Replication Data for: "The loss of Latin OV: Steps towards an analysis"

In this paper I offer a first attempt to analyse the loss of OV word orders in Latin/Romance. Assuming Yang's (2000, 2002) variational acquisition model of language change, I suggest that the eventual decline of the OV grammar can be traced back to an independent change in the grammar of Latin, namely one concerning the way in which the clausal EPP requirement is satisfied (Danckaert 2017). The crucial Late Latin innovation is argued to be the development of optional A-movement for subjects: as an indirect result of this, in the Late Latin period the VO grammar is more robustly cued than the OV grammar, despite the fact that the overall frequency of the order VO remains more or less constant over time (and in certain environments even decreases). Given this result, Yang’s model correctly predicts the VO order to oust the competing OV pattern. I also show that the rate of object extraposition (which yields linear VO sequences) remains fairly stable during the lifespan of the Latin language, which suggests that this phenomenon is unrelated to the spread of the VO order. Finally, the rate of object shift (i.e. a non-local type of OV) can in fact be shown to increase over time. This development too is suggested to be related to the new setting of the EPP parameter and the novel patterns of subject placement that follow from it.

本文作者首次尝试对拉丁/罗曼语系中主宾语词序的失落进行深入分析。基于杨（2000，2002）所提出的语言变迁的变分获取模型，作者提出，OV语法的最终衰落可以追溯到拉丁语法中的一种独立变化，即关于如何满足从句EPP（题元投射）要求的语法变化（Danckaert 2017）。有观点认为，晚期拉丁语的关键创新是主语可选的A-位移的发展：作为这一变化的间接结果，在晚期拉丁语时期，尽管VO语序的整体频率在时间上基本保持稳定（甚至在某些环境中有所下降），VO语序的语法结构相较于OV语序的语法结构更为稳固。鉴于这一结果，杨的模型准确地预测了VO语序将取代竞争的OV模式。此外，我还展示了在拉丁语的历史进程中，对象外置（产生线性VO序列）的频率保持相对稳定，这表明该现象与VO语序的传播无关。最后，可以观察到对象位移（即一种非局部性的OV）的频率实际上随时间增加。这一发展也被认为与EPP参数的新设定以及由此产生的新的主语放置模式有关。