К проблеме парадокса наблюдателя и переключение кодов
(On the “observer’s paradox” and code-switching)

In his article, the author gives some examples of different situations (drawn from his own recent experience) in which the presence of a linguist in fieldwork circumstances affects the informants' linguistic behaviour. These situations include: overt ‘audience design', in which the informant ‘performs' for the linguist; the use of meta-commentary, either in the minority language or the dominant language, that is directed at the audience; language purism, when the informant refrains from the use of loans and codeswitching where these would otherwise be naturally used; more permanent switching to the dominant language, either because the informant questions the language skills of the fieldworker or because the use of a minority language in the presence of strangers is unnatural to the informant. The author discusses several ways in which a conversation that has switched from the minority language to the dominant language can return to the minority language. Finally, the author discusses the possibilities for recording fully natural speech and, referring to Wertheim 2003, questions its existence.