The first time Jacqueline Brixey took a Choctaw language class, she was already pursuing a doctorate. A citizen of the Choctaw Nation of Oklahoma, Brixey grew up without fluent Choctaw speakers in her immediate family. Her great-grandfather was the last person she knew of in her family who spoke the language fluently, and it wasn’t until she moved to Los Angeles for graduate school that she had the opportunity to learn Choctaw in a classroom.
“It was truly the best part of living in Los Angeles for me,” she says, “being connected with community and then getting to finally learn the language.”
That experience helped shape her current research as an assistant professor at the Information School at UW–Madison, where she studies how technology can support Indigenous language learning and revitalization.
Brixey’s path to becoming a computational linguist was not a direct one, though each step gave her a new way to think about language. Her undergraduate studies in journalism and international studies led her abroad to teach English in Spain and France. Those years deepened her interest in languages, and she returned to the United States to pursue a master’s degree at the University of Texas–El Paso. Its location near the U.S.-Mexico border offered her the opportunity to study language in a bilingual community, while a course in computational linguistics showed her how she could combine her interests in language and computer science. During her doctoral studies in computer science at University of Southern California–Los Angeles, Brixey began applying those interests to her own experience, exploring how technology could support Choctaw learners like herself.
One outcome of that research was Masheli, a Choctaw-English bilingual chatbot designed to help people practice conversational Choctaw outside the classroom. Brixey built it around animal stories about wolves, ants, and turtles, introducing beginning learners to familiar vocabulary while drawing on traditional stories and knowledge.
Her work has since expanded to other questions about language technology. With other Indigenous researchers, Brixey evaluated how well large language models such as ChatGPT and Google Gemini handle Indigenous languages, testing their translation, vocabulary, and cultural knowledge of Choctaw, Muscogee, Cheyenne, Cherokee, and Hawaiian. She is also developing tools such as handwriting recognition software that can make historical records easier to search and speech recognition technology that allows users to search dictionaries by speaking a word aloud, helping fluent speakers who may not know how to write in Choctaw
For Brixey, however, the goal is not simply to build new technologies. She is equally interested in questions of data sovereignty, including who controls language data, where it is stored, and how it is used. Communities, she says, should decide for themselves what role technology plays in language preservation.
“AI is a tool,” she says. “We can decide which part of it that we’d like to use.”
That perspective also challenges common misconceptions about artificial intelligence. While public conversations often focus on large language models, Brixey notes that AI encompasses a much broader range of technologies, from speech recognition systems to recommendation algorithms and navigation software.
Brixey hopes to continue developing language technologies and sharing what she has learned with other Indigenous communities. Technology alone will not keep a language alive, but in the hands of the people who speak, teach, and learn it, she believes it can strengthen connections to language, culture, and community.
