”, Scientific American, 257 (Sept.), p. 86-93.

Horton, M.C., 1996, Shanga: The archaeology of a Muslim trading community on the east coast of Africa, London and Nairobi, British Institute in Eastern Africa.

Horton, M.C., 2004, “Craftspeople, communities, and commodities: Medieval exchanges between northwestern India and East Africa”, Ars Orientalis, 34, p. 62-80.

Horton, M.C., (forthcoming), Zanzibar and Pemba, Nairobi and London, British Institute in Eastern Africa.

Horton, M.C.,Blurton, T.R., 1988, “‘Indian’ metalwork in East Africa: The bronze lion statuette from Shanga”, Antiquity, 62, p. 11-23.

Horton, M.C., Middleton, J., 2000, The Swahili: The social landscape of a mercantile society, Oxford, Blackwell.

Insoll, T., Polya, D., Bhan, K., Irving, D., Jarvis, K., 2004, “Towards an understanding of the carnelian bead trade from western India to sub-Saharan Africa: The application of UV-LA-ICP-MS to carnelian from Gujarat, India, and West Africa”, Journal of Archaeological Science, 31, p. 1161-1173.

Jain, K.C., 1972, Malwa through the ages, Delhi, Motilal Banarsidass.

Jayasuriya, S.D.S., 2003, The African Diaspora in the Indian Ocean, Trenton, NJ, Africa World Press.

Juma, A., 2004, Unguja Ukuu on Zanzibar: An archaeological study of early urbanism, Uppsala, Uppsala Universitet.

Kennet, D., 2004, “The transition from Early Historic to Early Medieval in the Vakataka realm”, in H. Bakker (ed.), The Vakataka heritage: Indian culture at the crossroads, Groningen, Egbert Forsten, p. 11-17.

Kennet, D., 2013, “Reconsidering the decline of urbanism in late Early Historic and Early Medieval South Asia”, in J. Schiettecatte, C. Robin (eds), Les préludes de l’Islam: Ruptures et continuités dans les civilisations du Proche-Orient de l’Afrique orientale, de l’Arabie et de l’Inde à la veille de l’Islam, Orient & Méditerranée n° 11, Paris, de Boccard, p. 331-353.

Kenoyer, J., Vidale, M., Bhan, K., 1994, “Carnelian bead production in Khambhat, India: An ethnoarchaeological study”, in B. Allchin (ed.), Living traditions: Studies in the ethnarchaeology of South Asia, New Delhi, Oxford University Press, p. 281-306.

Kervran, M., 2013, “La mosquée de Ǧalāl al-Dīn Ḫwārazm Sāh à Damrilā, dans les bouches de l’Indus”, in C. Müller, M. Roiland-Rouabah (eds), Les non-dits du nom. Onomastique et documents en terres d’Islam: Mélanges offerts à Jacqueline Sublet, Bayreuth, Presses de l’Ifpo, p. 525-546.

Khan, F.A., 1963, Banbhore; A preliminary report on the recent archaeological excavations at Banbhore, Karachi, Department of Archaeology and Museums, Government of Pakistan.

Kusimba, C.M., 1993, The archaeology and ethnography of iron metallurgy on the Kenya coast, Unpublished PhD thesis, Bryn Mawr College.

Lambourn, E., 1999, “The decoration of Fakhr al-Din mosque in Mogadishu and other pieces of Gujarati marble carving on the East African coast”, Azania, 34 (1), p. 61-86.

Lambourn, E., 2003a, “La production de marbre sculpté à Cambaye au Gujarat et son exportation dans l’Océan Indien (xiiiexve siècles ap. J. C.)”, in J.M.d.S. Alves, C. Guillot, R. Ptak (eds), Mirabilia Asiatica. Produtos raros no comércio marítimo. Produits rares dans le commerce maritime, Wiesbaden and Lisbon, Harrassowitz Verlag and Fundação Oriente, p. 209-252.

Lambourn, E., 2003b, “Of jewels and horses: The career and patronage of an Iranian merchant under Shah Jahan”, Iranian Studies, 36 (2), p. 213-246.

Lambourn, E., 2014a, “Borrowed words in an ocean of objects: Geniza sources and new cultural histories of the Indian Ocean”, in K. Veluthat, D.R. Davis (eds), Irreverent history. Essays for M.G.S. Narayan, New Delhi, Primus Books, p. 215­242.

Lambourn, E., 2014b, “‘Describing a lost camel’ – Clues for West Asian mercantile networks in South Asian maritime trade (tenth – twelfth centuries CE)”, in M.-F. Boussac, J.-F. Salles, J.-B. Yon (eds), Harbours of the Indian Ocean. Proceedings of the Kolkata Colloquium 2011, New Delhi, Primus Books, p. 333-389.

Lambourn, E., 2015 (forthcoming), “Ali Akbar’s red horse – Collecting Arab horses in the early modern culture of Empire”, in C. Anderson (ed.), Early modern merchants as collectors, Aldershot, Ashgate Publishing.

Lowick, N., 1990, Coinage and history of the Islamic world, London, Variorum Publishing.

Martin, N., 2010, Les productions céramiques de l’océan Indien occidental: implications culturelles au carrefour d’influences, thèse de doctorat, Inalco.

Mason, R.B., Keall, E.J., 1991, “The ‘Abbasid glazed wares of Siraf and the Basra connection: Petrographic analysis”, Iran, 29, p. 51-66.

al-Mas’ūdī, 1861-1877, Kitāb murūj adh-dhahab wa ma’ādin al-jawhar, trad. C. Barbier de Meynard et A. Pavet de Courteille « Le livre des prairies d’or et des mines de pierres précieuses », Paris, Imprimerie nationale, 9 vol. ; nouvelle édition C. Pellat, Paris, Société asiatique, 1962-1971, 9 vol.

McPherson, K., 1993, The Indian Ocean: A history of the peoples and the sea, Oxford, Oxford University Press.

Mehta, R.N., 1968, Excavations at Nagara, Baroda, Maharaja Sayajirao University of Baroda.

Mehta, R.N., Sankalia, H.D., 1955, Excavations at Timbarva, Baroda District: April 1953, Baroda, Maharaja Sayajirao University of Baroda.

Miller, D., Desai, N., Lee-Thorp, J., 2000, “Indigenous gold mining in southern Africa: A review”, South African Archaeological Society Goodwin Series, 8, p. 91-99.

Mitchiner, M., 1977, The world of Islam: Oriental coins and their values, London, Hawkins.

Morin, E., 1990, Introduction à la pensée complexe, Paris, Seuil.

Morrison, H., 1984, “The beads”, in H.N. Chittick (ed.), Manda: Excavations at an island port on the Kenyan coast, London and Nairobi, BIEA, p. 181-189.

Mughal, M.R., 1992, “Early Muslim cities in Sindh and patterns of international trade”, Islamic Studies, 31 (3), p. 267-286.

Nanavati, J.M., Mehta, R.N., Chowdary, S.N., 1971, Somnath, 1956; Being a report of excavations, Baroda, Maharaja Sayajirao University.

Narrain, A.K., Roy, T.N., 1976-78, Excavations at Rajghat, 1957–1958, 1960–1965, Varanasi, Banaras Hindu University.

Nurse, D., Hinnebusch, T., Swahili and Sabaki: A linguistic history, Berkeley and Los Angeles, University of California Press.

O’Brien, P., 2006, “Historiographical traditions and modern imperatives for the restoration of global history”, Journal of Global History, 1, p. 3-39.

Oddy, A., 1995, “Gold foil, strip, and wire in the Iron Age of southern Africa”, in D.A. Scott, J. Podany, B.B. Considine (eds), Ancient & historic metals: Conservation and scientific research, Los Angeles, The Getty Conservation Institute, p. 183-196.

Patel, A., Rajesh, S.V., 2007, “Red Polished Ware (RPW) in Gujarat, Western India: An archaeological perspective”, Pragdhara, 17, p. 1-29.

Pearson, M., 1998, Port cities and intruders: The Swahili coast, India, and Portugal in the Early Modern Era, Baltimore and London, Johns Hopkins Press.

Pearson, M., 2003, The Indian Ocean, Abingdon, Routledge.

Piggot, D.W.I., 1941a, “Mafia – History and traditions. Collected by Kadhi Amur Omari Saadi (trans. by D.W.I. Piggott)”, Tanganyika Notes and Records, 12, p. 23-27.

Piggot, D.W.I., 1941b, “History of Mafia”, Tanganyika Notes and Records, 11, p. 35-40.

Pinto-Orton, N., 1991, “Red polished ware in Gujarat: A catalogue of twelve sites”, in V. Begley, R.D. De Puma (eds), Rome and India: The ancient sea trade, Madison, University of Wisconsin Press, p. 46-81.

Priestman, S., 2013, A quantitative Archaeological analysis of ceramic exchange in the Persian Gulf and western Indian Ocean, AD c. 400–1275, University of Southampton.

Rao, S.R., 1966, Excavations at Amreli: A Kshatrapa-Gupta town,Baroda, Maharaja Sayajirao University of Baroda.

Rao, S.R., 1973, Lothal and the Indus civilization, New York, Asia Publishing House.

Ray, H.P., Alpers, E.A., 2007, Cross currents and community networks: The history of the Indian Ocean world, New Delhi, New York, Oxford University Press.

Robertshaw, P., Wood, M., Melchiorre, E., Popelka-Filcoffd, R.S.,Glascocke, M.D., 2010, “Southern African glass beads: Chemistry, glass sources and patterns of trade”, Journal of Archaeological Science, 37 (8), p. 1898-1912.

Sankalia, H.D., Dikshit, M.G., 1952, Excavations at Brahmapuri (Kolhapur) 1945–1946, Poona, Deccan College Postgraduate and Research Institute.

Shastri, M.N.D., 1968, The Garuda-Purana, Varanasi, Chowkamba Sanskrit Series Office.

Sheriff, A., 2010, Dhow cultures of the Indian Ocean: Cosmopolitanism, commerce, and Islam, London, Hurst.

Sinclair, P., Ekblom, A., Wood, M., 2012, “Trade and society on the south-east African coast in the later first millennium AD: The case of Chibuene”, Antiquity, 86 (333), p. 723-737.

Sircar, D.C., 1962, “Chinchani copper-plates”, Epigraphia Indica, 32 p.

Smith, G.R., 2008, A traveller in thirteenth-century Arabia: Ibn Al-Mujawir’s Tarikh Al-Mustabsir, London, Ashgate.

Smith, V.A., 1889, “The coinage of the early or imperial Gupta Dynasty of Northern India”, The Journal of the Royal Asiatic Society of Great Britain and Ireland, New Series, 21 (1), p. 1-158.

SoundaraRajan, K.V., 1994, Kaveripattinam excavations 1963-73: A port city on the Tamilnadu coast, New Delhi, Archaeological Survey of India.

Spear, T., 2000, “Early Swahili history reconsidered”, The International Journal of African Historical Studies, 33 (2), p. 257-290.

Subbarao, B., 1953, Baroda through the ages (being the report of an excavation conducted in the Baroda area 1951–1952), Baroda, Maharaja Sayajirao University of Baroda.

Tolmacheva, M., 1993, The Pate Chronicle: Edited and translated from Mss 177, 321, 344, and 358 of the Library of the University of Dar Es Salaam, Ann Arbor, MI, University of Michigan Press.

Tomber, R., Cartwright, C., Gupta, S., 2011, “Rice temper: Technological solutions and source identification in the Indian Ocean”, Journal of Archaeological Science, 38 (2), p. 360-366.

Walker, J., 1939, “Some new coins from Kilwa”, Numismatic Chronicle, XIX, p. 223-227.

Wallerstein, I., 1974, The modern world-system, vol. I: Capitalist agriculture and the origins of the European world-economy in the sixteenth century, London, New York, Academic Press.

Walsh, M., 2010, “Deep memories or symbolic statements? The Diba, Debuli and related traditions of the East African coast”, in C. Radimilahy, N. Rajaonarimanana (eds), Civilisations des mondes insulaires (Madagascar, îles du canal de Mozambique, Mascareignes, Polynésie, Guyanes): Mélanges en l’honneur du Professeur Claude Allibert, Paris, Karthala, p. 453-476.

Walz, J.R., 2010, Route to a regional past: An archaeology of the Lower Pangani (Ruvu) Basin, Tanzania, 500–1900 C.E., University of Florida.

Wilson, H., 1835, Select specimens of the theatre of the Hindus, London, Parbury, Allen & Co.

Wilson, T.H., Omar, A.L., 1997, “Archaeological investigations at Pate”, Azania, 32 (1), p. 31-76.

Wink, A., 1990, Al-hind: The making of the Indo-Islamic world, Leiden, Brill.

Wolf, E., 1982, Europe and the people without history, Berkeley, University of California Press.

Wood, M., 2002, “The glass beads of Kaole”, in F. Chami, G. Pwiti, C. Radimilahy (eds), Southern Africa and the Swahili World, Dar es Salaam, University of Dar es Salaam Press, p. 50-65.

Wood, M., 2012, Interconnections: Glass Beads and Trade on Southern and Eastern Africa and the Indian Ocean: 7th to 16th centuries AD, Uppsala, Department of Archaeology and Ancient History, Studies in Global Archaeology 17.

Wood, M., 2015, “Divergent patterns in Indian Ocean trade to East Africa and southern Africa between the 7th and 17th centuries CE: The glass bead evidence”, Afriques [en ligne], 06. URL:

Wood, M., Dussubieux, L., Robertshaw, P., 2012, “The glass of Chibuene, Mozambique: New insights into early Indian Ocean Trade”, South African Archaeological Bulletin, 67 (195), p. 59-74.

Wynne-Jones, S., in press 2015, A material culture: Consumption and trade on the precolonial East African coast, Oxford, Oxford University Press.

Wynne-Jones, S., Fleisher, J.B., 2012, “Coins in context: Local economy, value and practice on the East African Swahili coast”, Cambridge Archaeological Journal, 22 (1), p. 19-36.

Haut de page


1 M.C. Horton, 1996.

2 M.C. Horton, T.R. Blurton, 1988.

3 Ibid., p. 19.

4 Ibid., p. 20.

5  L. Fouché, 1937.

6  D. Miller et al., 2000.

7  A. Oddy, 1995, p. 186.

8  M.C. Horton, 2004.

9  See also E. Lambourn, 1999.

10  Cf. E. Wolf, 1982; J. Blaut, 1993; P. O’Brien, 2006.

11  I. Wallerstein, 1974.

12  J. Abu-Lughod, 1989.

13  A. Wink, 1990.

14  J. Abu-Lughod, 1989, p. 266.

15 Ibid., p. 272.

16  P. Beaujard, 2005.

17 E.g. K. Ekholm, J. Friedman, 1982; F. Frank, B.K. Gills, 1993.

18  Also P. Beaujard, 2012, 2013.

19  Drawing on E. Morin, 1990.

20  P. Beaujard, 2005, 2012.

21  P. Beaujard, 2005, p. 241.

22  K.N. Chaudhuri, 1990.

23  M. Pearson, 2003. See also K. McPherson, 1993; H.P. Ray, E.A. Alpers, 2007.

24  As opposed to attempting to define an ‘Indian Ocean culture’, cf. K.N. Chaudhuri, 1990.

25  M. Pearson, 2003, p. 5.

26  Following M. Pearson, 1998.

27  P. Beaujard, 2013.

28  See also A. Sheriff, 2010.

29  R. Chakravarti, 2004.

30  E. Lambourn, 2014b.

31  E.g. E. Lambourn, 2003a, 2014a.

32  E. Lambourn, 2014b.

33  E. Lambourn, 1999, 2003a, 2003b.

34  F.B. Flood, 2009.

35 Ibid., p. 11.

36  E. Lambourn, 2015 (forthcoming).

37  S.D.S. Jayasuriya, 2003.

38  J. Hawkes, 2014.

39 al-Mas’ūdī, 1861-1877.

40  H.N. Chittick, 1984.

41  J.B. Fleisher, S. Wynne-Jones, 2011.

42  C. Hardy-Guilbert et al., 2004.

43  F.B. Flood, 2009, p. 21.

44  For instance, the 11th-century writings of al-Biruni, and the 13th-century accounts of al-Mujawir.

45  C. Allibert, 1991a, 1991b; M. Tolmacheva, 1993; T. Spear, 2000.

46  M. Walsh, 2010.

47  D.W.I. Piggot, 1941a, 1941b.

48  See M. Walsh, 2010 for a critical overview.

49  H.N. Chittick, 1965; M.C. Horton, 2004.

50  M.C. Horton, forthcoming, p. 40.

51  M. Walsh, 2010.

52  M.C. Horton, J. Middleton, 2000.

53  B.A. Collins, 1994; G.R. Smith, 2008.

54  M.R. Mughal, 1992; M. Kervran, 2013.

55  F.A. Khan, 1963.

56  S.P. Gupta et al., 2002; S.P. Gupta et al., 2003; S.P. Gupta et al., 2004.

57  D.C. Sircar, 1962.

58  H.W. Brown, 1996, p. 369.

59  H.P. Brown, 1996; M.C. Horton, 2004.

60  A.H. Dani, 2007.

61  V.A. Smith, 1889; I.E. Coins, 1998.

62  N. Lowick, 1990.

63  J. Walker, 1939; H.P. Brown, 1993; S. Album, 1999; S. Wynne-Jones, J.B. Fleisher, 2012.

64  H.W. Brown, 1996, p. 369.

65  F.A. Khan, 1963, p. 45.

66  M. Mitchiner, 1977, p. 131.

67  H.W. Brown, 1993, p. 14.

68  H.P. Brown, 1996, p. 372.

69  L. Dussubieux et al., 2008; P. Robertshaw et al., 2010; M. Wood et al., 2012; M. Wood, 2015.

70  V. Begley, 1996; V. Begley, 2004.

71  P. Francis, 2002.

72  H. Morrison, 1984; M.C. Horton, 1996; M. Wood, 2002, 2012.

73  M. Wood, 2015.

74  P. Sinclair et al., 2012; M. Wood et al., 2012.

75  T. Insoll et al., 2004.

76  A.J. Arkell, 1936.

77  P. Francis, 2002.

78  K.P. Gupta, 1972, p. 266.

79  S.R. Rao, 1973, p. 102.

80  A. Juma, 2004, p. 128.

81  M.C. Horton, 1996, p. 333.

82  H. Morrison, 1984, p. 184.

83  F. Chami, 1994, carnelian bead illustrated, p. 66.

84  J.R. Walz, 2010, p. 335.

85  L. Casson, 1989, p. 81.

86  H. Wilson, 1835, p. 85-86.

87  N.R. Banerjee, 1959.

88  K.C. Jain, 1972.

89  A.J. Arkell, 1936. For a valuable ethno-archaeological study of carnelian bead production in India, see J. Kenoyer et al., 1994.

90  For further discussion, see P. Francis, 2002.

91  J. Hawkes, 2015 (forthcoming).

92  M.N.D. Shastri, 1968, p. 207.

93  Cf. L. Gopal, 1965.

94  P. Francis, 2002, p. 106-107.

95  P. Francis, 1982.

96  M.C. Horton, 1987.

97 Ding Yu, pers. comm., 2014.

98  S. Priestman, 2013. See also C. Hardy-Guilbert et al., 2004.

99  For exception, see M.C. Horton, forthcoming.

100  H.N. Chittick, 1974.

101  S. Wynne-Jones, in press.

102  H.N. Chittick, 1974, p. 302.

103  D. Kennet, 2004.

104  H.N. Chittick, 1984.

105  A. Juma, 2004.

106  M.C. Horton, 1996.

107  C.M. Kusimba, 1993, p. 174.

108  T.H. Wilson, A.L. Omar, 2010, p. 57.

109  H.N. Chittick, 1974, p. 306.

110  S.R. Rao, 1966.

111  D. Kennet, 2004, 2013.

112  S. Subbarao, 1953, p. 56-58.

113 Ibid.; S.R. Rao, 1966.

114  S. Subbarao, 1953; N. Pinto-Orton, 1991; A. Patel, S.V. Rajesh, 2007.

115  For the most detailed typological survey of Red Polished Wares in India to have been published so far, see N. Pinto-Orton, 1991.

116  J.B. Fleisher, S. Wynne-Jones, 2011.

117  N. Martin, 2010. It should also be noted, however, that a small number of undiagnostic imported sherds (with a yellow, grainy porous fabric) from Dembeni have been identified as having a possible Indian origin, C. Allibert et al., 1990, p. 100. Cf. P. Beaujard, 2012, vol. 2, p. 120, cites C. Allibert, pers. comm.

118  B. Subbarao, 1953, p. 36, fig. 15.

119  V. Begley, 2004, p. 164-165.

120  R.N. Mehta, 1968, p. 52, 73, 75, figs 28, 41, 43.

121  S.B. Deo, M.K. Dhavalikar, 1968, p. 51-52, fig. 12.

122  R.C. Gaur, 1983.

123  A.K. Narrain, T.N. Roy, 1976-78.

124  R.N. Mehta, H.D. Sankalia, 1955.

125  K.V. Soundara Rajan, 1994.

126  J.M. Nanavati et al., 1971.

127  M.G. Dikshit, 1955.

128  H.D. Sankalia, M.G. Dikshit, 1952.

129  Cf. P. Beaujard, 2012.

130  E.g. R.B. Mason, E.J. Keall, 1991; L. Dussubieux et al., 2008; R. Tomber et al., 2011; M. Wood, 2012 and 2015.

131  For instance, Nurse and Hinnebusch have highlighted a number of linguistic innovations in Proto-Sabaki, between AD 350 and 650, such as the use of the words *mbalazi, for “pigeon pea” (Cajanus cajan), a plant of Indian origin, and *kilazi, for “potato” or possibly “yam” or “tuber”; see D. Nurse, T. Hinnebusch, 1993, p. 292. Beaujard has also argued for the Indian origins of *mbalazi and *kilazi; see P. Beaujard, 2012.

Haut de page

Table des illustrations

Titre Figure 1: The coast of eastern Africa, showing 7th–10th century sites mentioned in text
Crédits Illustration: Stephanie Wynne-Jones
Fichier image/jpeg, 132k
Titre Figure 2: Early Tana Tradition ceramics from Dakawa
Crédits Photograph: J. Fleisher.
Fichier image/jpeg, 116k
Titre Figure 3: Gold coins from Kilwa Kisiwani
Crédits Photograph: Helen Brown
Fichier image/jpeg, 96k
Titre Figure 4: Carnelian bead finds from 7th–10th century contexts across the Indian subcontinent
Crédits Illustration: Jason Hawkes
Fichier image/jpeg, 100k
Titre Figure 5: Red Polished Ware from India
Crédits Photograph: Jason Hawkes
Fichier image/jpeg, 48k
Titre Figure 6: Distribution of Red Polished Ware within India
Crédits Illustration: Jason Hawkes
Fichier image/jpeg, 104k
Titre Figure 7: Graph of bowls vs. jars at East African sites of 7th–10th centuries AD
Crédits Illustration: Stephanie Wynne-Jones
Fichier image/jpeg, 80k
Titre Figure 8: Examples of closed bowl forms from East African coastal sites
Crédits Illustration: Stephanie Wynne-Jones
Fichier image/jpeg, 24k
Titre Figure 9: Comparative examples of bowl forms from Indian sites
Crédits Illustration: Jason Hawkes (after V. Begley, 2004, R.N. Mehta, 1968 and B. Subbarao, 1953).
Fichier image/jpeg, 33k
Haut de page

Pour citer cet article

Référence électronique

Jason D. Hawkes et Stephanie Wynne-Jones, « India in Africa: Trade goods and connections of the late first millennium »Afriques [En ligne], 06 | 2015, mis en ligne le 21 décembre 2015, consulté le 22 février 2024. URL : ; DOI :

Haut de page


Jason D. Hawkes

Institute of Culture and Society, Aarhus University

Stephanie Wynne-Jones

Department of Archaeology, University of York et Department of Archaeology and Ancient History, Uppsala University

Haut de page

Droits d’auteur


Le texte seul est utilisable sous licence CC BY-NC-ND 4.0. Les autres éléments (illustrations, fichiers annexes importés) sont « Tous droits réservés », sauf mention contraire.

Haut de page
Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search