Towards a Composite Notion of ‘what is said’

Sammanfattning

If, when asked whether you want more beer, you utter, ‘I’ve had enough’, your interlocutor knows that you have had enough of beer. That is, from the literal content of ‘I’ve had enough’, your interlocutor infers that you say, ‘I’ve had enough [of beer]’. Language theorists convincingly claim that hearers almost always make similar inferences to figure out what speaker say. However, in doing so they also claim that interlocutors do not evaluate literal content for truth in everyday conversation – it is either true or false that you have had enough of beer, not just ‘enough’. That is, they claim that literal content is not useful to interlocutors. In this thesis, I defend Emma Borg’s (2019) claim that literal content actually is useful to interlocutors. First, I argue that Borg is correct in that asking for ‘speaker liability’ shows that interlocutors evaluate literal content for truth in cases of lying. Second, I both consider a challenge to speaker liability and an objection to how the Gricean distinction between saying and implicating, the composite notion of ‘what is said’, I (and Borg) argue for should be drawn.