Abstract

In interviews, it may happen that a respondent gives an answer which seems well formatted, but is not receipted as acceptable by the interviewer. In this article I examine one way in which interviewers display their diagnosis of the problem and act to bring about its solution. In the cases I describe, the interviewers defer revision of the question until they have established a new, more personalized basis for it, informed by their knowledge of the respondents' circumstances. There are three things of interest. The first is how this actually works conversationally; it seems to be structured as an insertion sequence and played out by presequential turns which are highly projective of the respondent's agreement. The second is that the scenario that the interviewer inserts is a positive example of what would have been an answer to the `failed' original question. The third is that there is a difference between the original, general question and its subsequent, specific revision. I argue that all these features manifest interviewers' solution of their dilemma in choosing between literal and sensitive questioning.