Madagascar exhibits a strong linguistic uniformity since all dialects are regional variants of the same language, which belongs to the Greater Barito East group of the Austronesian family [Houtman, 1603]. This was firmly established a long time ago in [Tuuk, 1864], while, more recently, [Dahl, 1951] pointed out a particularly close relationship between Malagasy and Maanyan of south-eastern Kalimantan [Dyen, 1953]. Nevertheless, Malagasy also bears similarities to languages in Sulawesi, Malaysia, Sumatra and Philippines, including loanwords from Malay, Javanese, and one (or more) language(s) of south Sulawesi [Adelaar, 2009] and Philippines. On the contrary, the east-African contribution to the vocabulary seems to be limited to few faunal names [Blench and Walsh, 2009].
Because of these linguistic relationships, it is widely accepted that the island was settled by Indonesian sailors after a maritime trek but dates and place of landing are still debated and it is also not clear whether there were multiple settlements or just a single one. The linguistic composition of the Austronesian settlers is also debated as well its consequences on the vocabulary of Malagasy dialects.
In this paper we review our research [Serva et al, 2012, Serva, 2012], which tries to shed new light on these problems. The key point is the application of a new quantitative methodology [Serva and Petroni, 2008, Petroni and Serva, 2008, Bakker et al, 2009] which is able to find out the kinship relations among languages (or dialects). New techniques are also introduced in order to extract the maximum information from these relations concerning time and space patterns [Blanchard et al, 2010a, Wichmann et al, 2010a]. We consider 23 Malagasy dialects plus Malay and Maanyan. The data concerning Madagascar were collected by one of the authors (M.S.) at the beginning of 2010 and consist of Swadesh lists of 200 items for each of the 23 dialects covering all areas of the island.