Friedrich-Alexander Universität Erlangen-Nürnberg, Germany.
CNRS & Université Paris Cité, LLF, France.
Cognition. 2025 Jan;254:105950. doi: 10.1016/j.cognition.2024.105950. Epub 2024 Sep 27.
The article presents four acceptability judgment experiments that evaluate novel predictions of the Focus-Background Conflict constraint (Abeillé et al. 2020, Cognition) with respect to the acceptability of long distance dependencies for so-called "subject islands" in English and French. In contrast with syntactic accounts, the Focus-Background Conflict constraint predicts differential behavior across different constructions. The current paper tests a novel prediction of this theory, in a construction that has not yet been tested experimentally: it-clefts. Experiment 1 shows that elements in clefted clauses are not uniformly backgrounded, contrary to a standard assumption in the syntax / discourse literature. Experiments 2-4 tested long-distance dependency relations in relative clauses and clefts. In both languages, there is strong evidence of a cross-construction difference when comparing the two constructions with each other: extraction of the subject complement out of a subject NP was super-additively difficult in clefts, but not in relative clauses, as predicted by the Focus-Background Conflict constraint.
本文呈现了四项可接受性判断实验,这些实验评估了焦点-背景冲突约束(Abeillé 等人,2020,认知)对英语和法语中所谓的“主语孤岛”的长距离依赖可接受性的新预测。与句法解释不同,焦点-背景冲突约束预测了不同结构的不同行为。本文在一个尚未经过实验测试的结构中检验了该理论的一个新预测:it-clefts。实验 1 表明,分裂句中的成分并非均匀地处于背景中,这与句法/语篇文献中的标准假设相反。实验 2-4 测试了关系从句和分裂句中的长距离依赖关系。在这两种语言中,当比较这两种结构时,有强有力的证据表明存在跨结构差异:与关系从句相比,主语 NP 中的主语补语的提取在分裂句中是超附加困难的,这与焦点-背景冲突约束的预测一致。