Abstract/Summary

Second language (L2) listening processes are generally thought to be under researched, especially in academic listening contexts, despite the importance of listening to the academic performance of international students in English-speaking countries. Additionally, research in the area is often not informed by theory from other fields. This research addresses both issues and investigates the word recognition processes of L2 listeners using methods informed by psycholinguistic theory. It explores the relationship between the words listeners’ report having heard and the evidence on which they rely, in the form of cues, within the speech signal. In addition, by ascertaining how much perceptual information is available to L2 listeners, the degree of their ability to build meaning and construct an accurate interpretation of the overall discourse can be implied, processes crucial in an English for Academic Purposes context. A paused transcription tasks is used, which is a method designed to induce the processes used in real-life listening. In this study, 171 university pre-sessional student participants listened to a recording taken from a pre-sessional listening course book which replicated a section of an academic lecture. Pauses were inserted irregularly into the recording at the end of target phrases. Whenever a pause occurred, participants reported immediately in writing the last four or five words heard. The outcome was a set of responses in the form of a series of words. Participants’ correct responses were analysed to establish the lexical features they may have relied on in their accurate word recognition processes, and their incorrect responses were analysed to establish which cues they made use of in their erroneous word recognition processes. This enabled insight into participants’ cognitive processes employed during the listening process. The findings were extensive and addressed several gaps in the knowledge of L2 listening. The major findings were i) word frequency was a key factor in accurate word recognition; ii) perceptual cues were more heavily relied on than contextual cues when listening broke down; iii) incorrect responses were informed by more than one single cue; iv) co-textual cues should be viewed separately from contextual cues; v) co-text is highly influential to L2 listeners; and vi) participants’ lack of confidence in their incorrect responses indicate they are behaving strategically. The findings support the notion that traditional teaching methods are not appropriate for L2 mid-level listeners, especially those preparing for academic study. Consequently, a call is made for a re-evaluation of listening training both in an academic context and more generally and suggestions for doing so are put forward.