Bomberai Tip as a language chain : A lexical similarity analysis of the relationship between Onin, Sekar and Uruangnirin

Using recently collected fieldwork data, this study investigates lexical similarity between the three closely related Austronesian languages Onin, Sekar and Uruangnirin (“Bomberai Tip”) as well as Arguni, a slightly more distant relative, all spoken on Bomberai Peninsula in West Papua, Indonesia. Two specific measurements of lexical similarity were used: estimated percentage of cognates and phonetic distance between probable cognates, both based on Normalized Levenshtein Distance. Data consisted of 275–456 lexemes for each language which were indexed according to meaning, enabling comparison of word pairs with the same meaning for each pairwise language combination. Probable cognates were identified computationally with an Oswalt Monte Carlo test. The test compares the distribution of phonetic distances for semantically aligned word pairs with the distribution of phonetic distances for word pairs in a large number of randomly matched lists. These provide a baseline of what similarity can be expected from chance, meaning that word pairs in the real data which had a lower phonetic distance than expected by chance could be identified and classed as probably cognate. Cognate percentage calculated from these probable cognates ranged from around 55–70% for intra-Bomberai Tip pairings and from 20–32% for pairings with Arguni. Average phonetic distance (on a scale 0–1 where 0 is identical and 1 entirely different) between probable cognates were in the 0.18–0.27 range for intra-Bomberai Tip pairings and 0.36–0.46 for pairings with Arguni. Compared to Arguni, it is clear that the Bomberai Tip languages have a similar lexicon, but the cognate percentage is too low to assume mutual intelligibility and thus they should continue to be considered separate languages. The lexical similarity measurements taken together with proposed sound changes within Bomberai Tip from Grimes & Edwards (forthcoming) suggest that Bomberai Tip is best viewed as three distinct languages in a chain-like relationship. Onin, spoken in the middle geographically, shares both higher lexical similarity and exclusive sound changes with Sekar and Uruangnirin, while they have the lowest lexical similarity within Bomberai Tip and do not share exclusive sound changes. Onin is not significantly more similar to either Sekar or Uruangnirin, meaning that this subgroup is best understood through the wave model rather than attempting to subgroup two of them in a bifurcating tree.