You get what you put in: Elicited production versus spontaneous verbal interaction in cross-linguistic studies of language use

Abstract

This examination illustrates weaknesses in the traditional research paradigm in cross-cultural studies of linguistic politeness – the use of elicited data, and the model tacitly underlying the construction of elicitation tools. This is done by discussing insights gained in a quantitative cross-linguistic study of situated, naturally occurring requests and proposals, which could not be gained using elicited data. After indicating the scale on which the current elicitation paradigm has been applied in cross-cultural pragmatics, the article summarizes the main critical points highlighted by previous research and introduces new points, such as a social pressure on the respondents to produce the kind of data expected by the researcher, and the researcher's lack of insight into the context models construed by the respondents. Finally, it is discussed how the choice of elicitation tasks is tacitly based on a dyadic model of interaction, in which the addressee is a single person and the speaker acts on his or her own behalf and only represents him- or herself. There follows a discussion of a corpus-based study of requests and proposals that compares data from British, German, and Polish versions of the reality TV show Big Brother. It is shown how quantitative study of natural data not restricted to dyadic interactions demands new categorizations of interaction parameters considered, and provides new insights into language-specific impact of contextual factors upon politeness strategies.