Navigation – Plan du site
Connexions commerciales, circulations et appropriation des objets

Divergent patterns in Indian Ocean trade to East Africa and southern Africa between the 7th and 17th centuries CE: The glass bead evidence

Modèles divergents dans le commerce de l'océan Indien avec l'Afrique orientale et l'Afrique australe entre le VIIe et le XVIIe siècles : le témoignage des perles de verre
Marilee Wood

Résumés

Les perles de verre forment une grande partie des éléments archéologiques attestant un  commerce de l'océan Indien avec l'Afrique de l'Est entre le viie et le xviie siècles de notre ère, mais les archéologues les ont généralement sous-utilisées dans leurs efforts pour comprendre et interpréter ce commerce. Au cours de la dernière décennie, la recherche a distingué sept types de perles ayant eu cours en Afrique australe pendant ce laps de temps, et l'analyse chimique a ouvert des perspectives sur les origines des verres utilisés pour les fabriquer. Associés, ces deux types de données aident à reconstituer l'évolution des modèles de relations commerciales explorées ici. En outre, les différences entre les assemblages de perles découverts dans l'intérieur et aux extrémités nord et sud de la côte montrent que ces deux régions semblent avoir connu des relations commerciales différentes. Les perles suggèrent que le commerce avec le Sud n'a jamais été totalement contrôlé par le Nord, et qu'il en a parfois été complètement distinct – du moins en ce qui concerne le commerce de perles.

Haut de page

Texte intégral

Introduction

  • 1  In this discussion, ‘eastern Africa’ denotes the coastal areas from southern Somalia to southern M (...)

1Studies of trade in the Indian Ocean between the 7th and 17th centuries have paid relatively little attention to the far western edge of that great sea. Yet the eastern African coast1 was an active and important participant in that trade, supplying raw materials that were sought after and traded as far away as China. This paper will explore that trade and how it differed between the northern and southern ends of the coast and the interior, based on glass beads from archaeological excavations.

2Glass beads are important in the study of long-distance trade in eastern Africa because they form a significant portion of its surviving evidence. In addition, the technical difficulties involved in glassmaking, along with lack of any evidence of its production in the region, make it reasonable to assume that glass artifacts represent foreign trade. The past few decades have seen technological advances that have encouraged a surge in glass studies aimed largely at sourcing the origins of the raw glass. This raw glass, as well as recycled scrap glass, was widely traded, so beads were often made at venues different from the glass itself, sometimes even on different continents. Thus, the methods by which the beads were made also need to be taken into account in attempting to determine their region of production.

Figure 1: Indian Ocean and sites mentioned in text

Figure 1: Indian Ocean and sites mentioned in text

Marilee Wood

Glass types in this study

3A few terms used in glass bead studies should first be explained, along with a brief discussion of glass analysis as it pertains to this work. The glass beads in this study were made with one of two processes: drawing or winding. Drawn beads are made by drawing (pulling) hollow gathers of glass into tubes. These tubes are cut or chopped into bead-length segments. Although they may be left as is—with untreated sharp ends—it is more common for the ends to be smoothed or rounded through reheating. The longer reheating is carried out, the more rounded the beads become.

  • 2  For a full discussion of beadmaking and classification, see M. Wood, 2011.

4Wound beads are made by winding molten glass onto a mandrel (usually an iron rod or wire) until the desired bead volume is reached. While still viscid, the beads can be shaped by marvering or manipulating with various tools; they can also be decorated with glass of different colors.2

  • 3  For a full discussion, see P. Robertshaw, M. Wood, E. Melchiorre et al., 2010 and L. Dussubieux, B (...)

5Apart from a few beads made of lead-silica glass (discussed below), all those found in eastern and southern Africa are made of soda-lime-silica glasses. Such glasses are made by melting either sand or crushed quartz. To accomplish this at temperatures achievable in antiquity, a flux was used to lower the melting point of the silica. The alkalis used as fluxes were derived either from mineral-soda or plant-ash (which was obtained by burning halophytic [salt tolerant] plants). Five basic soda-lime-silica glasses have been identified in the glass beads from this study: three are plant-ash glasses and two are mineral-soda.3 As Table 1 shows, mineral-soda and plant-ash glasses can usually be separated based on levels of magnesia, while levels of alumina and uranium are useful in separating Middle Eastern glasses from ones originating in South or Southeast Asia (most of these basic glass groups have sub-types as well, which accounts for five types being discussed here).

Table 1: Oxides useful in determining origins of soda-lime-silica glasses

Oxide

Mineral-soda glasses
MgO <1.5%

Plant-ash glasses
MgO c. 2.5-6%

Middle East

S & SE Asia

Middle East

K2O

<1.5%

>2%

2-5%

CaO

>5%

<5%

>4.5%

Al2O3

<4%

>5%

<4%

U3O8

>5ppm

<1.5ppm

  • 4  M. Wood, 2012, p. 35.
  • 5  P. Wheatley, 1975, p. 90; E. Dreyer, 2007, p. 83-93; and A. Li, 2012, p. 41-43.

6The rare lead-silica beads in this study were all found at East Coast sites, apart from a single example from Great Zimbabwe (see Figure 1 for a map of sites mentioned in the text). They are small wound beads of transparent-to-translucent glass. Most are ruby-red, but a few amber colored and one emerald green specimen have been recovered. Thirty-three were found at Kilwa; all other sites combined account for another 16. Based on the chemical composition of the glass and method of manufacture, these beads would have been made in China.4 They all come from sites that encompass the early 15th century, the period in which the fleets under the command of the Chinese admiral, Zheng He, are believed to have visited Africa’s East Coast (1417–1419 and 1421–1422)5. Although it is possible these beads reached East Africa indirectly, comingled with other trade goods from China such as ceramics, this seems unlikely given the very limited numbers that have been found in East Africa and the circumscribed time span in which they appear. If they were arriving along with other more usual trade goods, one might expect to find them in larger quantities and over a longer period of time. In addition, the Chinese beads were individually produced, making it unlikely they would have been able to compete with the mass-produced drawn beads from India, which were the stock-in-trade beads at this time.

The southern African bead series

  • 6  M. Wood, 2000, 2005 and 2012; P. Robertshaw, M. Wood, E. Melchiorreet al., 2010; and L. Dussubieux(...)

7Based on a combination of morphological characteristics and chemical composition, seven bead series have been identified for southern Africa (Figure 2).6

Figure 2: Southern African bead series

Figure 2: Southern African bead series

Marilee Wood

Chibuene series

  • 7  The term cylindrical is used to describe a shape between oblate (a bead with a fully rounded profi (...)
  • 8  M. Wood, L. Dussubieux, P. Robertshaw, 2012.
  • 9 Ibid., p. 71.
  • 10  E. Wilmsen, J. Denbow, 2010.
  • 11  J. Denbow, C. Klehm, L. Dussubieux, 2015.
  • 12  A. Daggett, M. Wood, L. Dussubieux, in review.

8The Chibuene series is characterized by small (2.5–3.5 mm diameter) drawn beads with slightly rounded ends. The translucent glass often has a satiny sheen and is softly colored in hues of gray that can be tinged with pale blue, yellow, green, or blue-green. Most are tubular to cylindrical in shape.7 This series has only recently been recognized8 at the port of Chibuene in southern Mozambique in deposits that probably place it sometime in the 7th to 8th centuries. The glass used to make the Chibuene series beads is a plant-ash type (v-Na-Ca 3) that most likely was produced in the Middle East, possibly the Persian Gulf region.9 Subsequent to this discovery, a number of beads made of this glass type have been identified in Botswana: a few from Nqoma in the Tsodilo Hills,10 which is 1,800 km inland from the coast, and several from both Kaitshàa on the Makgadikgadi Pans11 and from two small sites on the southern edge of the Sowa Pan.12 Morphologically, these beads from Botswana are more like Zhizo series beads (described next and which were also found at these sites) than those from the Chibuene site. These beads demonstrate that Indian Ocean trade reached far into the interior of southern Africa from an early date.

Zhizo series

  • 13  P. Robertshaw, M. Wood, E. Melchiorre et al., 2010 and M. Wood, L. Dussubieux, P. Robertshaw, 2012
  • 14  M. Wood, 2005, p. 41 and M. Wood, 2011, p. 72.

9Zhizo series beads are characterized by drawn tubes whose ends are normally not rounded by reheating (although a few rounded examples do occur). Many can be identified by striations that run parallel to the perforation (caused by rows of bubbles probably produced by working the glass at low temperatures). Transparent-to-translucent cobalt blue is the most common color, followed by opaque yellow. Opaque blue-green and green examples are relatively rare. These beads are prone to corrosion, which forms a thick whitish crust, so they are often misidentified as white. Zhizo beads are made of a plant-ash glass (v-Na-Ca 1) that is similar to that of the Chibuene series and probably came from the same region.13 They are found at many sites in the interior of southern Africa, including Zimbabwe, Botswana, and South Africa,14 as well as at the coast in southern Mozambique, beginning in the 8th century and—apart from the Chibuene series—are the only bead type present in the interior up to sometime in the second half of the 10th century.

K2 Indo-Pacific series

  • 15  L. Dussubieux, B. Gratuze, M. Blet-Lemarquand, 2010 and P. Robertshaw, M. Wood, E. Melchiorre et a (...)

10Shortly after the disappearance of the Zhizo series, the K2 Indo-Pacific series appeared. This series is characterized by small transparent-to-translucent turquoise and blue-green drawn tubes and cylinders with heat rounded ends. A few soft green examples are also found. They are made of a mineral-soda-alumina glass (m-Na-Al 2) that is very stable and often shiny, no doubt due to high levels of alumina. This type of glass was made in South Asia.15

East Coast Indo-Pacific series

  • 16 Ibid.

11A few decades after the arrival of the K2 Indo-Pacific series, a closely related group of beads, the East Coast Indo-Pacific series, appeared. It is made up of heat-rounded small to medium (2.5–4.5 mm diameter) drawn beads that range in shape from tubular to cylindrical. Colors include opaque yellow, green, brick-red, and black—there are no cobalt blue beads. Chemically they are closely related to the K2 Indo-Pacific beads and have similar origins.16

Mapungubwe Oblate series

  • 17  P. Robertshaw, M. Wood, E. Melchiorreet al., 2010.
  • 18 Ibid.

12Both of the Indo-Pacific bead series disappeared from circulation in the first to second quarter of the 13th century and were replaced by the Mapungubwe Oblate series. These beads are distinctly different, being characterized by small, uniform oblates (although larger cylindrical forms are found as well, especially in outlying commoner areas). Most of the glass is opaque, with black being the most frequent color; yellow, dull orange, blue-green, and green are common as well. A few transparent cobalt blue and plum-colored specimens are also found. The glass used to make these beads is unusual, being a plant-ash glass with elevated levels of alumina and low lime (v-Na-Al).17 The region where this unusual glass may have been made is uncertain.18

Zimbabwe series

  • 19 Ibid.

13Toward the turn of the 14th century, a closely related series, the Zimbabwe series, displaced the Mapungubwe Oblate series. The new beads are more cylindrical in shape and some colors have changed: one now finds a transparent emerald green as well as a translucent pale jade green, yellow is more translucent, cobalt blue less transparent and more common, and opaque brick-red beads, which are not found in the Mapungubwe Oblate series, occur. The chemical composition of this glass (v-Na-Al)19 indicates that it is closely related to the Mapungubwe Oblate series and surely originated from the same region.

Khami Indo-Pacific series

  • 20  M. Wood, 2009.
  • 21  P. Robertshaw, M. Wood, E. Melchiorre et al., 2010.
  • 22  M. Wood, L. Dussubieux, L. Wadley, 2009.

14Sometime in the first half of the 15th century, the Khami Indo-Pacific series began to appear in southern Africa20 and soon replaced the Zimbabwe series. Khami Indo-Pacific beads are mostly cylindrical and less uniform in size and shape than any previous series except the Zhizo series. More than half fall into the medium size category (3.5–4.5 mm diameter). Colors include opaque brick-red, yellow, orange, green, blue-green, and black. Cobalt blue beads occur as well, but they are opaque and paler in color than earlier cobalt beads from the Mapungubwe Oblate and Zimbabwe series. Also, the first beads that might be considered white are found, but they are off-white and somewhat translucent, so they cannot be confused with the opaque white beads that eventually arrived from Europe. Khami series beads signal a return to a mineral-soda-alumina type glass (m-Na-Al 2)21 and, thanks to Portuguese records (as well as glass chemistry), we know they were being produced in India.22

Glass bead types in East Africa

15The glass bead picture in East Africa is far more complicated than that in southern Africa. During any time period one finds beads from various origins rather than from a single one as in southern Africa. As part of this diversity, East Africa received not only the simple drawn beads as found in the south but also larger wound beads of various shapes, as well as polychrome decorated beads, both drawn and wound. In addition, the temporal boundaries in which bead types are found in East Africa are less clear, making it impossible, at least at this time, to develop a temporal sequence as has been done in the south. Because they cannot be seriated, the beads found in East Africa will be described and discussed in the section below based on the time period in which they occur (See Figure 3 for examples of these beads).

Figure 3: Comparison of southern African and East African bead types by period

Figure 3: Comparison of southern African and East African bead types by period

Marilee Wood

A glass bead perspective on Indian Ocean trade with Africa’s eastern seaboard

  • 23  See F. Chami, F. Le Guennec-Coppens, S. Mery, 2002 for a full discussion of early evidence of cont (...)
  • 24  See for example A. Crowther, M. Horton, A. Kotarba-Morley et al., 2014.

16Indian Ocean trade to the African coast was enabled by the monsoon system of winds and currents, but they function only north of the Equator. Other less well-known currents and winds occur in the south and make travel there more difficult; but it is probable that southerly routes, not involving the northern monsoon system, were in use as well. Although documentary evidence of Indian Ocean trade with East Africa early in the first millennium CE is found in the Periplus Maris Erythraei and Ptolemy’s Geographia, physical evidence of that trade remains elusive,23 and claims for such evidence to date have been challenged24 and so will not be dealt with here. As will become clear, trade patterns to the northern and southern ends of the coast differed significantly, so they will be approached separately. The islands of the Comoros and Madagascar will also be discussed.

7th to mid-10th centuries

Southern Africa

  • 25  P.J.J. Sinclair, 1982 and 1987, p. 86-90; A. Ekblom, 2004; and P.J.J. Sinclair, A. Ekblom, M. Wood(...)
  • 26  M.C. Horton, 2004, p. 64-65; M. Wood, 2012, p. 47; and M. Wood, L. Dussubieux, P. Robertshaw, 2012 (...)

17The site of Chibuene, on Vilanculos Bay in southern Mozambique, was the southernmost port on the African coast in this early period. It was excavated numerous times between 1977 and 200125 and provides a wealth of information about Indian Ocean trade to the region at this time. Over 2,800 glass beads recovered from Chibuene bear witness to the volume of trade that passed through this ancient port, and hundreds of shards from imported vessels (of both exotic ceramics and glass) provide evidence of foreign visitors and inhabitants who participated in this trade.26

  • 27  M. Wood, L. Dussubieux, P. Robertshaw, 2012.
  • 28 Ibid., Table 15.
  • 29  P. Robertshaw, M. Wood, E. Melchiorre et al., 2010.

18Both glass beads and glass shards from Chibuene have been chemically analyzed using laser ablation-inductively coupled plasma-mass spectrometry (LA-ICP-MS).27 As has been mentioned, an early, previously unknown bead series—the Chibuene series—was recognized and named after the site and may date from as early as the 7th century. But the vast majority of Chibuene’s beads belong to the Zhizo series. This series accounts for all but a few pre-mid-10th century beads that have been recovered from the southern African interior. Their presence, sometimes in large numbers, at over 25 sites28 testifies to the breadth and depth of the interior trading network and its connections with Chibuene. As has been discussed, both of these early bead series were probably produced in the Middle East, east of the Euphrates,29 and—as will be discussed below—were traded directly to the southern coast by merchants from Oman and the Persian Gulf region.

East Africa

  • 30  M.C. Horton et al., 1996.
  • 31  H. Morrison, 1984.
  • 32  H.N. Chittick, 1974.
  • 33  M.C. Horton et al., 1996, p. 13.
  • 34  M. Wood, 2002.
  • 35  A. LaViolette, J. Fleisher, 2009; J. Fleisher, 2010; and J. Fleisher, A. LaViolette, 2012.

19Until recently pre-second millennium Indian Ocean trade to East Africa, as viewed through the glass bead evidence, has been characterized by several observations. First, far fewer glass beads have been recovered from sites in East Africa than in southern Africa. For example, in pre-11th century contexts at Shanga, in the Lamu archipelago, Horton30 reports finding 33, while at Manda, also in the Lamu archipelago, 79 were recorded,31 and at Kilwa in southern Tanzania Chittick32 lists only 4. It is becoming obvious, however, that this disparity may be mostly due to excavation techniques. In the south, deposits at most sites have been sieved with mesh that is 3 mm or even smaller, while deposits at many of the East Coast sites have either not been sieved or were sieved with mesh measuring 5 mm or larger.33 Given that the bulk of glass beads traded at this time were smaller than 5 mm (some are as small as 1 mm in diameter), it is not surprising that few were recovered. Also, unlike the south, a wide variety of bead types have been recovered from East Coast sites, including wound and decorated beads, some of which probably came from Egypt and the eastern Mediterranean.34 Zhizo beads are rare in the north: to date, one has been identified from Shanga and 11 from Tumbe, a 7th to 10th century site on Pemba Island in northern Tanzania.35

  • 36 http://www.sealinksproject.com
  • 37  A. Juma, 2004, did extensive work at Unguja Ukuu, proposing it dated from ~ 500 to 900 CE with sub (...)
  • 38  L. Dussubieux, B. Gratuze, M. Blet-Lemarquand, 2010.
  • 39  L. Dussubieux, P. Robertshaw, M. Glascock, 2009, p. 159 and L. Dussubieux, B. Gratuze, M. Blet-Lem (...)
  • 40  M. Tampoe, 1989, p. 108-109 and J. Carswell, S. Deraniyagala, A. Graham, 2013.
  • 41  P. Francis, 2002, p. 228, note 21.
  • 42 Ibid.

20However, recent excavations by the Sealinks Project36 at Unguja Ukuu in southwest Zanzibar are filling in some of the gaps in our knowledge of first millennium CE trade to the East Coast.37 Large numbers of glass beads (over 1,500) that are securely dated to the first millennium have been recovered. In addition, the probable origin of the majority of these beads adds new insights into Indian Ocean trade patterns. It has been assumed that glass beads in this early period were coming from various sources, including India, the Middle East, and the Persian Gulf region, but the bulk of the Zanzibar beads appears to have come from Sri Lanka. The majority of the beads analyzed from the 2011 and 2012 excavations are made from a mineral-soda-alumina glass (m-Na-Al 1). This glass is similar to the m-Na-Al 2 glass from which the three Indo-Pacific bead series described earlier are made, but LA-ICP-MS analysis has revealed that the two glass types can be separated based on levels of several trace elements, especially uranium and barium.38 The m-Na-Al 1 glass (with more barium and less uranium) was made in Sri Lanka (and possibly south India) between the 5th century BCE and the 10th century CE.39 The chemical composition of the glass is not the only indication that Sri Lanka may be the source of these beads. Several rather bright pumpkin-orange drawn beads made of m-Na-Al 1 glass were found at Unguja Ukuu—I believe this is the first time such beads have been found on the African coast. These beads are common at Mantai, a major entrepôt and beadmaking center in northwestern Sri Lanka,40 and are believed to have been made there.41 They are also present at the Red Sea port of Berenike in 4th to 6th century contexts.42 Thus, it raises the possibility that a trading circuit that operated between Sri Lanka and the Red Sea also included Unguja Ukuu in this early period.

  • 43  M. Wood, L. Dussubieux, P. Robertshaw, 2012.
  • 44  B. Chaisuwan, 2011.
  • 45  P. Francis, 2002, p. 44, p.  97-98 and 2013, p. 357.
  • 46  An article detailing evidence of this and other information about the Unguja Ukuu glass bead assem (...)
  • 47  P. Francis, 1993, p. 14.
  • 48  J. Lankton, L. Dussubieux, 2006, p. 135 and 2013, p. 431.
  • 49  B. Bronson, C. Ho, 1991, p. 18.

21Most of the remaining glass beads at Unguja Ukuu are made of a plant-ash glass (v-Na-Ca). Three types of this glass have been identified on the eastern African coast;43 the type found at Unguja Ukuu is called v-Na-Ca 1 and is believed to have been made in the Middle East, east of the Euphrates. Zhizo series beads are made of this glass, but most of the v-Na-Ca 1 beads at Unguja Ukuu are morphologically unlike Zhizo beads, both in terms of some glass colors and details of production technique. For example, in addition to the colors found in the Zhizo series (cobalt blue, yellow, blue-green, and green), the Zanzibar beads appear in brick-red, white, and black. While all of the Zhizo beads are drawn, some v-Na-Ca 1 beads from Unguja Ukuu are wound and even decorated. And while the ends of Zhizo beads are left untreated, almost all of the drawn Zanzibar ones have been reheated in an unusual manner: they have been reheated on a flat surface, resulting in beads with one somewhat rounded end while the other is flat. The Unguja Ukuu examples are similar to beads found at the site of a 9th to 11th-century trading and beadmaking center on the southwest coast of Thailand, known as Thung Tuk44 [but sometimes identified as Takua-Pa]. Francis45 identified several bead types that might have been made at Thung Tuk which are similar to ones at Unguja Ukuu, including a type of eye bead (known as a Takua Pa eye bead) and cobalt blue beads with white stripes.46 He also noted that Thung Tuk “did not heat their beads very long, and most have a squared rather than rounded profiles”47—a description that fits beads reheated on a flat surface. LA-ICP-MS analysis of these bead types from Unguja Ukuu shows that they are all made of v-Na-Ca 1 glass. Lankton and Dussubieux48 record that both beads and raw glass of this type are found at many South and Southeast Asian sites and note that most of the beads are drawn, suggesting they were locally made and that much of the glass was imported in raw or cullet form to be used in beadmaking (or sometimes for making bangles). Thung Tuk, just across the Bay of Bengal from Mantai, was particularly active in the 9th century in conjunction with the port of Laem Pho on the Gulf of Thailand side of the Kra Isthmus opposite Thung Tuk.49 The two ports acted as transfer stations where goods from east and west were transported overland, thus avoiding the lengthy journey around the Malay Peninsula and the pirate-infested waters of the Straits of Malacca. Thus, if some of the v-Na-Ca 1 beads at Unguja Ukuu were made at Thung Tuk, they would probably have arrived in Zanzibar via Mantai along with m-Na-Al 1 beads made there. Although it is obviously possible that these beads were made elsewhere, the similarities between v-Na-Ca 1 beads at Unguja Ukuu and Thung Tuk are thought-provoking and suggest further research into the parallels could be worthwhile.

The Comoros and Madagascar

  • 50  H. Wright, 1984.

22Although substantial numbers of imported ceramics have been recovered from pre-11th-century sites in the Comoros, glass beads are rare and in total number fewer than 20.50 This suggests that the inhabitants either did not have much access to these beads or did not favor them. In either case, it is unlikely these islands acted as a transit point in the trade that brought glass beads to southern Africa.

  • 51  C. Radimilahy, 1998.
  • 52  P. Robertshaw, B. Rasoarifetra, M. Wood et al., 2006, p. 106.
  • 53  H. Wright, F. Fanony, 1992, p. 32-33.
  • 54  P. Robertshaw, B. Rasoarifetra, M. Woodet al., 2006, p. 93.

23Few sites dating to this early period in Madagascar have produced glass beads. Although Radimilahy51 reported finding glass beads in Phase Ia (late-9th to 10th century) contexts at Mahilaka in northwest Madagascar, subsequent research has indicated that none of the beads were found to pre-date the end of the 10th century.52 In another case, the site of Sandrakatsy in northeast Madagascar53 produced 37 drawn yellow beads from an undated pit. Their chemical composition and associated non-glass beads indicate they probably date to the more recent end of the site’s proposed time span (~ 8th to 14th century).54

Politics and power in the western Indian Ocean

  • 55  M. Pearson, 2010.
  • 56  D. Whitehouse, A. Williamson, 1973, p. 48.
  • 57  G.S.P. Freeman-Grenville, 1962, p. 16.
  • 58  M. Tampoe, 1989, p. 113.

24From the 7th to mid-8th century, the Umayyid Caliphate ruled from their capital at Damascus. During this period, trade in the western Indian Ocean was dominated by the Persian Gulf region. When political power was transferred to Baghdad in the mid-8th century, with the ascendancy of the Abbasid Caliphate, trade increased significantly55 and thus could have played a role in the increase seen in trade to southern Africa in the form of Zhizo beads. From the early 9th century, commerce was driven by large-scale trade conducted between Persian Gulf region merchants and China,56 a trade that included goods from southern Africa, including large ivory tusks that al-Mas’udi recorded were in demand in China for numerous uses, including constructing luxury palanquins.57 With the rise of the Fatimid Caliphate and their move to Fustat (ancient Cairo) in 969, power in the Middle East shifted westward and western Indian Ocean trade became more focused on the Red Sea and eastern Mediterranean. The Persian Gulf region’s dominance over western Indian Ocean trade declined, although trade continued;58 and beads made of glass from that region (v-Na-Ca 1) disappeared from circulation.

  • 59  J.S. Trimingham, 1975, p. 122 and p. 135; H.N. Chittick, 1977, p. 192; G.F. Hourani, 1995, p. 148; (...)
  • 60  G.S.P. Freeman-Grenville, 1962, p. 15.
  • 61  J.S. Trimingham, 1975, p. 121.
  • 62  M. Tampoe, 1989, p. 102.

25In this phase of Indian Ocean trade with eastern Africa it is clear that voyages to the south as well as the north were made directly by ships and merchants from the Persian Gulf region. Al-Mas’udi, who visited Qanbalu (thought to be on Pemba Island)59 in about 915 AD on a return voyage from China and India aboard a Persian ship, recorded that Sofala (at this time the name refers not to a single port but more-or-less to the entire coast of modern-day Mozambique) was the “furthest limit of the land and the end of the voyages made from Oman and Siraf [a port in the Persian Gulf] on the sea of Zanj.”60 He stated that the ships of Oman and Siraf made the voyage to Sofala, indicating that they sailed directly from the Gulf to Qanbalu before heading south. Of great interest here, al-Mas’udi reported that both enormous ivory tusks and gold came from Sofala. This is the earliest evidence we have of the gold trade out of southern Africa. He also stated that Sofala was a commercial destination for ships from the Gulf region61 and that voyages to Sofala continually increased in number owing to an increase in the demand for luxury goods at the Buyid court in Shiraz.62

  • 63  P. Robertshaw, M. Wood, E. Melchiorre et al., 2010, p. 1903.
  • 64  J.C. Wilkinson, 1981, p. 282 and M. Tampoe, 1989, p. 44.
  • 65  M. Tampoe, 1989, p. 106.
  • 66  The likelihood that artisans moved around the Indian Ocean to ply their trades has been argued by (...)
  • 67  M.C. Horton, 2004, p. 68.
  • 68  F.A. Khan, 2005, p. 45.
  • 69 Ibid., p. 18.

26In the search for the origins of Zhizo beads, their chemistry suggests the glass used to make them came from the Middle East, east of the Euphrates,63 but the method used to manufacture them is an Indian technology. So where could these beads, which were most likely made by Indian artisans using glass from the Iran–Iraq region and were then carried to southern Africa on ships from the Persian Gulf region, have been manufactured? The history of the port of Sohar, in Oman near the mouth of the Gulf, may provide clues. Sohar was the dominant port in the region from the 8th century up to the ascendancy of Siraf in the mid-9th century. But Sohar continued to be an active participant in the trade up to the mid-10th century, when it was attacked by the Buyids.64 Thus, the period of Sohar’s prime fits the time span during which Zhizo beads were made and traded. Tampoe65 has noted that evidence of glassmaking was unearthed during excavations at Sohar, raising the possibility that Zhizo beads might have been made there; and if so, it is likely that South Asian artisans were there to produce them.66 Tampoe unfortunately did not provide references for this information, and all attempts by the author to locate further evidence about glassworking at Sohar have been fruitless to date. Another possible manufacturing site for Zhizo beads could be Banbhore at the mouth of the Indus. It is often suggested to be the ancient city of Daybul [or Debal];67 and indeed in the preliminary report on excavations at Banbhore, Khan states that “[t]he question of the identification of Banbhore with Debal has now been solved beyond and [sic] reasonable doubt.”68 The report also tantalizingly states that a mound was found in the industrial area “with a concentration of ashes, charcoals, and kiln slag and glass pieces [which] appears to indicate the site of an ancient glass factory.”69 Glass beads were also found at the site, but unfortunately it has not been possible to find any further details about glassmaking, beadmaking, or even images of the beads.

Mid-10th to mid-13th centuries

Southern Africa

  • 70  P. Robertshaw, M. Wood, E. Melchiorre et al., 2010; M. Wood, 2011.
  • 71  L. Dussubieux, B. Gratuze, M. Blet-Lemarquand, 2010, p. 1650.

27In the mid-10th century, dramatic changes took place in Indian Ocean trade to southern Africa. Chibuene lost its place as the major port of entry for the southern interior (its successor has not been located), and Zhizo series beads disappeared. They were replaced by two closely related series, the K2 Indo-Pacific series (which makes up 90% of most assemblages) and the East Coast Indo-Pacific series. Both are made of mineral-soda-alumina (m-Na-Al 2) glass.70 As has been noted, glasses of this type are believed to have originated in India,71 indicating that trade providing glass beads for the western Indian Ocean had shifted from the Persian Gulf region to South Asia. Concurrently, the volume of beads entering southern Africa increased dramatically.

East Africa

  • 72  A. Kanungo, 2004, p. 45 and p. 51.
  • 73  For discussions about southern routes, see G. Shepherd, 1982; M. Tampoe, 1989 p. 109, p. 121, and (...)

28Beginning in the 11th century, trade to East Africa also expanded dramatically, and glass beads arrived in large numbers across the region. Beads from the Middle East (but not necessarily from east of the Euphrates) still formed part of the trade, but most beads belong to the East Coast Indo-Pacific series. However, unlike the south—where only drawn beads are found—large numbers of wound beads (made of glass similar to that used to make the drawn beads) form part of assemblages. K2 Indo-Pacific beads are present in the north but only in small numbers. The differences in the distributions of the two Indo-Pacific series may indicate that the two ends of the coast were receiving beads from different regions of India via different trade routes—had they all come from the same source one would expect to find some wound beads in southern Africa. Kanungo72 has pointed out that in India wound beads are mainly found at archaeological sites in the north, while small drawn beads are found in the south and the Deccan. Thus, it is possible the East Coast was receiving beads shipped mainly from northern India, while those arriving in southern Africa may have come mostly from further south in India, possibly by a more southerly route.73

The Comoros and Madagascar74

  • 74  Because it is difficult to disentangle time frames here and few sites are involved, this section w (...)
  • 75  H. Wright, 1992, p. 105.

29At sites in the Comoros that date from the 11th to the 15th centuries, Wright’s excavations produced only nine glass beads.75 Based on his descriptions, it appears they all probably belong to the East Coast Indo-Pacific series.

  • 76  C. Radimilahy, 1998.
  • 77  P. Robertshaw, B. Rasoarifetra, M. Wood et al., 2006.

30In Madagascar the main site that has produced glass beads in this period is Mahilaka.76 About 2,300 beads were recovered from the site, of which 99% are glass. A detailed study,77 which included chemical analysis, of 180 of the beads concluded that the earliest beads to arrive at Mahilaka belong to the K2 Indo-Pacific series. Mapungubwe Oblate and Zimbabwe series beads are present but only in small numbers. The majority of the beads belong to the East Coast Indo-Pacific and Khami Indo-Pacific series, with wound beads making up 9% of the assemblage. Taken as a whole Mahilaka’s bead assemblage is more closely related to assemblages on the East Coast than to those in southern Africa, suggesting that northwestern Madagascar was articulated with northern trade circuits rather than those to southern Africa.

Mid-13th to late-15th centuries

Southern Africa

31Another major change in trading patterns occurred in the mid-13th century, at about the same time that the regional capital moved from K2 to Mapungubwe. The K2 Indo-Pacific and the East Coast Indo-Pacific series disappeared from circulation and were replaced by the Mapungubwe Oblate series. These beads are made of a plant-ash glass with elevated alumina and low lime (v-Na-Al), so the glass they are made of must have come from a different geological zone than the Indo-Pacific beads. In addition, they arrived in staggering numbers, suggesting the trade had increased yet again. Mapungubwe’s position of power was short-lived, however, and at the end of the century both the site and its hinterland were largely abandoned. Great Zimbabwe became the new regional capital and center of control over trade with the Indian Ocean. A change in the bead series being traded occurred at about the same time Mapungubwe was abandoned, but the chemical composition of this new series, the Zimbabwe series, differs only slightly from the Mapungubwe Oblate series, so a shift in trading patterns is not evident; it rather suggests a slight shift in manufacturing location. Enormous numbers of beads from excavations at Great Zimbabwe bear witness to continued or even elevated levels of trade with Indian Ocean commerce that continued into the mid-15th century. The origins of these two bead series are elusive. Although the Middle East is known for producing plant-ash glasses, the levels of alumina in these glasses are rather high for Middle Eastern glass; in addition, that region is not known for producing small drawn beads.

East Africa

  • 78  M. Horton et al., 1996, p. 417.
  • 79  R.L. Pouwels, 2002, p. 400.
  • 80  H.N. Chittick, 1974, p. 464.
  • 81  Qin Dashu of Peking University has reported finding early Ming imperial kiln shards of the type us (...)

32From about the mid-13th century, changes in trading patterns in East Africa become evident in the archaeological record; for example, Chinese ceramics increased in volume and Islamic ceramics came mostly from Arabia rather than the Gulf.78 This period saw the height of the gold trade out of southern Africa spurred on by increasing demand for gold in Europe and elsewhere. Trade to the south was segmented, and the Swahili town of Kilwa, in southern Tanzania, is often considered to have controlled the trans-shipment of goods between north and south, as well as trade that arrived directly from southern India via the Maldives.79 Bead assemblages on the East Coast once again differ from those in southern Africa. Although a range of bead types are found, East Coast Indo-Pacific series beads dominate, and Mapungubwe Oblate and Zimbabwe series beads are rare. It is during this period that one finds the unusual small wound beads made of lead-silica glass discussed above. The only large number (30) of them came from the Husuni Kubwa palace complex at Kilwa.80 Because of their rarity and limited distribution, it seems unlikely they were actual trade items; they, along with some ceramics,81 might possibly have been gifts from Admiral Zheng He’s Chinese fleet, which reportedly reached the African coast in the first quarter of the 15th century.

Late 15th to mid-17th centuries

Southern Africa

  • 82  M. Wood, 2009.
  • 83  G.M. Theal, 1898, II, p. 303; W.G.N. van der Sleen, 1956, p. 28 and 1958, p. 212; J.F. Schofield, (...)

33Another major change in patterns of Indian Ocean trade with southern Africa is evident with a shift from the Zimbabwe series to the Khami Indo-Pacific series in the mid-to-late 1400s.82 This shift signals a return to trade between South Asia and southern Africa. Although this trade was originally conducted by Indian and Swahili traders, it was eventually largely overtaken by the Portuguese. They initially attempted to introduce European beads into the market, but African consumers refused to accept them, thus forcing the Portuguese to purchase beads for the African trade in India.83

East Africa

34During this period, bead assemblages in East Africa are for the first time similar to those in southern Africa. Indo-Pacific beads continue to dominate East African assemblages. Wound beads occur less frequently than in earlier periods, but their continued presence suggests that East Africa and northern India maintained trade contacts that were absent or infrequent in the south, where wound beads are very rare.

Discussion and conclusion

  • 84  In this period at Kilwa, more than 50% were reportedly wound (H.N. Chittick, 1974); at Manda it wa (...)
  • 85  M. Wood, 2000, p. 81, 2005, p. 30, and 2011, p. 75.

35Close examination and comparison of glass bead assemblages from southern Africa and the East Coast bring new insights into Indian Ocean trade to the African coast. Up to the late 15th century, the beads indicate that Indian Ocean trade in glass beads to southern Africa differed from that to East Africa. In southern Africa, bead series appeared one at a time and, except for brief transition periods, only one series was present during any period. All but a very few beads were monochrome, drawn, and small-to-medium in diameter. In contrast, in all periods examined East African bead assemblages included beads from diverse sources, which at times included Sri Lanka, India, Egypt, the Persian Gulf, the Middle East, and possibly Southeast Asia. Many of them were wound, some were multicolored and decorated, and many were large. Although it has often been assumed that all trade to southern Africa passed through and was controlled by ports in East Africa, the bead evidence demonstrates this is unlikely to have been the case—rather because of the relative scarcity of southern African bead types in East Africa than the other way around. Between the 7th and mid-10th centuries in southern Africa, the trade that carried beads (along with other products) appears to have been conducted directly with merchants from the Persian Gulf region, while East Africa received beads from Sri Lanka (and perhaps south India), the Middle East, and possibly Thailand (via Sri Lanka). Tumbe on Pemba Island (the area believed to be the ancient site of Qanbalu, which was a stopping point for Gulf ships headed south) is the only northern site identified thus far to have even small numbers of the Zhizo beads that were ubiquitous in the south. From the mid-10th to the mid-13th centuries, East Africa became more active in Indian Ocean trade but received few of the K2 series beads that made up 90% of assemblages in the south, suggesting the two ends of the coast maintained somewhat different trading contacts. Because a sizeable proportion of wound beads, which were made in the north of India, are found on the East Coast84 but are absent in the south, it is possible southern Africa traded through southern routes with sources further south in India where wound beads were not produced. One cannot rule out local tastes in these trends, but it seems unlikely that no one in southern Africa would have accepted wound beads if they had been available. In fact, during this period wound beads (known as Garden Rollers) were being locally made in the Shashe-Limpopo area by melting and reworking the imported K2 series beads.85 Then, between the mid-13th and the late 15th centuries, the trade that brought beads to southern Africa diverged significantly from that in the north. Even during this period in which Kilwa was supposedly controlling the gold trade from the south, the near absence of Mapungubwe Oblate and Zimbabwe series beads in Kilwa’s assemblages suggests the beads were traded by a different route (although excavation techniques could play some role in their absence). It is perhaps possible that the main commodity involved in Kilwa’s trade to the south was cloth and that the beads were part of another—as yet unidentified—trading circuit. The resolution of this issue lies with finding the source of the Mapungubwe Oblate and Zimbabwe series beads. Finally, in the late 15th to mid-17th centuries, the bead assemblages at both ends of the coast converged; the arrival of European traders in the Indian Ocean contributed to this as they attempted to control trade on the entire eastern African coast, which they supplied, reluctantly, with beads from India.

36The bead series developed for southern Africa can tell us a great deal about trade connections between that region, the East Coast, and the greater Indian Ocean, including recreating trade routes and identifying possible participants involved in the trade. This study has demonstrated that southern Africa was an active and independent participant in Indian Ocean trade from the second half of the first millennium AD and that the East Coast was not in control of all trade to the south as has often been assumed. From the perspective of imported glass beads, the East Coast’s trading contacts included Sri Lanka, the Persian Gulf region, India (particularly the northwest), Egypt, the eastern Mediterranean, and possibly Thailand, whereas the south shows almost no links to trade north of the mouth of the Red Sea but had strong links to the Persian Gulf before the second millennium and after that to India (particularly the southern regions).

  • 86  M. Horton, 2004, p. 64.
  • 87 Ibid., p. 65.

37Finally it might be useful to note that the distribution of glass beads is largely unrelated to the distribution of exotic ceramics in eastern Africa. Ceramics are found almost exclusively in coastal venues. They probably first arrived as personal possessions of Indian Ocean traders who lived in communities for extended periods while waiting for favorable winds for their onward voyages. As Horton86 has suggested, they may have become part of gift exchanges to cement relations between traders and eventually become socially embedded commodities with symbolic meanings about the prestige and status of their owners. They thus “are not direct indicators of trade but denote membership in a common Indian Ocean culture.”87 Although it may be true, as many have suggested, that people in the interior did not desire imported ceramics, it also seems probable that they would not have been available to interior trade. The glass beads, on the other hand, were true trade objects that were an important constituent of the goods sent into the interior in exchange for products there. Although far more beads are found in southern Africa than in the East African interior, it is likely that this discrepancy will close at least somewhat as more archaeologists sieve deposits with fine mesh.

Haut de page

Bibliographie

Axelson, E., 1973, The Portuguese in South-East Africa 1488–1600, Johannesburg, Struik.

Beaujard, P., 2011, “The first migrants to Madagascar and their introduction of plants: Linguistic and ethnological evidence”, Azania, 46 (2), p. 169-189.

Bronson, B., Ho, C., 1991, “A Silk Road in southern Thailand”, In the Field: The Bulletin of the Field Museum of Natural History, September/October: 1 & 18. URL: https://archive.org/stream/field62D#page/18/mode/2up.

Carswell, J., Deraniyagala, S., Graham, A., 2013, Mantai: City by the sea, Aichwald, Linden Soft Verlag.

Chaisuwan, B., 2011, “Early contacts between India and the Andaman Coast in Thailand from the second century BCE to eleventh century CE”, in P.-Y. Manguin, A. Mani, G. Wade (eds), Early interactions between South and Southeast Asia: Reflections on cross-cultural exchange, Singapore, ISEAS Publishing, p. 83-112.

Chami, F., Le Guennec-Coppens, F., Mery, S., 2002, “East Africa and the Middle East relationship from the first millennium BC to about 1500 AD”, Journal des Africanistes, 72 (2), p. 21-37.

Chittick, H.N., 1974, Kilwa. An Islamic trading city on the East African coast, 2 vols, Nairobi, British Institute in East Africa Memoir, n° 5.

Chittick, H.N., 1977, “The East Coast, Madagascar and the Indian Ocean”, in J.D. Fage, R. Oliver (eds), The Cambridge history of Africa, 3, Cambridge, Cambridge University Press, p. 183-232.

Crowther, A., Horton, M., Kotarba-Morley, A., Prendergast, M., Quintana Morales, E., Wood, M., Shipton, C., Fuller, D.Q., Tibesasa, R., Mills, W., Boivin, N., 2014, “Iron Age agriculture, fishing and trade in the Mafia Archipelago, Tanzania: New evidence from Ukunju Cave”, Azania, 49 (1), p. 21-44.

Daggett, A., Wood, M., Dussubieux, L., (in review), “Glass trade beads at Thaba Di Masego, Botswana: Analytical results and some implications”.

Denbow, J., Klehm, C., Dussubieux, L., in press, “The glass beads of Kaitshàa: New insights on early Indian Ocean trade into the far interior of southern Africa”, Antiquity.

Dewar, R.E., 1988, “The earliest colonization of Madagascar”, Colloquium on the Indian Ocean in Antiquity at the British Museum.

Dreyer, E., 2007, Zheng He: China and the oceans in the early Ming Dynasty, 1403–1433, New York, Pearson/Longman.

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, p. 797-821.

Dussubieux, L., Robertshaw, P., Glascock, M., 2009, “LA-ICP-MS analysis of African glass beads: Laboratory inter-comparison with an emphasis on the impact of corrosion on data interpretation”, International Journal of Mass Spectrometry, 284, p. 152-161.

Dussubieux, L., Gratuze, B., Blet-Lemarquand, M., 2010, “Mineral soda alumina glass: Occurrence and meaning”, Journal of Archeological Science, 37, p. 1646-1655.

Ekblom, A., 2004, “Changing landscapes: An environmental history of Chibuene, Southern Mozambique”, Global Studies in Archaeology 5, Uppsala, Department of Archaeology and Ancient History, Uppsala University.

Fleisher, J., 2010, “Swahili synoecism: Rural settlements and town formation on the central East African coast, AD 750–1500”, Journal of Field Archaeology, 35 (3), p. 265-282.

Fleisher, J., LaViolette, A., 2013, “The early Swahili trade village of Tumbe, Pemba Island, Tanzania, AD 600–950”, Antiquity, 87 (338), p. 1151-1168.

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

Francis, P. Jr., 1993, Advanced bead identification, Lake Placid, NY, Lapis Route Books.

Freeman-Grenville, G.S.P., 1962, The East African Coast: Selected documents from the first to the early 19th century, Oxford, Clarendon Press.

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

Horton, M.C., Brown, H.W., Mudida, N., 1996, Shanga. The archaeology of a Muslim trading community on the coast of East Africa, London, British Institute in Eastern Africa Memoir, n° 14.

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

Hourani, G.F., 1995, Arab seafaring, expanded edition, Princeton, NJ, Princeton University Press.

Juma, A., 2004, Unguja Ukuu on Zanzibar. An archaeological study of early urbanism, Doctoral thesis, Uppsala University. URL: http://uu.diva-portal.org/smash/record.jsf?pid=diva2%3A164832&dswid=2470.

Kanungo, A., 2004, “A database of glass and glass beads in India”, Man and Environment, 29 (1), p. 42-102.

Khan, F.A., 2005, Banbhore: A preliminary report on the recent archaeological excavations at Banbhore, fifth edition, Pakistan, Department of Archaeology and Museums.

Lankton, J., Dussubieux, L., 2006, “Early glass in Asian maritime trade: A review and an interpretation of compositional analyses”, Journal of Glass Studies, 48, p. 121-144.

Lankton, J., Dussubieux, L., 2013, “Early glass in Southeast Asia”, in K. Janssens (ed.), Modern Methods for Analyzing Archaeological and Historical Glass, New York, John Wiley & Sons, p. 415-443.

Lankton, J., Dussubieux, L., Rehren, T., 2008, “A study of mid-first millennium CE Southeast Asian specialized glass beadmaking traditions”, in E.A. Bacus, I.C. Glover, P.D. Sharrock (eds), Interpreting Southeast Asia’s past: Monument, image and text, Singapore, NUS Press, p. 335-356.

LaViolette, A., Fleisher, J., 2009, “The urban history of a rural place: Swahili archaeology on Pemba Island, Tanzania, 700–1500 AD”, International Journal of African Historical Studies, 42 (3), p. 433-455.

Li, A., 2012, A history of overseas Chinese in Africa to 1911, New York, Diasporic Africa Press.

Morrison, H., 1984, “The beads”, in H.N. Chittick (ed.), Manda. Excavations at an island port on the Kenyan Coast, Nairobi, British Institute in Eastern Africa Memoir, 9, p. 181-189.

Patel, A., 2004, “Communities and commodities: Western India and the Indian Ocean, eleventh–fifteenth centuries”, Ars Orientalis, 34, p. 7-18.

Pearson, M., 2010, “Islamic trade, shipping, port-states and merchant communities in the Indian Ocean, seventh to sixteenth centuries”, in D.O. Morgan, A. Reid (eds), The new Cambridge history of Islam, vol. 3, The eastern Islamic world, eleventh to eighteenth centuries, Cambridge, Cambridge University Press, p.317-65.

Pouwells, R.L., 2002, “Eastern Africa and the Indian Ocean to 1800: Reviewing relations in historical perspective”, The International Journal of African Historical Studies, 35 (2/3), p. 385-425.

Radimilahy, C., 1998, Mahilaka: An archaeological investigation of an early town in northwestern Madagascar, Studies in African Archaeology 15, Uppsala, Department of Archaeology and Ancient History.

Ray, H.P., 2004, “The beginnings: The artisan and the merchant in early Gujarat, sixth–eleventh centuries”, Ars Orientalis, 34, p. 39-61.

Robertshaw, P., Rasoarifetra, B., Wood, M., Melchiorre, E., Popelka-Filcoff, R., Glascock, M., 2006, “Chemical analysis of glass beads from Madagascar”, Journal of African Archaeology, 4 (1), p. 91-109.

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

Robinson, K.R., 1959, Khami Ruins: Report on excavations undertaken for the Commission for the Preservation of Natural and Historical Monuments and Relics, Southern Rhodesia 1947–1955, Cambridge, Cambridge University Press.

Schofield, J.F., 1958, “South African beads”, in R. Summers (ed.), Inyanga. Prehistoric Settlements in Southern Rhodesia, Cambridge, Cambridge University Press, p. 180-229.

Shepherd, G., 1982, “The making of the Swahili”, Paideuma, 28, p. 129-47.

Sinclair, P.J.J., 1982, “Chibuene: An early trading site in southern Mozambique”, Paideuma, 28, p. 150-64.

Sinclair, P.J.J., 1987, Space, time and social formation: A territorial approach to the archaeology and anthropology of Zimbabwe and Mozambique c. 0–1700 AD, Uppsala, Societas Archaeologica Upsaliensis, Aun 9.

Sinclair, P.J.J., 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, p. 723-737.

Tampoe, M., 1989, Maritime trade between China and the West: An archaeological study of the ceramics from Siraf (Persian Gulf), 8th to 15th centuries AD, Based on DPhil thesis, Oxford, BAR International Series 555.

Theal, G.M., 1898, Records of South-Eastern Africa, vols. 1 & 2, Cape Town, Government of the Cape Colony.

Trimingham, J. S., 1975, “The Arab geographers and the East African coast”, in H.N. Chittick, R.I. Rotberg (eds), East Africa and the Orient: Cultural synthesis in pre-colonial times, New York & London, Africana Publishing Company, p. 115-146.

Van der Sleen, W.G.N., 1956, “Trade-wind beads”, Man, 56, p. 2-29.

Wheatley, P., 1975, “Analecta Sino-Africana recensa”, in H.N. Chittick, R.I. Rotberg (eds), East Africa and the Orient: Cultural synthesis in pre-colonial times, New York & London, Africana Publishing Company, p. 76-114.

Whitehouse, D., Williamson, A., 1973, “Sasanian maritime trade”, Iran, 11, p. 29-49.

Wilkinson, J.C., 1981, “Oman and East Africa: New light on early Kilwan history from the Omani sources”, The International Journal of African Historical Studies, 14 (2), p. 272-305.

Wilmsen, E., Denbow, J., 2010, “Early villages at Tsodilo: The introduction of livestock, crops, and metalworking”, in L. Robbins, A. Campbell, M. Taylor (eds), Tsodilo, Mountain of the Gods, East Lansing, Michigan State University Press, p. 72-81.

Wood, M., 2000, “Making connections: Relationships between international trade and glass beads from the Shashe-Limpopo area”, South African Archaeological Society Goodwin Series, 8, p. 78-90.

Wood, M., 2002, “The glass beads of Kaole”, in F. Chami, G. Pwiti, C. Radimilahy (eds), Southern Africa and the Swahili World. Studies in the African Past 2, Dar-es-Salaam, Dar-es-Salaam University Press, p. 50-65.

Wood, M., 2005, Glass beads and pre-European trade in the Shashe-Limpopo region, MA thesis, Faculty of Humanities, Johannesburg, University of the Witwatersrand.

Wood, M., 2009, “The glass beads from Hlamba Mlonga, Zimbabwe: Classification, context and interpretation”, Journal of African Archaeology, 7 (2), p. 219-238.

Wood, M., 2011, “A glass bead sequence for southern Africa from the 8th to the 16th century AD”, Journal of African Archaeology, 9 (1), p. 67-84.

Wood, M., 2012, Interconnections: Glass beads and trade in southern and eastern Africa and the Indian Ocean—7th to 16th centuries AD, Studies in Global Archaeology 17, Uppsala, Department of Archaeology and Ancient History, Uppsala University.

Wood, M., Dussubieux, L., Wadley, L., 2009, “A cache of ~5000 glass beads from the Sibudu Cave Iron Age occupation”, South African Humanities, 21, p. 239-261.

Wood, M., Dussubieux, L., Robertshaw, P., 2012, “Glass finds from Chibuene, a 6th to 17th century AD port in southern Mozambique”, South African Archaeological Bulletin, 67 (195), p. 59-74.

Wright, H.T., 1984, “Early seafarers of the Comoro Islands: The Dembeni Phase of the IXth–Xth centuries AD”, Azania, 19, p. 13-59.

Wright, H.T., 1992, “Nzwani and the Comoros, XIth–XVth centuries”, Azania, 27, p. 81-128.

Wright, H.T., Fanony, F., 1992, “L’évolution des systèmes d’occupation des sols dans la vallée de la rivière Mananara au nord-est de Madagascar”, Taloha, 11, p. 16-64.

Haut de page

Notes

1  In this discussion, ‘eastern Africa’ denotes the coastal areas from southern Somalia to southern Mozambique (including offshore islands). ‘East Africa’ and ‘East Coast’ refer to Kenya and Tanzania and their off-shore islands, whereas ‘southern Africa’ refers to the interior including the Zimbabwe Plateau / Shashe-Limpopo Basin and Botswana as far west as the Tsodilo Hills. Coastal southern Africa is represented only in the early period (up to the late 10th century) because, unfortunately, the only known port site in the periods under discussion is Chibuene in southern Mozambique.

2  For a full discussion of beadmaking and classification, see M. Wood, 2011.

3  For a full discussion, see P. Robertshaw, M. Wood, E. Melchiorre et al., 2010 and L. Dussubieux, B. Gratuze, M. Blet-Lemarquand, 2010.

4  M. Wood, 2012, p. 35.

5  P. Wheatley, 1975, p. 90; E. Dreyer, 2007, p. 83-93; and A. Li, 2012, p. 41-43.

6  M. Wood, 2000, 2005 and 2012; P. Robertshaw, M. Wood, E. Melchiorreet al., 2010; and L. Dussubieux, B. Gratuze, M. Blet-Lemarquand, 2010.

7  The term cylindrical is used to describe a shape between oblate (a bead with a fully rounded profile and a length smaller than diameter) and tubular (a straight-sided bead with flat or slightly rounded ends). In other words, cylindrical beads have ends that are rounded but small sections of the body may be straight. Most poorly shaped drawn beads fall into this category.

8  M. Wood, L. Dussubieux, P. Robertshaw, 2012.

9 Ibid., p. 71.

10  E. Wilmsen, J. Denbow, 2010.

11  J. Denbow, C. Klehm, L. Dussubieux, 2015.

12  A. Daggett, M. Wood, L. Dussubieux, in review.

13  P. Robertshaw, M. Wood, E. Melchiorre et al., 2010 and M. Wood, L. Dussubieux, P. Robertshaw, 2012.

14  M. Wood, 2005, p. 41 and M. Wood, 2011, p. 72.

15  L. Dussubieux, B. Gratuze, M. Blet-Lemarquand, 2010 and P. Robertshaw, M. Wood, E. Melchiorre et al., 2010.

16 Ibid.

17  P. Robertshaw, M. Wood, E. Melchiorreet al., 2010.

18 Ibid.

19 Ibid.

20  M. Wood, 2009.

21  P. Robertshaw, M. Wood, E. Melchiorre et al., 2010.

22  M. Wood, L. Dussubieux, L. Wadley, 2009.

23  See F. Chami, F. Le Guennec-Coppens, S. Mery, 2002 for a full discussion of early evidence of contact between East Africa and the rest of the world. See also M.C. Horton, J. Middleton, 2000.

24  See for example A. Crowther, M. Horton, A. Kotarba-Morley et al., 2014.

25  P.J.J. Sinclair, 1982 and 1987, p. 86-90; A. Ekblom, 2004; and P.J.J. Sinclair, A. Ekblom, M. Wood, 2012.

26  M.C. Horton, 2004, p. 64-65; M. Wood, 2012, p. 47; and M. Wood, L. Dussubieux, P. Robertshaw, 2012, p. 72.

27  M. Wood, L. Dussubieux, P. Robertshaw, 2012.

28 Ibid., Table 15.

29  P. Robertshaw, M. Wood, E. Melchiorre et al., 2010.

30  M.C. Horton et al., 1996.

31  H. Morrison, 1984.

32  H.N. Chittick, 1974.

33  M.C. Horton et al., 1996, p. 13.

34  M. Wood, 2002.

35  A. LaViolette, J. Fleisher, 2009; J. Fleisher, 2010; and J. Fleisher, A. LaViolette, 2012.

36 http://www.sealinksproject.com

37  A. Juma, 2004, did extensive work at Unguja Ukuu, proposing it dated from ~ 500 to 900 CE with subsequent occupations in c. 1100 and 1500. He recovered many glass beads but did not publish detailed records, so it has not been possible to use them in bead studies.

38  L. Dussubieux, B. Gratuze, M. Blet-Lemarquand, 2010.

39  L. Dussubieux, P. Robertshaw, M. Glascock, 2009, p. 159 and L. Dussubieux, B. Gratuze, M. Blet-Lemarquand, 2010.

40  M. Tampoe, 1989, p. 108-109 and J. Carswell, S. Deraniyagala, A. Graham, 2013.

41  P. Francis, 2002, p. 228, note 21.

42 Ibid.

43  M. Wood, L. Dussubieux, P. Robertshaw, 2012.

44  B. Chaisuwan, 2011.

45  P. Francis, 2002, p. 44, p.  97-98 and 2013, p. 357.

46  An article detailing evidence of this and other information about the Unguja Ukuu glass bead assemblage is in preparation.

47  P. Francis, 1993, p. 14.

48  J. Lankton, L. Dussubieux, 2006, p. 135 and 2013, p. 431.

49  B. Bronson, C. Ho, 1991, p. 18.

50  H. Wright, 1984.

51  C. Radimilahy, 1998.

52  P. Robertshaw, B. Rasoarifetra, M. Wood et al., 2006, p. 106.

53  H. Wright, F. Fanony, 1992, p. 32-33.

54  P. Robertshaw, B. Rasoarifetra, M. Woodet al., 2006, p. 93.

55  M. Pearson, 2010.

56  D. Whitehouse, A. Williamson, 1973, p. 48.

57  G.S.P. Freeman-Grenville, 1962, p. 16.

58  M. Tampoe, 1989, p. 113.

59  J.S. Trimingham, 1975, p. 122 and p. 135; H.N. Chittick, 1977, p. 192; G.F. Hourani, 1995, p. 148; and M.C. Horton, J. Middleton, 2000, p. 66.

60  G.S.P. Freeman-Grenville, 1962, p. 15.

61  J.S. Trimingham, 1975, p. 121.

62  M. Tampoe, 1989, p. 102.

63  P. Robertshaw, M. Wood, E. Melchiorre et al., 2010, p. 1903.

64  J.C. Wilkinson, 1981, p. 282 and M. Tampoe, 1989, p. 44.

65  M. Tampoe, 1989, p. 106.

66  The likelihood that artisans moved around the Indian Ocean to ply their trades has been argued by M.C. Horton, 2004; H.P. Ray, 2004; and specifically for beadmakers by P. Francis, 2002, p. 35-36.

67  M.C. Horton, 2004, p. 68.

68  F.A. Khan, 2005, p. 45.

69 Ibid., p. 18.

70  P. Robertshaw, M. Wood, E. Melchiorre et al., 2010; M. Wood, 2011.

71  L. Dussubieux, B. Gratuze, M. Blet-Lemarquand, 2010, p. 1650.

72  A. Kanungo, 2004, p. 45 and p. 51.

73  For discussions about southern routes, see G. Shepherd, 1982; M. Tampoe, 1989 p. 109, p. 121, and p. 123 citing R.E. Dewar, 1988; M.C. Horton et al., 1996, p. 418; R.L. Pouwels, 2002, p. 400; and P. Beaujard, 2011, p. 170-173.

74  Because it is difficult to disentangle time frames here and few sites are involved, this section will include the span from the 11th through the 17th centuries.

75  H. Wright, 1992, p. 105.

76  C. Radimilahy, 1998.

77  P. Robertshaw, B. Rasoarifetra, M. Wood et al., 2006.

78  M. Horton et al., 1996, p. 417.

79  R.L. Pouwels, 2002, p. 400.

80  H.N. Chittick, 1974, p. 464.

81  Qin Dashu of Peking University has reported finding early Ming imperial kiln shards of the type used only by the imperial court at a coastal site in Kenya (abstract of paper for the 14th Congress of the Pan-African Archaeological Association/Society of Africanist Archaeologists, Johannesburg, July 14-18, 2014).

82  M. Wood, 2009.

83  G.M. Theal, 1898, II, p. 303; W.G.N. van der Sleen, 1956, p. 28 and 1958, p. 212; J.F. Schofield, 1958, p. 183; E. Axelson, 1973, p. 46; and M. Wood, 2009.

84  In this period at Kilwa, more than 50% were reportedly wound (H.N. Chittick, 1974); at Manda it was 25 % (H. Morrison, 1984); and at Shanga over 50% (M.C. Horton et al., 1996).

85  M. Wood, 2000, p. 81, 2005, p. 30, and 2011, p. 75.

86  M. Horton, 2004, p. 64.

87 Ibid., p. 65.

Haut de page

Table des illustrations

Titre Figure 1: Indian Ocean and sites mentioned in text
Crédits Marilee Wood
URL http://journals.openedition.org/afriques/docannexe/image/1782/img-1.jpg
Fichier image/jpeg, 180k
Titre Figure 2: Southern African bead series
Crédits Marilee Wood
URL http://journals.openedition.org/afriques/docannexe/image/1782/img-2.png
Fichier image/png, 308k
Titre Figure 3: Comparison of southern African and East African bead types by period
Crédits Marilee Wood
URL http://journals.openedition.org/afriques/docannexe/image/1782/img-3.jpg
Fichier image/jpeg, 546k
Haut de page

Pour citer cet article

Référence électronique

Marilee Wood, « 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 | 2015, mis en ligne le 21 décembre 2015, consulté le 18 décembre 2017. URL : http://journals.openedition.org/afriques/1782 ; DOI : 10.4000/afriques.1782

Haut de page

Auteur

Marilee Wood

School of Geography, Archaeology and Environmental Studies, University of the Witwatersrand

Haut de page
  • Logo Institut des mondes africains
  • Logo DOAJ – Directory of Open Access Journals
  • OpenEdition Journals