Abstract

The problem of the fictional truthfulness of deviant evaluations is a description of a paradox that has become the subject of debate in the philosophy of literature of the analytical tradition in the last thirty years. Philosophers such as Walton, Tanner, Moran, Gendler, and others have constructed “mini-stories” in order to show that the authors of fiction, regardless of the almost unlimited powers of what they can make true in their fictional storyworlds cannot successfully construct a fiction in which an objectively false evaluation would nevertheless be fictionally true in the story. For example, someone comes up with a fictional account in which a brutal unmotivated murder takes place. The author merely postulates such an event, and then readers accept it as true in the story. But let us imagine that the description of the murder also contains the evaluation of such an act as a morally right thing to do. We do not accept an evaluation that we consider “deviant” in relation to our current (moral) beliefs as an integral part of the fictional world, even though the author asks us to do so. The paradox is that it seems to be a basic convention of fiction that all third-person narrative statements that cannot be attributed to fictional characters are always fictionally true. This seems to constitute a major restriction to the narratorial authority prompting us to question our usual understanding of the phenomenon of fiction. However, by analysing specific “mini-fictions”, the article argues that (i) there is no good reason to assume that deviant evaluations in these mini-stories are uttered by some thirdperson narratorial voice that cannot be identified with a fictional character; (ii) that deviant evaluations must always be voiced by a personal narrator; (iii) that such personal narrators are never sensu stricto “omniscient” and that we do not need to trust them unquestioningly when, for example, they evaluate a murder as “the right thing to do”. The whole paradox, we claim, is just a pseudo-problem generated by ignoring concrete reading practices and a simplified understanding of narrative conventions.