about
anna björklund
I am a linguitics PhD candidate at UC Berkeley with a focus on the Wintuan languages of the northern Sacramento Valley (California, USA). My research is situated in the intersection between phonetics, language revitalization, and deep learning. I have worked with the Paskenta Band of Nomlaki Indians since 2020 and the Wintu Tribe of Northern California. The Wintuan languages (Wintu, Nomlaki, and Patwin) in the Wintuan languages, traditionally spoken in the northern Sacramento Valley of California. The Wintuan languages (Wintu, Nomlaki, & Patwin) are among the least documented in North America, with Nomlaki having neither a grammar nor dictionary and less than an hour of spoken audio. My primary research interests are in leveraging archival recordings for the phonetic and phonological analysis of these and similarly under-documented languages, as well as exploring how deep learning can be used to supplement revitalization, particularly text-to-speech. I have worked as a linguistic consultant for the Paskenta Band of Nomlaki Indians since 2020 and the Wintu Tribe of Northern California since 2022. I am also currently a Data Science Fellow at the UC Berkeley D-Lab.
My recently completed second qualifying paper (in preparation) focuses on developing a ToBI style system of transcription for Patwin (PaToBI). PaToBI is then used to identify common intonational patterns across text and elicitation, which are compared with those suggested by computational methods such as Kaland (2021).
My advisor is Keith Johnson.