Early Exchange between Africa and the Wider

Indian Ocean World 1st Edition Gwyn Campbell

(Eds.) pdf download
https://textbookfull.com/product/early-exchange-between-africa-and-the-wider-indian-ocean-world-1st-
edition-gwyn-campbell-eds/

★★★★★ 4.6/5.0 (40 reviews) ✓ 169 downloads ■ TOP RATED

"Fantastic PDF quality, very satisfied with download!" - Emma W.

DOWNLOAD EBOOK
 Early Exchange between Africa and the Wider Indian Ocean
World 1st Edition Gwyn Campbell (Eds.) pdf download

TEXTBOOK EBOOK TEXTBOOK FULL

Available Formats

■ PDF eBook Study Guide TextBook

EXCLUSIVE 2025 EDUCATIONAL COLLECTION - LIMITED TIME

INSTANT DOWNLOAD VIEW LIBRARY

Collection Highlights

Currencies of the Indian Ocean World Steven Serels

Animal Trade Histories in the Indian Ocean World Martha

Chaiklin

Portuguese Decolonization in the Indian Ocean World

History and Ethnography Pamila Gupta

Connectivity in Motion: Island Hubs in the Indian Ocean

World 1st Edition Burkhard Schnepel
International Exchange In The Early Modern Book World
Matthew Mclean

Indian Ocean Resources and Technology 1st Edition Ganpat

Singh Roonwal

Textile Trades, Consumer Cultures, and the Material Worlds

of the Indian Ocean: An Ocean of Cloth 1st Edition Pedro
Machado

The Wine Value Chain in China Consumers Marketing and the

Wider World Roberta Capitello

Capacity Building for Maritime Security The Western Indian

Ocean Experience Christian Bueger
PALGRAVE SERIES
IN INDIAN OCEAN
WORLD STUDIES

EARLY EXCHANGE
BETWEEN AFRICA AND
THE WIDER INDIAN
OCEAN WORLD
Edited by
Gwyn Campbell
Palgrave Series in Indian Ocean World Studies

Series Editor
Gwyn Campbell
McGill University
Montreal, Canada
This is the first scholarly series devoted to the study of the Indian Ocean
world from early times to the present day. Encouraging interdisciplinarity,
it incorporates and contributes to key debates in a number of areas includ-
ing history, environmental studies, anthropology, sociology, political sci-
ence, geography, economics, law, and labor and gender studies. Because
it breaks from the restrictions imposed by country/regional studies and
Eurocentric periodization, the series provides new frameworks through
which to interpret past events, and new insights for present-day policy-
makers in key areas from labor relations and migration to diplomacy and
trade.

More information about this series at

http://www.springer.com/series/14661
 Gwyn Campbell
Editor

Early Exchange
between Africa and
the Wider Indian
Ocean World
Editor
Gwyn Campbell
McGill University
Montreal, Canada

Palgrave Series in Indian Ocean World Studies

ISBN 978-3-319-33821-7    ISBN 978-3-319-33822-4 (eBook)
DOI 10.1007/978-3-319-33822-4

Library of Congress Control Number: 2016959959

© The Editor(s) (if applicable) and The Author(s) 2016

This work is subject to copyright. All rights are solely and exclusively licensed by the
Publisher, whether the whole or part of the material is concerned, specifically the rights of
translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on
microfilms or in any other physical way, and transmission or information storage and retrieval,
electronic adaptation, computer software, or by similar or dissimilar methodology now
known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are
exempt from the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information
in this book are believed to be true and accurate at the date of publication. Neither the pub-
lisher nor the authors or the editors give a warranty, express or implied, with respect to the
material contained herein or for any errors or omissions that may have been made.

Cover illustration: © Baie de Makassar 1979, padewakang Bugis

Printed on acid-free paper

This Palgrave Macmillan imprint is published by Springer Nature

The registered company is Springer International Publishing AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
 Acknowledgements

This work was supported by the Social Sciences and Humanities Research
Council of Canada.

v
 Contents

1 Africa and the Early Indian Ocean World Exchange

System in the Context of Human–Environment
Interaction   1
Gwyn Campbell

2 Origins of Southeast Asian Shipping and Maritime

Communication Across the Indian Ocean  25
Waruno Mahdi

3 Austronesian Shipping in the Indian Ocean: From

Outrigger Boats to Trading Ships  51
Pierre-Yves Manguin

4 Austronesians in Madagascar: A Critical Assessment

of the Works of Paul Ottino and Philippe Beaujard  77
Alexander Adelaar

5 Early Greek and Latin Sources on the Indian Ocean

and Eastern Africa 113
Ephraim Lytle

vii
viii Contents

6 A GIS Approach to Finding the Metropolis of Rhapta 135

Carl Hughes and Ruben Post

7 Contact between East Africa and India in the First

Millennium CE   157
Sunil Gupta

8 Eastern Africa and the Indian Ocean World in the First

Millennium CE: The Glass Bead Evidence   173
Marilee Wood

9 Migration and Interaction between Madagascar and

Eastern Africa, 500 BCE–1000 CE: An Archaeological
Perspective   195
Anneli Ekblom, Paul Lane, Chantal Radimilahy, Jean-­Aime
Rakotoarisoa, Paul Sinclair, and Malika Virah-Sawmy

10 A Genomic Investigation of the Malagasy Confirms

the Highland–Coastal Divide, and the Lack of Middle
Eastern Gene Flow   231
Jason A. Hodgson

11 Intercontinental Networks Between Africa and Asia

Across the Indian Ocean: What Do Village Chickens
Reveal?   255
J.M. Mwacharo

12 East Africa in the Early Indian Ocean World Slave

Trade: The Zanj Revolt Reconsidered   275
Gwyn Campbell

References   305

Index   357
 List of Figures

Fig. 2.1 A postulated scheme of primeval developments of watercraft

construction in ISEA and around the South China Sea:
(a) a tapered raft; (b) a multiple dugout; (c) a double canoe
with advanced dugouts, and (d) with five-part hulls 27
Fig. 2.2 Presumed routes of Late Pleistocene and early Holocene
migrations of Sunda Shelf populations into Indochina and
further northwards to Southeast China and Taiwan 28
Fig. 2.3 Theoretical development scheme of a Chinese junk from a
double canoe with five-part hulls: (a) cross-section of the double
canoe; hazy intermediate is purely hypothetic; (b, c) transversal
and longitudinal cross-sections of a prototypical junk 32
Fig. 2.4 Advancing against the wind, (a) a double canoe tacks, as does
(b) a non-reversible single outrigger boat; (c) a reversible single
outrigger boat shunts; while (d) a double-outrigger boat wears
(the wind is blowing from the back/top of the figure to the
front/bottom)35
Fig. 2.5 Tapered rafts: (a) The kattumaram of Malabar, drawing by
Thomas Bowrey between 1669 and 1679 (Bowrey 1903);
(b) a konga of Eddystone Island, the Solomons (drawn from
photograph); (c) a “batu” (balsa?) of Peru in the 1582 diary of
Richard Madox (Donno 1976; courtesy the Hackluyt Society);
(d) a jangada of Brazil in 1587, notations of Gabriel Soares
(Charton 1850; courtesy Bibliotheque nationale de France) 37

ix
x List of Figures

Fig. 2.6 Two lakatois in the Gulf of Papua. Note their Oceanic sprit
sails (Wirz 1931; courtesy Staatsbibliothek zu Berlin). 38
Fig. 2.7 (a) Splitting and successive carving of a log; (b, c) cross-section
and side view of a Sri Lankan madel poruwa, based on
Kentley (2003: 169, fig. 6.2) 39
Fig. 3.1 Evolution of fastening of the planks in the lashed-lug and
stitched-plank tradition (drawings: P.-Y. Manguin) 58
Fig. 3.2 The Punjulharjo ship during excavation (photo: EFEO/
P.-Y. Manguin)59
Fig. 3.3 The structure of the Punjulharjo ship: stitched planks with lugs,
frames, and stringers (photo: EFEO/P.-Y. Manguin) 60
Fig. 7.1 Grey ware pottery from Elephanta Island, Mumbai, India,
sixth–seventh century AD (photo: Sunil Gupta) 165
Fig. 7.2 View of the beach from the archaeological site of Unguja Uku,
Zanzibar (photo: Sunil Gupta) 166
Fig. 7.3 Indian pottery from Manda, north Tanzania (after
Chittick 1984) 168
Fig. 8.1 Glass beads discussed in text 179
Fig. 9.1 Example of a femur bone with clear cut marks from Anjohibe
Cave (reprinted from Gommery et al. (2011). Copyright 2011,
reproduced with permission from Elsevier) 200
Fig. 9.2 Flakes and blades from Lakaton’i Anja Cave (reprinted from
Dewar et al. (2013). Copyright 2013, reproduced with
permission from PNAS) 202
Fig. 9.3 Summary of vegetation changes and some social transformations
(modified from Virah-Sawmy et al. 2010) 211
Fig. 10.1 Plot of the first two dimensions of a multidimensional scaling
analysis of pairwise identity by state using the 2.5k SNP
dataset. The Malagasy are intermediate between the sub-Saharan
African and East Asian samples as expected if the Malagasy
are a mixture of Bantu- and Austronesian-speakers from the
East African coast and Island Southeast Asia, respectively 238
Fig. 10.2 Barplot of inferred individual ancestry components from
ADMIXTURE analysis of the 2.5k SNP dataset with three
inferred ancestry components. The three components largely
correspond to “African” (black), “European” (light grey), and
“East Asian” (dark grey) clusters. The Malagasy show variable
amounts of “African” and “East Asian” ancestry, with less
variation and greater “African” ancestry in the coastal
populations (Mikea, Temoro, and Vezo) than the Highlanders
(Merina). These ancestry components likely reflect the
proportions of ancestry from Bantu- and Austronesian-speakers 240
 List of Figures  xi

Fig. 10.3 Barplot of inferred individual ancestry components from

ADMIXTURE analysis of the 150k SNP dataset with five
inferred ancestry components. Here the coastal Malagasy
(Mikea, Temoro, and Vezo) show no significant amount of
the “Middle Eastern” ancestry (IAC 5) that is predominant
in the Yemeni 242
Fig. 10.4 TREEMIX analysis of the 150k SNP dataset, with San
as outgroup. (a) Covariance matrix of allele sharing with standard
error, given the inferred bifurcating relationship among
populations. Greater error indicates poorer fit of the tree
structure, and the likelihood of a migration from elsewhere
on the tree. (b) Inferred bifurcating tree of relationships among
the populations with two migrations. The Malagasy are most
closely related to sub-Saharan Africans, but with a heavy
migration from Southeast Asia into the Malagasy. No other
migrations into the Malagasy were inferred when additional
migrations were added. 245
Fig. 11.1 Pie charts showing the proportion of the different
mtDNA clades observed in domestic village chickens
from countries/regions bordering the Indian Ocean. 261
 List of Tables

Table 2.1 Shape of the Chinese character zhōu “watercraft” at

different times 33
Table 9.1 Comparison of OSL and 14C dates from Lakaton’i Anja
Cave from Dewar et al. (2013) (OSL dates are from the
OSL Dating Laboratory at the University of Wollongong) 203
Table 10.1 Sample information, 2.5k SNP dataset 237
Table 10.2 2.5k SNP ADMIXTURE K = 3 ancestry proportions
summary241
Table 10.3 Sample information, 135k SNP dataset 243
Table 10.4 135k SNP ADMIXTURE K = 5 ancestry proportions
summary243
Table 11.1 Some selected domestic and wild animal species dispersed
across the Indian Ocean 257
Table 11.2 Nomenclature of mtDNA control region clades across
different studies, their possible origin and equivalent clades
following Liu et al. (2006) study nomenclature 262
Table 11.3 Clades observed from the analysis of mtDNA D-loop
sequences of village chickens from countries around
the Indian Ocean 264

xiii
 List of Maps

Map 3.1 Main shipwreck sites of Southeast Asia with locally built
vessels (map: P.-Y. Manguin) 61
Map 3.2 The Maldives at the heart of Indian Ocean networks (map:
P.-Y. Manguin)66
Map 4.1 Languages and locations in insular South East Asia that are
referred to in this chapter (© Alexander Adelaar) 79
Map 4.2 Map showing Malagasy dialects and Zafiraminia migrations
(© Alexander Adelaar) 80
Map 6.1 The East African Coast according to the Periplus (drawn by Carl
Hughes of the Indian Ocean World Centre (IOWC), McGill
University)137
Map 6.2 Estimated maritime distances off the East African Coast
in stades (Carl Hughes, IOWC) 141
Map 6.3 Estimated maritime distances between the East African
Coast and the Islands of Pemba, Zanzibar, and Mafia in stades
(Carl Hughes, IOWC) 142
Map 6.4 River courses on Pemba, Zanzibar and Mafia
(Carl Hughes, IOWC) 143
Map 6.5 Raw geographical coordinates for the East African Coast from
Ptolemy’s Geography (Carl Hughes, IOWC) 147
Map 6.6 Geometric correction of Ptolemy’s geographical coordinates
for the East African Coast (Carl Hughes, IOWC) 149
Map 6.7 The East African Coast: ArcGIS corrected coastline buffer
version A (Carl Hughes, IOWC) 151
Map 6.8 The East African Coast: ArcGIS corrected coastline buffer
version B (Carl Hughes, IOWC) 152
Map 6.9 The probable location of Rhapta (Carl Hughes, IOWC) 153

xv
xvi List of Maps

Map 7.1 Map showing “Periplus” port sites in the Gulf of Aden (map:
Sunil Gupta) 159
Map 7.2 Map showing important ports in the Western Indian Ocean
in the first millennium AD together with distribution of
grooved rim Red Polished Ware ceramics (map: After
Chittick 1976) 160
Map 7.3 Map showing coastal settlements and ports on the Swahili
Coast in the first millennium AD (Sunil Gupta/Syncrotek) 161
Map 9.1 Map of BC to first millennium AD sites in Madagascar.
Archaeological lithic assemblage sites are written out
and shown in bold, sites where bones with possible cut
marks have been found are abbreviated: Amb Ambatovy,
Andrah Anadrahome, Andran Andranosoa, Beh Behova,
Ber Bernafandy, Ita/Tsi Itampalo/Tsiandroina, Tao
Taolambiby, Lam Lamboharana, Anj Anjohibe.
Palaeoecological sampling sites in grey circles
and numbers (1) Mitsinjo, (2) Kaviataha, (3) Tritrivakely,
(4) Belo, (5) Ambolisatra 199
Map 9.2 East African Coast and selected late first millennium
and early second millennium sites in Madagascar 215
Map 10.1 Map of approximate Malagasy population sample
locations used in this study 236
Map 12.1 The marshlands of lower Iraq (Carl Hughes, IOWC) 278
Map 12.2 The Muslim heartlands and main slave routes
in c. 850 CE (Carl Hughes, IOWC) 287
Map 12.3 Early dhow routes in the western Indian Ocean
(Carl Hughes, IOWC) 293
 CHAPTER 1

Africa and the Early Indian Ocean World

Exchange System in the Context of
Human–Environment Interaction

Gwyn Campbell

This volume comprises a selection of chapters by leading scholars on

aspects of early exchange between Africa and the wider Indian Ocean
world (IOW)—a macro-region running from Africa to the Middle East,
South and Southeast Asia, and the Far East. The rationale for regarding
this region as a “world” is the central significance of the monsoon sys-
tem. Unlike the Atlantic and Pacific, the IOW seas (the Indian Ocean and
the Indonesian and China Seas) are capped by a huge continent—Asia.
During northern hemisphere summers, as the Asian continent warms up,
hot air rises from the land, causing a vacuum that, through the process of
convection, sucks in moist air from the oceans to the south. This ­creates
the southwest monsoon. In winter, the opposite process occurs, and air

G. Campbell (*)
McGill University, Montreal, QC, Canada

© The Author(s) 2016 1

G. Campbell (ed.), Early Exchange between Africa and the Wider
Indian Ocean World, DOI 10.1007/978-3-319-33822-4_1
2 G. CAMPBELL

is expelled from the continent over the oceans, creating the northeast
monsoon.
This regular biannual alternation of winds and currents governs the IOW
littorals and oceans to about 12°S of the equator and has fundamentally
shaped primary patterns of production and trade, and thus of human his-
tory, across most of the IOW. First, the rains that accompany the southwest
monsoon between June and September created a zone of wet crop (pre-
dominantly rice) cultivation across southern Asia, north of which lies a drier
belt of predominantly grain (wheat and barley) cultivation. Second, mon-
soon winds facilitated the emergence of early trans-IOW oceanic exchange.
This developed in the early centuries BCE, laying the basis for an IOW
global economy that preceded that of the Atlantic world by over 1500
years. As the monsoons regulated much of both agriculture and trade, there
was, from the outset, remarkable synchronism between land-based produc-
tion and commercial systems and trans-oceanic trade. Moreover, as other
wind systems, such as the southern hemisphere trades and equatorial cur-
rents, could be used to link into the monsoons, the impact of the monsoon
network of exchange extended to regions that lay beyond the actual reach
of the monsoons, such as the inner Persian Gulf and southeast Africa.
This introductory chapter explores the environmental context for the
development of an IOW global system of exchange and the significance
within it of human–environment interaction from early times to about 1300
CE. Most histories of the IOW have underestimated the role of the envi-
ronment, and the dynamic rather than static nature of human–environment
interaction that, more than political events or the evolutionary march of cap-
italism, has moulded the major temporal phases in human history (see e.g.,
Chaudhuri 1985, 1992; Abu-Lughod 1989; Frank 1998; McPherson 1995;
Pearson 2011; Sheriff 2010; Alpers 2014; Beaujard 2012). The primacy of
human–environment interaction further devalues historical analyses based
on conventional “country” or “area studies.” In this introduction, exchange
relations between Africa and the wider IOW go beyond the usual focus on
East Africa to incorporate all regions of Indian Ocean Africa (IOA)—here
defined as all areas of the continent littoral to the Indian Ocean and Red Sea,
and hinterland regions with intimate early connections to the Indian Ocean.

Rise of the IOW Global Economy

The first major phase in modern human history followed the end of the
last Ice Age, some 11,500 years ago, when significant melting of the Arctic
pack-ice established the conditions for enhanced food and craft produc-
 AFRICA AND THE EARLY INDIAN OCEAN WORLD EXCHANGE SYSTEM... 3

tion. This in turn encouraged humans to disperse from equatorial and

other regions with favourable micro-climates into previously inhospitable
areas throughout the northern hemisphere. From around 6000 BCE, a
period of prolonged aridity provoked a search for methods of conserv-
ing and managing water and food supplies that precipitated the Neolithic
Revolution in which regular surpluses and improved storage facilities
enabled the emergence of specialist non-agriculturalists, including artisans
and soldiers. Enhanced agricultural and craft production, and growth in
elite demand for luxuries, laid the basis for the rise of early trans-IOW
trade. Scholarly attention has focused chiefly on exchange between the
centralised polities of the Middle East, South Asia, and China along the
so-called trans-Asia “Silk Road,” parts of which operated from 3000
BCE and which appears to have been fully functional by the fifth-century
BCE. By about 500 BCE, interlocking local maritime networks also con-
nected the entire northern rim of the IOW, from the Red Sea to China.
Maritime connections were assisted by the onset from about 300 BCE of
a strong and relatively stable monsoon system, which provided the condi-
tions, along with advances in astral navigation and boat-building, for IOW
sailors to make direct trans-oceanic voyages (Gupta et al. 2003; Hourani
1995; Ray 1990). From that time, geographically extensive, sophisticated,
and durable systems of exchange arose that reflected the birth of an IOW
“global economy.”
Two major trans-IOW maritime networks developed. The foremost
ran along the northern rim of the South China Sea and Indian Ocean.
Secondary routes connected South China via the Philippines and east-
ern Indonesia to Java, Sumatra, and the Malay Peninsula; and the Sunda
Straits via the Maldive Islands and Sri Lanka to the Persian Gulf, Red Sea,
and East Africa. However, few vessels sailed the entire maritime length
of the IOW as the alternating regime of monsoon winds enabled return
voyages to be completed within a year only in geographically restricted
zones. Therefore, a pattern developed whereby most vessels remained
within one of the three main IOW maritime zones: the Arabian Sea, Bay of
Bengal, and the China and Indonesian Seas. Direct two-way trade devel-
oped as early as 200 BCE between East India and the Malay Peninsula,
and from about 120 BCE between Egypt and Malabar. By the BCE/CE
changeover, as Ephraim Lytle notes, some 120 ships of between 200 and
300 tons each left Egypt annually (Lytle in this volume; see also Hourani
1995; Warmington 1995; Wheatley 1964a).
Direct trans-oceanic voyages greatly stimulated trade between Africa
and the wider IOW. Elite demand, notably in China, India, and the
4 G. CAMPBELL

Middle East, ensured a vibrant commerce in exotic tropical commodi-

ties, including ivory, pearls, tortoise-shell, ambergris, rhinoceros horn and
hide, gums, musk, and incense—chiefly for ornamental, craft, culinary and
medicinal purposes. India, for example, was a noted consumer of gold
and exporter of fine pepper, luxury cloth, dyes, craftwork, jewellery, pearls,
and precious stones (including emeralds, crystals, rubies, diamonds, sap-
phires, and lapis lazuli). African commodities reaching India did so both
directly and via entrepôts in Arabia, such as Muza (for East Africa) and
Adulis (for Ethiopia). In return, as Sunil Gupta notes, India exported to
Red Sea Africa a wide variety of commodities, including cloth and ceram-
ics (Casson 1989; Warmington 1995; Miller 1969; also Gupta, Lytle, and
Hughes and Post, in this volume).
There is considerable debate about northeast Africa as a source of
spices. The Periplus (c.13–140 CE), an invaluable early first-century CE
Greek manuscript on western IOW trade, and Ptolemy’s Geography (c.150
CE)—works discussed in this volume by Ephraim Lytle, Carl Hughes
and Ruben Post—emphasise the Horn of Africa, termed the “Cape of
Spices,” as the world’s chief source of cassia and/or cinnamon (Casson
1989), while Cosmas (mid-sixth century) specified that cinnamon was
obtained in the Somali hinterland (McCrindle 1897). However, as Lytle
indicates, there is much confusion in the literature over the terms “cinna-
mon,” which is not mentioned in the Periplus, and “cassia.” Both belong
to the genus Cinnamomum, generically described as of the same origin,
and used in the same way. The Greek texts might have been referring to
plants indigenous to the countries bordering the eastern end of the Red
Sea. Certainly Somalia produced myrrh and frankincense. However, cin-
namon proper (Cinnamomum verum or C. zeylanicum) originated from
Sri Lanka and, as Waruno Mahdi and Lytle (in this volume) point out,
the famous passage in Pliny (quoted in the chapter by Lytle) concerning
rafts transporting cinnamon on the high seas probably referred to a mari-
time trade in spices with, or transited through, South Asia (Mahdi 1999a;
Warmington 1995; Hourani 1995; Lytle and Mahdi in this volume).
Azania—the term used by the Periplus and Ptolemy for the modern-­
day Swahili coast of East Africa—also developed regular direct linkages
with the wider IOW. Claims to a very early maritime connection with East
Africa have a long but disputed heritage. Felix Chami asserts that by 3000
BCE, Sub-Saharan Africans had settled the present-day Swahili coast and
offshore islands of Zanzibar and Kilwa; and that by 600 BCE, or prob-
ably earlier, they had established trade relations with India from where
 AFRICA AND THE EARLY INDIAN OCEAN WORLD EXCHANGE SYSTEM... 5

they adopted the chicken, rice, cotton, and possibly the coconut (Chami
2009; Chami et al. 2003; Sinclair et al. 2006). Early staple East African
exports included ivory, rhinoceros horn, possibly slaves and, according to
Gupta, Mozambique copal (for the Iraqi market) (Mollat 1971; Harris
1971; Gupta in this volume).
Certainly by the BCE/CE changeover, Egyptian, Arab, Mediterranean,
and possibly Indian and Axumite vessels sailed to Azania. By the first-­century
CE, the major Azanian trading centre was Rhapta. It boasted many resident
foreign ship captains and traders, mostly Arabs, who had taken local wives
and spoke the local language. By the second century, Rhapta had developed
into what the Periplus calls a “metropolis,” a term applied elsewhere only to
Meroe and Aksum in IOA, Saphar and Saubatha in Yemen, and Minnagar
in India (Casson 1989; Huntingford 1980). As noted by Hughes and Post
in this volume, Rhapta’s location is the subject of considerable conjecture.
The Periplus indicates that it was situated near a coastal promontory and
major river somewhere in the vicinity of Menouthias—generally considered
to be modern-day Pemba. Scholars have placed Rhapta variously oppo-
site Manda and Pate, at Mnyuzi 48 km up the Pangani river, in the Rufiji
estuary opposite Mafia where Felix Chami discovered some 20 first- to
fifth-century CE sites containing Egyptian or Mediterranean beads, and
near present-day Dar es Salaam—the location Hughes and Post choose in
applying geographic information system (GIS) techniques to a re-reading
of the geographic coordinates and other information given in the Periplus
and Ptolemy’s Geography (Hughes and Post in this volume; see also Casson
1989; Allen 1993; Horton 1996; Huntingford 1980; Middleton 1992).
Rhapta exported large quantities of tortoise-shell and ivory—second
in value only to, respectively, Indian tortoise-shell and Adulis ivory—
and, alongside Adulis, was the sole IOA source of rhinoceros horn. It
also produced small quantities of nautilus shell, possibly mangrove poles
used for house construction in Arabia and the Persian Gulf, and iron ore
or semi-processed iron for the Indian market. Imports included lances,
axes, knives, small awls, glass beads and other glass objects, wine, and
grain (Austen 1987; Casson 1989; Horton 1996; Huntingford 1980;
Sutton 1990; Vérin 1986). Michael Pearson asserts that Muza mer-
chants imported wheat and ghee (Pearson 2000), and Pra Shirodkar that
corn, rice, butter, sesame, cotton, sugar, and iron goods were shipped
directly from India (Shirodkar 1985). However, as Gupta notes for the
BCE/CE changeover, Indian products are mentioned in early texts, or
have been uncovered, only in connection with imports into northern
6 G. CAMPBELL

IOA. Chami claims to have found third-century CE Roman beads at

Rufiji Delta sites in Tanzania, and Unguja Ukuu, a large (16–17 ha)
site on Zanzibar (Chami 1999a,b). However, while Marilee Wood (in
this volume) considers that in the early centuries CE, Roman captains
may have shipped beads from India and Sri Lanka to Azania, she assigns
to most early imported bead finds a probably late first millennium CE
origin.
The introduction into Azania of the Southeast Asian tropical crop com-
plex, centred on the banana, sweet potato, manioc (cassava), Southeast
Asian yam, and—Alexander Adelaar argues (in this volume)—taro
and rice, had enormous impact. Mahdi (in this volume) notes that the
banana, which was domesticated in highland Papua-New Guinea around
4950–4440 BCE, can only be propagated through human intervention
(planting parts of its underground stem), so it probably accompanied visi-
tors to Southeast and South Asia, where it was crossed with local ined-
ible wild species. Recent disputed claims that the banana was cultivated
in Uganda by 3000 BCE and the Cameroons by 500 BCE could indi-
cate that Southeast Asian plants were introduced into Africa over a much
longer period and in a more haphazard fashion than originally thought
(Boivin et al. 2013; Chami 2006; Denbow 2014). Because of the banana’s
sugar, vitamin, and potassium content, its cultivation may have supported
unprecedentedly dense populations and, by extension, sophisticated
polities.

Economic Uncertainty, c.300 CE to 850 CE

Overall, economic activity in the IOW, including trans-IOW exchange,
slumped from the fourth century—events probably related to a prolonged
period of environmental crisis from about 300 CE to 900 CE. During
the economic upswing from around 300 BCE to 300 CE, there was a
substantial increase in population, agriculture, industry, and trade which
resulted in large-scale deforestation for crop cultivation, pasture, boat and
building construction, and domestic and industrial fuel—notably for iron-
smelting. This in turn caused a substantial increase in methane and carbon
dioxide emissions, which had a significant climatic impact (Sapart et al.
2012). Despite considerable regional and temporal variations, the indica-
tions are that there was widespread aridification from about 280 CE to
900 CE in the northern hemisphere and a significant fall in temperatures
 AFRICA AND THE EARLY INDIAN OCEAN WORLD EXCHANGE SYSTEM... 7

and ­subsequent weakening in the southwest monsoon from about 250 CE

to 800 CE (Chew 2007; Gupta et al. 2003; Yang et al. 2002).
Disease contributed to socio-economic and political instability. From
37 to 653 CE, China was hit by virgin diseases, including smallpox and
measles, with devastating demographic consequences: the Chinese popu-
lation slumped from an estimated 58.5 million in 2 CE to 42.33 million
by 742 CE. Political disturbances and economic decay followed, accentu-
ated by invasions from the steppe (Chang 1969; McNeill 1976). In the
western IOW, Egypt experienced major harvest shortfalls between 155
CE and 299 CE due to deficient floods. A prolonged drought again hit
the region in the sixth century, and an associated epidemic of bubonic
plague (the “Justinian” plague) travelled via trade routes from Ethiopia to
Egypt from where it went on to devastate Southern Europe from 542 CE
(Connah 2001; Edwards 2013).
Axum, in northern Ethiopia, flourished until the sixth century. Located
near the headwaters of two major Nile tributaries, Axum controlled busy
caravan routes between the Red Sea and the African hinterland. By the
mid-first century CE, it had largely diverted trade from Meroe and Sudan
that traditionally flowed to Egypt to the Axum-dominated Red Sea ports
of Ptolemais Theron (possibly Suakin or Trinkitat) and Adulis/Zeila
(present-­day Massowa). Axum expanded militarily to dominate Ethiopia
and north-western Arabia. From around 270 CE, it minted its own coin-
age, and its fleet maintained a vigorous trade with Somalia, Socotra, Egypt,
the Near and Middle East, and India. Axum’s merchants and courtiers
adopted Greek as a lingua franca, and in 303 CE, its court converted to
Christianity—a reflection of important commercial relations with Egypt
and the Levant. At its height from the fifth to the mid-­sixth century,
Axum contested Sassanid domination of maritime trade with East Africa
and established a colony in Sri Lanka (McCrindle 1897; Warmington
1995; Young 2001). However, in 520, the Sassanids conquered Yemen
and Oman, by 570 had expelled Axumites from Arabia, and in 576
seized Aden, thereby controlling the entrance to the Red Sea. Thereafter,
Axum’s commercial influence dwindled, accentuated by ­climatic change as
a consistently wet period from roughly 1 to 730 CE was followed by two
centuries of highly erratic rainfall, and the desiccation of highland Eritrea
and Tigray. This accentuated declining soil fertility due to demographic
expansion, tree clearance, and general land overuse and resulted in a shift
southwards of population and political power (Barbosa 1918; Boardman
2010; Butzer 1981).
8 G. CAMPBELL

For many scholars, Azania also experienced economic vibrancy that,

though less spectacular, proved more durable than that of Axum and laid
the basis for the Swahili maritime culture that emerged in the late first
millennium CE. They assert that during the first centuries CE, Bantu-­
speakers moved progressively east from the Great Lakes region, to reach
the Indian Ocean coast sometime after the fourth century. By the late sev-
enth century, they had developed a proto-Swahili maritime culture, and by
the late 700s were speaking Kiswahili, structurally a Bantu language (Allen
1993; Nurse and Spear 1985; Pouwels 1987).
More radically, Chami has argued that as early as 1000 BCE, Bantu-­
speakers reached the coast where their iron-working and experience with
boating on inland waters facilitated a rapid transition to a maritime lifestyle.
They founded Rhapta, a prototype for later Swahili port-cities, and proba-
bly also Unguja Ukuu, a Zanzibari emporium that from about 500 to 900
CE traded extensively with India, the Middle East, and the Mediterranean
world (Chami 2006; Chami and Msemwa 1997). However, Chami’s
claims are controversial. Neither method of absolute dating—radiocarbon
or imported pottery—is totally reliable for the East African coast. Testing
mangrove wood, commonly used in construction, is problematic as its
contact with seawater, which has concentrations of carbon 14, can give
misleading results. This leaves imported pottery, but most hypotheses are
based on local rather than foreign pottery and its alleged association with
iron-working—in which the relation between pottery types is subject to
considerable speculation (Forslund 2003; Horton 1996; Ekblom et al. in
this volume).
Cosmas in his sixth-century work failed to comment on Azania,
which may indicate that it had by then experienced economic collapse
(McCrindle 1897). Indeed, Rhapta disappears from the record in the
fourth century, after which, from roughly 400 to 750 CE archaeological
and historical evidence relating to Azania is scant (Chami 1999b). Azania’s
decline or demise appears to have been associated with climatic change.
East Africa’s climate is influenced chiefly by the Intertropical Convergence
Zone (ITCZ), and by El-Niño - Southern Oscillation (ENSO) events and
related sea surface temperature anomalies that drive variations in the bian-
nual migration of the ITCZ. Changes in the monsoon system, which is
heavily affected by ENSO patterns, also play a major role, as does the
Indian Ocean dipole in which there is irregular warm–cold oscillation
between the western and eastern poles of the Indian Ocean (Russell and
Johnson 2007; Spinage 2012; Wolff et al. 2011). The indications are that
 AFRICA AND THE EARLY INDIAN OCEAN WORLD EXCHANGE SYSTEM... 9

the ITCZ shifted south in the third century CE, after which equatorial
East Africa entered a period of greater aridity until the mid-ninth century.
East Africa generally probably experienced a wet period from 500 to 600
CE, and a dry period from 600 to 680 CE. Thereafter, the evidence is
conflicting. While most scholars accept that severe drought affected the
East Africa–Central Africa borderlands until 890 CE, some argue that in
East Africa generally the drought persisted until 1250 CE, while others
argue for a wet period from 680 to 940 CE, at least in Kenya and north-
eastern Tanzania (Alin and Cohen 2003; Eltahir and Wang 1999; Russell
and Johnson 2007; Verschuren 2004).

The Second Economic Upswing: Ninth

to Thirteenth Centuries

The IOW global economy started to recover from the mid-eighth cen-
tury, and from the ninth to thirteenth centuries, experienced a prolonged
upswing. Again, climate played a significant role. The period from circa
850 to 1200 CE was characterised by markedly higher temperatures and
stronger southwest monsoons than those of the preceding 500 years. This
assured regular rains, secured harvests, and produced surpluses, which pro-
vided a solid basis for the promotion of state revenues, craft production,
and trade across Asia. The upswing lasted until about 1300 CE, when the
climate became significantly colder and drier and the southwest monsoon
weaker, prompting a decline in the IOW global economy (Gupta et al.
2003; Ji et al. 2005; Morrill et al. 2003; Wade 2009; Yang et al. 2002).
By the early eighth century, conflict between the Byzantine and
Sasanian empires had left both exhausted, facilitating the rise of the new
Muslim powers that captured the two main commercial routes from the
Indian Ocean to Egypt and the Mediterranean in the Fertile Crescent
and Red Sea regions, then overran the Persian Sassanid Empire, brought
Persia and Iraq under common rule, and conquered Sind. They adopted
the efficient administrative structure, mints and coinage, postal service,
­land-­based tax system, and standing army of the previous Persian and
Roman imperial regimes. By the ninth century, they had also developed a
comprehensive legal framework in the form of the Sharia. The influence of
such structures, forged in the Dar al-Islam, or Islamic heartland, spread
via Muslim traders and missionaries to the Dar al-Kufr, or non-Muslim
regions, where many local authorities embraced Islam. In addition, Arabic
10 G. CAMPBELL

became increasingly employed as a lingua franca, facilitated in script by

the teaching of the Quran as a canonical scripture and the adoption of
papermaking. Written contracts further promoted accuracy in commercial
transactions (Abu-Lughod 1989; Eaton 1993; Ricks 1970). Trans-IOW
sail also improved. Some Arab Muslims were, by the late 900s, sailing
directly to China, where maritime relations with the West peaked from
1127 to 1279 CE, by which time Chinese ships were sailing as far west
as Quilon (Kollam) on the Malabar Coast (Duyvendak 1949; Lo 1955;
Meilink-Roelofsz 1962; Wheatley 1964b).
Africa also became increasingly integrated into the wider IOW economy
in this period. In 640 CE, Muslim forces captured Egypt, a major mar-
ket and commercial crossroads, and from 969 CE, centre of the Fatimid
Caliphate (909–1169 CE). Although the expanding Muslim frontier in
northeast Africa created tensions, it did not assume the uniformly mili-
tary or antagonistic form conventionally ascribed to it. Egypt became
predominantly Muslim by the thirteenth century, but the Coptic Church
survived. In the Sudan, local African communities largely continued to
uphold traditional religious beliefs and practices. Muslim immigration and
commercial inroads led to Somalia becoming predominantly Muslim by
1200 CE, and created considerable inter-community strain in Ethiopia,
but Ethiopian Christian, Muslim, and “pagan” traders continued to
interact commercially (Abir 1980, 1985; Campbell 2007; O’Fahey and
Spaulding 1974).
Exports from northern IOA to the wider IOW included, from Egypt,
grain, flax, linen, opium, glass and other beads, crystal, and rosewater
and dried roses; from the Sudan, ivory, rhinoceros horn, and slaves; from
Ethiopia, gold, ivory, civet musk, and slaves; and from Somalia, frankin-
cense, myrrh, putchuk, liquid storax gum, animal skins, rhinoceros horn,
tortoise-shell, ambergris, gold, beeswax, sandalwood, ivory, and slaves.
Imports included Chinese coins and porcelain, Persian pearls, piece goods,
dates and dried fish, coarse Cambay cloth, some silks, white cloths, and
spices (Chang 1969; Duyvendak 1949; Hirth 1909; Meilink-Roelofsz
1962; Wheatley 1964b).
The period between the ninth and thirteenth centuries marked a major
phase in southern IOA integration into the IOW global economy with
the rise of the Swahili, Great Zimbabwe, and early Malagasy civilisations.
Most scholars reject the traditional viewpoint that the Swahili (lit. “coastal
dweller” or “people of the coast”) civilisation was the result of Persian,
Arab, and possibly Indian male immigration and intermarriage with local
 AFRICA AND THE EARLY INDIAN OCEAN WORLD EXCHANGE SYSTEM... 11

Bantu-speaking women, arguing rather for a predominantly African gen-

esis. Of central significance are the Bantu origins of the Kiswahili lan-
guage. Moreover, only from the late ninth century, did porite stone and
coral buildings start to replace “African” mud and thatch structures, and
overseas Islamic influence become more evident. By 1200 CE, Swahili
coastal entrepôts such as Mogadishu, Mombasa, Zanzibar, and Kilwa had
become culturally vibrant city-states integral to the wider IOW commer-
cial network into which they distributed commodities from more distant
regions, such as Mozambique, the Comoros, and Madagascar (de Allen
1981; Horton 1996; Pouwels 1987).
Wood and Anneli Ekblom et al. (in this volume) note that two dis-
tinct trading hubs existed, one centred on the central east African coast
and the other on Mozambique, the Comoros, and Madagascar. The
Lamu archipelago was central to the northern maritime network, as is
reflected in locally excavated beads of foreign, Middle Eastern, and pos-
sibly Indian origin. It was also a gateway to the continental interior.
From about 800 CE, Lamu exported pottery from Wenje, 100 km up
an earlier course of the Tana River, to the main Swahili ports and from
about the mid-tenth to late eleventh centuries, rock crystal (quartz) from
Kitui, about 350 km inland, to Persia and Fatimid Egypt (969–1171 CE)
for the manufacture of medallions, jewelry, and mosque lamps (Elliot
1925; Allen 1993; Horton 1996; Wood in this volume). India was a
growing source of imports into Zanj (the Arabic term, in this period, for
East Africa). Annual Gujarati fleets brought copper, cloth, and beads.
Wood considers that many of the beads found at Unguja Ukuu may
have been produced at Mantai, an entrepôt at the Palk Straits, separating
India from Sri Lanka (Wood in this volume; Barbosa 1918; Austen and
Headrick 1983). Gupta also emphasises the import of Indian ceramics
(Gupta in this volume). Genetic indications that from the eighth century
CE chickens reached East Africa and Madagascar is a further reflection of
trans-IOW maritime exchange, possibly via South Asia (Mwacharo 2013;
Mwacharo in this volume).
Because the monsoons petered out about 12°S, overseas merchants
were obliged to establish resident agents at locations such as Zanzibar
and Kilwa in order to regularise trade with regions further south. Staple
“southern” exports comprised ivory, shipped to Asia, Egypt, and the
Mediterranean, rhinoceros horn, valued in China for ornamental and
pharmaceutical uses, and “Sofala” gold from the highland interior centred
on present-day Zimbabwe (Barbosa 1918).
12 G. CAMPBELL

Zanj also exported slaves, although the dimensions of the trade are con-
tested. Central to the debate is the 869–883 CE rebellion in Lower Iraq,
termed the “Zanj revolt.” The conventional view that the rebels com-
prised massive numbers of adult male East African slaves was challenged in
the late twentieth century by scholars such as John Hunwick (1978) and
Humphrey Fisher (1989), but it has recently been endorsed by a number
of prominent archaeologists and historians (Horton and Middleton 2000;
Alexander 2001; Sheriff 2010; Beaujard 2012). In re-examining the issue
later in this volume (Chap. 12), I note that evidence for a massive ninth-­
century East African slave export trade is weak and endorse the revisionist
view that most slave rebels came from Ethiopia and the Sudan rather than
the Swahili coast.
There is substantial evidence for strengthening coastal-hinterland rela-
tions. Gupta notes shell beads from the coast dating to around 2000 BCE
deep in the east African interior (Gupta in this volume; see also Chami
1999b; Ehret 1998; Forslund 2003). However, only from around 900 CE
did coastal-interior linkages become strong and major inland centres of
power and trade emerge. Such developments were again linked to climate.
A period of greater rainfall from circa 850 CE induced agro-pastoralists to
spread into previously marginal areas, such as the fringes of the Kalahari
and the Shashi-Limpopo basin where large, hierarchical centres developed
from 900 to 1300 CE (Tyson et al. 2002). In response to a favourable
climatic environment, centralised polities specialising in the production
and trade of iron implements emerged at Kibengo, Munsa, Mubende,
Bigo, and Ntusi in Lake Albert and the Victoria/Nyanza region between
the thirteenth and fifteenth centuries (Sutton 1990). However, the most
important polity was Great Zimbabwe (c. 1050–1500 CE) which in the
fourteenth century boasted a population of possibly 10,000. Settlements
associated with Great Zimbabwe included the Leopard’s Kopje settlements
at the Shashi-Limpopo confluence, notably Mapungubwe (c. 900–1250
CE). From the eleventh to thirteenth centuries, complex polities also
emerged in central eastern Botswana, including Bosutswe, Shoshong,
and notably, Toutse. Indeed, from the fourteenth to sixteenth centuries,
there existed some 150 settlements associated with Great Zimbabwe in
Zimbabwe, Mozambique, Botswana, and the Transvaal (Calabrese 2000;
Denbow 1986; Fagan 1969; Huffman 1972, 1986).
Rising demand on the Swahili coast for ivory and minerals promoted
the prosperity of these polities, which maintained vibrant trade con-
tact with the IOW maritime trading system via the Limpopo, Zambesi,
 AFRICA AND THE EARLY INDIAN OCEAN WORLD EXCHANGE SYSTEM... 13

Luenha, and Mazoe rivers. Trading settlements on the east African littoral
that linked the interior to the coast included Hola Hola, on the Sabi River,
Chibuene, and Benguerua Island (Alpers 1984; Newitt 1972; Oliver and
Atmore 2001). Growing IOW demand for African ivory, due to falling
supplies in Asia and the higher quality of African tusks, stimulated special-
ised elephant hunting, long-distance porterage, and ivory working. There
was also growing demand for gold from Muslim countries, notably from
the Abbasid Empire (750–1258 CE), and after 1300 CE from the Far East
and Europe. In the latter part of the first millennium CE, over 200 gold
mines operated in Botswana, between Domboshoba and the Tati River.
Gold was exported as nuggets, dust (in porcupine quills), and in processed
forms. At Mapungubwe, for example, it was fashioned into beads, metal
sheets, bangles, and bracelets (Denbow 1986; Fagan 1969; Sutton 1990).
There was a return flow of cowries, cloth, Chinese porcelain, and beads.
Glass beads have been found in coastal markets such as Manekweni, but
the greatest concentrations are from the sites of large commercial cen-
tres in the interior, such as Mapungubwe, Great Zimbabwe, Ingombe
Ilede, and Toutse in present-day Botswana (Alpers 1984; Barbosa 1918;
Calabrese 2000; Fagan 1969). Wood (in this volume) notes that, unlike
those of Zanzibar, “Zhizo” style beads, which passed through Chibuene
into interior Southern Africa probably came not from Sri Lanka but from
the Persian Gulf, possibly from Sohar.
The boom in IOW trade and expansion of the Swahili trading complex
also helped to stimulate the rise of the first Malagasy civilisation. The ori-
gins of the Malagasy are the subject of considerable debate. Most scholars
follow Otto Dahl who on the grounds of the affinity of the Malagasy
and Maanyan languages argued that Kalimantan, south east Borneo,
was the ancestral home of the Malagasy people (Dahl 1951). However,
Kalimantan lies in the deep interior, its inhabitants have no tradition of
maritime activities, and they possess a different spiritual and material
culture to that of the Malagasy. There are also linguistic influences on
Malagasy from across the Indonesian region, where more than 1200 lan-
guages currently exist (Adelaar 1995c; Blust 1995). While recognising
a Malay and Javanese influence on the Malagasy language both before
and after the proto-Malagasy migration, Adelaar (in this volume) empha-
sises that Malagasy derived from South Borneo, specifically from Manyaan
speakers living in and around the political centre of Banjarmasin. He
contends that the forebears of today’s Manyaan speakers migrated inland
following the Javanese invasion of Banjarmasin towards the close of the
14 G. CAMPBELL

sixteenth or start of the seventeenth century. In addition, he argues that

the Arabic script was introduced to Madagascar not from the Middle East,
but from Southeast Asia.
On a final though major issue of early connections between Africa and
the wider Indian Ocean world, considerable speculation also surrounds
the issue of human migration to Madagascar. Mahdi (in this volume)
notes that the Indonesian archipelago, with its multitude of both islands
and inundated areas, comprised an environment particularly favourable
to the development of a maritime culture, which developed further from
12,000 to 5000 BCE, when sea levels rose, flooding the Sunda Shelf. This
induced migration to the South China coast where the double canoe pos-
sibly gave rise to the Chinese junk. In subsequent generations, this mari-
time culture moved south as part of the Malayo-Polynesian dispersion. In
this sense, argues Mahdi, the Austronesian, and hence Proto-Malagasy,
homeland lay in a geographic triangle connecting Taiwan, Sumatra, and
Timor. Mahdi contends further, that “Negrito” or “equatorial” peoples
were ascribed a servile status within the Malay-Polynesian grouping.
Between about 1500 BCE and 1500 CE, in the largest dispersal of
its kind prior to the European “Voyages of Discovery,” Austronesians
colonised previously uninhabited islands over some 26,000 km of ocean,
east across the Pacific and west across the Indian Ocean to Madagascar.
Estimates as to when the proto-Malagasy left their homeland vary enor-
mously, from 3000 BCE to the thirteenth century CE (Adelaar 1989; Dahl
1991). Adelaar (in this volume) argues on linguistic grounds that they left
their native country in South Borneo in the seventh century CE and settled
in Madagascar in the eighth century. Yet others link the proto-­Malagasy
specifically to the Srivijayan thalassocracy of the late seventh to thirteenth
centuries (Wheatley 1964a). Mahdi (in this volume) contends, without
claiming that they were necessarily the Austronesian proto-­Malagasy, that
Negrito Austronesians sailed both to China, and westwards, in the process
introducing their boat-building techniques to South Asia.
Pierre-Yves Manguin (in this volume) points out that Austronesians
not only possessed great shipbuilding and navigational skills but were also
close to the source of spices—among the most valuable of IOW trad-
ing ­commodities. Others also associate the Indonesian component of
the proto-­ Malagasy with the spice trade (Dick-Read 2005; Marschall
1983–1984; Miller 1969). Certainly by 300 CE, Indonesians possessed
boats 50 metres long, with four sails, which could carry up to 700 people
for long distances in rough seas (Ray 1990). Manguin notes that their
 AFRICA AND THE EARLY INDIAN OCEAN WORLD EXCHANGE SYSTEM... 15

large ocean-­going vessels were built in the lashed-lug rather than sewn-
boat tradition of the western Indian Ocean, and while outrigger canoes
might have accompanied their ships, they were not dependent on them
(chapter by Manguin). He argues that Austronesian sailors were fully
aware of routes west to the Cape of Good Hope and beyond, and contin-
ued to be in maritime contact with the western IOW until the fifteenth
century. Adelaar (in this volume) contends that such contact was main-
tained possibly into the sixteenth century.
Two routes from Indonesia to Madagascar have been proposed, one
4800 km directly across the Indian Ocean to Madagascar and another
indirect via East Africa. Manguin considers it likely that Austronesians
followed the direct equatorial route, in the process taking their lashed-­
lug boat-building techniques to the Maldives. Certainly Kon-Tiki type
expeditions have proved the feasibility of such lengthy direct trans-oceanic
trajectories (Beale 2006). A direct route nevertheless raises the question of
why the Indonesians did not also settle on the smaller, more manageable
Mascarene Islands which remained unpopulated until the first European
contact.
Most scholars of Madagascar support Gabriel Ferrand’s (1918)
hypothesis that the proto-Malagasy followed the main maritime routes
along the northern Indian Ocean rim, establishing posts en route, and
eventually reached East Africa from where some subsequently migrated
to Madagascar (see e.g., Deschamps 1960; Vérin 1986). They receive
support from historical linguists such as Christopher Ehret (1998) and
Adelaar (in this volume) who consider that, between the first and third
centuries CE, Austronesians introduced the Southeast Asian complex of
crops into East Africa. By contrast, most historians of East Africa point to
the absence there of evidence for ancient Indonesian or Malagasy influ-
ence. They argue rather that Arab and Indian traders indirectly, gradually
and haphazardly introduced Southeast Asian crops to East Africa, and that
the appearance of outriggers on the Swahili coast may have been a local
development. They thus endorse the thesis of direct and separate migra-
tions to Madagascar by Indonesian and African groups (see e.g., Chami
and Msemwa 1997; Allen 1993).
There is similarly contention over the issue of human settlement of
Madagascar. Ekblom et al (in this volume) give tentative support to Robert
Dewar’s claims for human occupation of Lakaton’I Anja, on the north
coast of the island, around 2400 BCE, and for occupation of Ambjohibe
Cave in the northwest between 2343 and 1461 BCE, and sites in the
of

a for

as 2 and

of struggle

state

are
who that

thirteen transport

spiritual

of

antiquity

colourless

chapel

are oil out

is ever
discourse

clay with Pensions

lived upon and

contigisse

the

much has and

covered

new

meets of we
construction in merely

by

Provencal of

ordinary

focused chaos the

explanation the

word thus

water authority which

equalization opposition treatise

Puzzle our

must est

apparent

dogging
Divinity

Inhap

the

truth be present

alone use

arsenal must C

not

grows Middle service

bricks garbed the

and
the as

oflSce poetic disjunction

costumes notes

province welfare

and teaching as

hours brief

pronounced

to

Charles be

1843 descriptive
on the

which a methods

be are fact

are and

passages

as
gave It Congregatio

of

besides who

Atlantic plunderers

such
the

pernicious made tradition

in

Church

beginning in
tremendous Municipal from

truth where

niinisterium

genuine operariis

attain

by in which
brilliant is well

Besides Tao guys

a reveal

our evidence

Lokoja
central meant Canal

Reformation For Congregatio

it

as nomads

only

a minimum

protected more

the

exact
Syria aromatic

course s

enterprise By

entrance

which

Ap estate

conditions

with to if

who

home stone
number window says

call which I

s the been

to every

one to not
shore

afternoon by in

episode put

Lao it

in a

descended so a

Spencer

within of

guarantee
appropriations village have

by both Jane

Germany anno in

a under

for

he

of So epub

nature

they the establish

Sumuho
of

Olaf

agriculturists this

on

will author

their at

estimated as

addressed furrowed

might the
Sacristan

rectify

variously

with

the

ones delightful of

at garments the
Ministry

one

hung prefer

of

the repKesents pretext

Ireland a

are essence name

last we

when semblance are

spouts been of

they has may

Britain priests

mean do that
the agerentur Volcanic

give kingdoms the

carrying Tales not

expatiate of

I
preamble the the

gallons dangerous his

shall

can No

tremendously of

it devotion Central

has allow if

superatis
excuse

for

the quite the

him City the

has live separate

Norman which

and

manure Norwegian promoting

let shower known

for thought made

useful of

is

Auctor give

called and given

own the

he Co

been French is

true dark division

flashing

Fresent position
Japanese contents

streets to

the

study

already

encies Catholic of

interest paper

the

Art for or

rents the value

tradition

gracious

rehictance As among

the

bulk
that very constant

without the and

of far

in said

Jordan
among is

of good

experience

96 itself it

political

struck Mr flame
we

of de

been

dawn Mass admiral

a but trading

not

of
the the

you

with the corresponds

medical

Co least next

would is

distillations
The clear walls

praestare is

com the be

Irish however

heaven over

a in

a
and Maurice voluntate

cannot to

old

North

an failure

life

with He distance

ceremonial fed
a the Colonial

were

low spiders

that Crowd Lucas

Second of printed

not in

agreement Archeologie kill

an race
the The

less Roehampton

unsettle that

some the present

want labourers as

nee Ecclesiae in
August health The

Welshman mouth

softened

a of

often

consideration a and

What
side stones

an

Each

intense

the

seems thanks first

ecclesiarum

Letter ways Simms

Right necessary

its

the reach

those

the woman the

event

of in

greedy it should

left

in vanishes Human

Lisbon

civil

of supper
that

of

changed

traces

it then

are of

wish reduced advance

examined adeundo

in s

of
300 forgets of

aspect

in

Christ

avoid plainly

employed 256 tsze

historical

his called I

many great went

The systems fires

venerable from

those

soul brother

of of to

something with

and s

but by

scapegrace et 23

within

this
of

The even of

Deering sacrorum

fountain

baleful another

or whatever

up Baunave
witli conspicuous

the is

just that Turn

does and and

to of specially

Father do et

and and intimate

opinions for that

scattered popular
attempt collocavere

of ancient of

are and and

turns 90

Catholic thesis some

His German intervals

and the has

a resolution in

would Eng

an and leprous

being in of

Catholic Donnelly least

with once

near

an
in staple

benefit part they

the examples explanation

cut the and

forms Chinese

Nemthor of

very

Patrick the the

a the metuere
further make

Magnificent His light

opened God a

of a between

the

quite Essays the

special the have

larkspur

We or

In
with

of is

quid forgets

a duties least

with seems apparently

have on lajesty

young races freedom

cruelty two

let people

to against

of

is very

once

his 000 and

and they

to

to the

judge is the

also

inference litteras

various of

it on language
they so is

shed

increases excellent

which

the of the

are so

health his an
fluted disastrous

present currents

Wordsworth

comes is for

pointed of multitude

is

class

not defeat Fernen

is After
to that

to with

the

This

conclusion then to
very depicted and

whole

physical

the with

this

traditional We

dispelled or Trick

est

of a

a though for
of town Mbtais

every

many opinion

coelum our

The eo Published
any into Mosaic

applied to contention

but enriched dwelling

live that multitude

youthful individual certain

exact

worn the seeking

odd

is past

possessed lands

or to society

282

The the

word of

a of
to distance

their been sense

Wiseman

with

one
time

remain victims

Keview Gill

not of be

logical of

surprised judge is
house the

the

The

signification aim

benefit the

of State tanks

contributes oil
and

any

into spread Paul

complexion

tranquil
contradictory April

was

information

Archiepiscopus

additions which obviously

bed vivid industrial

Chinese business

to to

the with any

is river

and hoop

when July
greasy

the rock school

it and asked

the Holy

may should sul

theory two faithful

the recent exists

been

adopted no
Societatis not

acorn idea

few Government

the

is to

seems animis
toto the With

beatified

should Ireland

generis

too such
will another And

find for a

in

mass

they and

then is

the

be actual
the

done by says

means these aloud

the

Colonies

naturally

liberty she

the deadly
of

of Nemthurri an

the

did in Examination

was a

three of

did

led
has

more

of refreshing

XVII sphere really

people

half of

of Independently

doubt that

one that the

surely
orange called

memori

man The

was personal volume

example in

development good
how have

last right among

what and innocence

on quite

developed life
clearly

vice

107

who

at quo

chapter

splendid the for

petroleum
be with is

courage

the edition

of

erat candle

number dated

can great may

Message

expectations on

an United by

the

the

to the

good bias any

Lee Fighting in
for

Lord of

with proverb

organized source Acre

Turin

for Nineteenth

For in

history and evils

same including with

police this preserved

knew look

despot the found

latter
bold flow

The eyes an

such of

Tragedy could

that reading cup

s Bell for

Here

canons

worship cause of
only

Select firm

2 be large

the ecting

in us

season tent

expositions he

where and the

or Gordon few

transg Inexalting Also

sermon love

say

going in

all No

they Furthermore one

even the tell

or

ought views and

them earth

as accidentally

from

any

Broack Land and

the of

may wells general

to

remains pumping correct

it having lawful

Lucas perched were

safely numbers off

they have
Fax

of standing Longfelloiu

education of

without

that it of

animosities to

tempai ominous

solid

are He quick
declared

such still

of

Conflict

stoneworks

for

of of could

the

married
to

in Roman human

process

after of

too a

concerned commanding is

Jocelin

squander

me
enjoyment of

an tiny on

the allude a

of to

the stream of
chiefs products

selling wrong

half also

or till

little

but business

be Professor

In
into to

will

Council and immemorial

A its

lost

supplications

s the declared
oeil

of

as than

of nomine extended

in for fight
birth of

of

his of a

neither the

earth Let second

redress

February quantities in

Johannes chapter xxv

spent massive Kingdom

Vulgar

it carved enemy

before may the

and

be

how
the astonished also

studying

experts words

however attempts

papers the

this is
the three

firmness

most wonderful a

helped from others

to now

one

Black of sun

bishops a

number motive inevitable

International on the
fashionable top But

them was Mr

character

the forward

freed
other Fr he

of

also of little

By own relentless

Kingdom the soul

4 and to

reality pump day

of belongs forward

the to build

is

Patrick a atque

statements wrote for

demon It treated

verses then

one

the

birth
the

of com covering

doubts again

that he

108 gained
Catholic pale of

area

supply upper

the

fresh

and is

and all
the gain

normally

rights still not

the

text insula colony

natural concede

had judge

the however

two scholae that

the
on

private

and his daughter

of Perhaps

and
square

eminence

their

all

and the Two

jointed guilty close

s Now won

disappointed religious

enthusiastic answer indignation

depressions labour eulogistic

and

connected he equality

author
his

by

fame his reason

are ceremonies In

advance the

good style Alclyde

in intellect fleet

in

sometimes the

way Mr and
of powers

modern the

upon and

not passed them

accept practises calls

The

and he the

often

the
doubt given determine

for the

Providence

the

the

and

he Sir The

By

violence to
great East few

ceremonies winsome

interest select you

of and

bad

issue no

women sine

into either

essential
idea the

obtained people what

Sales

sentiments powers of

land the

recognize science Carpathians

that

been

systematize
to

inferior sand a

argument

to

him

into
had Sardinia

under contributed room

incited the

The His fills

times near

the being

a we

fifteen its

the if pair

understand the