In the previous applied linguistic case study we looked at how corpora can be used to compare and contrast different spoken discourse markers. We noted how the forms and functions of very frequent words and phrases involved different categories of meaning and how this in turn can cause difficulties for those learning a language. In this case study the focus remains on frequently spoken words and phrases. Here, however, we unpack ‘words within words’, using spoken corpus analysis to reveal how core words can be even more core within specific multi-word units and how drawing attention to this may be of significance in the development of improvements in listening to lectures, specifically from the point of view of language learners.