Bonatti Luca, Frot Emmanuel, Zangl Renate, Mehler Jacques
Université de Paris VIII, France.
Cogn Psychol. 2002 Jun;44(4):388-426. doi: 10.1006/cogp.2002.0779.
How do infants individuate and track objects, and among them objects belonging to their species, when they can only rely on information about the properties of those objects? We propose the Human First Hypothesis (HFH), which posits that infants possess information about their conspecifics and use it to identify and count objects. F. Xu and S. Carey [Cognitive Psychology, 30(2), 111-153, 1996] argued that before the age of 1 year, infants fail to use property information. To explain their results, Xu and Carey proposed the Object First Hypothesis (OFH), according to which infants under 1 year of age have only the general concept of physical object to identify and count objects. We show that infants have a more extensive knowledge of sortals than that claimed by the OFH. When 10-month-olds see one humanlike and one non-humanlike object, they successfully identify and count them by using the contrast in their properties, as predicted by the HFH. We also show that infants succeed even when they make a decision based on differences between two close basic-level categories such as humanlike objects and doglike objects, but fail when they have to use differences within the human category. Thus, infants treat "human" as a basic sortal, as predicted by the HFH. We argue that our results cannot be accounted for by general purpose mechanisms. Neither the strong version of the OFH and its explanation in terms of object indexing mechanisms [A. M. Leslie, F. Xu, P. Tremoulet, & B. J. Scholl, Trends in Cognitive Sciences, 2(1), 10-18, 1998] nor explanations in terms of task demands [T. Wilcox & R. Baillargeon, Cognitive Psychology, 37(2), 97-155, 1998] are sufficient to explain our results.
当婴儿只能依靠有关物体属性的信息时，他们如何区分和追踪物体，尤其是属于自己物种的物体呢？我们提出了“人类优先假设”（HFH），该假设认为婴儿拥有关于同种个体的信息，并利用这些信息来识别和计数物体。F. 徐和S. 凯里[《认知心理学》，第30卷第2期，第111 - 153页，1996年]认为，1岁之前的婴儿无法利用属性信息。为了解释他们的研究结果，徐和凯里提出了“物体优先假设”（OFH），根据这一假设，1岁以下的婴儿只有物理物体的一般概念来识别和计数物体。我们表明，婴儿对类别词的了解比OFH所声称的更为广泛。正如HFH所预测的那样，当10个月大的婴儿看到一个类人物体和一个非类人物体时，他们会利用它们属性上的差异成功地识别并对其进行计数。我们还表明，即使婴儿基于两个相近的基本层次类别（如类人物体和类狗物体）之间的差异做出决策时也能成功，但当他们必须利用人类类别内部的差异时则会失败。因此，正如HFH所预测的那样，婴儿将“人类”视为一个基本类别词。我们认为，我们的研究结果无法用通用机制来解释。无论是OFH的强版本及其基于物体索引机制的解释[A. M. 莱斯利、F. 徐、P. 特里穆利特和B. J. 肖尔，《认知科学趋势》，第2卷第1期，第10 - 18页，1998年]，还是基于任务需求的解释[T. 威尔科克斯和R. 拜拉吉昂，《认知心理学》，第37卷第2期，第97 - 155页，1998年]，都不足以解释我们的研究结果。
