Fieldwork on language in interaction: Benefits and barriers
Amalia Horan Skilton, Cornell University
Includes joint work with Claire Bowern (Yale), Sophie Pierson (UT Austin), and Sunny Ananthanarayan (U. Washington)
This talk examines the role of data from informal social interaction, such as conversation and child-caregiver interaction, in language documentation. Although interactional data can be transformative for both linguistic analysis (Sacks et al. 1974; Schlegoff et al. 1977) and language reclamation (Baldwin et al. 2018; Hermes & King 2019), it is rarely included in language documentation projects (Rosenblum & Berez-Kroeker 2018; Dingemanse & Floyd 2014). I outline the most common technical, institutional, ethical and (inter)personal barriers which prevent linguists from collecting interactional data. Some of these barriers, such as the time demands of transcription (Rossi et al. 2020), are unavoidable. But many barriers, I argue, are not intrinsic to interactional data. Instead, they reflect that the technical infrastructure for language documentation — including the software, recording specifications, and archives that we use — has been built around the theoretically significant assumption that all language data is monologic. This assumption is related to, and often surfaces in the same infrastructure as, the equally significant assumptions that all data is monolingual and monomodal. Following this discussion, I present some of the strategies which have allowed me to navigate the barriers to fieldwork on interaction, including a workflow for collecting and analyzing interactional data using common documentation tools (and a few add-ons). I illustrate the talk with successes and failures from my own fieldwork with speakers of Ticuna, an Indigenous Amazonian language.