Although it is widely assumed that the linguistic description of events is based on a structured representation of event components at the perceptual/conceptual level, little empirical work has tested this assumption directly. Here, we test the connection between language and perception/cognition cross-linguistically, focusing on the relative salience of causative event components in language and cognition. We draw on evidence from preschoolers speaking English or Turkish. In a picture description task, Turkish-speaking 3-5-year-olds mentioned Agents less than their English-speaking peers (Turkish allows subject drop); furthermore, both language groups mentioned Patients more frequently than Goals, and Instruments less frequently than either Patients or Goals. In a change blindness task, both language groups were equally accurate at detecting changes to Agents (despite surface differences in Agent mentions). The remaining components also behaved similarly: both language groups were less accurate in detecting changes to Instruments than either Patients or Goals (even though Turkish-speaking preschoolers were less accurate overall than their English-speaking peers). To our knowledge, this is the first study offering evidence for a strong—even though not strict—homology between linguistic and conceptual event roles in young learners cross-linguistically.

中文翻译：

---

在致因事件中代表代理、患者、目标和工具：早期语言和认知的跨语言调查

尽管人们普遍认为事件的语言描述是基于在感知/概念层面对事件成分的结构化表示，但很少有实证工作直接测试过这一假设。在这里，我们跨语言测试语言和感知/认知之间的联系，重点关注语言和认知中因果事件成分的相对显着性。我们从说英语或土耳其语的学龄前儿童那里汲取证据。在图片描述任务中，说土耳其语的 3-5 岁儿童比说英语的同龄人更少提到特工（土耳其语允许主题放弃）；此外，两个语言组都比目标更频繁地提到患者，而工具比患者或目标更频繁。在改变盲测任务中，两个语言组在检测代理的变化方面同样准确（尽管在代理提及方面存在表面差异）。其余部分的表现也相似：两个语言组在检测仪器变化方面的准确度都低于患者或目标（尽管说土耳其语的学龄前儿童总体上不如说英语的同龄人准确）。据我们所知，这是第一个提供证据证明跨语言的年轻学习者的语言和概念事件角色之间存在强烈的（尽管不是严格的）同源性的研究。两个语言组在检测仪器变化方面的准确度都低于患者或目标（尽管说土耳其语的学龄前儿童总体上不如说英语的同龄人准确）。据我们所知，这是第一个提供证据证明跨语言的年轻学习者的语言和概念事件角色之间存在强烈的（尽管不是严格的）同源性的研究。两个语言组在检测仪器变化方面的准确度都低于患者或目标（尽管说土耳其语的学龄前儿童总体上不如说英语的同龄人准确）。据我们所知，这是第一个提供证据证明跨语言的年轻学习者的语言和概念事件角色之间存在强烈的（尽管不是严格的）同源性的研究。