The Tunica language is a language isolate whose last known native speaker died in 1948. Three separate linguists, working with three different Tunica speakers, documented the language at three different time periods from 1886 to 1939. This dissertation’s first goal is to quantify language change across these three periods of time, as well as in 2017, during Tunica’s revival effort. To accomplish this, a Tunica language corpus was created from Tunica language documentation to examine language change during the period of Tunica language obsolescence and revival. The dissertation uses frequency analysis to analyze potential loci of language change, including the verb aspectual system, nominal gender marking, use of evidentials, syncope in possessive noun prefixes, and word order. The differences between the subcorpora of documented historical Tunica yielded no conclusive evidence of gradual language change over time, revealing instead that the three Tunica speakers varied widely in their use of the language, but in no clear direction. Comparing these three subcorpora to the Modern Tunica subcorpus reveals certain language changes that would be expected in both obsolescence and early revival, including a higher use of less complex and more analytical verb forms and much lower use of marked structures. This dissertation’s second goal is to provide information and tools that will enable greater access to and easier analysis of all Tunica language documentation. Tunica is a heritage language of the Tunica-Biloxi Tribe and has been the subject of community-based revival efforts since the 1970s. In 2010, the tribe partnered with Tulane University and created a more formal language revival organization based on principles of community-engaged scholarship. Language revival is a herculean task requiring community buy-in and the marshaling of as many resources as possible. In any language revitalization effort, easily accessible documentation is valuable for language workers, language learners, and the wider community. The work that underlies this dissertation has resulted in the creation of a more complete corpus of parsed Tunica texts, as well as methods and procedures that make it easier to investigate questions about the language.