Austronesians in East Africa

7/28/2019 Austronesians in East Africa

1/12


2/12


3/12

239

Evidence for the Austronesian Voyages in the Indian Ocean

Chapter 19

Evidence for the Austronesian Voyagesin the Indian Ocean

origin from the Barito-speaking area of southeastBorneo;

f) the Malay selements on the East African coasttransported captured mainland African popula-tions from the Sabaki-speaking area to Madagascar,primarily for agricultural labour, between the hand seventh centuries ;

g) that other Island Southeast Asian peoples may alsohave followed these established trade routes toEast Africa, accounting for a residue of non-MalayAustronesian items in the Malagasy lexicon;

h) that the Malay impact on Barito society was indi-rectly responsible for the evolution of the Samalicpeoples, the sea nomads of the region betweenBorneo and the southwest Philippines;

i) that similarly, on the East African coast, the transferof nautical technology to coastal Iron Age cultiva-tors stimulated the development of Swahili mari-

time culture;j) that the expansion of Arab shipping in the IndianOcean from the tenth century onwards obscuredthe Austronesian origins of local seafaring throughthe replacement of boat types and maritime termi-nology;

k) nally, if the Indian Ocean was criss-crossed byexperienced Austronesian navigators from anearly period, then selement would be expectedon many Indian Ocean islands. Although mostislands were apparently unoccupied at rst Euro-pean contact, they may still have been reached byAustronesians and that more extensive archaeo-logy will reveal this.

A story this complex can only be summarized in thepresent synthesis, which aims to direct the reader toplaces where the argument is made more fully andonly presents detail in areas so far not covered inmaterial published or in print.

Roger Blench

Despite the presence of undoubted Austronesianspeakers on the island of Madagascar, the Austrone-sian world is very much characterized by the Pacicand the dramatic narrative of its selement. Althoughthere is credible evidence for the presence of Austro-nesians in India (Mahdi 1999), Japan (Summerhayes &Anderson in press), in the Arabian Gulf and in North,Central and South America1 the present-day absenceof Austronesian speakers has tended to relegate theseremarkable voyages to footnotes. Similarly, the move-ment of Austronesian navigators across the IndianOcean still excites lile comment, in part becausearchaeologists and prehistorians tend to keep to theirappointed spheres. As a consequence, standard textsstill repeat an increasingly outdated narrative. Recentresearch has altered existing perspectives on the set-tlement of the East African coast and Madagascar(Adelaar 2006; in press; Blench 2007; 2008; 2009a,b;

Walsh 2007; Beaujard 2003; 2007a,b). It now seemslikely that:a) Madagascar was rst seled, not by Austronesians,

but by hunter-gatherers migrating from the EastAfrican mainland prior to 300 ;

b) Madagascar was also reached by Graeco-Romantrading ships, which may have been tradingtortoise-shell with the resident foragers and wereresponsible for the translocation of commensalmurids;

c) there was regular contact between Island SoutheastAsia and the East African coast prior to 0 by anunknown people using outriggers and trading inspices;

d) aer a gap, precursors of the modern Malay es-tablished a raiding and trading culture based inselements along the East African coast from theh century onwards;

e) Malay ships had pressed crews of non-maritime


4/12

240

Chapter 19

The earliest selement of Madagascar

Archaeology in Madagascar has so far uncoveredno site earlier than the h century (Dewar 1996;1997). It is generally assumed that these early sitesrepresent incursions by trans-Oceanic populationsfrom Island Southeast Asia. But evidence frompalaeozoogeography suggests a very dierent paernfor the primary selement of Madagascar (Blench2007). Bones with cut-marks indicating butcheryare now dated to approximately 2300 , pointing tosignicantly earlier human selement of Madagascar.Burney (Box 5 in Godfrey & Jungers 2003) summarizesthe sequence as it now appears. Following the initialarrival of humans, large herbivores were intensivelyhunted, reducing their density and leading to anincrease in plant biomass. This increased the incidenceof res (intentionally set by hunters). This would

reduce the diversity of habitats, in particular extendinganthropic grasslands, resulting in a catastrophic lossof megafauna.

Part of the evidence for vegetation change is theearliest occurrence of Cannabis/Humulus pollen in thecentral highlands at Tritrivakely, at an interpolatedage of 2200 cal. (Burney 1987). More directly linkedto extinctions is a decline in incidence of the sporesof the coprophilous fungus Sporormiella, a proxy formegafaunal density (Burney et al. 2003). The mostsecure date for this change is 172040 (230410 cal.) but a less certain earlier date of 199050 (100cal. 110 cal. ) has been recorded.

Vegetational change, faunal extinctions andnumerous bones with cut-marks all point to the pres-

ence of low-density hunter-gatherers from about 400300 onwards. If so, then Madagascar may have beenrst colonized by low-density foraging populations,similar to the modern-day Hadza of Tanzania, whowould have crossed from present-day Mozambique.Beginning in the southwest, they would have rapidlyeliminated many species of large lemur, and begunthe process of habitat conversion in the highlands.Upon the arrival of Austronesian colonists, they werelargely assimilated, surviving as the residual hunter-gatherers, now known as the Mikea or Vazimba.

Rats, mice and Graeco-Roman contact

Records of the East African coast exist dating from theclassical period, although these are oen dicult tointerpret. The main sources are shown in Table 19.1.

In the Periplus, a rst-century seamans guide tothe East African coast, Madagascar perhaps makes anappearance in the text as the Great Island of Menuth-ias, a source of mountain tortoises. Menuthias had

no wild animals except crocodiles and was inhabitedby people who use dugouts and sewn boats and whocatch sh in wicker traps. However, Menuthias wassaid to be some 300 stades from the mainland (c. 67km) and from Madagascar to the mainland at theclosest point is 400 km, making the Pemba-Zanzibararchipelago another equally likely possibility. Thetext of Geography of Ptolemy was rst compiled around 150; the version has come down to us representsthe geographical knowledge of Byzantium in thefourth century. For Ptolemy, Menuthias is certainly

Madagascar, since it was east of Prason [=Delgado?],a promontory south of Rhapton, itself south of Rhapta(Freeman-Greville 1962, 4).

Archaeology now conrms that the maritimecultures of the Mediterranean were reaching theEast African islands during the era of the texts. Juma(1996) reported nds of late Roman poery on Zan-zibar and Chami (1999a,b) recorded Roman beads inthe Ruji Delta. A possible proxy for Graeco-Romancontact is the spread of Eurasian commensal murids inMadagascar. Both Raus raus andMus musculus arecommon on the island today and have penetrated evenremote forest areas, where they out-compete endemicsmall mammals (Ganzhorn 2003; Hingston et al. 2005).Vasey2 & Burney (2007) show that between 2480 and

1760 at Andrahomana Cave (in the southeast) pro-portions of endemic mammals such asMacrotarsomyspeeri in bone assemblages decrease dramatically,while both Raus raus andMus musculus rise sharply.Despite some stratigraphic mixing, a clear paern offaunal turnover still emerges. They claim that datapresented here show that these taxa were introducedprehistorically by the earliest human selers but thisis unlikely. A probable source would be either theGraeco-Roman traders or other players in the Arabianend of the Indian Ocean network, who were certainlyactive during this period. It is even possible such shipswere trading with the resident populations rather thanpicking up turtles directly.

Early Austronesian contacts with theEast African coast

Textual evidenceThe literature on the earliest Austronesian contactswith the East African coast has been somewhat con-

Table 19.1. Classical sources for the Indian Ocean.

Book Date Source

On the Erythrean Sea


5/12

241


fused by conation with the evidence for later set-tlement of Madagascar. But it is now clear that theseare two quite separate events. The East African coastmay have been visited by Austronesian marinersfrom an early period, probably prior to 0 (Adelaar2006; Blench 1996; 2008; 2009b; Beaujard 2007a). Pliny(Natural History, Book XII: cap. 42) describes the Trog-lodytae, who buy cinnamon from their neighboursand carry it over vast tracts of sea, upon ras [eitheran actual ra, such as are used in other parts of theIndian Ocean or an approximate description of anoutrigger]. Moreover, they choose the winter season,about the time of the equinox, for their voyage, forthen a south easterly wind is blowing; these windsguide them in a straight course from gulf to gulf, adescription of the use of the equatorial counter-currentfor seasonal travel.

There is no direct archaeological evidence for

these early period contacts, but other indications arehighly suggestive. These draw from textual, maritimetechnology, plant and animal transfers, disease andother aspects of material culture.

DiseaseGeneral: Populations isolated from one another forlong periods usually develop immunities to particulardiseases and become carriers of subclinical pathogens.When they encounter wholly unfamiliar populations,disease transfers can oen have disastrous conse-quences. The most well-known of these stories is theimpact of European voyages on the New World; highmortality among Amerindian populations as a con-sequence of their lack of resistance to a spectrum of

pathogens. If Austronesians reached the East Africancoast, then it is probable that similar disease exchangesoccurred, and indeed we have evidence for this. Itseems that African malaria reached southeast Asia atan early period, and by return elephantiasis spreadthrough central Africa.

Malaria: African malaria would have made a markedimpact on incoming voyagers from southeast Asia.The indigenous malaria in southeast Asia, Plasmodiumvivax, was probably introduced to New Guinea beforethe end of the Pleistocene (Groube 1993, 168, 171).However, African malaria, P. falciparum, is more viru-lent than that transmied by P. vivax and the resultant

illness, although non-relapsing, is oen fatal. As theparasite cannot survive within its host beyond theduration of its initial infection, it requires large hostpopulations moving rapidly for successful migration.When African malaria was carried across the IndianOcean its impact on island populations with limitedresistance was very serious. There is some evidence

in the archaeological record for demographic lacunaewhich might well represent the impact of incomingP. falciparum in the islands of southeast Asia and thePacic.

Malaria would have been equally serious forthose arriving on the coast with the intention ofremaining for a protracted period. Europeans in WestAfrica stayed oshore for three centuries aer traderst began, only venturing inland when quinine cameinto use in the middle of the nineteenth century. Pre-sumably any aempt to sele in East Africa wouldhave been similarly lethal in the rst phase. So theearly wave of Austronesians must soon have learntto limit their time onshore. The selement of Mada-gascar, with its disease-free highland areas, may wellhave been stimulated by high morbidity on the coast.

Elephantiasis: Elephantiasis, or lymphatic lariasis, is

an extremely visible condition which originated inthe Pacic or Island Southeast Asia. The thread-like,parasitic larial worms Wuchereria bancroi and Bru-gia malayi that cause lymphatic lariasis live almostexclusively in humans. Elephantiasis is, however,widespread in Africa and must have been broughta long time ago, because of its distribution andembeddedness. Laurence (1968) notes this infec-tion is placed in the area of evolution of the Malay-Polynesian-Malagasy [Austronesian] language-groupand it is conceivable that the disease was introducedinto Africa by movements of people belonging to thesame linguistic group. Elephantiasis must be transmit-ted by movements of infected people. As it happens,there is at least one piece of archaeological evidence

aesting to its antiquity in West Africa, a Nok terra-coa (Fagg 1977, pl. 121). Fagg (1977, 11) notes otherstatues with possible representations of the diseaseand Hoeppli (1969) gives some more recent examplesin African bronzes and terracoas. The Nok culture istypically dated within a window of 500 500.For elephantiasis to have moved across the continentin this way must have taken some centuries, pointingto a likely introduction in East Africa prior to 0.

Plant and animal transfersFrom southeast Asia to Africa: There are at least threestaple food-crops in Africa that are considered to comefrom southeast Asia and the Pacic (Blench 1996).

Plantains Musa paradisiacaWater-yam Dioscorea esculentaTaro Colocasia esculenta

These are deeply embedded in African staple ecolo-gies and are very unlikely to be recent introductions


6/12

242

Chapter 19

(Blench 2009b). Indeed, phytolith evidence for culti-vated Musaceae goes back to the rst millennium in the forest of southern Cameroon (Mbida et al. 2000;2001). On the East African coast, there is some evi-dence that Austronesian banana names were actuallyborrowed into Tanzanian languages. For example, theShambala and Bondei languages of the coastal regionhave (hu)ti for cooking bananas which may well relateto proto-Malayo-Polynesian *punti and Malagasyfontsy (Beaujard 2003). The presence of a wild-seedingdiploid (AA) banana on Pemba island is another clearanthropic introduction from Island Southeast Asia,although it is impossible to date (Williams 1949, 373;Simmonds 1966).

Taro and water-yam are similarly important inmany parts of humid Africa, although no immediatedate can be aached to their transmission (Blench1996). However, it is likely that all three plants were

brought to the East African coast directly, prior to 0.They were adopted by cultivators who then transmit-ted them westwards across the continent.

From Africa to southeast Asia: Crops known to have beentransported in the opposite direction to Asia are muchscarcer, but there is one very well-documented case,that of the winged bean, Psophocarpus tetranoglobus(L.) DC. The winged bean is so well embedded in theagriculture of southeast Asia and New Guinea thatearlier writers situated its domestication in this regionand evolved a series of unlikely hypotheses to accountfor the absence of wild relatives. However, Harder &Smar (1992) have argued conclusively for its domes-tication in east-central Africa, although no dates have

yet been aached to its movement across the IndianOcean. Another possible candidate is the Africanyam bean, Sphenostylis stenocarpa (Hochst. ex A. Rich.Harms). This was domesticated in tropical Africa butseems to have made its way early to southeast Asiaand India, where it is widely cultivated (Poer 1992;Poer & Doyle 1992).

Chickens are now known to have three centresof domestication, India, China and Island SoutheastAsia (Han Jianlin pers. comm.).3 African chickens inEthiopia and across a wide zone of West Africa descendfrom the Indian breeds. However, many of the chick-ens of eastern and southern Africa are derived fromIsland Southeast Asia and were not introduced via any

identied intermediary location. Whether these werebrought to Madagascar and transmied onwards to themainland is as yet unclear. However, Malagasy termsare borrowed from terms in the Bantu languages of thecoast (and indeed across the Bantu domain) notAustro-nesian. So it is likely that the chicken was established onthe coast by the time Madagascar was seled.

The Malay selement of the East African coastand Madagascar

Textual evidenceAlthough Malagasy is accepted as a member of Aus-tronesian, its precise genesis has been much debated.It belongs genetically to the Barito languages, todayspoken in southeast Kalimantan (Dahl 1951; 1991;Hudson 1967; Simon 1988) but has undergone con-siderable inuence from Malay (Adelaar 1995; 1996;2006; in press). One aspect of transformation of Baritointo present-day Malagasy is the presence of Malayterms for cardinal directions4 and other nautical terms(Adelaar 1996). This suggests either that the Baritotravelled in Malay ships as crew or a separate migra-tion of a Malay-speaking population (Beaujard 2003).Whether the Barito were crew by their own choice orwere pressed remains to be determined. However,

their previously non-sea-going culture suggests thelaer, as does the presence of various socially sensitiveMalay loanwords in Malagasy, arguing that the Baritowere not in control of the migration process (Adelaarpers. comm.).

Malay ships may not have been simply sailing toMadagascar but participated in an active raiding andtrading culture along the East African coast. Medi-eval Arab sources point to possible semi-permanentIndonesian trading outposts. Ferrand (1907) was therst writer to propose southeast Asian identities forthe islands mentioned in the Arab geographers. TheEast African coast was considered important enoughfor the Waqwaq raiders and traders from Sumatrato mount a raid on Qanbalu (an East African island

as yet unidentied) in 945, according to Buzurgibn Shahriyar, Book of the Wonders of India (Freeman-Grenville 1981). The Waqwaq seem also to haveseled on the Sofala coast in the early tenth century(al-Masudi, in Freeman-Grenville 1962, 14). Al-Idrisi,writing in 1154 suggests that the coastal Bantu didnot develop seagoing vessels for long-distance tradeuntil quite late:

The Zenjs [the people of the East African coast southof Cape Guardafui] have no ships for voyaging ... Thepeople of the isles of Ziibag [here Ziibag = WesternIndonesia] come to the country of the Zenjs in largeand in small ships. They trade with them and exportthe Zenj merchandise, for they understand each

others language. (Al-Idrisi, in Ferrand 1907)

As Hornell (1936) observed, the statement that theIndonesians understood the language of Zenj onlymakes sense if we assume there were Austronesian-speaking selements on the East African coast, notmerely on Madagascar.


7/12

243


Another piece of evidence comes from anunlikely source; the large canoes of Lake Victoria.Hornell (1928) undertook a detailed description ofthese canoes and showed that in a number of detailsof construction they closely resemble the small coast-ing vessels of Java and Madura. His conclusions areworth quoting in detail.

In view of these facts and of a number of otherconsiderations, prominent among which is the factthat the common shing canoe of the east coast ofAfrica, from Mozambique to Somaliland, thoughdiering in details, is unquestionably derived fromthe same type as the outriggers of Madagascar, andthat this type is known nowhere else than in Java, Ican come to no other conclusion than that Indonesianselements at one time existed upon the east coastof Africa at the time of the Indonesian colonizationof Madagascar; further that the origin of the canoes,equally with the double outriggers of the coast, is to be

traced to Indonesian culture exercised upon the Bantutribes of this region by Javanese selements along thecoast-selements subsequently obliterated in the sameway as was that of the Portuguese at Mombasa in theseventeenth century. (Hornell 1928, 3)

2. There are striking similarities in the tuning ofxylophones in Africa and southeast Asia, whereequiheptatonic scales occur.

3. Claims the xylophones could have been broughtfrom southeast Asia to Africa are falsied by theoccurrence of all stages in the evolution of the

xylophone in Africa. Such intermediate stages areconspicuously absent in southeast Asia where therst representations (at Borobudur) show frame-xylophones were already complete in design.

4. It is therefore more likely that the xylophone wascarried from Africa to southeast Asia, probably bythe Zenj enslaved by Malay raiders on the EastAfrican Coast in the seventhtwelh centuries.

Other musical instruments: Groanelli (1947, 173) rstpointed to the distribution of transversely-blownconches which occur only in some Pacic islands andEast Africa. Since end-blown conches are widespreadthroughout the region, this is suggestive but notconclusive. The at-bar stick-zither, sese, occurring in

both East Africa and Madagascar, similarly originatesin the Indonesian islands (Kaudern 1927; Sachs 1938;Blench 1984; 1996).

Crossbow traps: Lagercrantz (1950, 112) discusses thedistribution of crossbow traps in Africa. Figure 19.1shows a typical crossbow trap from the coast of Tan-zania, although they are also found on Zanzibar andalong the east coast of Madagascar. Crossbow trapsare found in large parts of Asia, but particularly inmainland and Island Southeast Asia (Mrite 1942).

The transport of mainland populationsto Madagascar

In addition to Austronesian inherited vocabulary, Mal-agasy has a considerable number of terms of demon-strably Bantu origin, but their exact source within theBantu domain has never been properly identied. Themost comprehensive aempt to consider this issueis Beaujard (1998), whose rich dictionary of Tanala

Figure 19.1. Ruvuma crossbow trap from Tanzania.

Swahili oral traditions recorded early in the twentiethcentury by Gray (1954) talk of a cruel people, theWadiba, who built quadrilateral houses and wereassociated with the introduction of the coconut palm.They are later supplanted by the Wadebuli, whoseidentication is not clear but may well be pre-Omanitraders. The Wadiba could have been the Indonesiansresponsible for the introduction of the coconut andits distinctive methods of processing characteristic ofcoast today.

Material-culture transfersTransfers of material culture from southeast Asia toMadagascar are numerous, as are paerns of socialorganization and aspects of religion. Traces of Indone-sian selement on the mainland are far less numerousas they have oen been overwrien by the subsequentspread of Arab culture. This section summarizes a fewintriguing and controversial cases, which also suggestthat inuences spread across the Indian Ocean in bothdirections.

Xylophones: The origin and spread of the xylophonehas been subject to considerable discussion, not all

of it well founded (e.g. Jones 1971; Blench 1982). Theshort version is as follows:1. Xylophones occur with any signicant time-depth

in only two regions of the world, sub-SaharanAfrica and Island Southeast Asia. All other occur-rences (e.g. in Eurasia and the New World) arerecent spreads.


8/12

244

Chapter 19

contains many etymological speculations. Recentobservations on the historical origins of Malagasymake it possible to develop a more denite model forthe origin of Bantu loanwords. Blench (2008) focuseson the terminology for domestic and translocatedanimals, and considers some other areas of vocabularyin less detail.

Swahili is the dominant language today along thecoast facing Madagascar as far down as Mozambique,with a variety of Bantu languages related to it in theimmediate interior. Even within Mozambique, anisolated language such as Ekoti appears to originatefrom selements of pre-Swahili speakers (Schadeberg& Mucanheia 2000). Swahili is part of a larger group ofcoastal lects oen referred to as Sabaki and Nurse &Hinnebusch (1993) contains both a history of Swahiliand a lexicon of proto-Sabaki reconstructions. Of theoshore islands, the most important are the Comores.

Walsh (2007) has recently studied the Malagasy termsfor wild animals and the great majority also derive,not from Austronesian, but from Bantu languages. Aswith livestock names, almost all are from Swahili andlanguages of the Sabaki group, not from the Bantulanguages of Mozambique, nearest to Madagascar.

This is somewhat contrary to expectation. Histor-ical accounts of the formation of the Malagasy popula-tion, such as Kent (1970) point to intensive interactionwith Mozambique coastal peoples. However, it seemsthat the earlier interaction of Malay/Barito crews withtheir coastal interlocutors must have been much morethan casual trade. To explain the dominance of Sabakilanguages in the Malagasy lexicon and the rich bor-rowings in the eld of ora and fauna, it must be that

substantial numbers of coastal Bantu were carried toMadagascar, and were primary observers of the natu-ral world. This might be because they were engagedto herd livestock or were captured and set to work toproduce food. Whatever the explanation, this pointsto intensive interactions between the Malay/Baritoand the precursors of the Swahili on the coast prior tothe selement of Madagascar proper. Malay nauticalterms borrowed into Swahili independently suggestsexternal inuences played an important role in theevolution of Swahili maritime culture.

The impact of the Malay on Sabaki coastal cultureThe Swahili peoples are presently identied by their

maritime culture and this is usually assumed toderive from Omani sources (Horton & Middleton2000). The Swahili, especially Zanzibaris, like to tracetheir ancestry to Oman, which functions as a prestigeorigin for cultural traits. But indirect evidence pointsto contact with Java as a major stimulus to nauticalevolution along the East African coast. We know from

the testimony of Al-Idrisi quoted earlier that the Zenjhad no ocean-going ships as late as the early twelhcentury, but that they were in intensive contact withSumatra. It therefore seems credible that the initialtransformation of the Swahili from land-based cultiva-tors to seafarers can be aributed to the Malay contact.Curiously enough, a parallel evolution occurred inChina; as Manguin (1980, 274) points out, China onlybegan to build an oceangoing navy aer contact withlarge southeast Asian vessels in the eighth and ninthcenturies.

Subsequently, nautical technology on the coastseems to have undergone a revolution, due to theinuence of Arab and Indian commerce and the rapidadoption of new cra. Jewell (1976) describes themany ship types on the waterfront at Mombasa, whichin the 1960s were still undertaking a circular voyagebetween the Gulf of Aden, the west coast of India,

Nossi B in northwest Madagascar and the Comoresand on to East Africa. When Vasco da Gama arrived,all these ships were constructed entirely withoutnails, an ancient practice characteristic of the entireIndian Ocean region. At any rate, from this periodonwards, the expansion of commerce stimulated thedevelopment of shipyards all along the coast and thevirtual replacement of whatever older shipping formswere in use, with the exception of the mtepe or sewnboat (Hornell 1941). As a consequence, it has becomedicult to reconstruct older Austronesian maritimeinuence, even as the disappearance of dhows in theyears since Jewells book was published have mademuch more recent trade paerns less researchable.

Islands in the streamRecent research points ever more strongly to earlyand persistent Austronesian contacts across theIndian Ocean, both via outriggers as part of the spicetrade and with the rigid ships of the Malay sea-borneempire. The later rise of Arab shipping has largelyobscured this narrative in the broader history ofthe region and had a negative impact on archaeo-logical survey. One of the more curious aspects of thearchaeology of trans-Indian Ocean voyaging is thatthere is so lile evidence for Austronesian (or other)incursions on the intermediate islands. The ability ofAustronesian navigators to nd very small islandsin large expanses of open ocean is well documented,

yet it seems that almost all the Indian Ocean islandswere uninhabited during the rst period of Europeancontact. Anderson (2002) in a survey of Pacic islandsin remote Oceania found no less than thirty whichwere apparently reached by Austronesian navigatorsbut were devoid of inhabitants during the period ofEuropean exploration.5 Explanations are various:


9/12

245


typhoons, disease, lack of sustainable food and watersupplies are all probable causes. The situation may bereplicated in the Indian Ocean: Mauritius or the Sey-chelles may have been reached, but then abandonedagain for a variety of reasons. Indeed, the computersimulation of trans-Indian Ocean voyages by Fitz-patrick & Callaghan (2008) suggests that it is veryunlikely the voyagers did not touch the interveningislands. One reason for their abandonment may havebeen the lack of easily exploitable food resources. Itis notable that European sailors tended to rapidlyconsume any readily caught resource (e.g. the dodo)

and bring in and release oen destructive species suchas the goat in order to ensure future food supplies. Inthe case of the Maldives, no Austronesian language isspoken there today. However, as both Hornell (1920,230) and Manguin (1985, 12) argue, constructionaltechniques in boat-building point unambiguously toearly Austronesian contact.

It seems increasingly likely that the lack of evi-dence for Austronesian landings is an artefact of thepatchy archaeology. Creating a checklist of possiblelanding sites or failed colonization aempts may helpto recover this intriguing period. Table 19.2 is intendedto highlight the extremely uneven record of archaeo-logy in this region and to establish the parameters of aprogramme to put their selement and colonization ona sounder footing. Mitchell (2004) has surveyed exist-ing knowledge of the prehistory of Africas oshoreislands and this summary is indebted to his resultsfor the relevant ones.

The southern Indian Ocean islands consist ofAmsterdam Island, Crozet Islands, Heard Island,Kerguelen Islands, Marion Island, McDonald Islands,Prince Edward Islands and Saint-Paul Island. Theseare generally extremely cold and windswept andalthough Austronesian selement is not impossible,it seems extremely unlikely.

Table 19.2. Possible Indian Ocean sites for Austronesian and other selement.

Name Possible Austronesiancontact

Possibleother contact

Aldabra Yes Arabs?

Bassas da India No NoBazarutoArchipelago

Yes Arabs?

Cargados Carajos Yes Arabs?Chagos Archipelago Yes Arabs?Comoros Currently part-seled by

Austronesian speakersNumerous

Diego Garcia Yes Arabs?Europa Island No DoubtfulGlorioso Islands Yes DoubtfulJuan de Nova Island Yes DoubtfulLakshadweepArchipelago

Yes Numerous

Lamu Archipelago Yes NumerousMadagascar Currently seled by


Maa Island Yes NumerousMaldives Yes NumerousMauritius Yes NumerousMayoe Yes NumerousPate Yes NumerousPemba Yes NumerousQuirimbasArchipelago

Doubtful Numerous

Runion Yes Arabs?Rodrigues Yes Arabs?Seychelles Yes Arabs?Shanga Yes NumerousSocotra Islands ? NumerousTromelin Island ? DoubtfulZanzibar Yes Numerous

Name Possible Austronesiancontact

Possibleother contact

Eastern Indian Ocean

Andaman Islands Yes NumerousAshmore & CartierIslands

Yes Unlikely

Christmas Island Yes but negative survey UnlikelyCocos & KeelingIslands

Yes Unlikely

Dirk Hartog Island Yes UnlikelyHoutman Abrolhos Yes UnlikelyLangkawi Islands Currently seled by


Mentawai Islands Currently seled byAustronesian speakers

Numerous

Mergui Archipelago Currently seled byAustronesian speakers

Numerous

Nias Island Currently seled byAustronesian speakers

Numerous

Nicobar Islands Yes; currently seled byAustroasiatic speakers Numerous

Penang Currently seled byAustronesian speakers

Numerous

Phi Phi Islands Yes; currently seled byThai speakers

Numerous

Phuket Yes; currently seled byThai speakers

Numerous

Simeulue Island Currently seled byAustronesian speakers

Numerous

Weh Island Currently seled byAustronesian speakers

Numerous

Sri Lanka Yes NumerousWestern Indian Ocean

Agalega Yes Arabs ?Albatross Island Yes Arabs ?


10/12

246

Chapter 19

Synthesis and conclusions

An accumulation of evidence suggests that there weredirect Austronesian contacts with the East Africancoast prior to 0 unconnected with the selement ofMadagascar. This includes textual sources, maritimetechnology, plant and animal transfers, disease andother aspects of material culture. If we accept Plinysaccount, then these were direct voyages by outriggerthat brought spices and took back Graeco-Roman man-ufactures, making use of the equatorial counter-current.

This should be distinguished from later contactwith the coast, initiated from the Malay-speakingarea which probably begins from the sixth centuryonwards. Despite the absence of Austronesian-speak-ing populations on the mainland today, it is likely thata trading and raiding culture existed between the sixthand twelh centuries. It would have been responsible

for the transport of the Barito- and Sabaki-speakingpeoples to Madagascar, originating the complex cul-ture that exists there today.

In archaeology, you nd what you seek, andevidence for the Austronesian crossings of the IndianOcean has hardly been sought. Further work shoulduncover evidence for both selements on the EastAfrican mainland and traces of temporary landings onthe many intervening islands apparently unoccupiedat rst European contact.

Notes

1. Contact between Polynesia and the Americas remainscontroversial. For California see Jones & Klar (2005),

Klar & Jones (2005), Anderson (2006) and responseJones & Klar 2006. For Central America, see Baudouin& Lebrun (2008). For South America, see Storey et al.(2007) and response Gongora et al. (2008).

2. Thanks to Natalie Vasey who forwarded a poster versionof this presentation and corrected my summary of it.

3. This section has been completely rewrien in the light ofas yet unpublished research by ILRI, based on a sampleof >500 chickens worldwide, presented by Han Jianlinat the SAFA meeting in Frankfurt, September 2008.

4. Albeit turned around 90.5. These were approximately: five in the Pitcairn-

Henderson Islands region, eight in the New Zealandregion, one in New Caledonia, one in Tonga, three inthe Cooks-Societies, eightnine in the Equatorial islandsregion, four in Hawaii.

Acknowledgements

I would like to thank Sander Adelaar, Atholl Anderson,Philippe Beaujard, Laurie Godfrey, Han Jianlin, Paul Sinclair,Mahew Spriggs and Martin Walsh for discussions relatingto various ideas in this paper, although none should be held

responsible for the nal outcome. A very early version waspresented in Manila at the XVIIIth IPPA in 2006, where myaendance was funded by the University of Uppsala.

References

Adelaar, K.A., 1995. Malagasy cultural identity from anAsian perspective. Bdragen tot de Taal-, Land- en Vol-kenkunde 151-III, 32556.

Adelaar, K.A., 1996. Malagasy culture-history: some lin-Malagasy culture-history: some lin-guistic evidence, in The Indian Ocean in Antiquity ed. J.Reade. London & New York (NY): Kegan Paul/BritishMuseum, 487500.

Adelaar, K.A., 2006. The Indonesian migrations to Madagas-car: making sense of the multidisciplinary evidence,inAustronesian Diaspora and the Ethnogenesis of Peoplein the Indonesian Archipelago. Proceedings of the Interna-tional Symposium, eds. T. Simanjuntak, I.H.E. Pojoh &M. Hisyam. Jakarta: LIPI Press (Indonesian Instituteof Sciences), 20532.

Adelaar, K.A., in press. Towards an integrated theory aboutthe Indonesian migrations to Madagascar, in AncientHuman Migrations: a Multidisciplinary Approach, eds.P.N. Peregrine, I. Peiros & M. Feldman. Salt Lake City(UT): University of Utah Press.

Anderson, A.J., 2002. Faunal collapse, landscape changeand selement history in Remote Oceania. World

Archaeology 33, 37590.Anderson, A.J., 2006. Polynesian seafaring and American

horizons: a response to Jones and Klar. AmericanAntiquity 71, 75963.

Baudouin, L. & P. Lebrun, 2008. Coconut (Cocos nucifera L.)DNA studies support the hypothesis of an ancientAustronesian migration from Southeast Asia toAmerica. Genetic Resources of Crop Evolution 56(2),25762.

Beaujard, P., 1998. Dictionnaire Malgache-Franais. Paris:lHarmaan.

Beaujard, P., 2003. Les arrives austronsiennes Madagas-car: vagues ou continuum? (partie 1, 2). tudes OcanIndien 356, 59147.

Beaujard, P., 2007a. LAfrique de lEst, les Comores et Mada-gascar dans le systme-monde eurasiatique et africainavant le 16e sicle, in LAfrique et Madagascar, eds. D.Nativel & F. Rajaonah. Paris: Karthala, 29102.

Beaujard, P., 2007b. East Africa, the Comoros islandsand Madagascar before the sixteenth century: on aneglected part of the world system.AzaniaXLII, 1535.

Blench, R.M., 1982. Evidence for the Indonesian origins ofcertain elements of African culture: with special ref-erence to the arguments of A.M. Jones.African Music6(2), 8193.

Blench, R.M., 1984. The morphology and distribution ofsub-Saharan musical instruments of North-African,Middle Eastern and Asian origin.Musica Asiatica IV,15591.

Blench, R.M., 1996. The ethnographic evidence for long-distance contacts between Oceania and East Africa. inThe Indian Ocean in Antiquity, ed. J. Reade. London &New York (NY): Kegan Paul/British Museum, 46170.


11/12

247


Blench, R.M., 2007. New palaeozoogeographical evidencefor the selement of Madagascar.Azania XLII, 6982.

Blench, R.M., 2008. The Austronesians in Madagascar andtheir interaction with the Bantu of East African coast:

surveying the linguistic evidence for domestic andtranslocated animals. Philippines Journal of Linguistics18 (Special issue), 1843.

Blench, R.M., 2009a. Bananas and plantains in Africa:re-interpreting the linguistic evidence. EthnobotanyResearch & Applications 7 (Special issue), 36380.

Blench, R.M., 2009b. Remapping the Austronesian expan-sion, in Discovering History through Language: Papers inhonour of Malcolm Ross, ed. Bethwyn Evans. Canberra:Pacic Linguistics, 3559.

Burney, D.A., 1987. Pre-selement vegetation changes atLake Tritrivakely, Madagascar. Palaeoecology of Africa18, 35781.

Burney, D.A., G.S. Robinson & L.P. Burney, 2003. Sporormiellaand the late Holocene extinctions in Madagascar.Proceedings of the National Academy of Sciences of the

USA 100, 10,800805.Burstein, S.M., 1989.Agatharchides of Cnidus: On the Eryth-raean Sea. London: The Hakluyt Society.

Casson, L., 1989. The Periplus Maris Erythraei. Princeton (NJ):Princeton University Press.

Chami, F.A., 1999a. Graeco-Roman trade link and the Bantumigration theory.Anthropos 94(13), 20515.

Chami, F.A., 1999b. The Early Iron Age on Maa island andits relationship with the mainland. Azania XXXIV,110.

Dahl, O.Ch., 1951. Malgache et Maanjan, une comparaisonlinguistique. Oslo: Egede Instituet.

Dahl, O. Ch., 1991.Migration from Kalimantan to Madagascar.Oslo:Norwegian University Press.

Dewar, Robert E., 1996. The archaeology of the early sele-ment of Madagascar. in The Indian Ocean in Antiquity

ed. J. Reade. London and New York: Kegan Paul/Brit-ish Museum, 471-86.Dewar, R.E., 1997. Were people responsible for the extinc-

tion of Madagascars subfossils and how will weever know?, in Natural Change and Human Impactin Madagascar, eds. S.H. Goodman & B.D. Paterson.Washington (DC) & London: Smithsonian InstitutePress, 36477.

Fagg, B., 1977. Nok terracoas. Lagos: Ethnographica.Ferrand, G., 1907. Les les Rmny, Lmery, Wkwk, Komor

des gographes arabes et Madagascar. Journal Asia-tique X, 433566.

Fitzpatrick, S.M. & R. Callaghan,2008. Seafaring simulationsand the origin of prehistoric selers to Madagascar,in Islands of Inquiry: Colonisation, Seafaring and the

Archaeology of Maritime Landscapes, eds. G. Clark, F.Leach & S. OConnor. (Terra Australis 29.) Canberra:Australian National University Press, 4758.

Freeman-Grenville, G.S.P. (ed.), 1962. The East African Coast:Select Documents from the First to the Earlier NineteenthCentury. Oxford: Clarendon Press.

Freeman-Grenville, G.S.P., 1981. Book of the Wonders of India.Four Corners: East-West Publications.

Ganzhorn, J.U., 2003. Eects of introduced Raus raus on

endemic small mammals in dry deciduous forestfragments of western Madagascar.Animal Conservation6, 14757.

Godfrey, L.R. & W.L. Jungers, 2003. The extinct sloth lemurs

of Madagascar. Evolutionary Anthropology 12, 25263.Gongora, J., N.J. Rawlence, V.A. Mobegi et al., 2008. Indo-European and Asian origins for Chilean and Pacicchickens revealed by mtDNA. Proceedings of the National

Academy of Sciences of the USA 105(30), 10,30813.Gray, J., 1954. The Wadebuli and the Wadiba. Tanganyika

Notes and Records 36, 2242.Groanelli, V.L., 1947. Asiatic inuences on Somali culture.

Ethnos 4, 15381.Groube, L.M., 1993. Contradictions and malaria in Mela-

nesia and Australian prehistory, in A Community ofCulture: the People and Prehistory of the Pacic, eds. M.Spriggs, D. Yen, W. Ambrose, R. Jones, A. Thorne& A. Andrews. Canberra: Department of Prehis-tory, Research School of Pacic Studies, AustralianNational University, 16486.

Harder, D.K. & J. Smar, 1992. Further evidence on theorigin of the cultivated winged bean, Psophocarpustetragonolobus (L.) DC. (Fabaceae): chromosome num-

bers and presence of a host-specic fungus. EconomicBotany 46, 18791.

Hingston, M., S.M. Goodman, J.U. Ganzhorn & S. Sommer2005. Reconstruction of the colonization of southernMadagascar by introduced Raus raus. Journal ofBiogeography 32, 154959.

Hoeppli, R., 1969. Parasitic Diseases in Africa and the WesternHemisphere. (Acta tropica supplementum, 10.) Basel.

Hornell, J., 1920. The origins and ethnological signicanceof Indian boat designs.Memoirs of the Asiatic Societyof Bengal 7, 139256.

Hornell, J., 1928. Indonesian culture in East Africa.Man 28, 1.Hornell, J., 1936. Indonesian inuence on East African

culture. Journal of the Royal Anthropological InstituteLXIV, 30532.Hornell, J., 1941. The sea-going mtepe and du of the Lamu

archipelago.Mariners Mirror 27, 5468.Horton, M.C. & J. Middleton 2000. The Swahili, the Social

Landscape of a Mercantile Society. Oxford: Blackwell.Hudson, A.B., 1967. The Barito Isolects of Borneo. (Southeast

Asia Program (Department of Asian Studies), DataPaper 68.) Ithaca (NY): Cornell University Press.

Jewell, J.H.A., 1976. Dhows at Mombasa. Revised edition.Nairobi: East African Publishing House.

Jones, A.M., 1971. Africa and Indonesia: the Evidence of theXylophone and Other Musical and Cultural Factors. 2ndedition.Leiden: Brill.

Jones, T.L. & K.A. Klar, 2005. Diusionism reconsidered:linguistic and archaeological evidence for prehistoricPolynesian contact with southern California.American

Antiquity 70, 45784.Jones, T.L. & K.A. Klar 2006. On open minds and missed

marks: a response to Atholl Anderson. AmericanAntiquity 71, 76570.

Juma, A.M., 1996. The Swahili and the Mediterraneanworlds: poery of the late Roman period from Zan-zibar.Antiquity 70, 14854.


12/12

248

Chapter 19

Kaudern, W., 1927.Musical Instruments in the Celebes. Gote-borg: Elanders Boktryckeri A.B.

Kent, R.E., 1970. Early Kingdoms in Madagascar, 15001700.New York (NY): Holt, Rinehart & Winston.

Klar, K.A. & T.L. Jones, 2005. Linguistic evidence for a pre-historic Polynesiasouthern California contact event.Anthropological Linguistics 47, 369400.

Lagercrantz, S., 1950. Contribution to the Ethnography ofAfrica. (Studia Ethnographica Upsaliensia I.) Lund:Hkan Ohlssons.

Laurence, B.R., 1968. Elephantiasis and Polynesian origins.Nature 219, 536.

Mahdi, W., 1999. The dispersal of Austronesian boat formsin the Indian Ocean, in Archaeology and Language vol.III, eds. R.M. Blench & M. Spriggs. London: Routledge,14479.

Manguin, P.-Y., 1980. The Southeast Asian ship: an historicalapproach.Journal of Southeast Asian Studies 11, 25369.

Manguin, P.-Y., 1985. Late medieval shipbuilding techniquesin the Indian Ocean. Areappraisal.Moyen Orient et

Ocan Indien 2(2), 130.Mbida, C.M., W. Van Neer, H. Doutrelepont & L. Vrydaghs,2000. Evidence for banana cultivation and animalhusbandry during the rst millennium in the for-est of southern Cameroon. Journal of ArchaeologicalScience 27, 15162.

Mbida, C.M., H. Doutrelepont, L. Vrydaghs et al., 2001.First archaeological evidence of banana cultivationin central Africa during the third millennium beforepresent. Vegetation History and Archaeobotany 10, 16.

Mrite, ., 1942. Les piges. Paris: Payot.Mitchell, P., 2004. Towards a comparative archaeology of

Africas oshore islands.Journal of African Archaeology2(2), 22950.

Nurse, D. & T.J. Hinnebusch 1993. Swahili and Sabaki: aLinguistic History. (University of California Publica-

tions in Linguistics.) Berkeley (CA): University ofCalifornia Press.

Poer, D., 1992. Economic botany of Sphenostylis. EconomicBotany 46, 26275.

Poer, D. & J.J. Doyle, 1992. Origins of the African yam bean(Sphenostylis stenocarpa (Hochst. ex A. Rich.) Harms):

evidence from morphology, isozymes, and chloroplastDNA. Economic Botany 46, 27692.Rackham, H., 1942. Pliny: Natural History, Books IIVII.

(Loeb Classical Library.) Cambridge (MA): HarvardUniversity Press.

Sachs, C., 1938. Les Instruments de Musique de Madagascar.Paris: Institut dEthnologie.

Schadeberg, T.C. & F.U. Mucanheia 2000. Ekoti: the Maka orSwahili language of Angoche. Cologne: Rdiger Kppe.

Simmonds, N.W., 1966. Bananas. London: Longmans.Simon, P.R., 1988. Ny teny fahizany: reconstitution et prio-

disation du malgache ancien jusquau XIV sicle. Paris:INALCO.

Stevenson, E.L. (trans. & ed.), 1932. Claudius Ptolemy: theGeography. New York (NY): New York Public Library.

Storey, A.A., J.M. Ramrez, D. Quiroz et al., 2007. Radiocar-

bon and DNA evidence for a pre-Columbian intro-duction of Polynesian chickens to Chile. Proceedingsof the National Academy of Sciences of the USA 104(25),10,3359.

Summerhayes, G.R. & A. Anderson, in press. An Austrone-sian presence in southern Japan: early occupation inthe Yaeyama islands.Asian Perspectives.

Vasey, N. & D.A. Burney, 2007. Subfossil Rodent SpeciesAssemblages from Andrahomana Cave, SoutheasternMadagascar: Evidence of Introduced Speciesand Faunal Turnover. Unpublished poster at theconference Rats, Humans, and their Impacts onIslands, Hawaii, 2007.

Walsh, M., 2007. Island subsistence: hunting, trapping andthe translocation of wildlife in the western IndianOcean.Azania XLII, 83113.

Williams, R.O., 1949. The Useful and Ornamental Plants in Zan-zibar and Pemba. Zanzibar: Zanzibar Protectorate.