Trinch's article provides a thought-provoking consideration of the risks involved for abuse victims in telling their story. In bringing together two seemingly different types of narrative, Trinch compels us to consider what they share. First, there are similarities in the interactional production of the written versions of Rigoberta Menchú's published 'testimonio' and the stories which survivors of domestic abuse tell to paralegals. In both situations, the stories were/are told to an interested interlocutor who works to produce a written account in the first-person narrative of the storyteller. Second, Trinch's analysis of the controversy surrounding Menchú's 'testimonio' and of the process by which the domestic abuse survivors' written affidavits are produced highlights the risks involved for people reporting abuse. At the immediate level is the risk of antagonising their abusers, and on another level are the risks in the ways in which their story is transformed in the recontextualisation process, and the subsequent risks for the narrators involved in being seen as not 'telling the truth', or as embellishing the 'facts'. It is this second level of risks which Trinch addresses, exposing powerful language ideologies about narrative truth relevant to the reception of these two different types of narrative. Building on her important (2003) book, Trinch's work here on the interactional and entextualised nature of narrative production advances both the sociolinguistic analysis of narrative and linguistic anthropological work on language ideologies. Here, I take up Trinch's point about the role of what she has referred to as the 'ideology of narrator authorship' (Trinch 2003, pp. 49–50) in the risks faced by victims of abuse in telling their story.