iThere was some discussion in the panel I was on about fieldwork and endangered languages and the type of data that sociolinguists want. We didn’t have much time for it but let me put down some points that got made more or less in the session.
- Language documentation these days does not (and/or should not) involve just work from one speaker, where possible. We all know of cases where documentation from one speaker is better than from no speakers, but where there is a speech community, good data can be gkeaned from multiple members of that community.
- Documentation of a language also involves documenting who says what. That is, documenting a language is not the same thing as writing a grammar of that language.
- Some people in the audience criticised the grammar, dictionary, texts model of fieldwork as leaving out the sociolinguistics, but I can’t resist pointing out that my grammar-dictionary-texts description of Yan-nhaŋu was an ideal way to get information about variation in the speech community; it was a partially structured set of survey materials for studying difference, in effect. It also provided a reason to be in the community and a an excuse to be linguistically nosey.
- Some discussion was also had about whether it’s necessary to speak the language in order to do sociolinguistic work on the language. It’s probably not absolutely necessary, but I think that in any fieldwork doing whatever it takes to learn the language is a pretty good idea, and that applies to any type of language work.