摘要：The general purpose of this work is to make some theoretical considerations about the inferential aspect of the Logic of Intelligent Dialogues (LID). This shall be done with the help of the conceptual framework provided by commonsense logics of artificial intelligence. More specifically, I try to show that, in order to perform its job, the LID must have a way to represent nonmonotonic inferences. To support this I shall argue that nonmonotonicity is an intrinsic feature of language use and understanding, which in its turn is supported by some considerations about the non-literal use of language.
关键词：Logic of intelligent dialogues; nonmonotonic reasoning; non-literal use of language.
