This chapter reports an investigation into areas of miscommunication in interactions between first and second language speakers of English at a New Zealand university. Specifically, it focuses on the way that problems may be triggered by the use of noun phrases that are referentially ambiguous or misleading in context. The focus of the chapter is to report on how stimulated recall (SR) was used to identify such problems.

However, identifying miscommunication is inherently problematic. For example, many studies rely on the analysis of transcripts (e.g. Cook-Gumperz & Gumperz, 2002; Verdonik, 2010), but a serious methodological limitation of such approaches is that miscommunication often leaves no clear verbal trace, particularly when it goes unnoticed by both interlocutors. Unrecognized miscommunication is also an obvious limitation for studies involving self-reported data (e.g. Tzanne, 2000).

The approach of the present study was to access the hearer’s mental representation of the discourse and compare this to the speaker’s intended message. To achieve this aim, a film retelling task was used to elicit linguistic data, enabling strong grounds from which to identify the speaker’s intended meaning. This procedure was followed by a Stimulated Recall (SR) interview in which the hearers described how their mental model of the discourse developed. This approach potentially offers much richer insights into the hearer’s interpretation of learner speech than text-based approaches.