Documenting Semantic Distinctions in Highly Endangered Languages

Abstract

Semantic research often targets implicit grammatical knowledge characteristic of highly proficient speakers. This knowledge is invaluable for language revitalization and for furthering linguistic theory. Yet in highly endangered language contexts, traditional semantic elicitation techniques such as the Contextual Acceptability Judgment Task can be difficult to apply successfully. In this paper, I discuss the challenges I faced in using Contextual Acceptability Judgment Tasks, Implication Judgment Tasks, and Storyboards in three communities with highly endangered languages. I discuss which implementations of the tasks proved successful across a range of speakers and communities. Employing the methods described in this paper allowed for subtle distinctions in temporal semantics, modality, and evidentiality to be successfully investigated in Chickasaw (cic), Mvskoke (mus), and Koasati (cku). I provide details and recommendations for how these tasks can be implemented. The outcome of documentation and language research using these methods resulted in engaging revitalization materials and insightful linguistic analysis.