What conditions tone paradigms in Yukuna: Phonological and machine learning approaches

Yukuna is an understudied Arawak language of North-West Amazonia with a privative tonal system. In this system, roots are underlyingly specified for tone, whilst affixes toneless. However, affixation interacts leading to many variations in surface patterns. This paper puts forth qualitative analysis Yukuna’s and provides data-driven evidence favor using machine learning methods. More precisely, we use decision trees random forests assess quantitatively the predictions phonological analysis. A manually annotated corpus verbal paradigms was split into training testing set. We trained computational classifiers on first tested their second. found that they predict majority patterns support Additionally, suggest avenues enhancing analysis, by providing ranking variables highlight statistical tendencies within Besides its contribution understanding systems general particular, our work also suggests such approaches might become part complex theoretical methodological toolkit needed description linguistic theory development.