Truth-conditions are systematically determined when they are the output of an algorithmic procedure that takes as input a set of semantic and (optionally) contextual features.Truth-conditional sceptics have cast doubts on the thesis that truth-conditions are systematic in this sense. Against this form of scepticism, Schoubye and Stokke (2016) and Dobler (2019) have provided systematic analyses of utterance truth-conditions. My aim is to argue that these theories are not immune to the kind of objections raised by truth-conditional sceptics. In particular, I argue that the use of Questions Under Discussion (Schoubye and Stokke) and ways of being (Dobler) is problematic.

Pragmatic modulation makes contextual information necessary for interpretation. This poses a problem for homophonic reports and inter-contextual communication in general: of co-situated interlocutors, we can expect some common ground, but non-co-situated interpreters lack access to the context of utterance. Here I argue that we can nonetheless share modulated contents via homophonic reports. First, occasion-unspecific information is often sufficient for the recovery of modulated content. Second, interpreters can recover what is said with different degrees of accuracy. Homophonic reports and inter-contextual communication are often successful because the reporting context does not demand full accuracy.

No abstract