| Abstract | Linguistics has increasingly prioritised greater sample sizes, experimental approaches, and standardising methods, whilst also acknowledging that greater representation is needed from smaller languages and a variety of sociocultural contexts. There is a tension inherent to these observations with little clear guidance in the scholarship around how best to apply increasingly complex methodologies in contexts of language endangerment. This methodology-focussed discussion article draws on the experiences of three researchers who have deployed experimental sociolinguistic methods in Baba Malay, Francoprovençal, and Scottish Gaelic-speaking communities. We argue that such communities present particular challenges to the fieldworker that are born of sociohistorical and ideological reflexes, with implications for research design, from sampling, through to stimuli production and data elicitation. These challenges are especially pronounced when designing sociolinguistic perception experiments, and they demand substantive responses rather than ad hoc workarounds. Methodological and ethical dimensions are treated here as inseparable, since experimental sociolinguistic research in language shift operates within contexts where power relations differ markedly from those assumed by standard laboratory or survey-based paradigms, and researchers must attend to these carefully. To this end we offer a toolkit that might better serve linguists collaborating with and working for endangered-language communities. |
|---|