This chapter analyses conversations from a corpus of literature in order to uncover the extent to which these conversations contain typical features of spoken language, such as vague language and discourse markers. Such features have long been identified as key features of spoken language in corpora based on unscripted conversations (Carter and McCarthy ). The extent to which naturally occurring spoken language is similar to and different from literary conversations has been researched within the field of stylistics (e.g. Hughes ; Semino and Short ), but the extent to which conversations in literature could provide a useful model for learners of English as a Foreign Language and English as a Second Language (EFL/ESL) is less clear. Drawing upon corpus data from the CLiC Dickens corpus (Mahlberg et al. ) and the BYU-BNC spoken corpus (Davies ), this chapter seeks to explore how often common features occur and whether the frequency of occurrence is significant in comparison to data from unscripted conversations. We also explore the data qualitatively to examine whether the functions of common spoken language features differ or are similar. In doing so, we hope to uncover the extent to which conversations in literature can offer a plausible model of spoken English for EFL or ESL learners.