AbstractA domain pertinent to knowing in interaction is evidentiality, but documenting evidential markers can be challenging. Among methodologies, direct elicitation and questionnaires offer the advantages of efficiency and cross-linguistically comparable data. They can, however, miss markers that are below the level of speaker consciousness, as well as significant discourse and social factors. Experimental tasks can provide cross-linguistically comparable data complete with discourse context, and in some cases evidence of the role of differential knowledge states of participants. A single task might miss genre-specific markers, however. Documentation of extensive unscripted speech in a variety of genres, much of it interactive, can provide a foundation for identifying the full sets of markers to be investigated and for uncovering functions beyond specifying the source of information. Insights from speakers can then take us further, potentially shedding light on subtle circumstances underlying choices among alternatives, particularly those reflecting social factors. But we need to know how to listen. Effective collaboration depends crucially on recognition of the variability of speaker consciousness of the markers. If this is kept in mind, speakers can serve as important co-analysts, scouting through their lifetime experiences to provide hypotheses about the contexts in which alternative constructions would be appropriate, meanings they can add, and social and cultural factors influencing their use. Resulting hypotheses can then be tested against the documented material and refined until they account well for the data. These points are illustrated with material from Central Pomo, indigenous to California.
Interrogating psychocentrism and the role of a white researcher
