Only half right: comment on Regier and Kay.
Only half right: comment on Regier and Kay.
复制标题
只对了一半：对雷吉尔和凯的评论。
DOI：
10.1016/j.tics.2009.10.004
复制
发表时间：
2009
影响因子：
19.9
通讯作者：
Roberson D
中科院分区：
文献类型：
--
作者：
Roberson D
我们接受Regier和Kay的论点，即语言学术语影响右侧而不是左侧视野（LVF）的颜色分类[1]。我们也同意左右半球可能包含不同的色彩处理系统。我们不同意，但是，与建议，右半球包含一组语言前的颜色类别，影响知觉时，刺激在左心室纤维。他们的论点建立在研究[2，3]的基础上，这些研究似乎显示了语言前婴儿对颜色的分类感知。关键的发现是，婴儿从边缘绿色背景到蓝色目标的眼球运动比到更中心的绿色目标的眼球运动更快。不幸的是，这些研究中的婴儿只接受了两对刺激的测试。因此，结果可能反映了婴儿对蓝色而不是绿色目标的简单颜色偏好[4]。这一结论得到了最近研究结果的支持[5，6]，在这些研究中，婴儿对绿色色调的注视时间与注视其他色调的平均时间相比显示出显著的负偏差。（关于婴儿类别知觉（CP）的早期研究要么存在重大的方法论缺陷，要么没有被复制，正如[7]中所讨论的那样，所以我们在这里不考虑它们。）如果婴儿仅仅因为对颜色的偏好而将目光更快地移向蓝色，而不是中央的绿色，那么就不需要解释右半球的颜色类别。在我们得出结论之前，我们需要对来自不同语言社区的参与者使用各种不同颜色类别的婴儿行为进行更多样化的调查，以证明语言前婴儿对颜色的CP反映了语言使用者的影响。人类可以在大约200万种不同的颜色之间进行精细的感知区分。我们最近的研究结果表明，歧视在类别边界上并不比类别内更敏感[8]。因此，我们相信存在一个非分类的颜色处理系统（可能但不一定限于右半球），它可以对颜色进行非常精细的区分，并评估两种颜色是否相同。我们不相信这个系统“知道”关于两种色调之间的相似性和差异的精确信息（例如，一种比另一种更明亮或更饱和，或者两种不同的色调共享相同的名称）。一个非分类系统可能会更好地进行精细的感知区分，因为它不是分类的。语言使用者的合作原则很可能是由知觉系统和范畴系统的同时运作而产生的。例如，当判断两种不同深浅的蓝色是否相同时，来自左半球的分类系统的信息表明它们具有相同的名称，而来自知觉系统的信息表明它们不同。对不同颜色类别的物品做出决定更快，不是因为对类别边界的感知变化的敏感性增加，而是因为分类和感知系统都表明它们不同。在较慢的反应时间，CP发生在两个视野中的刺激[9，10]，因为呈现给LVF的信息有时间进入左半球的分类系统。因此，如果一个处理系统是基于语言的和分类的，而另一个系统是感知的，但不是分类的，那么颜色的CP就很容易解释。
We accept Regier and Kay’s argument that linguistic terms affect color categorization in the right but not the left visual field (LVF)[1]. We also agree that left and right hemispheres might contain qualitatively different color processing systems. We disagree, however, with the suggestion that the right hemisphere contains a set of prelinguistic color categories that influence perception when stimuli are presented in the LVF. Their argument rests on studies [2, 3] that seem to show categorical perception of color by prelinguistic infants. The key finding was that infants showed faster eye-movements from a borderline green background to a blue target than to a more central green. Unfortunately, infants in these studies were tested on only two pairs of stimuli. Consequently, the results might have reflected simple color preference [4] by the infants for the blue rather than the green target. Such a conclusion is supported by results from recent studies [5, 6] in which infants showed a significant negative bias for looking times to a green hue compared to the average time spent looking at other hues.(Earlier studies of infant categorical perception (CP) either had major methodological flaws or have not been replicated, as discussed in [7], so we do not consider them here.) No explanation in terms of color categories in the right hemisphere is required if infants shift their gaze faster towards blue than to central green on the basis of simple color preference. More varied investigations of infant behavior using a variety of different color categories with participants from different linguistic communities are required before we can conclude that prelinguistic infants demonstrate CP for color that mirrors the effects found in language users. Humans can make fine perceptual discriminations between approximately two million different shades of color. Our recent findings indicate that discriminations are no more sensitive at category boundaries than within categories [8]. We therefore believe there to be a noncategorical color processing system (possibly, but not necessarily, restricted to the right hemisphere) that can make extremely fine discriminations between colors and assess whether two colors are identical. We do not believe that this system ‘knows’ precise information concerning similarities and differences between two shades of color (eg that one is brighter or more saturated than another, or that two different shades share the same name). A non-categorical system might work better at making fine perceptual discriminations because it is not categorical. CP in language users probably arises from the concurrent operation of the perceptual and categorical systems. When judging, for example, whether two different shades of blue are identical, information from the left-hemisphere categorical system indicating that they share the same name conflicts with information from the perceptual system that they differ. Decisions for items from different color categories are made more quickly not because of increased sensitivity to perceptual change at category boundaries, but because both categorical and perceptual systems indicate that they differ. At slower response times, CP occurs for stimuli in both visual fields [9, 10] because information presented to the LVF has had time to access the categorical system in the left hemisphere. CP for color can therefore be readily explained if one processing system is languagebased and categorical and the other system is perceptual but not categorical.