This paper contributes to an ongoing debate about the validity of interview data and the ways in which they are interpreted in the ‘interview society’. We understand the need for an extensive reliance on interviews and, at the same time, recognise the serious limitations that exist regarding access to the interviewee’s worldview, their motivations and orientations. A crucial problem in this regard and the main concern of our paper is how to interpret subjective accounts, such as arguments or everyday theories, interviewees hold about themselves. While ethnomethodologists suggest that the complete authorship for meaning depends on the interview setting, we argue that the interviewee’s practices of generating interview content are quite stable across various sequences that allows for a reconstruction of their agency dispositions based on interview transcripts. Taking Mannheim’s and Bourdieu’s idea of a formative or generative principle as a point of departure, we introduce the most recent variant of the documentary method of interpretation (DMI) that aims at the reconstruction of this principle’s manifestation (as an individual’s frame of orientation) and helps us then to understand everyday theories, subjective explanations and justifications presented by interviewees.