Languages differ in their complexity. One possible explanation for this observation is that differences in social factors influence linguistic complexity: languages that are used for communication in small-scale 'societies of intimates' exhibit greater complexity as a result of the communicative contexts in which they are typically employed. We used the techniques from referential communication studies across three experiments to assess the effects of two social group factors-group size and amount of communally shared knowledge-on the brevity and transparency of linguistic conventions. In Experiment 1, we explored the effects of a manipulation of group size, comparing the conventions which develop from the interaction of two speakers, with those which develop between three speakers. In Experiment 2, we manipulated the extent to which groups of three speakers share talk-relevant contextual information. While we found the conditions that involve larger groups and less shared background information initially resulted in longer labels and a greater reliance on more literal descriptive terms, there was no effect of either factor in the longer term. In Experiment 3, we investigated the transparency of the conventions of Experiments 1 and 2 by assessing how well they could be matched to their intended referents by naive individuals. We found no evidence to support the claims that communicative contexts involving communicating with more individuals, or individuals with whom less relevant information is shared, produce more transparent conventions. Our experiments ultimately provide no support for the idea that the structure of linguistic conventions is shaped by the groups in which they develop.

中文翻译：

---

社会群体对交际惯例的出现和语言复杂性的影响

语言的复杂程度各不相同。对这一观察结果的一种可能解释是，社会因素的差异会影响语言的复杂性：在小规模的“亲密社会”中用于交流的语言由于通常使用的交流环境而表现出更大的复杂性。我们在三个实验中使用参考传播研究中的技术来评估两个社会群体因素——群体规模和公共共享知识的数量——对语言惯例的简洁性和透明度的影响。在实验 1 中，我们探讨了操纵群体规模的影响，比较了从两个说话者的互动中发展起来的习惯与在三个说话者之间发展起来的习惯。在实验 2 中，我们操纵了三个发言者一组共享与谈话相关的上下文信息的程度。虽然我们发现涉及更大群体和较少共享背景信息的条件最初会导致更长的标签和更多地依赖更字面的描述性术语，但从长远来看，这两个因素都没有影响。在实验 3 中，我们通过评估实验 1 和实验 2 的约定与幼稚个体的预期参考对象的匹配程度来研究它们的透明度。我们没有发现任何证据支持这样的说法，即涉及与更多个人或与其共享较少相关信息的个人进行交流的交流环境会产生更透明的约定。