The Linguistic Origins of Malagasy and Convergent Research that Sheds Light on the Origins of Madagascar’s Original Indonesian Settlers - Maya Lorey

I did my previous blog post on the mystery of who the original Indonesian settlers of Madagascar were, focusing specifically on recent mtDNA analysis suggesting that the island was colonized by about 30 women in 830 AD. While researching the linguistic origins of Malagasy a few days later, I stumbled upon some cutting edge lexicostatistical analysis that sheds more light on the original colonization event and identity of Madagascar’s colonizers.
Let’s start with some basic information about about Malagasy, the official language of Madagascar. There are about 13 million first language speakers of Malagasy today, and while there are 23+ dialects, they are almost all mutually intelligible (70-90% lexical similarity). Malagasy is a member of the Austronesian Language Family, more specifically, part of the East Barito Group of the Malayo Polynesian Branch. Its closest relative is Ma’anyan, which is spoken today in inland southeastern Kalimantan by a small population of Dayak. The two languages share 45% of their basic vocabulary and are more closely related to each other than they are to any other languages. The remarkable similarity between Malagasy and Ma’anyan—first officially diagrammed by linguist Otto Christian Dahl in 1951—lead many researchers to postulate that that the Indonesian ancestors of the Malagasy people sailed from the southeastern region of Kalimantan where Ma’anyan is spoken today. As mentioned in my previous post, this hypothesis is challenged by the reality that the community has thus far offered no archaeological, ethnographic, or historical evidence for maritime tradition during the range of time Madagascar was most certainly settled. It is plausible that sailors and ships of the Srijava empire, a maritime and commercial empire centered on Sumatra, could have reached Madagascar. The Buddhist empire traded with China and India, controlled the strait of Malacca, and was powerful between the 7th and 13th centuries. The Srivijaya spoke proto-Malay. Tellingly, the Malagasy words for compass points, wind, sea, boats and other nautical/directional terms are all Malay loan words. The Malagasy orientation system, which is defined by fixed cardinal points, is also typical of Malay and Javanese maritime orientation systems and distinctly un-Bornean. This may suggest that whoever was captaining the vessel that reached Madagascar was indeed a Srijava sailor (and his crew). Given the small founding population of 30 women, who likely would not have been aboard a trading vessel, some have postulated that an errant Srijava slave ship carrying Bornean Ma’anyan speaking peoples wrecked on Madagascar. Others, who favor the idea that Madagascar was settled in waves, believe that the linguistic composition of Malagasy reflects multiple different “founding populations”. In 2012, researchers (Serva et al.) used new quantitative methodology that can find the kinship relationships among languages to analyze 23 dialects of Malagasy in addition to Malay and Ma’anyan. They were primarily looking to find more evidence about whether there was one founding population of Madagascar or many, and if there was only one founding population, as suggested by the recent mtDNA analysis, when and where on Madagascar they landed. The major findings are as follows:
 1.All of the dialects have almost the same lexical distance from Malay to distance from Ma’anyan ratio. This indicates that the underlying linguistic makeup is the same for all of the dialects and therefore that they all originated from the same founding population. These findings suggest that there was only one Indonesian founding population, made up of people speaking some mix of Malagasy and Malay – i.e there were no waves of migration to the island. These findings are consistent with the findings of Cox et. al (mtDNA analysis).
 2.The variance in the dialects suggests that the landing in Madagascar occurred around 650 AD. This is slightly earlier than Cox et al.’s estimate of 830 AD, but not too far off.
 3.They used an incredibly complicated algorithm to estimate the “linguistic homeland” of Malagasy. In essence, it draws upon the basic premise that the homeland of a biological species or language group corresponds to the current area of greatest diversity; “for each language variant a diversity index is calculated as the average of the proportions between linguistic and geographical distances from the given language variant and to each of the other language variants.” The model suggests that the initial landing occurred between Mahanoro and Ambovobe, where Farafanga is (southeastern coast of the island).
 4.The authors point out that there is some unexpected supporting evidence for this landing location. In 1883, after Mount Krakatoa erupted, pumice washed ashore on Madagascar between Farafanga and Mahanoro—right where the linguistic analysis puts the landing area. During WWII, wreckage from ships that had been sailing between Java and Sumatra washed up in the same location. This suggests that at certain times of year, there is an ocean current connecting Sumatra and southeastern Madagascar. I am looking further into this now, and wonder if currents have changed since 650-830 AD.
 In conclusion—there are still so many questions! One last interesting point: The Srivijaya empire was Buddhist, and then subject to much Hindu influence. The Malagasy are neither Buddhist nor Hindu. Neither are the Ma’anyan speaking Dayak, who are known for ancestor worship and elaborate funeral rituals… sound familiar?

- Maya Lorey