This paper examines the availability of single-conjunct agreement in number and gender in Bosnian/Croatian/Serbian. Reported are the results of an experiment in which coordinated singulars are included, as well as disjunction and negative-concord conjunction, next to the typically examined conjoined plurals. The research shows that, contra the general assumptions in the literature (Marušič, Nevins, and Saksida 2007, Marušič, Nevins, and Badecker 2015, Bošković 2009) but in line with earlier research (Moskovljević 1983, Bojović 2003), single-conjunct agreement does occur with coordinated singulars, especially in gender, even if less frequently. This paper shows that (i) first-conjunct agreement in gender preverbally and even last-conjunct agreement postverbally are produced above error level, and that the availability of collective interpretations for the coordinated subject influences the acceptability of the different agreement patterns available, and (ii) number and gender agreement do not have to target the same constituent. The findings shed light on the relation between the features of number and gender with regard to the issues of their bundling and simultaneous agreement, where the experimental results suggest that, while number tends to agree in a pattern that fits either semantic agreement or agreement with the entire conjunction, gender prefers to target single members of coordination, the first or the last. We speculate that a degree of “attraction” obtains, whereby number may attract gender to agree with the entire conjunction or gender may attract number to agree with a single conjunct. The results are used to compare two analyses offered in the literature—Marušič, Nevins, and Saksida 2007/Marušič, Nevins, and Badecker 2015 and Bošković 2009—showing that our empirical findings are problematic for both, but give a certain advantage to Marušič and his co-authors.

中文翻译：

---

论协调主体协议中的数-性别（独立）依赖性

本文研究了波斯尼亚语/克罗地亚语/塞尔维亚语中在数量和性别方面的单一连词一致性的可用性。报告了一项实验的结果，其中在通常检查的连体复数旁边包括并列单数，以及分离和否定协和连词。研究表明，与文献中的一般假设相反（Marušič, Nevins, and Saksida 2007, Marušič, Nevins, and Badecker 2015, Bošković 2009）但与早期研究（Moskovljević 1983, Bojović 2003 单一协议）一致，确实与协调的单数一起出现，尤其是在性别中，即使不那么频繁。本文表明（i）在性别上的第一连词一致甚至口头后最后连词都产生了高于错误水平，并且对协调主题的集体解释的可用性影响可用的不同协议模式的可接受性，以及 (ii) 数量和性别协议不必针对相同的组成部分。研究结果阐明了数量和性别特征之间在捆绑和同时一致性问题方面的关系，其中实验结果表明，虽然数量倾向于符合语义一致或一致的模式整个连接，性别更喜欢针对协调的单个成员，第一个或最后一个。我们推测获得一定程度的“吸引力”，即数字可能吸引性别与整个连词一致，或者性别可能吸引数字与单个连词一致。