Chapter 6 is concerned with the different multimodal formats of reported clauses in the 1978–1988 and 2003–2013 data sets: direct, indirect and 'in between' speech. It is discussed how these formats have changed with respect to their forms and distribution over turn types and speaker roles, and shown that speakers from both periods are strikingly conservative in the contextualisation of the quotations, with indirect and ‘literalised’ direct speech representing the two dominating practices. While indirect speech is most frequent in both data sets, the 2003–2013 sample shows a rise of ‘literalised’ direct speech across turn types and speaker roles. It is argued that this development is indicative of a general tendency to authentication and authorisation in reported speech, which is achieved through the visual manipulation of (original) documents, and the use of the verbal formula (AND) I QUOTE. The latter also serves to perform mixed quotations, a practice not found in 1978–1988. It is concluded that the comparison between the 1978–1988 and 2003–2013 points to a general tendency towards greater credibility enhancement, and a more interpersonal style in quotations.