The developmental origins of object categories: how motion cues and labels bootstrap early categorization skills

将世界划分为有意义的类别是一项具有挑战性的任务，特别是考虑到在这一过程中可能会招募到大量信息(Cohen&Lefebvre，2017)。早期出现的分类技能使孩子们能够理解他们环境中各种各样的物体和实体。特别是，将一个特定的实体归类为有生命的或无生命的被认为是人类感知和认知的一个基本过程(例如，Opfer&Gelman，2010)。为了形成这样的类别，儿童必须认为实体是相似的，也就是说，使用某种类型的相似规则将它们分组在一起。婴儿如何成功地将世界组织成有意义的类别？近年来，在研究婴儿形成类别的条件以及他们在此过程中考虑的信息类型方面取得了巨大的进展(参见Pauen，Birgit，Hoehl&Bechtel，2015和Poulin-Dubois&Pauen，2017)。使用适合于说话前婴儿的研究程序，我们发现婴儿在生命早期对动物、车辆和家具等全球类别的区分早于每个全球类别中更细粒度的基本类别，如猫和狗、椅子和桌子或汽车和卡车(Pauen，2002)。自从发现这一转变以来，研究人员一直试图了解言语前年龄类别歧视背后的心理过程，从而专注于婴儿在语言开始发挥作用之前是否发展出基于知识的类别表征(Poulin-Dubois&Pauen，2017)。然而，人们对儿童将注意力集中在与分类最相关的东西上的机制知之甚少。婴儿如何确定哪些对象要相互比较？他们如何确定要关注这些对象的哪些属性？哪些运动提示最先与有生命的生物联系在一起？生物运动是这种区别的特权暗示吗？最后，语言的出现如何促进基本层面等包容性较低层面的宾语范畴化？关于最后一个问题，标签被认为可以通过允许婴儿专注于共同点来促进物体分类(Ferguson&Waxman，2017)。当同一名词一致地用于一组不同的对象时，婴儿和成年人都更有可能将它们表示为同一对象类别的成员(Keates&Graham，2008；Lupyan，Rakison&McClelland，2007；Waxman&Booth，2001；2003)。然而，标签的确切功能以及双语儿童在形成类别时是否同样受益于语言，这一点并不清楚，因为他们的策略是将一个以上的单词映射到一个物体。将在当前研究方案范围内进行的一系列研究试图填补这些空白。

Partitioning the world into meaningful categories is a challenging task, especially considering the vast amount of information that could be recruited in the process (Cohen & Lefebvre, 2017). The early emergence of categorization skills allows children to make sense of the vast diversity of objects and entities in their environment. In particular, to categorize a given entity as being either animate or inanimate is thought to be a fundamental process characterizing human perception and cognition (e.g. Opfer & Gelman, 2010). In order to form such categories, children must perceive entities as alike, that is, employ some type of similarity rule to group them together. How do infants successfully organize the world into meaningful categories? In recent years, tremendous strides have been made toward investigating the conditions under which infants form categories, as well as the types of information that they consider in doing so (see Pauen, Birgit, Hoehl, & Bechtel, 2015 and Poulin-Dubois & Pauen, 2017 for recent reviews). Using research procedures that are appropriate for preverbal infants, it has been found that infants discriminate global categories like animals, vehicles, and furniture earlier in life than more fine-grained basic-level categories within each global class, such as cats and dogs, chairs and tables, or cars and trucks (Pauen, 2002). Since discovering this shift, researchers have been trying to understand the mental processes underlying category discrimination at a preverbal age, thereby focusing on whether infants develop knowledge-based representations for categories before language comes into play (Poulin-Dubois & Pauen, 2017). However, little is known about the mechanisms that allow children to focus their attention on what is most relevant to categorization. How do infants determine which objects to compare to each other? How do they determine which properties of those objects to focus on? Which motion cues are first associated to animate beings? Is biological motion a privileged cue for such distinction? Finally, how does the emergence of language promote categorization of objects at less-inclusive level, such as the basic level? Regarding the last issue, labels are considered to boost object categorization by allowing infants to focus on commonalities (Ferguson & Waxman, 2017). When the same noun is applied consistently to a set of distinct objects, both infants and adults alike are more likely to represent them as members of the same object category (Keates & Graham, 2008; Lupyan, Rakison & McClelland, 2007; Waxman & Booth, 2001; 2003). However, the exact function of labelling is unclear as well as whether bilingual children benefit as much from language when forming categories given their strategy to map more than one word to an object. The series of studies to be conducted in the context of the current research program attempts to fill these gaps.**