【作者】 蒲玥； 韩世辉；

【Author】 Yue Pu;Shihui Han;School of Psychological and Cognitive Sciences,Peking University;

【机构】 北京大学心理与认知科学学院；

【摘要】 人们面对动物时,通常只需识别其属于哪一类别,而不用区分同一类别的不同个体:种类信息已足够用于绝大部分行为决策。生物分类学中,人类也属于广义的动物,但人类面对同类很少进行"他/她是人类"的类别加工,而更注重区别个体间的不同。本研究试图通过行为实验,证明对动物与人类的这一识别策略差异,并通过脑电实验探究这一认知加工的神经过程。行为实验借鉴了视觉搜索范式:我们假设,若对某一类别倾向于采用不区分个体的分类加工,那么该类别作为搜索分心物时的数量对于搜索速度影响较小;而倾向于对类别内个体进行区分的类别,随着数量增多,作为分心物对搜索的干扰效果更强。实验中我们将人类的轮廓图片与黑猩猩、狗、白头雕三种动物的轮廓图片进行组合,每种组合中人类与动物互为分心物/搜索目标,分心物的数量有7/11/15三种水平,形成3(动物种类)×3(分心物数量)×2(人类/动物是搜索目标)的组内设计。结果发现,人类作为分心物时,搜索动物的速度显著受到分心物数量的影响;搜索人类的速度则几乎不受动物分心物的数量影响。脑电实验采用了重复抑制范式:对于被分类加工的类别,类别中的不同个体连续呈现将使类别信息被重复加工,相比非单一类别交替呈现,类别相关的神经过程将被抑制。与行为实验类似,我们将人类与上述三种动物分别组合;动物与人类都有单类别重复呈现序列,以及两种类别交替呈现序列,形成3(动物种类)×2(呈现序列)×2(人类/动物)的组内设计。结果发现,动物均在P2这一ERP成分上引发显著的重复抑制效应,人类则未引发此效应。我们将图片全部倒置进行对照实验:准确识别客体变得困难,但依然能大致区分不同物种。人类图片则也引发P2的重复抑制效应。行为与脑电实验一致证明了"分类加工动物、差异化加工人类"的这一识别策略差异,并且对人类的差异化加工依赖于有效提取其差异化信息。这一认知过程的对应神经活动大概在目标出现后200ms左右发生。更多还原

【Abstract】 When faced with animals, people usually categorize them into certain species, but not individualize them:the category information is enough for decision-making in the most cases. In biological taxonomy, human beings also belong to the generalized animals. We seldom process somebody with the classification of human beings, but prefer to distinguish different individuals. The current study aimed to demonstrate this distinction of recognizing strategies with behavioral and EEG experiments, and in the meantime, explored the temporal neural underpinnings of the processing. The behavioral experiment drew on the paradigm of visual search. We hypothesized that the increase of the number of distractors would not slow down the search performance, if subjects preferred to categorize the distractors rather than individualize them. On the contrary, if people liked to distinguish different individuals within the category, which served as the distractors, the increasing number of distractors would bring more interference to the search task. In the experiment, we paired human beings with chimpanzees, dogs and eagles respectively. All of them were represented by black contours. In each pair, human beings and animals were targets/distractors for each other, and we set 3 levels for the number of distractors, 7, 11, and 15. Therefore, the experiment was a 3(animal species) × 3(levels of the distractor number) × 2(humans as targets/distractors) within-subject design. The results showed that when human beings acted as the distractors, the speed of animal searching was slowed down with the increase of the distractor number. In contrast, the human searching was not significantly influenced by the number of animal distractors. The EEG experiment used the repetition suppression paradigm: if individuals within one category were treated only as examples of this category, the repetitive presentation of different individuals belonging to one category would suppress the related categorical processing, compared with the alternative presentation of several categories. Similar to our behavioral design, 3 nonhuman animal species were respectively combined with human beings, each of which included repetitive and alternative sessions for both animals and humans. This EEG experiment was a 3(animal species) × 2(repetitive/alternative sessions) × 2(humans/animals) within-subject design. The results demonstrated that 3 kinds of animals evoked significant repetition suppression effect on the P2 component, but this effect didn’t come to humans. In addition, we used inverted pictures to conduct one control experiment, with the totally identical design to the above. When observing inverted contours, it was hard for subjects to clearly recognize the objects, but the intact contour information enabled subjects roughly distinguish different species. In the control EEG experiment, human beings also evoked the repetition suppression effect on P2, just like nonhuman animals. To sum up, our behavioral and EEG experiments proved that people take distinct strategies to recognize animals and human beings: categorical for animals, but individualized for human beings. And the corresponding neural processing takes place about 200 ms after the target onset.更多还原