How to doubt about a conditional

Abstract

Psychological studies of reasoning with simple conditional arguments have shown that about one half of the participants do not consider the conclusion as certain when some specific information is added to the premises, explicitly or implicitly. This nonmonotonic effect is explained by generalising Mackie's [8] analysis of conditionals within the framework of Relevance theory (Sperber & Wilson [12]): Conditionals are uttered with a ceteris paribus assumption of normality; calling in question this assumption induces doubt in the conditional: This is what characterises additional premises used in the experiments mentioned above as well as in ordinary conversation.