This article explores the proposal that a phase can be at least partially spelled out and then subsequently accessed by heads from a higher phase. I provide morphological evidence for this claim, by looking at examples in which a morphological (i.e., postsyntactic) process feeds a later narrowly syntactic one. Main evidence comes from the interaction between syntactic head movement and postsyntactic affixation in the Niger‐Congo language Wolof, which confirms an important prediction of the analysis. Namely, if a postsyntactic process can feed a syntactic one, we also expect the postsyntactic process to be bled if the structural conditions for its application are not met, resulting in surface opacity effects. This is precisely what we find in Wolof, where the past‐tense morpheme oon is postsyntactically affixed onto the verb in a particular structural configuration and carried along with it to C but is stranded by the verb below C if the structural requirements for its affixation are not met.

中文翻译：

---

句法和后句的交织：句法运动之前的拼写

本文探讨了可以至少部分拼出一个阶段，然后由更高阶段的负责人访问的提议。通过查看一些例子，其中形态学（即后句法）过程提供了后来的狭义句法过程，我为这一主张提供了形态学证据。主要证据来自尼日尔-刚果语言Wolof中句法头部运动与句法后缀之间的相互作用，这证实了这一分析的重要预测。即，如果后句法过程可以提供句法，则我们也期望如果不满足其应用条件，则后句法过程将放血，从而导致表面不透明效果。这正是我们在沃洛夫语，这里的过去时态的语素找到OON 在语法上以特定的结构配置附加到动词上，并随其带到C，但如果不满足其附加的结构要求，则在C之下将动词束缚住。