In this paper, I argue that while truth-conditional semantics in generative linguistics provides lots of good semantic explanations, truth-conditions do not play an important role in these explanations. That is, the fact that expressions have the particular truthconditional contents (extensions or intensions) they have does not even partly explain facts about semantic phenomena. Rather, explanations of semantic phenomena appeal to extra-truth-conditional properties attributed to expressions via their lexical entries. Focusing on recent truth-conditional work on gradable adjectives and degree modifiers by Kennedy and McNally (Language 81:345–381, 2005), I show that the best explanations of semantic anomaly and entailment for these expressions are non-truth-conditional—they do not depend on the fact that these expressions have the truth-conditional contents they have. I then provide reasons for thinking that the point generalizes beyond gradable adjectives and degree modifiers to other expressions, and beyond anomaly and entailment to other semantic phenomena. Truth-conditional content, generally, is extrasemantic.

中文翻译：

---

真值条件对真值条件语义有多重要？

在本文中，我认为虽然生成语言学中的真值条件语义提供了许多很好的语义解释，但真值条件在这些解释中并没有发挥重要作用。也就是说，表达式具有它们所具有的特定真值条件内容（外延或内涵）这一事实甚至不能部分解释有关语义现象的事实。相反，语义现象的解释诉诸于通过词汇条目归因于表达式的超真实条件属性。专注于肯尼迪和麦克纳利最近关于可分级形容词和学位修饰符的真值条件工作（语言 81：345-381，2005），我表明对这些表达式的语义异常和蕴涵的最佳解释是非真条件的——它们不依赖于这些表达式具有它们所具有的真条件内容这一事实。然后，我提供了理由，认为该点可以推广到其他表达的可分级形容词和程度修饰符之外，也可以推广到其他语义现象的异常和蕴涵。通常，真值条件内容是超语义的。