With fewer than five first-language speakers remaining, Hupa (Pacific Coast Athabaskan) is a critically endangered language traditionally spoken in Hoopa Valley in Northern California. Although Hupa morphophonology has been described, very little work has been done on syntax, semantics or discourse phenomena. Working with a fluent speaker, we propose to contribute to a complete description of the language by developing a multimedia text corpus and an integrated description of clause structure and interclausal relationships. The materials produced will be accessible through a simple web interface, suitable for use by language learners and linguistic researchers. Primary investigator: Amy Campbell