A detailed investigation of the Rote-Meto subgroup attests to the complex spread of Austronesian languages in Wallacea, in which contact with pre-Austronesian substrates and other Austronesian languages have played a role. The segmental inventory of Proto-Rote-Meto has been transformed according to regional norms. This has come about through acquisition of new vocabulary, as well as large-scale unconditioned splits with multiple interacting causes. The “non-Austronesian” nature of the segmental inventory contrasts with the relatively “Austronesian” character of the lexicon. No single explanation accounts for the Rote-Meto data. This highlights the need for multiple lines of investigation across multiple domains in order to understand language history in this region