Neural regionalization of knowledge access: preliminary evidence.
Neural regionalization of knowledge access: preliminary evidence.
复制标题
知识获取的神经区域化：初步证据。
DOI：
10.1101/sqb.1990.055.01.098
复制
发表时间：
1990
期刊：
影响因子：
--
通讯作者：
Brandt,JP
中科院分区：
文献类型：
--
作者：
Damasio,AR;Damasio,H;Tranel,D;Brandt,JP
我们之前已经注意到，一些双侧枕颞皮质损伤的患者在视觉上无法识别大量的自然实体（Damasio et al. 1982, 1990; Damasio, 1990）。一个例子是病人EH-034，他除了不能识别面孔、建筑物、动物和物体的独特身份外，也不能识别出这些实体中的许多实体是一个类别的成员。面对狐狸或狮子的图片，病人说这是“动物”，但具体的类别她想不起来。给她看一只知更鸟，她说那是一只“鸟”。她无法指出这种未被识别的物种的大致大小、栖息地或典型行为。正如我们在其他地方指出的那样，失败并没有涵盖整个概念范畴（所有动物或所有生物）。例如，一些动物很容易被识别出来，例如大象和长颈鹿，这表明，尽管属于一个独特的物种，但形状更相似的动物比那些具有全局或局部“异常”形状的不太典型的动物表现得更好，从而形成了视觉上“模糊”的分组。相比之下，人造实体大多很容易辨认，例如，工具和器具不构成问题。然而，即使在这些人中，也有例外。例如，一些乐器，它们像任何花园工具或厨房用具一样是人造的和可操作的，通过视觉路线无法识别，超出了它们作为乐器的上级任务。这与它们对声音的具体而迅速的识别形成了对比。其他实验室的研究也在我们认为神经系统受损的患者身上发现了几乎相同的结果（Warrington and Shallice 1984; Warrington and McCarthy 1987; Young et al. 1989），但对这些发现的解释并不相同。例如，沃灵顿提出，证据支持专门用于概念类别的系统的存在（例如，生命/非生命和有生命/无生命），这些系统将被复制到每一种感觉形态。如下所述，我们以不同的方式解释这些发现（即使我们的解释与实体感官特征的重要性一致）。对我们来说，迄今为止可用的数据表明，实体背后的概念的调用过程是基于不同的神经系统，这取决于每个实体的性质以及与该实体相关的学习和交互历史（Damasio 1990）。特定神经系统的损伤损害了对某些实体背后概念的理解，但对其他实体却保留了对概念的完整理解。在同一系统受损的患者中，识别受损或完整的实体之间的分割线似乎不尊重直觉概念类别的边界。相反，受损和完整的实体似乎根据各种因素进行分组，在这些因素的基础上可以形成几个子类别，例如，相似形状和操作的组合。有时，这些子类别与传统的概念类别重叠，但它们不一定是共延的。
We have previously noted that some patients with bilateral damage in occipitotemporal cortices fail to recognize visually a large number of natural entities (Damasio et al. 1982, 1990; Damasio, 1990). An example is patient EH-034, who, in addition to not recognizing the unique identity of faces, buildings, animals, and objects, also did not recognize many of those entities as members of a category. Confronted with the picture of a fox or a lion, the patient said it was" an animal," but the specific category eluded her. Shown a robin, she said it was a" bird." She was unable to indicate the approximate size, habitat, or typical behavior of the unrecognized species. As we have pointed out elsewhere, the failure did not cover entire conceptual categories (all animals, or all living things). For instance, some animals were readily recognized, eg, elephant and giraffe, suggesting that less prototypical animals with either global or local" outlier" shapes fared better than those which, despite belonging to a distinctive species, had a greater similarity of shapes and thus formed a visually" ambiguous" grouping. In contrast, man-made entities were for the most part readily recognized, eg, tools and utensils posed no problem. Even among those, however, there were exceptions. For instance, several musical instruments, which are as manmade and manipulable as any garden tool or kitchen utensil, were not recognized through the visual route, beyond their supraordinate assignment as musical instruments. This stood in contrast to their specific and rapid recognition from sound. Work from other laboratories has revealed virtually identical findings in patients whom we presume to have damage in comparable neural systems (Warrington and Shallice 1984; Warrington and McCarthy 1987; Young et al. 1989), but the interpretation of those findings has not been identical. For instance, Warrington has suggested that the evidence favors the existence of systems dedicated to conceptual categories (eg, living/nonliving and animate/inanimate), which would be duplicated for each sensory modality. As noted below, we account for the findings in a different way (even if our explanations coincide in the importance accorded to the sensory characteristics of entities). To us, the data available so far have suggested that the process of evoking the concept behind an entity is based on different neural systems, depending on the nature of each entity and the history of learning and interaction relative to that entity (Damasio 1990).Damage within a given neural system impairs access to the concepts behind some entities, but leaves intact access to concepts for other entities. In patients with damage to the same system, the breakdown line between entities whose recognition is impaired or intact does not appear to respect the boundary of intuitive conceptual categories. Instead, impaired and intact entities appear to be grouped according to a variety of factors on the basis of which several subcategories can be formed, eg, a combination of similar shapes and operations. On occasion, those subcategories overlap with traditional conceptual categories, but they are not necessarily coextensive.