A Classification of the Nicobarese languages

In this paper, I report the results of a computational phylogenetic experiment based on a modified Swadesh 100 list intended to provide indications of the internal branching of the Nicobarese languages. To date, little work has been done on the classification of the small Nicobarese group, which appears to consist of approximately seven distinct languages spoken across an island chain. Three of the languages have published dictionaries and partial descriptions, while the others are fragmentarily documented, and the materials can be highly problematic to work with. The findings of the experiment suggest three principal sub-groups: Northern and Southern groups consisting of single languages (Car and Shompen, respectively) and a complex Central group that includes all other documented lects and may be regarded as a dialect chain. The Central languages appear to be rather close lexically and grammatically, although some phonological diversity is evident. There are textual claims that the various islands have been in mutual contact historically, and the phylogenetic results perhaps provide some indications of the patterns of this contact. The data supporting the study, and Nexus file used to generate the analyses, are made available as a Zenodo archive for transparency and reproducibility (10.5281/zenodo.5508350)