Unlike belief and desire reports, intention reports (e.g., John intends to leave soon) are not well studied in formal semantics. This paper aims to begin to fill this gap, focusing on empirical similarities and differences that intention reports bear in relation to other attitude reports and to other expressions that involve intentional action. These empirical properties are shown to follow from the view that an intention report a intends p denotes true iff a has a maximally ranked ACTIONRELEVANT or EFFECTIVE PREFERENCE (in the sense of Condoravdi and Lauer 2016) that a bears the RESPONSIBILITY relation (in the sense of Farkas 1988) to p. Also discussed are intention reports that do not involve syntactic control (e.g., John intends for Bill to leave soon), which in some previous literature have been argued to involve coercion. The proposed semantics for intend enables a coercion-free analysis of such sentences, and issues surrounding the choice between a coercion-free and a coercion-based approach are discussed.

中文翻译：

---

意向报告的逻辑

与信念和愿望报告不同，意图报告（例如John打算很快离开）在形式语义学上没有得到很好的研究。本文旨在填补这一空白，侧重于意向报告与其他态度报告以及涉及意向行动的其他表述在经验上的异同。这些经验属性是从以下观点来看的：意图报告a意图p表示正确，前提是a具有最高等级的行动相关或有效偏好（在Condoravdi和Lauer 2016的意义上），且承担责任关系（在Farkas 1988）至第 还讨论了不涉及句法控制的意图报告（例如，John打算让Bill离开），在先前的一些文献中，这种意图报告被认为涉及强迫。