Most of the examples of dialogue I have included in the book will not fit neatly into the three categories of ‘cumulative’, ‘disputational’ and ‘exploratory’ talk. These are idealizations, models of ways of using language which may rarely be found in any pure form. No system of categories could ever really do justice to the natural variety of language, and even short stretches of dialogue may have characteristics of more than one of each of these types of talk.35 But this categorization is nevertheless useful for making sense of the messy, category-defying reality of conversation. We can hold the three models of talk up against examples of actual conversation and see whether they show the characteristics of one or more of the types. We can use ‘cumulative’, ‘disputational’ and ‘exploratory’ as concepts for discussing ‘discussion’.