Abstract Peter Hanks and Scott Soames have developed a theory of propositions as structured cognitive event types, as have I in earlier works. They use the theory to offer similar accounts of transparent propositional relation reports, and very different accounts of opaque reports. For both, the sentences used to report propositional attitudes or speech acts are semantically unambiguous. Hanks invokes context-sensitivity, Soames pragmatics, to account for the different interpretations. I raise problems and offer solutions. Their accounts succumb to the non-compositionality of transparent reports, and wrongly predict that all propositional relation reports have both transparent and opaque interpretations. Soames’s pragmatic enrichment account of the opaque interpretation is unfounded, and forces him to conclude that competent speakers do not know what the sentences they use mean. The notion of an “object-dependent” or “bare” proposition is both problematic and unnecessary. I offer a new account, on which propositional relation reports have the semantic ambiguity characteristic of idioms, with the transparent interpretation being highly but not completely compositional.

中文翻译：

---

命题关系报告的三个说明

摘要彼得·汉克斯（Peter Hanks）和斯科特·索阿姆斯（Scott Soames）提出了作为结构化认知事件类型的命题理论，正如我在较早的著作中一样。他们使用该理论为透明的命题关系报告提供相似的说明，为不透明的报告提供非常不同的说明。对于两者而言，用于报告命题态度或言语行为的句子在语义上都是明确的。汉克斯援引上下文相关性，即Soames语用学来解释不同的解释。我提出问题并提供解决方案。他们的陈述屈服于透明报告的非组合性，并错误地预测所有命题关系报告都具有透明和不透明的解释。Soames对不透明解释的实用丰富描述是没有根据的，并迫使他得出结论，称职的演讲者不知道他们使用的句子是什么意思。“依赖对象”或“裸露”命题的概念既有问题，又没有必要。我提供了一个新的观点，命题关系报告具有成语的语义歧义特性，透明的解释是高度但不是完全构成的。