Friday, 4 February 2011


Bajau child at sea (source:
One of the enduring mysteries of Indian Ocean history is how and why speakers of a Bornean language came to settle on the island of Madagascar. The Dutch explorer Frederik de Houtman (1603) drew attention to the similarities between Malay and Malagasy more than two centuries before Wilhelm von Humboldt (1836-39) proposed the name Malayo-Polynesian for the language family to which they both belong, now classified as a branch of the Austronesian language phylum. Subsequent scholars speculated on the nature of ancient contacts between Indonesia and East Africa, and a various theories were advanced about the settlement of Madagascar. The puzzle deepened when Otto Christian Dahl (1951) concluded that Malagasy is most closely related to Ma'anyan and the East Barito languages of southeastern Borneo, a hypothesis that is now widely accepted. How did the Bornean ancestors of the Malagasy, members of a group that has no apparent maritime tradition, reach Madagascar? Were they brought on the vessels of Malay (Srivijayan) or Bajau ('sea nomad') traders and/or raiders visiting the coast of East Africa, as some have suggested? How did they pick up the smattering of Bantu vocabulary and linguistic habits that are evident in Malagasy? Was Madagascar already settled by Africans before their arrival, as Dahl argued?

Festschrift for Professor Claude Allibert
These and other questions about the nature of Austronesian voyages and settlement in the western Indian Ocean remain unanswered, though our knowledge is advancing all the time. Last weekend I received my copy of Civilisations des mondes insulaires (2010), a collection of papers written in honour of Professor Claude Allibert, best known for his archaeological and historical research in the Comoros and Madagascar. I was pleased to find that a number of the thirty-plus contributions to this weighty volume have a direct bearing on the Austronesian problem. These include a review of the cultivated plants introduced by the first Austronesian migrants to Madagascar by Philippe Beaujard, and a discussion of the genetic evidence for the settlement of the Comoros and Madagascar by Axel Ducourneau. The first paper I read, though, was 'The Maldives connection: pre-modern Malay world shipping across the Indian Ocean' by Pierre-Yves Manguin, not just because it was in English, but because of its relevance to my own contribution about a particular set of historical traditions on the East African coast.

Aerial view of the Maldives (from
These traditions, about a foreign people called Diba (Wadiba), Debuli (Wadebuli), and variants thereof, have been recorded in scattered locations between the Lamu archipelago in the north and Kilwa (and thereabouts) in the south. The Diba et al. are sometimes described as unwelcome slave-raiders and invaders, and in other sources as the bearers of new cultural practices, technologies (maritime, architectural), and agricultural crops. There is no consensus on who the Diba and Debuli were: commentators have identified them as visitors of Arabian, Persian, Indian, Laccadive or Maldive origin. Harold Ingrams (1931: 47-48) considered but rejected the possibility, raised by Archdeacon Dale (1920: 13), that they might have been Malay. Some traditions as well as one version of the ethnonym (Dubuki) imply a connection with Madagascar, while the name Diba (ultimately from Sanskrit dvipa, 'island') could refer to inhabitants of any of the islands in the Indian Ocean, not just the Laccadives and Maldives. These and other features of the traditions suggest that the Diba et al. might have been Austronesians, though the evidence for this is equivocal. It is clear that the recorded traditions conflate different historical events, and in my paper I concluded as follows:

Unless we read the Diba/Debuli traditions very selectively, we cannot interpret them as evidence for interaction with a single group of visitors to the East African coast. And if they contain a deep memory of Austronesian contacts and trade, then this has evidently been overlain by other memories of interaction with outsiders, Persians, Arabs, Indian, and Portuguese included. (2010: 469).

Aerial view of the Maldives (from
Nevertheless, I remain attracted by a selective reading of the Diba traditions. Sacleux (1939: 538) defined Swahili Mdiba as the name for an indigenous inhabitant of the Maldives, and Lydekker (1919: 90-91) was led to believe that the Wadiba of Bajuni tradition came from the Laccadive islands. In his paper, Pierre-Yves Manguin reviews the documentary evidence for historical connections between insular Southeast Asia and these continguous archipelagos, which are often treated as one in the early literature. The evidence is impressive, and includes the suggestion that the Maldivian (Dhivehi) name for the Chagos archipelago, Folovahi (and variants), is derived from Malay Pulo Weh, an island off the northwestern tip of Sumatra which is on the old sailing route to the Maldives. Manguin argues that in addition to the direct and well-known route between the Maldives and the coast of Somalia, Austronesian voyagers may also have used Diego Garcia and the Chagos islands as a stepping stone to reach Madagascar and the East African coast. In any event these different journeys involve passing through the Maldives.

A boxfish in the Maldives (from
Other kinds of evidence, from maritime technology (Manguin 2000) and comparative linguistics (Tom Hoogervorst, personal communication, April 2010), support the hypothesis that the Maldives were an important staging point for Austronesian exploration and expeditions into the western Indian Ocean. We know from Buzurg ibn Shahriyar's Book of the Marvels of India that Indonesians were raiding and trading on Pemba and the mainland coast (down to Sofala) in the 10th century (Mauny 1965: 7-8), and they may well have been doing so for some time. Two centuries later Al-Idrisi reported that merchants from Sumatra were still trading with the East African coast, where "they understand one another's language" (Mauny 1965: 14). These facts fit well with some of the Diba and Debuli traditions, including one Pemban historian's repeated reference to the people of "Diba and Jawa" together, as though they were a single entity (Ingrams 1931: 125). Jawa, of course, is the island of Java (Yawadvipa in the Sanskrit epic Ramayana). Can it be that his garbled history and others like it are showing us a glimpse of events that happened so long ago? I'm tempted to say yes. The Maldives connection certainly deserves a closer look.

Aerial view of the Maldives (from

Beaujard, Philippe 2010. Les plantes cultivées apportées par les premiers migrants austronésiens à Madagascar. In Radimilahy and Rajaonarimanana (2010). 357-385.

Dale, Godfrey 1920. The Peoples of Zanzibar: Their Customs and Religious Beliefs. Westminster: Universities' Mission to Central Africa.

Dahl, Otto Christian 1951. Malgache et Maanjan: une comparaison linguistique. Oslo: Egede Institutett.

Ducourneau, Alex 2010. Approche phylogéographique du peuplement de l'océan Indien occidental. In Radimilahy and Rajaonarimanana (2010). 855-871.

de Houtman, Frederik 1603. Spraeck ende woord-boeck, inde Maleysche ende Madagaskarsche Talen met vele Arabische ende Turcse woorden. Amsterdam.

Aerial view of the Maldives (from
von Humboldt, Wilhelm 1836-39. Über die Kawi-Sprache auf der Insel Java (3 vols.). Berlin.

Ingrams, Harold 1931. Zanzibar, its History and its People. London: H. F. & G. Witherby.

Lydekker, C. J. W. 1919. The "mtepe" dhau of the Bajun islands. Man 19 (46): 88-92.

Manguin, Pierre-Yves 2000. Les techniques de construction navale aux Maldives originaires d'Asie du Sud-Est. Techniques & Culture 35-36: 21-47.

Manguin, Pierre-Yves 2010. The Maldives connection. In Radimilahy and Rajaonarimanana (2010). 261-284.

Mauny, Raymond 1965. The Wakwak and the Indonesian invasion in East Africa in 945 A.D. Studia 15: 7-16. 

Radimilahy, Chantal and Narivelo Rajaonarimanana (eds.) 2010. Civilisations des mondes insulaires (Madagascar, îles du canal de Mozambique, Mascareignes, Polynésie, Guyanes): Mélanges en l'honneur du Professeur Claude Allibert. Paris: Karthala.

Sacleux, Charles 1939.Dictionnaire Swahili-Français. Paris: Institut d'Ethnologie.

Walsh, Martin 2010. Deep memories or symbolic statements? The Diba, Debuli and related traditions of the East African coast. In Radimilahy and Rajaonarimanana (2010). 453-476.


  1. Je ne suis pas à l'aise pour écrire en anglais. Pour la lecture, ça va!
    Je viens de découvrir votre blog qui est particulièrement interessant et sérieux (compte tenu de la bibiographie que vous mentionnez). Que pensez-vous du "complexe agro-alimentaire asiatique" de Kwale développé par Hornelle et pratiquement ignoré par la suite, sauf quelques rares références.
    Merci d'avance

    1. Many thanks for your interesting comment and query. It's an important hypothesis, and we still haven't answered all of the questions raised by Hornell and Murdock about the introduction of crops of southeast Asian origin, though we now have better tools than ever to do so. Among others, a number of archaeologists, geneticists, and linguists associated with The Sealinks Project in the University of Oxford have been working on different aspects of this subject, as has my good friend Roger Blench in Cambridge. I'm looking forward to hearing some of the latest results of this work at a workshop in Oxford next week (East Africa in the Indian Ocean World II, 22-23 March 2012).

  2. there is a genetic research about Maldivian population. it has given no hints of any genetic link with Maldives and Madagascar.

    1. Thanks for this observation -- please do provide references if you can.

  3. Very interesting hypothesis. I have read an article on the DNA of the Malagasy people. There seems to be a link with Eastern Indonesia (Sulawesi, Maluku and Timor (Nusa Tenggara) as well.

    Wonder if the name Swahili is derived (a corruption of) from Tawaili (Sulawesi); which is directly next to the island Borneo (from where the Ma'anyan and Barito came). The Swahili have the same type of Outrigger Canoes.
    One of the (Demi-)gods in East Africa is "Muluku" perhaps there is a connection with Maluku there.

    In case of the Maldives Chagos archipelago: Folovahi (Folhavahi, Hollhavai) could indeed be derived from Pulo Weh (Pulau Weh). The local residents refer to Pulo Weh as "Sabang" (the Capital city of the island).
    So this could also mean that Folhavahi and / or Hollhavai is a corruption of Pulo Havahi / Havai which is Pulo Saba(ng).