Examines real-life speech data from the British National Corpus to show how language is used in natural conversation. This monograph describes the composition, annotation and transcription of the corpus, as well as providing a discussion of the methodology used in corpus analysis.