Moro (2009) proposes that Italian inversion structures provide evidence for one of the predictions of the Labeling Algorithm (LA, Chomsky 2013, 2015). This is that in structures in which labeling is ambiguous, one of the constituents must be moved to a higher position, instantiated in Italian inversion as an intermediate focus position. However, Italian provides little independent evidence for movement to such a position and the evidence it does provide is limited to symmetrical structures containing two DPs. In this paper, I argue that Kirundi Object‐Verb‐Subject (OVS) inversion structures and transitive expletive constructions (TECs) (Ura 1996, Ndayiragije 1999) provide more direct evidence for the type of movement that Moro proposes for Italian. Because of a confluence of Kirundi idiosyncrasies, Kirundi shows unambiguously that when the external argument does not move to Spec, TP, it cannot stay in Spec, vP. Importantly, this ban on remaining in Spec, vP cannot be accounted for by appealing to relativized minimality (Rizzi 1990). Kirundi thus instantiates a key prediction of the LA.

中文翻译：

---

标记算法和基隆迪反演结构

Moro (2009) 提出意大利反演结构为标记算法的预测之一提供了证据（LA, Chomsky 2013, 2015）。这是在标记不明确的结构中，必须将其中一个成分移到更高的位置，在意大利倒置中实例化为中间焦点位置。然而，意大利语几乎没有提供移动到这种位置的独立证据，它提供的证据仅限于包含两个 DP 的对称结构。在本文中，我认为 Kirundi 宾语-动词-主语 (OVS) 倒置结构和及物动词结构 (TEC)（Ura 1996，Ndayiragije 1999）为 Moro 为意大利语提出的运动类型提供了更直接的证据。由于基隆迪语特质的融合，Kirundi 明确地表明，当外部参数没有移动到 Spec, TP 时，它不能停留在 Spec, vP 中。重要的是，这种禁止保留在 Spec, vP 中的原因不能通过诉诸相对化的极小性来解释 (Rizzi 1990)。因此，基隆迪实例化了洛杉矶的一个关键预测。