Realization as a Device for Remedying Problems of Affiliation in Interaction

Abstract

This article investigates the use of the Danish response token nåja in naturally occurring conversation. The article shows that nåja functions as a change-of-state token with which a speaker claims to have just now realized matters relevant for the prior talk. Whether caused by the speaker's recollection, recognition, or understanding of relevant matters, the realization token functions as a claim that (a) there were problems in the speaker's prior turns due to her lack of understanding, and (b) that these problems are now solved due to a change in the speaker's epistemic access. The realization token can therefore be used to remedy prior turns that are potentially or actually disaffiliative in a way that minimally involves questions of character or relationship.

Notes

We are indebted to Charles Antaki, Hanne-Pernille Stax, Jakob Steensig, Tanya Stivers and two anonymous reviewers for their helpful suggestions and insightful comments.

¹We are indebted to Jakob Steensig, who kindly searched his corpora of Danish interactions for instances of nåja to supplement our own collection.

²The term “independent” is in parentheses here to reflect the way Betz and Golato (2008) refer to it. By contrast, Heritage (1984b, p. 338) explicitly states that indexations of recollection such as oh that's right treat the information provided in prior turns as “independently known information.” Our analysis in the following sections support the idea that realization tokens are employed to index independent epistemic access (i.e., access independently of the coparticipant's prior turn) and hence we have chosen, in contrast to Betz and Golato, not to put the term “independent” in parenthesie.

³In particular, caution should be applied when comparing languages such as German and Danish on the one hand, and English on the other. Speakers of German and Danish have a large variety of modal particles at their disposal, that can be used, among other things, to index epistemic relations, whereas in English much of this work has to be done through syntactic or phonetic demarcations (Heinemann, 2009b; Schubiger, 1972).

⁴Neither Betz and Golato (2008) or Heritage (1984b) discuss whether achja and oh that's right can be used to index understanding, but we do not preclude that the change-of-state tokens can be employed in that way in German and English, as it can in Danish.

⁵Mette's response to this answer may in a way reflect the implication that an assessment is relevantly missing. By stating that “she would rather have the father (than the son)” she backs down from her earlier potential implication that it was “good news” that her friends would be going to school with the son, thus aligning with Line's lack of assessment that could imply that she does not find this “good news.”

⁶The realization token is, in this case, produced in an elongated fashion, but nevertheless as one intonational unit. We have found no interactional differences between this an other instances of the realization token and hence treat them phonological variation of the same generic practice.